Monday, February 04, 2008

The Superbowl, David Hume and Global Warming

Why were the Patriots so heavily favored to win the Superbowl? Of course, they're a very good team, but let me suggest that part of the answer was an inductive fallacy. The Patriots came into the game undefeated. However, they won some of their games---including their playoff games and their final regular-season game against the Giants---by small margins. Had a few balls bounced differently, they might not have been undefeated; indeed, they might not even have been in the Superbowl.

Nonetheless, bettors and pundits looked at the 18-0 record and thought that it was almost a law of nature that, even if the score was going to be close, the Patriots would win. Why? Because that's how it worked out in all of their previous games this season. And there you have the inductive fallacy. As the great Scottish philosopher (and more) David Hume explained in the 18th century, the past never predicts the future, except if one assumes that the future will continue to be like the past, and there is no basis for that assumption other than the assumption itself (i.e., the past has predicted the future in the past).

Hume was not, however, a nihilist. In many ways, he was the father of modern science. And so the rational response to the problem of induction is to look for patterns and explanations that explain why the past has been the way it has, and then ask whether those patterns continue. To be sure, this answer depends on notions of causality that Hume also critiqued, but practical-minded people can take a more practical lesson from Hume: You need a reason to think that the future will be like the past. In the case of the New England Patriots, it's fair to say that the reason they won so many of their close games was some combination of toughness under pressure and luck. Given that the Giants, of late, had also shown toughness under pressure, good empiricists should have expected that the game had a decent chance of going either way, i.e., that luck would determine the outcome. As indeed it did. Had Eli Manning---not generally known for his scrambling skills---not somehow slipped away from that near-certain sack, the Patriots probably would have won the game.

Inductive fallacies are most dangerous when applied to non-linear phenomena. Part of what has thus far made many Americans resistant to calls for serious efforts to curb greenhouse gases, I suspect, is the apparent evidence of their senses. We've been pouring CO2 into the atmosphere for decades and yet they don't see that much change. And that's because global warming depends on feedback loops that reach tipping points. E.g., Siberian permafrost melts, releasing greenhouse gases and reducing reflectivity, thus increasing heat absorption, leading to higher temperatures, leading to more permafrost melting, and so on. The phenomenon starts out slowly, and no one pays much attention until the effects are dramatic, by which point it's too late.

Jared Diamond's book Collapse provides numerous historical examples of whole societies not noticing how they're destroying their natural environment until the present dramatically stops being like the past. By that point, it's typically too late to change.

Is it a hopeful sign or utterly pathetic that as his prize for winning the Superbowl MVP, Eli Manning received a gigantic Cadillac SUV that happens to be a hybrid?

Posted by Mike Dorf


hahamoudi said...

Mike, perhaps I might offer a more postmodern critique? Concerning what you describe as the "rational" way to consider the Patriots-Giants outcome, that is, to search for causes as to why the past might explain the future, and finding such a cause for past Patriots success in "toughness and good luck", I have a question. Is this really "rational"; that is, is it based in reason and not just itself in the best case a repetition of the same fallacy we are supposed to avoid?

That is, suppose I had said that the Patriots had won their games because there's a fellow who is a shaman who happened to watch them each week, and they've won every game he's seen, and he missed the last game they lost last year to the Colts. People should have realized the shaman wasn't going to see the game, I might argue, and so the Patriots were doomed. Didn't I do what you recommend and search, and find a cause that escaped the inductive fallacy, and provide a reason that the past might explain the future other than it just worked in the past? (To use a real life example, every bowl game that my friend Sanjay has seen Ohio State play in live in the last eight years, they've won, and every one he hasn't seen, they've lost. There is empirical evidence for this. Maybe he's a shaman.)

Surely the New York liberal cocktail circuit would dismiss this as some sort of religious superstitious voodoo, but why? What is the basis for the conclusion when the empirical data is equally solid, when the cause is just as external to the fallacy. What did I do wrong? It seems that in the search for causes, there are "legitimate" causes, and "illegitimate" ones, and there is no real way to separate those two except by the inductive fallacy that we are supposed to abandon, i.e. I've never seen a shaman successfully work the spirit world, and therefore the shaman is unlikely to affect this football game.

I suppose you could say one can test my idea to see if the shaman theory holds up, but reality is that nobody does. Nobody thinks it will, and they don't look for causes like that so for all we know there is such a shaman out there, and his presence explains the outcomes perfectly. Instead we base our ideas on what did and didn't lead to Patriots success solely on those combination of factors that might have led to success in the past (toughness? speed? agility? past competition?), and we test those. But that reduction is still inductive fallacy, it's all inductive fallacy in my view.

(And that by the way is on the generous assumption that we can even determine which factors truly led to success in the past, as opposed to merely subconsciously inventing past reasons in a manner that makes sense to our dispositions and construing the data to confirm those inventions, but I'll avoid that point since we are talking about inductive fallacy.)


AF said...

Not much of a football fan are you? The Patriots were not favored because they were undefeated. They were undefeated because they were great, and that is why they were favored.

During the regular season the Pats set an NFL record for points scored and averaged almost 20 pts a game more than their opponents. They did not win an abnormal number of close games. Only 4 of their 18 wins leading up to the Super Bowl were by less than a touchdown.

It is not "fair to say" that the Patriots won so many games because of "some combination of toughness under pressure and luck," if you mean toughness and luck as opposed to superior talent and execution. By every indication, the Patriots were historically good, not just lucky.

You are right, of course, that good empiricists should have expected that the game could go either way, ie, that there was a non-trivial chance the Giants would win. But that is because on "any given Sunday," a good team can beat a great team. It is not because of the inductive fallacy.

Unknown said...

And that by the way is on the generous assumption that we can even determine which factors truly led to success in the past, as opposed to merely subconsciously inventing past reasons in a manner that makes sense to our dispositions and construing the data to confirm those inventions

Actually, Haider, there is nothing particularly "postmodern" about your critique at all - it's precisely the account Hume himself gives of our tendancy to presume that the future will resemble the past - it's a mere psychological tic or habit. Hume, of course, thought this a perfectly satisfactory solution to his skepticism, largely because he believed everyone shared the same basic dispositions. I guess the postmodern spin, then, is that Hume was wrong on this point. Whether Hume and the post-moderns are right on the deeper point, however, is a further question. Clearly it is logically possible that one and the same event could have any of an infinite number of possible explanations. This is just to point out that induction is not deduction. To say, however, that therefore there is no empirical difference between claiming some phenomenon was the result of well-verified natural processes rather than the color of the socks your shaman of choice happens to be wearing when the sun, earth, and moon are in proper alignment is a bit of a stretch.

Michael C. Dorf said...

I'll leave the philosophy to the other commenters, but I must defend my honor against af's scurrilous attack on my football knowledge. I have been wasting time watching football, and the Giants especially, for most of my life. The Patriots were a great team for the first half of the season. Then, in Week 9 they squeaked by their nemesis of recent years, the Colts. After that, they won one more blowout---against the Bills, a team that ended up sub-500---then squeaked by the Eagles and the Ravens in successive weeks. The Patriots' only commanding win over a quality opponent in the second half of the season came over the Steelers. The Giants, with nothing to play for but pride, nearly beat the Patriots in the last week of the regular season, and the Patriots won solid but hardly dominating victories in their two prior playoff games. Yes, the Patriots scored lots of points but anyone paying attention over the last two months could not have possibly thought they were the sort of dominant that the Bears were in 85 or the Steelers teams were in the 70s. At least not against the Giants, who had already shown that they matched up well with and were not intimidated by the Patriots. That all suggests that the belief in the Patriots' destiny was based in substantial part on the fact that they had not lost any games prior to the Superbowl.

AF said...

Apologies for impugning your football knowledge, and congratulations to the Giants!

But the errors at issue here are hindsight bias (on your part), and perhaps a touch of overconfidence bias (on the pundits' part). The inductive fallacy has nothing to do with it.

Jamison Colburn said...

You have to have a reason to think the future will be like the past. Indeed. So will the Celtics implode or will the Knicks pull themselves up by their bootstraps? Or neither? ;)

Paul Scott said...

The result did not surprise me. I expected a close game. Not because the Patriots were not dominant this year, but because of a combination of factors that made the match-up likely to be fair (or close enough to fair to make luck play a decent role).

Mike has done a decent job on the Patriots half of the equation, but there is also the Giants half.

In the last 8 games they played they barely lost to the second best offense in football (Dallas), beat the third best offense (Green Bay), barely lost to the best offense (Patriots), then beat the second best offense (again, Dallas). The games were all close, but they were also games played to the Giant's late-found defensive strength.

There was no way the Giant's were going to win in a shoot-out, but with both defense's playing exceptionally well (even for them) that Sunday, the Giant's had a good chance - if not an advantage. That happened and thus the result.

As a long-time (since I was five) Dallas fan, this is not some Giant-biased analysis. I was really hoping the Patriot's would win (in part because I like watching the Giant's lose and in part because I am really sick of the '72 Dolphins blabbing on about who great they were when there have very clearly been better teams in history - the 08 Patriots and 85 Bears being the two most obvious examples).

Anonymous said...

免費A片, ut聊天室, AV女優, 美女視訊, 免費成人影片, 成人論壇, 情色交友, 免費AV, 線上a片, 日本美女寫真集, 同志聊天室, 聊天室交友, 成人文章, 成人圖片區, 色情網站, 辣妹視訊, 美女交友, 微風成人區, 色美媚部落格, 色情影片, 成人影片, 成人網站, 免費A片, 上班族聊天室, A片,H漫, 18成人, a漫, av dvd, 一夜情聊天室, 微風成人, 成人圖片, 成人漫畫, 情色網, 日本A片, 免費A片下載, 性愛, 成人交友, 嘟嘟成人網, 嘟嘟成人網, 成人貼圖, 成人電影, 成人, 中部人聊天室, 080中部人聊天室, 成人貼圖, 成人小說, 成人文章, 成人圖片區, 免費成人影片, 成人遊戲, 微風成人, 愛情公寓, 成人電影, A片, 情色, 情色貼圖, 情色文學, 做愛, 成人遊戲, 成人影城, 色情聊天室, 色情小說, 一葉情貼圖片區, 情色小說, 色情, 寄情築園小遊戲, 色情遊戲, 成人網站, 麗的色遊戲, 色情網站, 成人論壇, 情色視訊, 情色電影, aio交友愛情館, 言情小說, 愛情小說, 色情A片, 情色論壇, 自拍, 癡漢, , 俱樂部, 豆豆聊天室, 聊天室, 色情影片, 視訊聊天室, 免費視訊聊天, 免費視訊, 視訊交友90739 情人視訊網影音視訊聊天室 免費視訊聊天室 視訊聊天 視訊交友 美女視訊 視訊美女 視訊 免費視訊 免費視訊聊天 視訊聊天室 辣妹視訊 一夜情 色情a片 aio交友愛情館 情色電影 情色視訊 色情遊戲 色情 情色小說 一葉情貼圖片區 色情小說 色情聊天室 情色交友 成人論壇 成人網站 色情網站 情色論壇 小高聊天室 女同志聊天室 6K聊天室 080苗栗人聊天室 080聊天室 聊天室尋夢園 UT男同志聊天室 男同志聊天室 尋夢園聊天室 UT聊天室 聊天室 豆豆聊天室 A片 成人電影 成人貼圖 嘟嘟成人網 美女交友 本土自拍 成人交友 成人影片

Anonymous said...

I am so happy to get some hero gold and the hero online gold is given by my close friend who tells me that the hero online money is the basis to enter into the game. Therefore, I should buy hero gold with the spare money and I gain some hero money from other players.

Anonymous said...

酒店喝酒,禮服店,酒店小姐,酒店經紀,制服店,便服店,鋼琴酒吧,兼差,酒店兼差,酒店打工,伴唱小姐,暑假打工,酒店上班,日式酒店,舞廳,ktv酒店,酒店,酒店公關,酒店小姐,理容院,日領,龍亨,學生兼差,酒店兼差,酒店上班,酒店打工,禮服酒店,禮服店 ,酒店小姐,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,經紀 彩色爆米花,經紀人 彩色爆米花,酒店傳播,酒店經紀 彩色爆米花,爆米花,童裝,童裝拍賣,童裝大盤,童裝寄賣,童裝批貨,酒店,酒店,童裝切貨,酒店,GAP童裝,酒店,酒店 ,禮服店 , 酒店小姐,酒店經紀,酒店兼差,寒暑假打工,招待所,

Anonymous said... .
[url=]puma shoes[/url]
[url=]chaussures puma[/url]
[url=]nike air max ltd[/url]

Anonymous said...


Anonymous said...

酒店經紀人, 菲梵酒店經紀, 酒店經紀, 禮服酒店上班, 酒店小姐兼職, 便服酒店經紀, 酒店打工經紀, 制服酒店工作, 專業酒店經紀, 合法酒店經紀, 酒店暑假打工, 酒店寒假打工, 酒店經紀人, 菲梵酒店經紀, 酒店經紀, 禮服酒店上班, 酒店經紀人, 菲梵酒店經紀, 酒店經紀, 禮服酒店上班, 酒店小姐兼職, 便服酒店工作, 酒店打工經紀, 制服酒店經紀, 專業酒店經紀, 合法酒店經紀, 酒店暑假打工, 酒店寒假打工, 酒店經紀人, 菲梵酒店經紀, 酒店經紀, 禮服酒店上班, 酒店小姐兼職, 便服酒店工作, 酒店打工經紀, 制服酒店經紀,,