10 Luftballoons Go High November 2, 2009 10:21 PM   Subscribe

The Darpa Network Challenge: Let's talk strategy.

I think there was some consensus that we were going to actually make an attempt at this in the thread, so let's start planning. We only have a month. I think we honestly can make a decent go of it, and will probably get some news coverage if we get into the top tier of teams (and I don't see why we couldn't).

I think we'll need both people on the streets, but also a team of people scouring the internet for tips and sitings and getting people out on the road to confirm it, so everybody should be able to participate, no matter where you are.

So -- what are the first steps? Promotion here obviously -- but we also need a private website to coordinate, and a small team of volunteers (besides Cortex) willing to get stuff organized.

What else?
posted by empath to MetaFilter Gatherings at 10:21 PM (303 comments total) 22 users marked this as a favorite

(by consensus, I obviously just mean that Cortex mused out loud about the possibility).
posted by empath at 10:33 PM on November 2, 2009


I say we just buy our own damn balloons and put them wherever the hell we want.
posted by Sys Rq at 10:34 PM on November 2, 2009 [4 favorites]


That's an option. Putting up easily spotted fake balloons in major metro areas. All it would take is a few of them in NYC, DC, San Fran and LA to spoil the game for a lot of teams who can't put boots on the ground to confirm.
posted by empath at 10:38 PM on November 2, 2009


this is going to be awesome.
posted by milestogo at 11:13 PM on November 2, 2009


Note that the real balloons will be accompanied by DARPA representatives so good decoys will need to be accompanied by good actors. I'd love to see any links to cheap 8' red balloons, though.
posted by contraption at 11:15 PM on November 2, 2009


$24 for an 8-footer here
posted by contraption at 11:23 PM on November 2, 2009 [2 favorites]


Have you tried image searching for DARPA? the first page of results is like a bucket of crazy.
posted by qvantamon at 11:24 PM on November 2, 2009 [6 favorites]


I'm game! Count me in. I'm in the San Francisco area.

I've been fiddling around with google wave and haven't the foggiest idea what to use it for. But This.. This would be a perfect way to use it. Alas, I only have a few invites left. Too bad it's still invite only.

I have a couple other ideas, but I think they should be kept to a private forum.
posted by trixare4kids at 11:29 PM on November 2, 2009


Until DARPA clarifies exactly how they will be accepting entries I think a lot of strategy is up in the air. Do you really only get one shot with up to ten slots? Or can you submit many guesses, and how fast, and at what penalty? Etc...
posted by Rhomboid at 11:30 PM on November 2, 2009


People needed:

1: City by city teams -- folks with cars, iphones and a willingness to drive around all day on potentially a wild goose chase.

2: Co-ordinators -- People who can collect names and numbers and all the relevant organizational stuff from volunteers and make sure the right people get called on when needed.

3: Google geniuses -- people that can scour twitter, blogs, technorati, whatever trying to find sitings, fake or otherwise.

4: Black Ops -- info on a need to know basis...
posted by empath at 11:47 PM on November 2, 2009


1: City by city teams -- folks with cars, iphones and a willingness to drive around all day on potentially a wild goose chase.

I think this will be less of a problem than identifying willing goose-chasers in less-populous areas. Metro areas will be nice because we can in theory cover a good three-digit radius outward by sending folks off in spokes, but rural areas far from major metros will be harder to cover.

As far as talking behind the scenes, it might be good idea for someone to set up a listserv or something similar. Invite only, sanity checked against mefi if we're going to be properly paranoid about infiltrators. Which I don't think is probably the biggest concern we have, since I personally expect clean planning and execution of a search to be the decisive factor for a winning team, but no harm done in taking care.
posted by cortex (staff) at 11:58 PM on November 2, 2009 [1 favorite]


Would probably be a good idea, just in this case, not to do it through google groups.
posted by empath at 12:08 AM on November 3, 2009 [1 favorite]


In fact, I'd prefer some sort of private forum software hosted on a secure box.
posted by empath at 12:08 AM on November 3, 2009


Worry, worry, super-scurry.
Call the troops out in a hurry.
posted by Ambrosia Voyeur at 12:32 AM on November 3, 2009 [7 favorites]


I'm interested in helping.
posted by one_bean at 12:33 AM on November 3, 2009


1: City by city teams -- folks with cars, iphones and a willingness to drive around all day on potentially a wild goose chase.

Two types of city teams; searchers & verifiers. I don't know if want to codify our organization at this point or am I jumping ahead?
posted by scalefree at 12:52 AM on November 3, 2009


Two things that concern me if you're planning on faking it:
1. If you do plan on faking it, give photo evidence of you setting up the fakery and send it to whomever is the MeFi contact point/admins.
2. Get some wardrobe to prepare to look like official DARPA. If the first one discovered is wearing a suit, then suit up! If I were DARPA, I would have everyone wearing matching T-shirts with the logo of DARPA/the challenge, so maybe have an inkjet printer and a white t-shirt handy (the nearest drugstore should have iron-on print sheets in their stationery section).
posted by amuseDetachment at 12:53 AM on November 3, 2009


Scalefree: I think searching and verifying are two different activities that could be done by the same person; same with other activities; it's useful to know what people will be doing but obviously not to restrict them.

Empath has mentioned this a couple of times, but it bears repeating - I think that a 'team of teams' model will work very well here. You might get even a few thousand Mefiers in the US participating, but it won't be enough to cover all the ground - but if they can recruit their own local teams that they can co-ordinate and verify, that'll really expand the reach. This will require really good communication and an ability to clearly explain what the competition is about and how to send in information.

More general thoughts - I see a lot of groups getting excited online, but not actually doing much. The guys who win this are the people who use the month's lead time wisely and get their infrastructure up and running, rehearse everything, get the word out, etc. *EVERY HOUR* is vital here.
posted by adrianhon at 1:02 AM on November 3, 2009


Incidentally, if you're really thinking about faking balloons, here's how to do it right: don't just send up one, send up ten around the country. If you're the only guy wearing an iron-on Darpa T-shirt, it'll be obvious you're the odd one out and word will quickly spread that it's a fake.

But if you have ten people spread across the country with official-looking T-shirts and ID cards, then you can legitimately claim that *everyone else* is a fake; at the very least, you'll slow a lot of people down.

I still think that it's a bit boring doing that though.
posted by adrianhon at 1:06 AM on November 3, 2009


The city teams can be recruited among non-Mefites as long as they have a member at the core acting as a repeater for info, or are we afraid of false reports from our own teams? There's some point where we draw the line, just where is it?
posted by scalefree at 1:10 AM on November 3, 2009


I have just purchased a slice for use on this project. I'm willing to give cortex and pb root if they want it. I'm also willing to volunteer a little time to set up a mailing list or some forum software. If somebody else would buy the domain, that'd be great. I've spent my $20 (sameasintown), now who else is chipping in?

Please don't memail me for details on getting access. I'm going to completely defer to cortex and pb on that front. Mail from other people will be ignored.
posted by tarheelcoxn at 1:12 AM on November 3, 2009


Given the paucity of information in the rules, it's worth the energy to try to reverse engineer the challenge. What will they be watching for? What are they trying to achieve? How would they place the balloons to best achieve it? This could help us focus in on strategies/locations/etc.

I've got some thoughts along those lines, but I'll wait until we've got a listserv up.
posted by wemayfreeze at 1:28 AM on November 3, 2009


Might I suggest moving some of the communication to Wave? Just to see if we can make something of it? I know not nearly everybody's on it but quite a few are. I want to be experimental, try some things just to get data back on how it worked.
posted by scalefree at 1:31 AM on November 3, 2009


I'm just going to open up a Wave just to start a channel there. Go to the Metafilter wave if you can't find it.
posted by scalefree at 1:32 AM on November 3, 2009


Sounds like a great use of the Wave. Any spare invites around?
posted by wemayfreeze at 1:34 AM on November 3, 2009


I live in Moscow Idaho. I will look for balloons. I am willing to bet a significant amount of money that I won't find any though. So go team I guess? Memail me if I can help in any way. (TBH I am also a member of the SA forums. So if we are not trusting people who have dual citizenship then probably nobody should memail me. Now the real question is did I say that because I'm trustworthy or because I figured anyone who did any checking would catch the dual citizenship? :) )
posted by Peztopiary at 1:37 AM on November 3, 2009


MeMail me if you want on the Wave. Nothing happening there yet. wemayfreeze, check your MeMail too.
posted by scalefree at 2:05 AM on November 3, 2009


Reverse engineering this contest? The best way I can come up with is I suspect that DARPA may want to play this all legal-like and contact the city for a permit. At the day-of, if we can confirm one, a determined group can start calling around cities to see whether they requested a permit. It'd be a lot faster than driving around.
posted by amuseDetachment at 2:10 AM on November 3, 2009


You need a Wave account already, which makes it an imperfect medium but we knew that.
posted by scalefree at 2:27 AM on November 3, 2009


I may have some wave invites - will have to check again when not on the crackberry. I'm not even in the US until early next year but am happy to donate some invites and any other effort that may be useful to the team.
posted by allkindsoftime at 3:43 AM on November 3, 2009




It's buck easy for DARPA to prevent griefing via false balloons. All they have to do is write a alphanumeric hash of a few dozen letters on each balloon and then provide a verification web site requiring location + hash. Or have the DARPA rep hand the hash out.

Being outside the states I'm not sure what help I can provide but I'm willing so please add me to any group lists.
posted by Mitheral at 4:59 AM on November 3, 2009


Ok Jessamyn did the original FPP, Cortex is gung ho, is this a conspiracy by the mods to make us fail and then make us feel bad about ourselves so we wont have so many flame outs?
posted by wheelieman at 5:05 AM on November 3, 2009


Maybe this is obvious, but surely many of the MeFites who live in cities didn't grow up in them. If we can wrangle relatives still in home rural-ish areas, that'd be useful, right?
posted by lauranesson at 5:06 AM on November 3, 2009


Teams will need to be equipped with GPS as well.

My experience so far with Wave is that it's slow, buggy, and difficult to use. I'd prefer to use some other tool to collaborate.
posted by bondcliff at 5:48 AM on November 3, 2009


TBH I am also a member of the SA forums. So if we are not trusting people who have dual citizenship then probably nobody should memail me. Now the real question is did I say that because I'm trustworthy or because I figured anyone who did any checking would catch the dual citizenship?

This kind of thing is going to be a problem for us. We need to get to a more secure place as quickly as possible. We also need to make some decisions about who has access to the secure place.
posted by Kwine at 6:10 AM on November 3, 2009


I suspect secrecy will be our biggest obstacle. Obviously strategy should be discussed in a private location, but who's allowed in? Anybody with a Mefi account? How do we keep out spies from competing teams? Too much restriction, on the other hand, will greatly limit the effectiveness of a team as big as this one.

For that matter, how do we put up decoys for the other teams without our fellow teammates falling for them too? If I saw a red balloon with an official-looking attendant, I could say a code word to identify myself as a Mefite (and allow him/her to confirm if it's fake). But if I know the code word, anyone else could know it too.

I think a vital tool for us is a private (if we can figure out what "private" means) list of confirmed fake coordinate pairs. There will be many claims floating around and we need a way to quickly track their authenticity. If the list is going to be reliable, it can only be updated by a fixed number of trusted people who can travel to all of the coordinates and personally verify them. If we don't have a mechanism for independently verifying every report, we'll never win this thing. The trick, again, is finding enough trustworthy authenticators in each of the 48 states, urban and rural alike. I'd guess that restricting to Mefi memberships older than a year would be a good, but not foolproof, first step (that also rules me out, incidentally).
posted by The Winsome Parker Lewis at 6:13 AM on November 3, 2009


A couple of thoughts:

(1) Anyone planning on planting one or more fake balloons, keep in mind the FAA regs on moored balloons if you want to be legal.

(2) We should have a plan for the money if we happen to win.
posted by exogenous at 6:26 AM on November 3, 2009 [1 favorite]


I suspect secrecy will be our biggest obstacle. Obviously strategy should be discussed in a private location, but who's allowed in? Anybody with a Mefi account? How do we keep out spies from competing teams? Too much restriction, on the other hand, will greatly limit the effectiveness of a team as big as this one.

How about this setup:

Coordinator(s): A very small set of people who are absolutely trusted to keep the information safe run the whole project. They receive all of the data from the rest of the team, but only share it on a need-to-know basis.

Researchers: Anyone with an Internet connection can be scouring various other sites looking for leads. This may involve having some people function as moles in other groups to steal leads from other teams. The information would only be forwarded to the coordinator, and not the group as a whole.

Spotters: People with access to vehicles and the ability to drive around an area for most of the time period actually looking for new balloons. It would probably be most effective if these people were from more rural parts of the country. Once they spotted a balloon, they would probably verify it themselves and send the information to the Coordinators.

Verifiers: Probably the most important role, these people would receive unverified leads from the Coordinators and travel to the location to verify them. They would need a camera and a good GPS unit to make sure that there is no doubt that the balloon is legit and that the location coordinates are correct within the acceptable margin of error. Again, the verification would only be sent to the Coordinators and not to the group as a whole. It would also probably be a good idea for these people to register their locations in advance with the Coordinators, so that unverified locations would only be shared with people that can travel there in time.

People could have multiple roles, but the point is that all of the members of the group would more or less just do their job and send the results up to the top. With that setup, there would probably be some overlap and wasted effort due to the fact that any given member of the team will know very little about what other members of the team are doing, but the I think the challenge is setup in a way that there doesn't need to be much information sharing within the group.
posted by burnmp3s at 6:37 AM on November 3, 2009


As for parker lewis' question, I think we should allow in anyone who already has a mefi account signed up before October 28th or so.

I am definitely interested to participate and willing to help out.

*cough* I do not have a wave invite, so if we're organizing through that, I'd need one *cough*.
posted by fizzix at 6:40 AM on November 3, 2009


This is definitely a neat idea -- I think it would be nice to figure out different levels of participation so that people can decide how much time they can/would like to commit to this project. It looks like we've already started setting out roles, but sorting that out seems to me like a good first step (on preview, I think burnmp3s' suggestion seems like a good one). I also think it might be a good idea to see if there are any roles available for people who'd like to help but aren't super technically saavy and don't want to make a huge time commitment (for example, I might be willing to man a decoy balloon).

Also, although this is sort of premature, if we DID win I think it would be worth considering the feasibility of setting up in the money in some sort of "MetaFilter Help" fund. This might not work at all because of taxes or something but there have been a few times where someone has had a problem with, say, six hundred dollars worth of hospital bills or something, and mathowie has paid them off with money from t-shirt sales and I have to say it made me very proud to be a member of MetaFilter. Forty thousand dollars isn't enough to do anything hugely ridiculous, but it is enough to alleviate quite a few unexpected and difficult situations in a way that would make life better for a lot of MeFites and hopefully alleviate some financial difficulties in what are already stressful times. There might well be some scams and sketchy situations but I can't imagine MetaFilter giving anything over, say, seven or eight hundred dollars, it's not like someone making up a story to cheat someone out of hundreds of thousands of dollars or anything (I also think that, while there are always some exceptions, people as a whole are surprisingly decent about stuff like this and not likely to make up sob stories just to cheat someone out of a few hundred dollars, although I respect that I might be wrong about this, especially on the internet). It might well be unmanagable and if we did do that it seems like it would be best not to publicize it too much but for those AskMes with questions about paying legal fees or anything like that I think it would be nice if we as a community could help with that (and even for some bigger things like anything that can be done to help Hoder). Anyway, possibly a completely unworkable thing to do and I do get that it could be a logistical nightmare but I think it's worth considering.
posted by Mrs. Pterodactyl at 6:50 AM on November 3, 2009 [1 favorite]


Maybe this is obvious, but surely many of the MeFites who live in cities didn't grow up in them. If we can wrangle relatives still in home rural-ish areas, that'd be useful, right?

Yep, that's a good idea, and plays to adrianhon's argument for teams-of-teams. The best recruits in this respect are people who have little inclination to otherwise get involved in the challenge, which at the end of the day is a gut call for any mefite recruiter based on what they know about who they know but probably accounts for the vast majority of potential recruits.

If your uncle is a bright-eyed entrepreneur with a cutting sense for business, maybe don't ask him to help.

I suspect secrecy will be our biggest obstacle. Obviously strategy should be discussed in a private location, but who's allowed in? Anybody with a Mefi account? How do we keep out spies from competing teams? Too much restriction, on the other hand, will greatly limit the effectiveness of a team as big as this one.

There's two sorts of things we need to be secret about:

1. What we're planning to do, and
2. Where our team members actually see balloons on the day of.

How to deal with (1) is tricky. Containing the outflow of information, especially with questions like dual-citizenship and teams-of-teams and varying levels of information discipline even among totally loyal participants, will be hard. Worth trying, but hard.

How to deal with (2) is simpler: sighting information travels directly up from sighter to a trusted figure. Fewer changes of hands = fewer chances for leaks; if everyone on a sighting mission just mefimails me sighting info themselves (or has their really-trusted assistant do so, if necessary) then the only two people who can fuck with that piece of info are the sighter and me.

A sighter could still defect by mis- or non-reporting of sightings, but that's only on them. We could try to establish a blind double-recording system for sightings to at least keep 'em honest in the denouement if there's tomfoolery, but I may be leaping ahead on that front here.
posted by cortex (staff) at 6:54 AM on November 3, 2009 [1 favorite]


Also, re: covering 11 million highway miles: we need to look at this in terms of priority. Which million miles do we need to cover most? Can we draw some reasonable if speculative conclusions about which roads DARPA is vs. isn't likely to make people drive down? Is proximity to secure gov't installations enough to make DARPA not try to bring along traffic? Are shitty roads unlikely to be used? Will they avoid placing in difficult weather? Or hard-winter locations less likely?

Setting aside any of that speculation, some roads are faster than others, so it may make sense to prioritize paths just based on how major a road is, favoring larger and more well-maintained roads just for better likely average speed covered and driver safety.
posted by cortex (staff) at 6:57 AM on November 3, 2009 [1 favorite]


Count me in as "eyes".

I'm located in Grandview, WA. I'll monitor this thread and the one in the blue for instructions.

My only thought is that these ballons will not be anywhere near a major airport.
posted by Jumpin Jack Flash at 7:01 AM on November 3, 2009


I'm interested.
posted by eriko at 7:07 AM on November 3, 2009


I'm in for eyes. My boyfriend and I are geocachers (albeit not particularly hardcore), so this is right up our alley.
posted by Madamina at 7:14 AM on November 3, 2009


You should consider the possibility that the locations are not randomly chosen. If the purpose of the contest is to mark the 40th anniversary of the Internet, then maybe the locations are chosen for their historical significance.

The first four nodes on ARPANET are listed here: http://en.wikipedia.org/wiki/ARPANET#Initial_ARPA_deployment

Maybe we could make a list of 50~60 likely targets?
posted by kc8nod at 7:14 AM on November 3, 2009 [5 favorites]


I'll send MeMail to someone if I spot any around where I am (halfway between DC/Baltimore).
posted by sperose at 7:17 AM on November 3, 2009


If sabotage by fake balloons is acceptable, what about sabotage by joining other teams? The month lead time could be used to join other communities likely to start teams, build some credibility and join the team and A) Throw them off B) Steal their balloon locations.

I was thinking this might be a job suitable for people not based in the US, but moderators in those communities could notice non-US ip addresses.
posted by If only I had a penguin... at 7:27 AM on November 3, 2009


i'm in. don't know what i'll do yet, but count me part of the team. west suburban chicago.
posted by lester at 7:27 AM on November 3, 2009


And I realize this is both crazy and ridiculously fuel-waste-y but I wonder if in some rural areas it might not be more efficient to search by air (with high-resolution cameras?). Any mefites with pilots licenses?
posted by If only I had a penguin... at 7:31 AM on November 3, 2009


I really think the purpose of the contest is to find clever and effective ways to leverage social networking strategies. For example, and ideal network for this might be all mail carriers, all truckers, pilots, Twitter dataminers, and some big community like MeFi.

Given that assumption, I think this could turn into a real land grab in terms of locking in your sources and network. In other words, if we say we'll give the money to charity we might be able to get the postal union onboard to help us. But........ Only one team can lock in that support. Thus, land grab. How many such organizations can we list and then quickly lock in?

Radio stations
Unions
Teachers

Also, can we make a viral network? Perhaps start a charity for some real child in need, make the goal to raise money for the cause (rather than the goal being to win the contest), then flood various channels with information about how to get the sighting to us? Maybe with a famous spokesperson doing a plea for help on YouTube?

Look how fast the nation's attention got focused by the balloon boy. In-place networks may be good, but some sort of "flash network" could be bigger. We only need a network for a very short time period.
posted by y6y6y6 at 7:37 AM on November 3, 2009 [1 favorite]


Penguin, I'm hazarding a guess that it would take longer to fly, take pictures, and analyze the pics afterward than the balloons will be in play with their DARPA chaperons. Which means there'd be no reliable way to tell whether the balloons photographed were legit (or even balloons at all). So it was a good idea but probably more trouble than it's worth.
posted by The Winsome Parker Lewis at 7:38 AM on November 3, 2009


I'd like to help. I'm in Chittenden County, VT—not sure if my location is of any value, but I'm open to doing whatever. (...And if we're using Wave, I'd be most grateful for an invite!)
posted by lovermont at 7:48 AM on November 3, 2009


I'm in Kansas, arguably within a day's drive of most of that state, a big chunk of Nebraska, and parts of Iowa, Missouri, and Oklahoma. I love a good road trip. Count me in for anything mid-midwesty.
posted by donnagirl at 7:50 AM on November 3, 2009


Note that the rules (you've all actually read the rules right?) state you'll need to pay taxes on the $40,000. If we set it up as a charity can we avoid the taxes? Or is Cortex/Matt going to have to somehow deal with a messy tax issue?
posted by y6y6y6 at 7:51 AM on November 3, 2009


Postal Workers are federal govt employees, and so disqualified from the contest.

I live in Moscow Idaho. I will look for balloons. I am willing to bet a significant amount of money that I won't find any though. So go team I guess? Memail me if I can help in any way. (TBH I am also a member of the SA forums. So if we are not trusting people who have dual citizenship then probably nobody should memail me. Now the real question is did I say that because I'm trustworthy or because I figured anyone who did any checking would catch the dual citizenship? :) )

A couple of things.

Darpa is as likely to put the balloon out in Idaho as anywhere else? I'd expect at least one or two balloons in some far out of the way location.

Also, on dual citizenship/trust:

Well, first I think we need to keep the locations of the balloons a secret from the people on the team, entirely. There should just be a small core of trusted people who everyone is reporting to and who are running the team. We can't do this on a wiki or something, because what's to stop someone from getting the info and reporting it first?

As far as SA members. I think we should discuss this on the forum when we get this set up, but I'll just say that I think there is a place for double agents, and I'm not that concerned about security on our end.

Whatever you do, do not mention to anyone at SA that you are on any other message board that might take part in this, they're pretty paranoid about spies over there (at least if Goonswarm proper gets involved)...
posted by empath at 7:52 AM on November 3, 2009 [1 favorite]


The an odd percentage of the rules also go into the possibility that not all of the balloons will be found. Which makes me think some of them might be hidden. There is also the fact that the submission deadline is 9 days after the launch date. What circumstance could make a deadline over 24 hours necessary? The balloons are gone after the first day, so is there some additional investigation that might be done?

Just how big an area is an arc-minute? If someone describes in a blog post where they saw a balloon could someone reading the post have a reasonable chance of guessing the lat-long coordinates?
posted by y6y6y6 at 7:58 AM on November 3, 2009


And I realize this is both crazy and ridiculously fuel-waste-y but I wonder if in some rural areas it might not be more efficient to search by air (with high-resolution cameras?). Any mefites with pilots licenses?

If you're flying high enough to cover a large area, you're too high to tell an eight foot balloon from anything else round and red. I don't think planes would be worth it unless any Mefite has his/her own airplane and just wants an excuse to fly around that day.

We don't even know how high these balloons will be floating, do we?
posted by bondcliff at 7:59 AM on November 3, 2009


I am about 100% convinced that all the balloons will be in locations that are significant to DARPA's history. After all, when the results are finally released, that's where DARPA gets the publicity windfall. I can already picture their website showing a balloon at each location with an accompanying paragraph about what that location contributed to the DARPA legacy. A random hodgepodge of balloons around the country really doesn't do that. So I'd be expecting them in Livermore, Washington DC, Los Alamos, and so forth.
posted by crapmatic at 8:01 AM on November 3, 2009 [1 favorite]


So if we win, who gets the cash-money??
posted by Grither at 8:15 AM on November 3, 2009


We don't even know how high these balloons will be floating, do we?

I'd wager that they are 150 feet or less above the ground in order to avoid the notice requirements of 14 CFR 101.15.
posted by exogenous at 8:16 AM on November 3, 2009 [1 favorite]


I think we should leave speculation about where balloons might be (including thoughts about types of sites, not just actual locations) out of the google-able thread. Furthermore, I propose the following:

1) That mods should delete such speculation. (Not because they're bad posts but because they're good ideas)
2) That mods add something to the html in this thread to prevent google from indexing it.
posted by If only I had a penguin... at 8:18 AM on November 3, 2009


On a more personal note I make another two proposals:

1) That somebody send me a google wave invite.
2) That somebody find work for mefites not in the US.

And I concede that my airplane idea was overly crazy, which I suspected from the start, but figured best to throw it out there and see if someone can work with it, just in case.
posted by If only I had a penguin... at 8:20 AM on November 3, 2009


Okeedoke. I am seriously IN. Put me in the "City-by-City Team" for NYC. We should nominate an NYC coordinator? Any nominees? (Not it. But I do have a Zipcar account, a working knowledge of the Northeast, and a stupid amount of optimism.)

Also, I wonder if Zipcar would give TeamMeFi-NYC their special DARPA Discount. Hm.
posted by functionequalsform at 8:23 AM on November 3, 2009


Forget Wave, guys, it doesn't scale for large numbers of users.

Also, a few more thoughts:

Are there any amateur pilots on MeFi?

We should donate the money to something like landmine removal, since this is a contest sponsored by the military.

We should set aside some percentage of the money to reward people who actually find and verify the balloons (half for first report, half for verifying the report.)
posted by empath at 8:24 AM on November 3, 2009 [3 favorites]


Or at least give away free mefi t-shirts or something.
posted by empath at 8:24 AM on November 3, 2009


I'm willing to do a little leg work on the day of the event. MeMail me if I can be of any assistance.
posted by stuart_s at 8:25 AM on November 3, 2009


Well, first I think we need to keep the locations of the balloons a secret from the people on the team, entirely. There should just be a small core of trusted people who everyone is reporting to and who are running the team.

Not just a small core; cortex. The only person who needs to know about the location of putative balloons is the person who finds it, and cortex. But we really need to move discussion to a more secure location immediately.

Grither, we had discussed in the other thread that we should agree up front that the prize goes to a charity. We can figure out which one as a community in the event that we win-until then it isn't worth worrying about, unless someone thinks that 'having a cause' will be useful to our efforts in some way and wants to do the legwork.
posted by Kwine at 8:26 AM on November 3, 2009 [1 favorite]


on lack of preview, landmine removal is a fantastic idea.
posted by Kwine at 8:27 AM on November 3, 2009


I am about 100% convinced that all the balloons will be in locations that are significant to DARPA's history. After all, when the results are finally released, that's where DARPA gets the publicity windfall. I can already picture their website showing a balloon at each location with an accompanying paragraph about what that location contributed to the DARPA legacy. A random hodgepodge of balloons around the country really doesn't do that. So I'd be expecting them in Livermore, Washington DC, Los Alamos, and so forth.

That is if you think Darpa is doing this for publicity. I suspect they are testing some Game Theory ideas, and are likely to make it difficult.

That said, it can't hurt to test your theory by collecting that sort of information (NOT HERE!) and making sure we have them covered.
posted by empath at 8:27 AM on November 3, 2009


I'd love to help. Just a cell phone, no gps. In the LA area. Not afraid to drive to the desert.
posted by klangklangston at 8:29 AM on November 3, 2009


If you're flying high enough to cover a large area, you're too high to tell an eight foot balloon from anything else round and red.

Having a list of everything round and red in the country would be USEFUL INFORMATION. We'll have people to verify anything.

Also, if we have people in the air at 12.01, we may be able to catch the balloons AS THEY ARE GOING UP.
posted by empath at 8:30 AM on November 3, 2009


> I have just purchased a slice for use on this project.

I think we'll need a lot more pizza than that.
posted by nowonmai at 8:32 AM on November 3, 2009 [3 favorites]


(we should delete this thread once the forum goes up)
posted by empath at 8:32 AM on November 3, 2009


Where's the MetaFilter wave?
posted by grouse at 8:37 AM on November 3, 2009


I just happen to have access to an eight-foot balloon.

This could be fun.
posted by mr_crash_davis mark II: Jazz Odyssey at 8:38 AM on November 3, 2009


This sounds fantastic, I'd love to help: St. Louis, MO, have car & cell phone, could obtain GPS for a day's use.

>Invite only, sanity checked against mefi.

Dammit!
posted by philotes at 8:40 AM on November 3, 2009


Just how big an area is an arc-minute?

From the Wikipedia page:
At sea level, one minute of angle (around the equator or a meridian) equals about 1.86 km or 1.15 miles, approximately one nautical mile (approximately, because the Earth is slightly oblate); a second of angle is one sixtieth of this amount: about 30 meters or 100 feet.
posted by odinsdream at 8:47 AM on November 3, 2009


I'm in for the twin cities/metro area. I also have a foolish amount of optimism and fear of what DARPA could do to me
posted by Think_Long at 8:48 AM on November 3, 2009


Secure discussion. Now. I also have a slicehost available with very low usage currently. I'll be glad to purchase a domain for this as well. Might want to even consider secure PGP communications via e-mail.
posted by odinsdream at 8:52 AM on November 3, 2009


*sigh*
This one's EASY.
1) hack DARPA computers*
2) find where they plan to put the balloons
3) win contest before it even starts

*may require 'sexy computer scientist' outfit and/or cudgel
posted by sexyrobot at 8:54 AM on November 3, 2009 [2 favorites]


I have it: a brilliant role for people outside the US and an even more brilliant sabotage plan that can be done by people living outside the US. To prevent this resulting in self-sabotage of our own team will require a programmer who can write up a script to filter google search results.

I actually think this is a good and non-crazy idea. Now who do I tell it to or what private place can I post it?
posted by If only I had a penguin... at 8:55 AM on November 3, 2009


I am interested in participating, DC-area.
posted by Partial Law at 8:57 AM on November 3, 2009


I honestly think a secure installation of vbulletin is the way to go. People already know how to use it, it's relatively easy to setup, etc.

It's what Goonfleet uses in Eve Online, and they've been the target of lots of hacking attacks, espionage, etc. Probably should be good enough for our purposes.
posted by empath at 8:58 AM on November 3, 2009


This is great and I'd be happy to help (Denver Metro area, but willing to drive out into the boonies).

But can I put in a vote to not spend time or energy on sabotaging other teams? Not only does it seem counter to the spirit of the challenge, but wouldn't our resources be better dedicated to actually finding the balloons? I'm all for sabotage prevention, but it just seems mean-spirited and counter-productive to try and instigate sabotage ourselves. It would also kind of taint winning in my opinion.
posted by Kimberly at 9:02 AM on November 3, 2009


Btw-- I think many nationwide tech companies (ibm, microsoft, etc), might be interested in putting a team together. Would like to have reports from anyone inside on what's being discussed.
posted by empath at 9:05 AM on November 3, 2009


if everyone on a sighting mission just mefimails me sighting info themselves (or has their really-trusted assistant do so, if necessary) then the only two people who can fuck with that piece of info are the sighter and me.

I just had a thought while walking over here, which is: we should split into regional teams. That is, my slicehost server that I just bought could act as home base for states in EST and two other volunteers could handle the Central/Mountain and PST. Then cortex/jessamyn would be the only people to have access to all three servers. We could do more or fewer servers depending. Even if I wanted to, I could then no longer leverage my access to the server I donate to the cause.... except by denial of service... which... I guess you'll just have to trust me not to tank my several years of personal investment in the site for a crappy stunt like that.

So yeah... cortex/jessamyn: please ping me if you want to put this slice to use. I'm holding off on installing stuff beyond basic hardening tools until I've heard that people really do want me managing the machine.

On preview:

> I think we'll need a lot more pizza than that.

Hardy-har-har. Seriously though: a slice in this case means a virtual private server. It's secure enough for the context, I promise.

Also: odinsdream... check your memail soonish.
posted by tarheelcoxn at 9:05 AM on November 3, 2009


But can I put in a vote to not spend time or energy on sabotaging other teams? Not only does it seem counter to the spirit of the challenge, but wouldn't our resources be better dedicated to actually finding the balloons?

I'm not sure that it IS counter to the spirit of the challenge. I think it's pretty safe to assume that DARPA is not doing this 'just for fun'. Perhaps they want to know whether polluting data is a viable tactic in whatever theoretical situation they're wargaming out here.
posted by empath at 9:07 AM on November 3, 2009


I'm not sure that it IS counter to the spirit of the challenge. I think it's pretty safe to assume that DARPA is not doing this 'just for fun'. Perhaps they want to know whether polluting data is a viable tactic in whatever theoretical situation they're wargaming out here.

I suppose that's a fair point. I hadn't really thought about it that way.
posted by Kimberly at 9:09 AM on November 3, 2009


I'd like to help. I'm located between Chicago and St. Louis.
posted by Sailormom at 9:15 AM on November 3, 2009



I'd wager that they are 150 feet or less above the ground in order to avoid the notice requirements of 14 CFR 101.15.


I agree that the rules only say that the balloons "will be in readily accessible locations, visible from nearby roadways" so it seems safe to speculate that balloons flying under 150 feet are probable -- but it does not rule out the possibility of balloons at a higher elevation.

Note that 14 CFR 101.13 tells us that we are unlikely to find balloons (1) Less than 500 feet from the base of any cloud or (2) More than 500 feet above the surface of the earth -- as those conditions would violate the regs. But you can legally float a moored balloon between 150 and 500 feet with disclosure to FAA ATC (24 hours beforehand).

So if there's a likelihood that balloons would fly between 150-500 feet, part of a search strategy might be to call the FAA ATC facilities to inquire about moored balloon notices. Or, as someone pointed out elsewhere, to pollute that channel with false positives.
posted by QuantumMeruit at 9:17 AM on November 3, 2009 [1 favorite]


Put me on the team list or whatever; not sure I'd be able to drive around but I have (modest) ideas.
posted by LobsterMitten at 9:19 AM on November 3, 2009


I've been trying to think of some realistic scenarios they might be wargaming for this.

10 things, easy to visually identify, likely to raise interest, not already tracked by law enforcement or government. Something a social network would be better able to quickly find than say satellites or radar. Random distribution perhaps?

The only thing I'm coming up with is space aliens.
posted by y6y6y6 at 9:19 AM on November 3, 2009


Kwine: While restricting information to a single person can prevent leakage, it also creates a single point of failure. What if Cortex is ill, or his internet goes down, or gets sabotaged? Especially since we've identified Cortex here already, he's definitely vulnerable - we need more than one person who's receiving data (and no, I'm not volunteering).
posted by adrianhon at 9:19 AM on November 3, 2009


If I see one in Rhode Island, I will let you know.

Also: Wave invite? Plz?
posted by grapefruitmoon at 9:20 AM on November 3, 2009


we need more than one person who's receiving data (and no, I'm not volunteering).

Seems like a relatively simple thing to get legally binding agreements from the (let's say) dozen or so people at 'HQ' that states that they agree to not take the money for themselves.
posted by empath at 9:26 AM on November 3, 2009


And seeing as an arc minute of error in coordinates translates into perhaps a mile that gives you a very solid shot at getting the coordinates without having a GPS unit at the scene. If someone can tell you where they saw it you can easily use Google maps to get the coordinates.
posted by y6y6y6 at 9:27 AM on November 3, 2009


y6y6y6: Absolutely; you could imagine someone giving a zipcode or street address or even just a city or twon name. On that note, ideally there should be a workflow that can process different types of input (e.g. SMS, MMS, email, voice, Tweet, etc) and automate things as much as possible; things that require human checking (e.g. an image) would go into a queue.

It's something I suspect our dear mods have some experience in :)
posted by adrianhon at 9:33 AM on November 3, 2009


I think many nationwide tech companies (ibm, microsoft, etc), might be interested in putting a team together.

How many tech companies are really nation-wide though? They usually have large offices in a few big cities so most of their employees are clustered geographically, which is a huge downside in this challenge. If a big company really wanted to encourage their employees to spend time on this project, it would probably be most effective if it was a company that had small retail locations all over the place (McDonalds, Starbucks, CVS, etc.).
posted by burnmp3s at 9:33 AM on November 3, 2009


Count me in for on the ground surveillence for Cedar Rapids/Iowa City, IA area. MeMail me with any instrustions...
posted by jpdoane at 9:36 AM on November 3, 2009


Kwine: While restricting information to a single person can prevent leakage, it also creates a single point of failure. What if Cortex is ill, or his internet goes down, or gets sabotaged? Especially since we've identified Cortex here already, he's definitely vulnerable - we need more than one person who's receiving data (and no, I'm not volunteering).

Chain of command. Can be managed off the record by core organizers; some mechanism for making sure that delivery of sighting information won't be disrupted by last minute changes in receiver. Something as simple as filtered automatic email forwarding to my backups on the day of could do it, presuming I'm willing to engage in said forwarding, which is mostly a question of trust on my part in those backups. I'm confident I'd have no trouble putting together a shortlist of people whose integrity and discretion I trust, as far as that goes.

So yeah... cortex/jessamyn: please ping me if you want to put this slice to use. I'm holding off on installing stuff beyond basic hardening tools until I've heard that people really do want me managing the machine.

I'm keenly interested in the brainstorming and discussion of all this and will be happy to be really, really attentive to the day-of efforts, but I don't want to be too involved in the architecture/execution of stuff and so I should probably explicitly decline any responsibility for a lot of the nitty gritty on that front. I don't want to be a potential point of delay/failure for getting things organized.

So if there's something that it's helpful for me to do, let me know. But I don't think there's anything at the moment that I need or any organizational/software/hardware questions that are things that a decision from me should have anything to do with.
posted by cortex (staff) at 9:37 AM on November 3, 2009


I can guarantee you that Google and Microsoft will have internal teams getting together to tackle this; I'm not sure whether they'd have corporate backing, but either way, they're going to be formidable competition since they'll already have secure(ish) networks and a lot of development experience.
posted by adrianhon at 9:38 AM on November 3, 2009


I would like to participate somehow. In eastern SD, not much mobility mobility though.
posted by 47triple2 at 9:38 AM on November 3, 2009


This MeTa thread is way more fun than that horrible Favorites one.
posted by jpdoane at 9:41 AM on November 3, 2009 [7 favorites]


So if there's something that it's helpful for me to do, let me know.

Manage trust. For example, without further comment from you, I'm just going to go with my gut and throw my lot in with odinsdream. Also empath, I guess. I think mostly odinsdream for now, though, since we can reasonably meet in person over a weekend. Though I guess I could hop a train to DC....
posted by tarheelcoxn at 9:47 AM on November 3, 2009


Which is another way of saying... people should now memail me if you want to be on the EST sub-team. Also, I'm getting a domain and setting up some tools on the slice now....
posted by tarheelcoxn at 9:51 AM on November 3, 2009


My time is very tight but I have some ideas that I would like to share once we have a secure forum or by mefimail, so please count me in as far as that goes.
posted by exogenous at 9:53 AM on November 3, 2009


I'm in for NYC... and willing to Zip Car to tri-state locations. I also have relatives in various rural parts of TN and Hawaii.
posted by kimdog at 9:53 AM on November 3, 2009


Also empath, I guess

"I have the locations of all 10 balloons. Unless I receive 20,000 favorites in the next hour, I will turn them over to 4chan"
posted by empath at 9:57 AM on November 3, 2009 [1 favorite]


Leaders, step forth. Project needs someone to coordinate information, assign tasks, and more. I am not this person. This will get attention in the news, so unemployed MeFites with Mad Skilzs, represent! This may be an opportunity.

In the other thread, I suggested that any prize money go to the Fistula Fund; perhaps there should be other suggestions for this, and some attempt at consensus. If Mattamyntexpinb wants to declare a charity, I have no objection.
posted by theora55 at 9:58 AM on November 3, 2009


Three hosts sounds problematic. Why not one host with three separate forums/wikis? or one forum with three areas? Admins/Cortex could have access to all areas, but users would have to pick a region/function. That setup would let you segment access to information is lots of different ways, and different trust levels, without having to manage several different userbases and physical hosts.
posted by y6y6y6 at 9:59 AM on November 3, 2009


I would volunteer my time to help set up the web forums/ hosting, etc. And of course to do the project as well.
posted by localhuman at 10:03 AM on November 3, 2009


I'm in. Boston area, with a car.
posted by rmd1023 at 10:07 AM on November 3, 2009


I would like to be a part of Team Saint Louis as eyes/wheels on the ground for the day of. I'll keep checking back for further instructions, or see if anyone MeMails me.
posted by slogger at 10:07 AM on November 3, 2009


I'm volunteering me and gingerbeer to drive around and look for balloons. We live in San Francisco, but we've got a favorite birding area that's been unvisited for too long centered here, near the delta town of Rio Vista. We don't have a GPS but we do have binoculars and a spotting scope. And a BlackBerry.
posted by rtha at 10:08 AM on November 3, 2009


Lest this thread become the defacto place before a secure zone is set up, please memail me or tarheelcoxn. I'm going to try to coordinate with him on managing trust relationships.
posted by odinsdream at 10:12 AM on November 3, 2009


Since this seems to be the "volunteer registration" thread at the moment, I'll register myself, too. I'm in the Columbus, OH, area, with access to a car, phone, and GPS. I have server space in a couple of places, although that seems to be taken care of, and I'm happy to share.

We should get some better registration document online promptly, one that allows us to catalog volunteer locations and numbers.
posted by aaronbeekay at 10:17 AM on November 3, 2009


With all this talk of PGP and secrecy and all that, remember that this game is mostly about getting large numbers of people to participate. This MetaTalk thread is bringing in a lot of people right now, but within a week most of the MetaFilter community will forget about it.

I would suggest posting a new MetaTalk post a day or two before the 5th with clear instructions for how people can help and what they should do the day of the event. And someone should be collecting contact information for everyone who has posted in this thread so that they can be sent updates/reminders between now and when the balloons go up.

Lest this thread become the defacto place before a secure zone is set up, please memail me or tarheelcoxn.

Just FYI, if you are trying to get a lot of people to join the project, asking them to memail you doesn't work too well. I just started organizing a MeFi writing project recently, and most the people who ended up participating didn't respond until I memailed them personally asking if they wanted to participate.
posted by burnmp3s at 10:17 AM on November 3, 2009 [3 favorites]


I don't think we'll really get going till we get a forum set up.
posted by empath at 10:21 AM on November 3, 2009


Assuming the money is going to a worthwhile general charity (on the very slim chance of success), I'll report any sighting in the Houston metro area to whatever mechanism you set up for such a thing.
posted by Burhanistan at 10:25 AM on November 3, 2009


In what way will this involve drinking and socializing?

Seriously, though- I will participate in the Twin Cities, MN area- and help organize and manage people (but not technology, thank you) if necessary.
posted by elmer benson at 10:25 AM on November 3, 2009


Strategy? Dust off and nuke it from orbit, it's the only way to be sure.
posted by electroboy at 10:42 AM on November 3, 2009


Another strategy might be to set aside the idea of winning and just decide to be the team that finds the most balloons. Secrecy puts a huge limt on how fast and how big your team can be. It also, and perhaps more profoundly, limits who will participate. It's possible to be so security conscious that you undermine the big opportunities.

For example, trying to take it viral won't work if there is a big secret security process to go through. If we get mail carriers to help, they aren't going to create user accounts on our secret website.

The easy to find balloons will be easy for any marginally networked effort. I think it's only the teams pulling in fringe networks that will succeed.
posted by y6y6y6 at 10:43 AM on November 3, 2009


Another loose knit group of watchers is trainwatchers. I believe that a couple of members here are into that hobby.
posted by Jumpin Jack Flash at 10:46 AM on November 3, 2009


Would teaming with another online community be a strategy we would want to consider? It would dilute the Team MetaFilter! aspect to it, but might significantly improve the chance of victory.
posted by jpdoane at 10:51 AM on November 3, 2009


If we get mail carriers to help, they aren't going to create user accounts on our secret website.

I think it's exceedingly unlikely that we'll get anyone to participate who is not fairly directly mefi related somehow.
posted by empath at 10:53 AM on November 3, 2009


Would teaming with another online community be a strategy we would want to consider?

I would consider this, but it would have to be a site with a pretty close connection to us.
posted by empath at 10:54 AM on November 3, 2009


I'd also like to help but am not on Wave. If anyone has an extra invite, toss it my way.
posted by allen.spaulding at 10:56 AM on November 3, 2009


to re-iterate -- Wave is not a solution for coordinating large groups (hell, it's bad for just 3 or 4 people), so I don't think we're going to use it.
posted by empath at 11:05 AM on November 3, 2009


(well, maybe on an ad-hoc basis.. i can see it being useful for group mapping, but man, i would hate to see what a wave with 1000 people looks like)
posted by empath at 11:06 AM on November 3, 2009


Not that I don't want a Wave invite *cough* but it sounds like that's not a scalable way to do this. Private forums or mailing lists (both formal and ad hoc) seem like the way to go.
posted by clauclauclaudia at 11:06 AM on November 3, 2009


Willing to participate however I can.
posted by bDiddy at 11:07 AM on November 3, 2009


Just a thought. If I were doing this as DARPA, I'd pick spots that were significant to the internet like:
1. UCLA
2. MIT
3. CMU
4. Sullivan, IL
5. University of California, Berkley
6. System Development Corporation in Santa Monica
7. Univeristy of Utah
8. University of California, Santa Barbra
9. Duke University
10. Stanford University
11. Digital Equipment Headquarters
12. Hewlitt-Packard Headquarters
13. NASA
14. NSF (National Science Foundation)
15. DOE (Dept of Energy)
16. Cisco
17. Network Information Center in Melno Park, CA
18. Network Solutions?
19. General Atomics (San Diego)
20. AT&T (Bell Labs)

These of course, were identified by rummaging through wikipedia... Just a thought - might want to do this intelligently...
posted by Nanukthedog at 11:09 AM on November 3, 2009


tarheelcoxn: "Which is another way of saying... people should now memail me if you want to be on the EST sub-team. Also, I'm getting a domain and setting up some tools on the slice now...."

Yeah, I think separate sub-servers is just going to unnecessarily complicate things. For now, just throw phpBB on your slice so real discussion can start. Maybe we'll quickly discover other tools will be necessary, but the most important thing right now is to quickly facilitate organization and recruitment.
posted by team lowkey at 11:22 AM on November 3, 2009


I suspect they are testing some Game Theory ideas, and are likely to make it difficult.

The real "game theory" research begins when the prize check goes to a single individual who has to decide how or whether to share the money. Unless it goes to cortex of course, who would have an obligation to share it as part of his employment. So ideally cortex should be the only guy to know every location.
posted by East Manitoba Regional Junior Kabaddi Champion '94 at 11:31 AM on November 3, 2009


I'm in San Francisco and I'd be interested in helping. I'm without a car, but I do have an iPhone and am underemployed at the moment. Please memail me as more secure forums are set up and more concrete planning begins.

If I were doing this as DARPA, I'd pick spots that were significant to the internet like:

Isn't this exactly the opposite of what DARPA would do? It's not really the kind of organization that likes to be cute. Frankly, I'm not sure how to rank the likelihood of balloon locations, but I wouldn't be surprised if the final map looks almost completely random.
posted by thebergfather at 11:36 AM on November 3, 2009


I think people should refrain from sharing useful ideas in the thread. Let's focus on really general strategy and organizing principles for now.
posted by empath at 11:37 AM on November 3, 2009 [1 favorite]


I'm in the Milwaukee area and am up for just about anything.

Also, I read this post yesterday, which meant I had a dream last night wherein I was part of an epic journey searching for all of the red balloons, only to find out that there was no prize, and the balloons were simply a viral ad for Darpa's non-existent sci-fi video.
posted by niles at 11:40 AM on November 3, 2009


Are we for or against faking people out? Because I think that, as much as flying fake balloons is a good idea, so is putting out fake info on Twitter, Facebook, Google Groups, etc. Not so much double secret agent stuff, but just putting it out there for other groups to happen upon?
posted by bDiddy at 11:45 AM on November 3, 2009


I'm another bay area volunteer.

What if Cortex is ill, or his internet goes down, or gets sabotaged?

That's easy, just have the trusted parties encrypt the results and post them to the project mailing list as they are collected. Make sure that two or three top level trusted people have the ability to decrypt. Then after the submission deadline post the key and then anyone that wants to can verify that what was submitted matched what was collected.
posted by Rhomboid at 11:49 AM on November 3, 2009


I'll keep an eye on this as we get closer, my participation depends on whether any other plans pop up for that day. I'm in Tallahassee, Florida.

As for disinformation, I think it'd be a waste of resources unless there are a significant number of people who would be unable to participate in searching and would otherwise like to contribute by spreading false information on social networks. I sort of suspect it would still be a wasted effort, though, simply because I think there will be enormous piles of disinformation spreading already from other sources. Better to focus on the actual search.
posted by empyrean at 11:50 AM on November 3, 2009


So if there's something that it's helpful for me to do, let me know.

I think it would be a good idea to decide on a charity early just to settle the discussion of what happens to the money.
posted by Rhomboid at 11:51 AM on November 3, 2009


Are we for or against faking people out? Because I think that, as much as flying fake balloons is a good idea, so is putting out fake info on Twitter, Facebook, Google Groups, etc. Not so much double secret agent stuff, but just putting it out there for other groups to happen upon?

I have lots of ideas in this direction, but I'd prefer not to talk about that publicly. Once we get the forum up and running, I'll get to work starting an intelligence operation.
posted by empath at 11:54 AM on November 3, 2009


Again, I think we should keep our ideas both about possible locations and strategies out of the public thread. Let's keep this for sign up and ideas about logistics of coordinating and take the logistics of finding and sabotaging private.
posted by If only I had a penguin... at 11:55 AM on November 3, 2009 [1 favorite]


I'm outside the of the US but would be happy to filter through information streams. Me-mail or e-mail both good.
posted by olya at 11:56 AM on November 3, 2009


Along the lines of disinformation, I forgot to mention this: I think monitoring social networking sites and identifying reliable reports to investigate in the time frame given is extremely problematic. It's so trivial to generate convincing photos and locations, and I suspect fakes will generate trending topics across social networking sites to such a degree that trying to filter them would be unproductive.

I really do think it'd be best to investigate information from only 1st person MeFi reports or their 2nd person trusted-sources (very close friends, relatives). I think the time frame demands it.
posted by empyrean at 12:00 PM on November 3, 2009


Also, I will start up a volunteer sign up sheet this evening.

Google Spreadsheet okay? I won't share it with anyone (well, except for the people who need to use it to coordinate volunteers), and will promise to delete it once the contest is over.
posted by empath at 12:01 PM on November 3, 2009 [2 favorites]


I'd be happy to spend a Saturday driving around the Memphis area looking for big red balloons. I can help cover West Tennessee, North Mississippi, and/or Eastern Arkansas.
posted by JeffK at 12:16 PM on November 3, 2009


Are there other public planning threads like this that we should be monitoring?
posted by milestogo at 12:17 PM on November 3, 2009


I am in the Twin Cities MN and would love to help.
posted by beandip at 12:20 PM on November 3, 2009


I think people should refrain from sharing useful ideas in the thread. Let's focus on really general strategy and organizing principles for now.

Organizing principles and strategy are the most useful ideas available right now. So, you shouldn't be sharing those either.

Also, the first person to volunteer setting up the secure site is probably the first person you should not trust. You have to distribute the trusted infrastructure so that no one person or one server has access to everything. I mean, you might be able to find the right person to set it up -- pb maybe -- but the more I think about it, the more I realise nobody really fits the bill.
posted by Chuckles at 12:20 PM on November 3, 2009


I think it would be a good idea to decide on a charity early just to settle the discussion of what happens to the money.

Yes. Whatever is the biggest slap in the face to the military industrial complex possible is what you want to choose. Land mine removal is a good candidate, but perhaps not good enough. An Iraq war resistors organization maybe? Anti-nuclear weapons group? Are there any organizations trying to talk engineers and scientists out of selling their souls to Lockheed Martin?
posted by Chuckles at 12:24 PM on November 3, 2009 [4 favorites]


If I were DARPA, I would choose locations that demonstrate either 1) where the internet started or 2) just how far the internet has reached into remote locations. I suspect that the locations will run the gamut between those. One or two super easy ones, a few quite difficult ones, and a few unbelievably remote and difficult ones in totally surprising and almost-never-traveled locations that happen to be visible from a remote, unpaved, rural road.
posted by The World Famous at 12:37 PM on November 3, 2009


Not to be pessimistic, but it would be a lot easier to wait until we lose to decide what the charity would have been.
posted by smackfu at 12:42 PM on November 3, 2009 [4 favorites]


Forget charities that make any heavy-duty political point. We want something everybody can get behind -- people here and people we might recruit from our real-world networks to help out. Rural relatives etc must be able to understand and endorse the charity. I think real-world networks will be crucial to success, because we're not going to have more than, say, 200 people participating directly on this site. Repeat, the charity is key, and it has to be broadly appealing.

Land mine removal is good.
Fistula charity is good but the topic may be too obscure/icky to discuss with distant relatives. (Horrible to say, but worth considering. We want buy-in from a maximally wide range of people.)
Funding for literacy or schools is good.
A charity that would achieve a concrete goal (eg building a school or stock a library, as xkcd recently did) with $40K is a good choice.
posted by LobsterMitten at 12:43 PM on November 3, 2009 [1 favorite]


Are we for or against faking people out?

I'll set up an army of inflatable army tanks and fake some radio traffic. They'll totally think the balloons are at Pas-de-Calais.

What? Why's everybody looking at me?
posted by marxchivist at 12:56 PM on November 3, 2009 [1 favorite]


I'm in Albuquerque, New Mexico and am totally interested. No car (though I might be able to borrow one if necessary).
posted by NoraReed at 1:02 PM on November 3, 2009


This is more idle speculation than Reptoid paranoia, but exactly how secure is the MeMail system? Do we know that SA or 4chan couldn't crack it? Should we be interested in a more secure way to pass sightings etc. to cortex?
posted by shakespeherian at 1:03 PM on November 3, 2009


Nth-ing landmine removal.

Also work + school = busy, so forgive me if it is tonight before I memail people to get you on the private forum. In the meantime please do recruit people by bringing them to this thread.
posted by tarheelcoxn at 1:03 PM on November 3, 2009




Clear Path International is a really great charity aimed at helping landmine survivors, and could use cash.
posted by Burhanistan at 1:09 PM on November 3, 2009 [1 favorite]


Sign me up, Northwest Wisconsin sector. HAve reasonably large contact base that can feed me sightings.
posted by cosmicbandito at 1:12 PM on November 3, 2009


I'm in the Baltimore / DC / Philly area with car and GPS. Count me in!
posted by youcancallmeal at 1:18 PM on November 3, 2009


This is more idle speculation than Reptoid paranoia, but exactly how secure is the MeMail system? Do we know that SA or 4chan couldn't crack it? Should we be interested in a more secure way to pass sightings etc. to cortex?

The question reduces to "do we know that SA or 4chan can't/won't compromise Metafilter, period". Another site/group taking it upon themselves to compromise the Metafilter server is something that goes well beyond the scope of my devotion to this project, in any case, and if that happened I would care very, very little about the secrecy of some balloon reporting.
posted by cortex (staff) at 1:22 PM on November 3, 2009 [4 favorites]


So... clarifying: we are up for recruiting non-MeFites? Or we're not?
posted by functionequalsform at 1:24 PM on November 3, 2009


So... clarifying: we are up for recruiting non-MeFites? Or we're not?

As long as information is segmented appropriately, then it doesn't hurt to recruit more people. The main thing is keeping the 'core' group small.
posted by empath at 1:31 PM on November 3, 2009


East Manitoba Regional Junior Kabaddi Champion '94: "The real "game theory" research begins when the prize check goes to a single individual who has to decide how or whether to share the money. Unless it goes to cortex of course, who would have an obligation to share it as part of his employment. So ideally cortex should be the only guy to know every location."

I don't think the game theory starts with awarding the prize. I think it ends there. Let me give an inductive proof:

Imagine a group has discovered and shared nine such locations; what stops the person with the 10th discovery from defecting? Yes, I know, you've all thought of that. It's common knowledge. So, if it's common knowledge to defect at the 10th balloon, why disclose the 9th balloon? And so on, until you ask why band together at all. And if you do have a motivation to form a group, when do you take that action, what are the terms and what are the requirements for joining? In summary the game theory starts at the beginning of collective action, not the end.

My theory is that most social networks will not bother with dividing the booty among ten thousand members, but that other systems will sprout. Some groups may try to charge information on the condition of membership., ie you can only join if you've spotted one of the balloons. For large existing networks like MeFi, I'm guessing the prize will be directed at social network maintenance costs, or when those are already covered, a charitable cause.

There is one other approach, but it's complicated. At some level, each balloon sighting is worth 4 thousand dollars. If you make a standing offer of 2 thousand dollars for every new balloon sighting (providional to winning), that's a 20k profit. It gets more complicated if you try to just do marginally better than everyone else. But the problem here is a lemon market for information; the seller of information knows more about its quality than the buyer. And woe unto the buyer who lets it be known he's got four sightings and needs a fifth; the marginal benefit is substantial.

The advantage to using GPS coordinates is that you can perform somewhat reliable comparison of datasets without disclosing the radius. Probably just reducing the GPS resolution is strong enough if you assume the balloons are evenly spaced, but I think there's even some algorithms published to do encrypted comparisons of location. That's an important point about social networking in the future: how do you balance privacy with locational awareness?

One last thought: Previous challenges have been high dollar, and 40k divides into 1m 25 times, although the rules explicitly make it clear there's only one prize. Perhaps the rest of the money will go to buy up data collected by these networks?
posted by pwnguin at 1:35 PM on November 3, 2009 [1 favorite]


IME, when somebody says that such-and-such "hacked" their email they really mean that they used a poor password (e.g. an English word and a year) and that party was able to guess it by brute force. That's really not got anything to do with hacking or site security and everything to do with people being stupid and using bad passwords, or using the same password for multiple sites.
posted by Rhomboid at 1:38 PM on November 3, 2009 [1 favorite]


Los Angeles, car, a phone that's ostensibly capable of doing 'GPS', lots and lots of time, and MetaFilter is (and always has been) my one and only. Let me know how I can help.

As for a forum, why not appropriate the travel subsite? In my imagination it's as easy as flipping a switch to make it secure.
posted by carsonb at 1:38 PM on November 3, 2009


How do I sign up? I'm in San Francisco, which already has lots of volunteers, but I wanna play too. Not sure how to make myself useful (no car, no GPS, cell phone is a piece of crap) but I'm game.
posted by Quietgal at 1:40 PM on November 3, 2009


Imagine a group has discovered and shared nine such locations; what stops the person with the 10th discovery from defecting? Yes, I know, you've all thought of that. It's common knowledge. So, if it's common knowledge to defect at the 10th balloon, why disclose the 9th balloon? And so on, until you ask why band together at all. And if you do have a motivation to form a group, when do you take that action, what are the terms and what are the requirements for joining? In summary the game theory starts at the beginning of collective action, not the end.

The solution to that is relatively simple. You don't report the balloon locations to the group.

Possible balloon locations are reported via a webform to a small core group of mefites at "HQ", who then dispatch people to confirm them.

Nobody besides the coordinators should know the locations of any balloons. The downside of that is multiple reports of the same balloons,
posted by empath at 1:40 PM on November 3, 2009 [1 favorite]


Forget charities that make any heavy-duty political point. We want something everybody can get behind -- people here and people we might recruit from our real-world networks to help out.

With $40k at our disposal, that will be the most awesome bike any child has ever received.
posted by Dumsnill at 1:41 PM on November 3, 2009 [1 favorite]


I googled it and apparently ponies cost about 1,000 dollars. Just sayin'.
posted by NoraReed at 1:44 PM on November 3, 2009 [1 favorite]


Imagine a group has discovered and shared nine such locations; what stops the person with the 10th discovery from defecting?

In our case, because the only person who could be in that position will be cortex and he wouldn't do that.

Also, the first person to volunteer setting up the secure site is probably the first person you should not trust.

Is it uncouth to worry here publicly that I don't recognize tarheelcoxn's username at all? Then so it is. Anybody willing to vouch for the guy, before we all go spilling our plans on his private server? I feel like I 'know' almost all of the rest of you.
posted by Kwine at 1:46 PM on November 3, 2009


Everyone's a superhero; everyone's a Captain Kirk with orders to identify, clarify and classify.

Note: Everyone needs a hug.
posted by not_on_display at 1:48 PM on November 3, 2009


Also, I will start up a volunteer sign up sheet this evening.

Google Spreadsheet okay? I won't share it with anyone (well, except for the people who need to use it to coordinate volunteers), and will promise to delete it once the contest is over.


Thumbs up from over here.
posted by functionequalsform at 1:50 PM on November 3, 2009


I recognize tarheelcoxn, and in my credulous fashion don't have any concerns about him being a liability. I will admit to not being willing to be sufficiently paranoid to satisfy those who ARE sufficiently paranoid, but as far as that goes he seems from what I know of him to be a smart, enthusiastic, and non-douchebaggy sort.

And again, not to discount the value of trying to keep our methods to ourselves, but (a) if a traditional search is going to be how we make our sighting work then that + direct reporting of sightings is going to be the most important part of this whole thing anyway and that can't really be "stolen" too easily, whereas (b) if a novel, non-large-crowd effort is what actually makes it work, then defection is a risk that's going to be really, really hard to guard against anyway.
posted by cortex (staff) at 1:55 PM on November 3, 2009 [1 favorite]


empath: "Nobody besides the coordinators should know the locations of any balloons. The downside of that is multiple reports of the same balloons,"

True, but you need to keep in mind the maximum market value for a balloon sighting is somewhere between 4k and 40k. Maybe you're a super charitable guy, but I'd kinda wonder about donating that much to charity when I've got loans to pay off. On the offhand chance I spot one, I'd have to think for a while on who to share it with. And is it evil if I share it with a charity group and sell it to someone else?
posted by pwnguin at 1:57 PM on November 3, 2009


I suspect that DARPA may want to play this all legal-like and contact the city for a permit.

If it's a temporary thing, they'll probably just set the balloons aloft at relatively safe locations. As such, I doubt (m)any cities, counties or other local jurisdictions would care much about one big balloon. Should someone float a large balloon in a questionable location and someone contacts the jurisdiction about it, the balloon will be down before anyone would be in on Monday to check up on it.

I'm intrigued. I'm in coastal California, north of Santa Barbara, and willing to drive around.
posted by filthy light thief at 2:04 PM on November 3, 2009


On the offhand chance I spot one, I'd have to think for a while on who to share it with. And is it evil if I share it with a charity group and sell it to someone else?

Good question. It depends on who you think would actually win. I think the 'correct' strategy would be to report as quickly as possible to every effort you can think of that pays off.
posted by empath at 2:09 PM on November 3, 2009


My wild-assed guess for a location of a balloon: Marconi Beach, on Cape Cod, site of the first trans-atlantic communication initiated from the US.
posted by rmd1023 at 2:10 PM on November 3, 2009


Count me in; waiting on secure server.
posted by RikiTikiTavi at 2:13 PM on November 3, 2009


Here's the Google Tech Talk I was thinking about when writing the earlier monolog. If someone wants to avoid the trusted third party, something like this is your weapon of choice. Privacy-Enhancing Technologies for Mobile Applications.

Here's his webpage if you want to check out his publications that might be relevant. But with only a month to go, there's not enough time to build community trusted decentralized versions.
posted by pwnguin at 2:15 PM on November 3, 2009


Good question. It depends on who you think would actually win. I think the 'correct' strategy would be to report as quickly as possible to every effort you can think of that pays off.

quoting myself here -- which is why I think we should set aside some money to pay off spotters. Just so that we are at least in the game in terms of getting the greed reports. Since the correct 'evil' strategy is to report early and often to all efforts which offer a payment, I don't think the amount matters as long as it's something non-trivial.
posted by empath at 2:28 PM on November 3, 2009


Maybe you're a super charitable guy, but I'd kinda wonder about donating that much to charity when I've got loans to pay off. On the offhand chance I spot one, I'd have to think for a while on who to share it with. And is it evil if I share it with a charity group and sell it to someone else?

It's sort of depressing that people like you exist, but I guess we need to plan for that. But not in this thread. In the secure one.
posted by Kwine at 2:30 PM on November 3, 2009 [1 favorite]


Is it uncouth to worry here publicly that I don't recognize tarheelcoxn's username at all? Then so it is. Anybody willing to vouch for the guy, before we all go spilling our plans on his private server? I feel like I 'know' almost all of the rest of you.

It's totally couth to not trust me, which is one of the reasons I suggested ways to avoid having me be a single point of failure. Maybe you don't trust that cortex could have enough confidence in me for this game based on one meetup and lots of IRC time? Other than that... I guess google me and decide for yourself if I'm reasonably trustworthy on this. Mefi web sleuths are pretty famous for digging up details on people, and there's plenty to start with in my profile. Actually it scares me a bit how findable I am now that I look. Oh, and if you find "me" on facebook please let me know, since I definitely should NOT have an account there. Ugh. I should update my blog.

Anyway, I'm currently waiting on a cert so we can do https on the server. If I don't get it in the next three hours or so I guess I'll make one via CAcert, which will require you folks to import their root cert into your browsers. Shouldn't have to do that, though. odinsdream got the ball rolling on a "real" cert.
posted by tarheelcoxn at 2:31 PM on November 3, 2009


cortex's voucher is good enough for me. We can't afford to go down that rabbit hole and we'll just have to trust each other at the end of the day. And hope pwnguin doesn't find a balloon, I suppose.
posted by Kwine at 2:42 PM on November 3, 2009


Is this the part of the thread where I can check to see if anyone recognizes my username, and thus inflate my own self-worth?
posted by shakespeherian at 2:46 PM on November 3, 2009


It's sort of depressing that people like you exist

Not to derail, but it's generally better to pay off one's own debts first before giving away to charity. You can do both, certainly. But in this case, there's nothing morally wrong there. When you owe someone else money, they can't use that for other things, and when you owe a bank money it's usually accruing interest. If the user in question spots a balloon he's not beholden to Metafilter to report it just because he chanced upon the link here. Let's not be overly herded in here.
posted by Burhanistan at 2:51 PM on November 3, 2009 [1 favorite]


Is this the part of the thread where I can check to see if anyone recognizes my username, and thus inflate my own self-worth?
posted by shakespeherian at 2:46 PM on November 3 [+] [!]


I have no doubt that you would smell just as sweet if you went by any other username.
posted by The World Famous at 2:52 PM on November 3, 2009 [1 favorite]


“1: City by city teams -- folks with cars, iphones and a willingness to drive around all day on potentially a wild goose chase.”

I would suggest zombifying that spotter manpower by organizing rather than DiY-ing and driving around oneself. So get your local folks who as a matter of course drive around a lot – mail carriers, Dept. of transportation folks, bikers, salespeople, truckers, etc. to do the spotting. A given mefite can verify from there and kick it up the chain.
(From there – whatever charity, etc, as purpose for those groups.)

The advantage is metafilter’s lack of centralization and identifiers as an organization per se. So individual members can start their own local groups of mailmen or whatnot and then group coordinate that as a local effort rather than being restricted the way Google or Microsoft will be by cooperation with a given outfit. (They wouldn’t be asking motorcycle clubs to help, for example. And the post office might have its own group. Or not. Either way the manpower can be subverted.)

The localization and to some degree compartmentalization of the effort on the ground as a series of local groups led by individual mefites rather than a more top down effort obviates the security problem of anyone but the coordinators knowing the location of the balloons.

Just a thought.
posted by Smedleyman at 3:01 PM on November 3, 2009


(I’ll add it also mitigates most of the costs that might otherwise be incurred by an individual user – gas, etc. So we’re not racking up debt other than what we might do as a hobby.)
posted by Smedleyman at 3:03 PM on November 3, 2009


(I’ll add it also mitigates most of the costs that might otherwise be incurred by an individual user – gas, etc. So we’re not racking up debt other than what we might do as a hobby.)

Seconding Smedleyman. It might be better to look for ways to leverage existing commutes/routes/trips etc of you and your own network of close contacts rather than just driving around aimlessly. That's a lot of gas just getting burned.
posted by Burhanistan at 3:06 PM on November 3, 2009


tarheelcoxn: Cert is ready. Let me know if you didn't get it.
posted by odinsdream at 3:21 PM on November 3, 2009


Somebody MeMail me with the secret "this is where you place your ideas" part. I got some doozies.
posted by Ironmouth at 3:26 PM on November 3, 2009


Ironmouth, tarheelcoxn and I are working on a secure server. It's not ready yet. We'll send details out soon.
posted by odinsdream at 3:35 PM on November 3, 2009 [1 favorite]


Good luck everyone! Honestly, I'd be shocked if Team MeFi doesn't win.
posted by Admiral Haddock at 3:37 PM on November 3, 2009


One word: Latex.
posted by blue_beetle at 3:40 PM on November 3, 2009


Available in Central Florida (East side from Disney area to Daytona Beach) with car, GPS, and iPhone. Plenty of time to follow leads in this area. MeMail if I can help.
posted by white_devil at 3:45 PM on November 3, 2009


I’ll add too, generally – some folks are thinking sabotage, which is logical, because it requires minimal manpower – but that only insures the other guys don’t win. We should be thinking decentralized execution with coordination in communications, not on the ground effort, which can be managed locally.
From there, someone doing coordination can figure out who’s got what backyard and area. Basic organization stuff.

As to what DARPA’s up to – some speculation.
posted by Smedleyman at 3:47 PM on November 3, 2009 [1 favorite]


"I suspect that DARPA may want to play this all legal-like and contact the city for a permit.

If it's a temporary thing, they'll probably just set the balloons aloft at relatively safe locations.
"

If I were DARPA, I'd just use the copious amounts of federal land available. Surely you enough spots located near highways to hide ten balloons. Social Security offices alone would be enough space without needing a "permit."
posted by pwnguin at 3:55 PM on November 3, 2009


Agreed, it would therefore be logical towards prioritizing federal land/highways over local/state roads. Better to make an excuse to visit national parks over state parks.
posted by amuseDetachment at 4:18 PM on November 3, 2009


Along the lines of disinformation, I forgot to mention this: I think monitoring social networking sites and identifying reliable reports to investigate in the time frame given is extremely problematic. It's so trivial to generate convincing photos and locations, and I suspect fakes will generate trending topics across social networking sites to such a degree that trying to filter them would be unproductive.

I really do think it'd be best to investigate information from only 1st person MeFi reports or their 2nd person trusted-sources (very close friends, relatives). I think the time frame demands it.

posted by empyrean at 12:00 PM on November 3 [+] [!]


Actually, I reckon the point of the exercise is precisely what you've ruled out - that ultimately there will be a flood of reports, and the real exercise here is to use the power of the social network to filter those reports and validate them. Think of this way - the pool of possible spotters is degrees of magnitude greater than the number of people in the Metafilter team (especially considering that you're likely to get noise from people spotting the balloons whether or not they know about the contest), which makes it highly likely that the balloons will be spotted by someone other than a Metafilter team member. This is assuming that they haven't deliberately hidden any of the balloons, but even if they have, a larger network (e.g Something Awful for example) is still more likely to uncover its location first and hopefully leak that information somehow.

In essence I think the point here is not to collect the data but to process it. I'd work out some strategy for monitoring all the available data sources online and work out how to capture and process the relevant information. This is where a Microsoft or Google team no doubt come into their own, seeing as they probably have the skills to automate this side of things.
posted by iivix at 4:19 PM on November 3, 2009 [4 favorites]


Actually, come to think of it, the long time between the balloon launch and the closing date does suggest that at least one balloon will be "hidden", which just reinforces my point - that it won't be spotting that balloon directly that will matter, but data mining online to reveal its location.
posted by iivix at 4:23 PM on November 3, 2009 [1 favorite]


I'm in. Eyes on the greater Buffalo area. Possibly al the way to Albany, depending on the date.
posted by kuujjuarapik at 4:49 PM on November 3, 2009


Possible balloon locations are reported via a webform to a small core group of mefites at "HQ", who then dispatch people to confirm them.

So we're basically talking about an overlord with an army of minions who exist only to do his bidding? Following orders without ever understanding them?

And if I want to join, I'm expected to be a minion?
The overlord position is already taken?
What about the second-in-command badass assassin-bodyguard position - is that available?
No? Just minions?

[sigh]... Fine. Where do I sign up?
posted by -harlequin- at 4:51 PM on November 3, 2009 [1 favorite]


iivix: "In essence I think the point here is not to collect the data but to process it."

I wholeheartedly agree the idea that processing the data is an extremely important factor here - perhaps more important than collecting it. However, if you *can* collect data, then you will have access to more information that will help in processing it; its true source, their reliability, original timestamp, etc.
posted by adrianhon at 4:54 PM on November 3, 2009


So we're basically talking about an overlord with an army of minions who exist only to do his bidding? Following orders without ever understanding them?

Well, assuming that we set aside some money for people who actually find/confirm the balloons, then it might be at least slightly worthwhile to go.

Basically -- it will be, hey, please go see if this thing is real and if it is and we're first, you get a few hundred dollars, and we also donate a lot of money to charity.
posted by empath at 4:57 PM on November 3, 2009


Should the mods close this thread before all the strategy discussion becomes a permanent internet record?
posted by acro at 4:59 PM on November 3, 2009


(though, btw, we should be sending multiple people to confirm each balloon -- i'd hate to get completely shafted by a double agent who goes, says the balloon isn't there, then gives it to some other team offering more money)
posted by empath at 5:00 PM on November 3, 2009


When the secure site is set up, could the mods maybe sidebar the announcement?
posted by Quietgal at 5:08 PM on November 3, 2009


Secure server with messaging will be ready in about an hour.
posted by odinsdream at 5:08 PM on November 3, 2009


shhhhhh. . . .
posted by Ironmouth at 5:14 PM on November 3, 2009


I'm in. The East Bay area going East to Davis, North to hwy 20.
posted by stirfry at 5:25 PM on November 3, 2009


A strange game. The only way to win is not to play.

Seriously, the cynical part of me can't see any logical outcome besides mutually assured destruction. That is, the noise will so greatly outweigh the signal that finding meaningful data will be impossible. It's engineered to stalemate.

That said, I want to play and provided I can pass whatever litmus test is required to join, I'm in. I really have no idea how this "secure server" is supposed to be an effective filter of trust, but whatever. Let's try it. If for no other reason than because this MeTa thread won't be on the front page forever.

I also have a suggestion I think is fairly decent: the buddy system. Cortex or whoever the gatekeeper is should not accept any coordinate pairs as verified unless two distinct members submit them. This requires a bunch of mini-meetups across the country (there's time to organize this!) where Mefites join together to go balloon hunting. Both members of each spotting team have to submit the exact same coordinates to the decimal, to verify they were recorded from the same GPS.

It's not foolproof but it could filter out a lot of noise. Add in some screening to prevent sockpuppets for extra security. Plus: Excuse for meetups! That's fun, win or lose. I'd love to go on a treasure-hunting roadtrip with other Mefites!
posted by The Winsome Parker Lewis at 5:31 PM on November 3, 2009 [1 favorite]


That said, I want to play and provided I can pass whatever litmus test is required to join, I'm in. I really have no idea how this "secure server" is supposed to be an effective filter of trust, but whatever. Let's try it. If for no other reason than because this MeTa thread won't be on the front page forever.

That's the first step. We can work on how to trust eachother if that ends up being a problem... on the alternate discussion site.
posted by odinsdream at 5:37 PM on November 3, 2009


Sorry I'm slow, folks. Messing with a few things so that I don't have service interruptions later to try to keep things locked down. I'll post back soonish when we're ready to start moving people over to the new forum.
posted by tarheelcoxn at 5:46 PM on November 3, 2009


Non-disclosure agreements for all.
posted by Ironmouth at 5:54 PM on November 3, 2009 [1 favorite]


I'm in for Phoenix, AZ and surrounding environments. I volunteer at an airfield some weekends, which might give me access to some interesting aerial intel. Not sure if I have the activity history to qualify, but if I do, I'm interested.
posted by Alterscape at 6:05 PM on November 3, 2009


Just tell us how you feel about favorites, and you'll qualify.
posted by kuujjuarapik at 6:07 PM on November 3, 2009 [1 favorite]


Looks like SA is taking this seriously (their page about the challenge went members-only, most recent public Google cache was around 14:30 GMT).

IMO, they're the biggest competitor, 4chan is not legitimate (many of the key people are actually also SA members and use 4chan as a vehicle for getting a lot of people to do their bidding). If there was a native 4chan intiative, they can't keep a secret and are worthless for any serious organization — they can seek out the balloons, but won't be able to use the information correctly.
posted by amuseDetachment at 6:23 PM on November 3, 2009


If I don't show up for this thing it's because Dragon Age: Origins is installing on my hard drive as we speak and I might DIE AT THE KEYBOARD sometime between now and December 5th. Excelsior!
posted by Justinian at 6:30 PM on November 3, 2009


I'm in Boston, I have a car and I'll be holding a Droid. This sounds awesome. I'd love to help out.
posted by SAC at 6:35 PM on November 3, 2009


As a non-US member, I'd be very interested in reading all the discussion that goes on via the secure server (after the competition is over - let's not encourage unnecessary security breaches by permitting live access to non-participants). So before it all gets taken apart, could somebody make sure archive.org gets a copy?
posted by flabdablet at 6:47 PM on November 3, 2009 [1 favorite]


Cortex or whoever the gatekeeper is should not accept any coordinate pairs as verified unless two distinct members submit them.

Me and my sockpuppet approve this idea.
posted by smackfu at 6:48 PM on November 3, 2009 [1 favorite]


Flabdablet, I think there may be a way for non-US members to participate ... talk to you on the secure server!
posted by Quietgal at 6:51 PM on November 3, 2009


their page about the challenge went members-only, most recent public Google cache was around 14:30 GMT

Nah, they do that from time to time. The whole forums are members only currently.
posted by empath at 7:11 PM on November 3, 2009


I'm in, though I'm in the already well-represented DC area.
posted by gsteff at 7:17 PM on November 3, 2009


If anybody wants to invite me to the Sooper Sekrit Spot What For Havin' Sekritz, be my guest. If there is something I can do to help, I will.

That's all I have to say about that, here.
posted by paisley henosis at 7:26 PM on November 3, 2009


Just wanted to add that I'm down for doing this in the Five College (western Mass.) area. Got no car but I do have free time! Send me a MeMail if anyone else is teaming up within this region!
posted by zer0render at 7:50 PM on November 3, 2009


I'd like to help, representing Lansing, MI.

Also, I kind of love it when Metafilter feels a little like Global Frequency. Just sayin'.
posted by Tesseractive at 8:00 PM on November 3, 2009


I'm in.
posted by beccaj at 8:20 PM on November 3, 2009


I'd like to be in on this.
posted by casarkos at 8:32 PM on November 3, 2009


I'm in-- Mid-Hudson Valley, NY
posted by exlotuseater at 9:54 PM on November 3, 2009


Me too, please, mid-Ohio area.
posted by HopperFan at 10:11 PM on November 3, 2009


Sign me up.

Also, nthing the landmine charity idea.
posted by hattifattener at 10:21 PM on November 3, 2009


Enlisting the help of (e.g. non-MeFite) acquaintances could massively expand the number of eyes and legs we draw on: Each MeFite becomes the funnel whereby a larger, non-geek network feeds balloon reports to us dorks who care. But this means the target charity needs to be essentially beyond politics, religion, free software, and other holy wars. Designate it for Iraq war protestors, and J. Random MeFite can forget about asking his red-stater uncle to keep his eyes peeled. But landmine removal is beyond reproach, probably even in the factory town.
posted by eritain at 10:30 PM on November 3, 2009


I'm another non-US mefite who's interested in helping, should I be able to contribute in any way. How do I get into the Secret Secure Server Supporting Sub-Stratospherical Sphere Seeking? *so proud of myself*
posted by destrius at 10:31 PM on November 3, 2009


I'm in.

We need a focused counter-strategy group. Cortex seems like his organizing the on-the-ground search. I think we need another prong to our strategy. SA and 4Chan have been named as competitors. I bet if larger corporate players like Google and Microsoft are going to enter in on this, they'll be making use of large private networks, and will likely have set up systems to monitor vast amounts of data from various social networking sites. Off the top of my head: Twitter, Facebook, Myspace, analysis from search engine queries, Youtube, Flickr etc. etc. etc. They'll probably also monitor their competitors (including us).

We may take an "on the ground" approach, but we should assume that our competitors have at least some sort of duel strategy: on the ground, and crowd source modeling.

We can't do anything about their on-the-ground strategy. But I think it would be easier to befuddle their crowd-sourcing attempts.

Obviously we shouldn't do anything unethical or illegal. But we could certainly figure out how to throw lots of false positives into the mix. Flickr pictures of weather balloon sightings, false twitter feeds, facebook reports, search engine queries, etc.

It seems that Cortex and others are in the process of building a system for on-the-ground reporting. Someone else should take charge of the counter-insurgency wing of MeFi. Someone who knows search engine optimization, how data is gathered, etc. That person should begin planning effective countermeasures. Others should volunteer to join in. The MeFi countermeasure group should create a secure forum to discuss their plans in.

I'm not going to volunteer because I don't really know much in the way of technical details, and don't think I'm fit to lead this effort. But someone who does should step up to the plate. Anyone?
posted by HabeasCorpus at 10:40 PM on November 3, 2009


Cortex seems like his organizing the on-the-ground search.

I'm happy to brainstorm about and be involved to some extent in said organization, but right now I'm not doing anything other than keeping half an eye on this thread, fwiw.
posted by cortex (staff) at 10:51 PM on November 3, 2009


I wouldn't count on cortex doing anything but being the trusted name to turn in the final results to so we can collect.

Once we get the forum going, we can start divvying up responsibilities, and I already have quite a few ideas along those lines.
posted by empath at 11:08 PM on November 3, 2009


This sounds really awesome - I'm in Atlanta, GA and have access to a car. I can round up a GPS as well.
posted by arcolz at 12:11 AM on November 4, 2009


What we need are drone aircraft.
posted by delmoi at 12:57 AM on November 4, 2009 [1 favorite]


Non-US users can flood twitter with fake balloon sightings.
posted by delmoi at 1:04 AM on November 4, 2009


I don't suppose that someone who doesn't live in the US and who doesn't get much of a chance do much but work in work hours can be very useful, but if there's anything at all I can do to help I'd like to.
posted by calico at 1:15 AM on November 4, 2009


Forum software is live on the new server. After I work out a few kinks with the help of people I've already memailed, I'll start inviting the rest of you who have expressed interest. Keep in mind that this is an early tool for general planning--more and better tools will come later.
posted by tarheelcoxn at 1:24 AM on November 4, 2009


I'm in San Diego and would be willing to lend a hand.
posted by The Pusher Robot at 1:55 AM on November 4, 2009


I have a car and a GPS and would be willing to help too, by the way.
posted by burnmp3s at 3:43 AM on November 4, 2009


instead of drone aircraft, how about balloon-photography rigs (i hear you can make them for under $150 now!) deployed to do surveillance.

better make the balloons on the rigs red, though. :-)
posted by rmd1023 at 5:25 AM on November 4, 2009


Count me in.
posted by Durin's Bane at 5:56 AM on November 4, 2009


I'm in Boston, willing to be local eyes.
posted by marginaliana at 6:09 AM on November 4, 2009


I don't really think this thing is going to play out the way many of you seem to think, but that's cool. I pretty much agree with others upthread that this is going to be a giant clusterfuck. It took MeFights all of 10 minutes to come up with the idea of fake ballons, so you can count on plenty of other folks doing the same. I think the amazing thing will be if anyone manages to win this thing at all.

That said, I assure you if I see a red balloon on my daily 60 mile commute along US 70 in eastern NC I will contact cortex. Good luck, all.
posted by Liver at 6:23 AM on November 4, 2009 [1 favorite]


I'm in Chicago and can mobilize a small eyes army in the western suburbs.
posted by shakespeherian at 6:24 AM on November 4, 2009


There's a whole shitload of interesting social dynamics that this brings out, but I wanted to focus on a couple things that people might want to keep in the back of their minds.

Tactically: The fact the balloons are there for only 24 hours, but there is 9 days to submit results may imply that the balloons are distributed in a pattern, and that by knowing some of the locations, one may be able to deduce other locations.

Strategically: If a large group is what you want, why focus on the competitive, rather than the cooperative? Why are we in competition with SA or other groups? Why not work with them to create a super-large, transparent, open system? The issue to solve in that situation is not keeping information secure, or preventing theft of the information, but rather to prevent a subset of the group attempting to take credit at the expense of the larger group. One approach, would be to create a legal framework and create terms around the use of the publicly available balloon information. This could be similar to any click through license agreement that one sees all the time. So if a subgroup submitted the information, they'd be in violation of the terms. How this would play out in the context of the competition, I have no idea. The main point being that a transparent open structure that protected the interest of the collective from the few would seem to be the most beneficial structure (and not just for this competition).


Regardless, I'll keep an eye out in Northern Virginia and help the group out if I see anything.

Aside:
An entire thesis could probably be written on how organizations involved in competitions quickly develop secret keeping organizations. I mean, change the context from social groups to nation states, and I see a lot of interesting similarities in approaches (the means, of course, are completely different between the two)

posted by forforf at 6:46 AM on November 4, 2009 [4 favorites]


Why not work with them to create a super-large, transparent, open system? The issue to solve in that situation is not keeping information secure, or preventing theft of the information, but rather to prevent a subset of the group attempting to take credit at the expense of the larger group. One approach, would be to create a legal framework and create terms around the use of the publicly available balloon information.

I think this is a really nice idea, but basically impossible to set up during the available time-frame.

The main point being that a transparent open structure that protected the interest of the collective from the few would seem to be the most beneficial structure

I agree. Avoiding prisoners dilemmas is the name of the game here. I hope that you'll take an active role in the planning-it seems that you have some good ideas.
posted by Kwine at 8:25 AM on November 4, 2009


Why are we in competition with SA or other groups? Why not work with them to create a super-large, transparent, open system?

Because all it takes is one person to give our info away to someone who steals the prize. The only way to do this is to have a large group funneling information to a small group of people and NOT sharing that information openly. I'm okay with the process being open, but posting balloon locations openly is a bad idea.

That said, a public wiki-style project would not be a bad idea, if only as a honeypot to get people to contribute information/distribute misinformation.
posted by empath at 8:33 AM on November 4, 2009 [2 favorites]


(Northern AZ - so in.)
posted by joshuaconner at 8:48 AM on November 4, 2009


Count me in as another non-US-mefite who would love to help doing any of the online-only work.

(I'm not terribly active, so dunno if that makes me suspicious as a possible mole. I've been here for a while tho! Oh, it was my 4th anniversary 2 days ago.)
posted by ClarissaWAM at 9:10 AM on November 4, 2009


We're still kicking the tires on the forum. I've made several other people admins over there, and I'm hoping I can start deferring to them sooner rather than later in terms of running that show. I'm looking at you, empath. ;)
posted by tarheelcoxn at 9:58 AM on November 4, 2009


I'm in too - southern CT, have car & phone, no GPS (though I'm sure I can scrounge one up).
posted by widdershins at 10:05 AM on November 4, 2009


I'm in! Raleigh, NC with phone and GPS and ideas.
posted by premortem at 10:14 AM on November 4, 2009


I haven't had a chance to log in yet (busy day at work), but i'll try to look later.
posted by empath at 10:21 AM on November 4, 2009


I am so in. Washington, DC, with GPS - but sadly, no car.
posted by harperpitt at 10:38 AM on November 4, 2009


I'm in, in Santa Cruz CA. No car or GPS, but hopefully I can contribute somehow.
posted by foodmapper at 10:42 AM on November 4, 2009


I haven't had a chance to log in yet (busy day at work), but i'll try to look later.

It's cool. Looks like odinsdream is picking up our slack. I'm stealing glances at it, but I really do have to focus on other stuff today.

If you've memailed me and not gotten a response, please pester again. I'm trying to pass names to odinsdream, so maybe if you were thinking of memailing me ping him instead.
posted by tarheelcoxn at 10:45 AM on November 4, 2009


That's a general you--people who are interested.
posted by tarheelcoxn at 10:48 AM on November 4, 2009


forforf: If a large group is what you want, why focus on the competitive, rather than the cooperative? Why are we in competition with SA or other groups? Why not work with them to create a super-large, transparent, open system?

empath: Because all it takes is one person to give our info away to someone who steals the prize. The only way to do this is to have a large group funneling information to a small group of people and NOT sharing that information openly.

If you consider 'the prize' to be forty large, empath makes a good point. But forforf reminds me of another benefit, perhaps more valuable than $40,000: access to (and maybe control of?) a very large organized information-gathering network.
posted by carsonb at 10:56 AM on November 4, 2009




I have some google wave invites. Please PM me your email address if you want want.
Just a caveat that they don't always send them right away.

...and how the heck would I find the metafilter wave for this...?
posted by trixare4kids at 10:59 AM on November 4, 2009


I work in Ann Arbor, MI, have GPS in the car and an iPhone. I'm interested.
posted by Kip at 11:03 AM on November 4, 2009


Definitely on board. Sign me up when there's something to sign up for.
posted by wemayfreeze at 11:37 AM on November 4, 2009


...and how the heck would I find the metafilter wave for this...?

Anybody with Wave can MeMail me & I'll add them. There's nothing going on there right now, I'm just building a resource in anticipation of needing it later.
posted by scalefree at 11:52 AM on November 4, 2009


IMPORTANT SECURITY WARNING
Please do not use your MetaFilter password for the Secure Forum! Password reuse is a Very Bad Thing.

Thank you for listening. We now return you to your regularly scheduled thread.
posted by scalefree at 12:32 PM on November 4, 2009


I'm in Pensacola, FL, have a car and GPS. I'd love to help out, either on the ground or on the net.
posted by pupperduck at 1:25 PM on November 4, 2009


Please do not use your MetaFilter password for the Secure Forum! Password reuse is a Very Bad Thing.

+1 Good call, scalefree. I should have said this myself earlier.

While we're talking security, I hope it's obvious that people shouldn't paste links to the private forum anywhere, including here. We will have more public resources later, but we're not there yet.

NOT sharing that information openly

Again, I think there will be room for being very open and public later, once we've planned our attack, so to speak. Pester me to join the forum if you'd like to help with planning.
posted by tarheelcoxn at 1:29 PM on November 4, 2009


I'm in the well-represented DC area too. I'll keep an eye out, and I'll be extra-schmoozy with any friends/acquaintances I have who work at DOD. ;)

The fistula charity is great in general, but land-mine removal might be more fitting.
posted by NikitaNikita at 1:58 PM on November 4, 2009


we have legal issues to work on too. not sure the tax issues are squared away here. Any tax attorneys here?
posted by Ironmouth at 3:03 PM on November 4, 2009




NO MORE WAVE INVITES!! Sorry for the all caps, shouting is necessary because I'm overwhelmed with requests.

I wish I could edit previous posts.
posted by trixare4kids at 3:09 PM on November 4, 2009


I think this'll be a fun way for me to explore nearby neighborhoods and towns that I don't know very well. Put me down for central NJ, and send me a memail when there's something I need to do.
posted by painquale at 4:27 PM on November 4, 2009


we have legal issues to work on too. not sure the tax issues are squared away here. Any tax attorneys here?

I think the tax issues are not that complicnated. Cortex is going to be the sole winner, not metafilter. I don't think that metafilter as an entity needs to be involved, which would complicate things.
posted by empath at 4:39 PM on November 4, 2009


Is the secure site down right now or is it just me?
posted by exogenous at 4:55 PM on November 4, 2009


Cortex is going to be the sole winner, not metafilter.

Cortex: "I'm happy to brainstorm about and be involved to some extent in said organization..."

So he's not even the nexus of the endeavor, here - and if we're talking about individuals winning, why not you, since you seem to have spearheaded the effort?

I also thought that the MeFi organization (and by that, I mean Matt) would be the recipient, and then perhaps would donate all or a portion of the winnings to charity, but I think I made that all up in my own head.

I'm probably misunderstanding how this will all play out.
posted by HopperFan at 5:16 PM on November 4, 2009


I think the point is to put both the sightings to be reported and, if that leads to a win, the money, in the hands of someone who is a minimal risk for betrayal.

Setting aside the question of my personal ethics, the social (and, fairly directly considering the nature of my job) financial cost to me for pulling a fast one would be tremendous. Ruining my reputation and losing hundreds of friends and acquaintances is not worth $40K to me; neither is losing my job. Simple as that: there's collateral.

Like I said, I'm fine with being very much in on this the day of, and I'll be chatting about it the interim. I just didn't want to be volunteered as a heavy-lifting organizer in setting it all up since I don't know that I'd either have the time or the best set of skills compared to any number of other folks here who are also enthusiastic about the project.

As far as the tax stuff on the money, that's a question for an accountant or a tax attorney, and a bridge we can plan to cross in the event that it comes up. I'd hope it'd be relatively straightforward to pass the money on to charity without serious tax ramifications, but it's a question for someone who knows what they're talking about.
posted by cortex (staff) at 5:29 PM on November 4, 2009


As far as the notion of paying portions to spotters, I'm pretty cool on the notion. If this is worth doing as a Team Mefi thing with the money off to charity in the event of a win, I think it'd be better to keep it that simple and not complicate things.

Obviously it's not totally analogous, but back when we did the meficomp project, there was some friction at the idea that artists on the album might get shares of sales. We decided against that, but even the discussion soured some folks a little on the whole thing, and when we decided to offer even a discount on bulk orders by participants, that brought the friction right back and remained a small sore spot.

I took from that experience a sense that it's far too easy to underestimate the effect that even a little bit of diluting or sharing-out of a charitable pot in a volunteer community venture can have. I'd rather aim to reward folks with the satisfaction of having been involved and helped out in a crazy effort and go from there.
posted by cortex (staff) at 5:35 PM on November 4, 2009 [4 favorites]


Cortex, that makes sense. I'd never suspect you of even thinking of "trying to pull a fast one," I apologize if my question implied that.
posted by HopperFan at 5:54 PM on November 4, 2009


No, not at all. More just me trying to explain my take on the reasoning for putting me (or someone else in a similar position, but I was around and enthusiastic that day) in the position of holder of those certain sets of keys, regardless of whether or not I do much else in the project.
posted by cortex (staff) at 6:02 PM on November 4, 2009


Count me in as another volunteer in the DC area. I won't be able to go out searching, but I can contribute in other ways that I will mention when we have a more secure site.
posted by bh at 6:43 PM on November 4, 2009


I'm in as a watcher. Car and GPS, great eyesight, Metro-ATL.
posted by pearlybob at 9:01 PM on November 4, 2009


My final thoughts on the charitable donations: its a bit sad that people are jumping to donate their time to go hunt for red balloons and donate the expected gain to charity, when there's plenty of organizations out there guaranteed to benefit from your time spent...
posted by pwnguin at 11:00 PM on November 4, 2009 [1 favorite]


I would suspect that most players are doing it for the fun of it, and the charity angle neatly resolves the problem that the man hours that need to be spent on this are going to be far mire than 40k would buy.
posted by empath at 11:21 PM on November 4, 2009


My final thoughts on the charitable donations: its a bit sad that people are jumping to donate their time to go hunt for red balloons and donate the expected gain to charity, when there's plenty of organizations out there guaranteed to benefit from your time spent...

It's a bit sad if you presume outright that none of these people would ever support charity in any other capacity, but that's itself a pretty uncharitable assumption. Let's not go there; we're all sitting in this thread typing at one moment or another when we could be down at the foodbank or running a needle exchange.
posted by cortex (staff) at 7:19 AM on November 5, 2009


I'd hope it'd be relatively straightforward to pass the money on to charity without serious tax ramifications, but it's a question for someone who knows what they're talking about.

This lawyer thinks that hope is the worst way to deal with tax questions. Knowing is far better. The more I think about this stuff, the more I start to think that maybe, from a legal standpoint, this isn't the brightest idea.

Also, I signed up, but never received confirmation.
posted by Ironmouth at 9:25 AM on November 5, 2009


Also, I signed up, but never received confirmation.

Check your spam folder. I thought the same thing but Gmail had marked it as spam, even though it almost never does that with forum registration emails for me.
posted by burnmp3s at 9:59 AM on November 5, 2009


Thanks burnmp3s, mine got spam dumped as well.
posted by Mitheral at 10:19 AM on November 5, 2009


If anyone is still reading here who would like to discuss on the alternate site and you've not received an invitation, please let me know.
posted by odinsdream at 11:43 AM on November 5, 2009


Hey folks. I need volunteers with LDAP and CRM experience. Please also ping me if you're a GIS whiz.

Is the secure site down right now or is it just me?

Oh poo. I guess I'll point Nagios at this thing. You can also memail me if you're having trouble. I'm trying not to spend too much personal money on this project, especially early on, but I can throw more resources at the forum if it's necessary.

Check your spam folder. I thought the same thing but Gmail had marked it as spam, even though it almost never does that with forum registration emails

Yeah sorry about that. We didn't have the DNS records in place that make you look legit when we started. That should be better now. If it's not, please let me know.
posted by tarheelcoxn at 1:08 PM on November 5, 2009


Thanks, tarheelcoxn, for setting that up. It was only down for a short while.
posted by exogenous at 2:14 PM on November 5, 2009


It's down for me. Found the emails in my spam, un-spammed them and trying to go, nothing.
posted by pearlybob at 3:22 PM on November 5, 2009


Never mind. Hit the link in my memail and I'm in. Thanks for setting all this up!!
posted by pearlybob at 3:29 PM on November 5, 2009


Oh, my mistake. It wasn't down before - I was trying to take a shortcut instead of using the link in the mefi mail.
posted by exogenous at 5:06 PM on November 5, 2009


I'm interested n that I'm always on the road, and should be able to keep my eyes peeled on that day, wherever I am.
posted by pjern at 1:17 AM on November 6, 2009


Darpa Challenge Signup form

Share as much or as little contact info as you feel comfortable. I'm only going to share it with the volunteer coordination team once we have it in place.
posted by empath at 9:49 PM on November 6, 2009


We're asking for volunteers for specific organizational roles over on the forum. If you're seriously interested we need your help. Please contact me or tarheelcoxn, or visit the forum for more details.
posted by odinsdream at 7:55 PM on November 17, 2009


« Older Pinning/sidebarring mod posts?   |   AskMe answer makes it to reddit. Newer »

You are not logged in, either login or create an account to post comments