Distributed Polling
The poll I did yesterday on Yahoo!’s closing price has been taken almost 2,200 times in about a day. How did it get so many takers? By being distributed on many blogs at the same time. The poll has a share tab with embed code and I asked my readers to embed it on their blogs. And I left comments on some of the most popular blogs suggesting they do so. As of this morning here’s the list of blogs that picked up the poll code and reposted it.
http://blog.hubdub.com/2008/05/05/what-price-will-yahoo-close-at-on-monday-may-5th/
http://gigaom.com/
http://gigaom.com/2008/05/03/microsoft-yahoo-bid-over/
http://hexia.net/frettavefur/technology
http://laurent.pierssens.com/2008/05/04/microsoft-walks-how-low-will-yahoo-close-tomorrow/
http://seekingalpha.com/article/75643-where-will-yahoo-close-today
http://www.alleyinsider.com/2008/5/play_the_where_will_yahoo_s_stock_go_now_game_yhoo_
http://www.collierclan.net/mark/index.php/2008/05/04/yhoo-stock-price-poll/
http://www.contentmatters.info/
http://www.contentmatters.info/content_matters/2008/05/microsoft-ends.html
http://www.stoweboyd.com/message
http://davemartin.blogspot.com/2008/05/if-all-else-fails-immortality-can.html
http://www.buzzya.com/2008/05/04/microsoft-to-yahoo-take-a-hike/
http://mojo-ramblings.blogspot.com/2008/05/microsoft-and-yahoo.html
http://amaappendagy.blogspot.com/2008/05/spliced-feed-for-social-media_04.html
http://technology.nuovoportale.com/microsoft-to-yahoo-take-a-hike/
http://swik.net/Technology-News/GigaOm/Microsoft+to+Yahoo:+Take+a+Hike!/b4tv2
http://nobosh.com/s/play-the-where-will-yahoos-stock-go-now-game-yhoo/57824/
http://www.mybloglog.com/buzz/members/fredwilson/newwith/me/?eid=1209917398_17_48186561
http://aware11.com/play-the-where-will-yahoos-stock-go-now-game-yhoo/
I love this concept. We can work together to figure stuff out. Instead of each of us hosting a small poll on our blogs, we can collaborate on one large one.
As for whether the poll was accurate, too soon to know. The wisdom of the crowd was around $22/share this morning before the market open. So I lobbed in a limit order before the market opened at $22 for the day. As of now, that order has not been filled.
We’ll see if the stock gets to my $26 prediction. As of now, it’s approaching $25. So much for the wisdom of crowds I guess.
Comments (Archived):
I like it! I just put in a (very small) buy at $22.50. Trying to be precise, even with the wisdom of the crowds on your side doesn’t usually result in filled orders. Either way, it’s not looking like it will be filled, at least not today, but there’s still over 3 hours left.
My guess on the diff between $22 and $25 is that the media today has an angle that MSFT might not be fully out, that this might still be a negotiating ploy. Over the weekend seemed like moer of a sure thing. I think if it becomes very clear that the deal is fully over that we’ll see the stock trade down significantly.
Here is the problem Fred. I dont trust the crowds you used. Davi winer was panicked on stocks two months ago and now gleefully adding. Om and TechCrunch readers are just firing off guesses without knowing the intracies of arbitrage.OI think polling like this is cool, but you really need to poll your trusted community at least I would and do, to get a picture of what the price could be and what I would be willing to trade MY money with.The crowd gets too bearish and bullish. always.
Howard, good points, though James Surowiecki and his Wisdom of Crowds disagrees. In fact, one of the key points he makes very early in his book is that having a diverse (even random) community who participate is crucial to a good prediction. Perhaps all of the distributed sites are/were not diverse enough, but his key argument is that selecting your crowd is the same as having an “expert” pick and that is worse than the crowd.
I hate that guy :)Long my dumb communรขยยty ๐
LOL
Good point Seth, however one caveat is that in general he was talking about a prediction market rather than a poll. Prediction markets tend not to be affected my the injection of ‘dumb money’ as the ‘smart money’ brings it back to its true value. Polls unfortunately have no self-correction mechanism.
You are wrong on this one Howard.And I am a nerd, that kind of people you don’t seem to like much ;-)http://vruz.tumblr.com/post…
Vruz…I am never wrong. ๐ .It looks like my commnity was exactly right in picking prices. The wisdomof a controlled crowd may be a better title for his book
i love your green head. excellent disqus choice.
you’re right on that one, Howard ๐
One thing for sure…disqus is great
I had a huge cluster of picks right around the opening price correct on my blog and time will tell on my groups closing prediction ability
That is a really cool concept for polling. Considering I don’t have a land line, I’m eager to see this kind of thing applied in a scientific fashion on the internet. Perhaps a Open Social / Facebook app that used the demographic information to normalize the results that are skewed from the mean in it being an online community.Note that the wisdom of crowds often works out because of people who don’t know anything that guess wildly. The poll didn’t allow for this, so it could be one reason it’s off.
Interesting that our prediction market had the $22 option as the favorite at the opening bell (at about 30%). We kept it open during trading and it (unsurprisingly) shifted very quickly to the $24 option. In general, it isn’t a good idea to use a play money market to predict a real money market. Fun, but unlikely to be a profitable strategy ๐
one thing off topic:could you please stop putting things on your feed that autostart video/audio? It makes using greader a pain in the a**(I’m talking about that gillette-thing above the text in this post in the feed)
I have no idea what you are talking about because I don’t read my feed (for obvious reasons). Is it an ad?
that’s weird. yeah, I guesssed it’s an ad.and I thought it was strangely placed (above and not below the article).Disqus schrieb:
LONG CDPs Quibblo and Disqus to gather and deploy folks across lots of sites.What a cool test of the javascript.
Yeah man… cross domain services…. they’re pretty cool.
great post