Skip

Red wins!
May 20, 2005 10:54 AM   Subscribe

Scientists have declared that red outfits give otherwise evenly matched athletes an advantage. Could this explain the history of success of the Red Wings, or last year's World Series victory of the Red Sox, or the 1970s dominance of the Reds? Uhh, most likely not. The nameless red-shirted Star Trek crewmembers could not be reached for comments.
posted by goatdog (30 comments total)

 
I hear if cops put violent prisoners in pink jail cells they calm down a lot more quickly too. But then they keep catching "teh gay".
posted by fenriq at 10:58 AM on May 20, 2005


Whoa, I parsed that as "history of success ... of the Red Sox."
posted by cl at 10:58 AM on May 20, 2005


I would say, even as a Sox fan, that the Red Sox history over the last 86 years specifically debunks that theory.
posted by horsemuth at 10:59 AM on May 20, 2005


That's why the Packers, Yankees, Lakers and Celtics have never won anything... oh wait.
posted by mania at 11:01 AM on May 20, 2005


The Red Sox, but for their predominantly red caps in the mid to late 70s have never really had all that much red in their uniforms.

It could explain the dominance of the CCCP in 1960s, 70s and 80s however. Well, that and the state sponsored pre-school through adulthood Soviet athletic academies.

Interesting to note that where there was a trend in the early 90s to reoutfit teams in black (and or teal) to appeal to cash in on merchandizing, maybe we'll be seeing more teams with red jerseys?
posted by Flem Snopes at 11:02 AM on May 20, 2005


They could probably get the Unkillable Mr. Leslie to do an interview...
posted by zoogleplex at 11:02 AM on May 20, 2005


Btw what was that line from that flick with Tom Hanks and Leonardo diCraprio?

"The Yankees win all the time because their opponents get hypnotized by the pinstripes."
"No they don't they win all the time because they've had Ruth, Gehrig, DiMaggio, Mantle, Mays, etc..."

I'll proscribe to that theory before the uniform color theory anyday.
posted by Flem Snopes at 11:05 AM on May 20, 2005


MaysMaris d'oh!
posted by Flem Snopes at 11:06 AM on May 20, 2005


Note the study used Olympic athletes as the sample.

Duh.

As a scientist(-in-training), I gots to say that's some shitty experiment design.
posted by docgonzo at 11:09 AM on May 20, 2005


Don't give the Leafs any more excuses, please.
posted by orange swan at 11:23 AM on May 20, 2005


It sounds odd, but I think there's something to it. Look at football teams that share a city - invariably, the team in red does better. Lisbon (Benfica/Sporting), Liverpool (Liverpool/Everton), Rome (Roma/Lazio), London (Arsenal/Tottenham), Munich (Bayern/1860) or Manchester (United/City). This is more than a coincidence.
posted by salmacis at 11:28 AM on May 20, 2005


As soon as I heard about this latest attempt at fifteen minutes last night, all I could think of was the Yankees. And I'm from Boston. So no, no, again no, and finally nope.
posted by jsavimbi at 12:00 PM on May 20, 2005


I have to agree Docgonzo, as a fellow scientist in training - this is just a tiny, correlationary study. Let's just pretend these idlers didn't make any silly hypotheses based on a tiny sliver of data, which they did not organize, control, or probably really look that close at.
posted by BlackLeotardFront at 12:25 PM on May 20, 2005


I don't really think that the American teams named in the initial post have all that much red in their strips. Certainly not like the football examples given by salmacis.

Continuing on into Rugby we've got Munster from Ireland, Llanelli from Wales, Toulouse in France and it does seem to hold to a certain extent there too!
posted by daveirl at 12:37 PM on May 20, 2005


Continuing on into Rugby we've got Munster from Ireland, Llanelli from Wales, Toulouse in France and it does seem to hold to a certain extent there too!
posted by daveirl


...and it explains why the All Blacks suck....

No, wait...
posted by Floydd at 12:59 PM on May 20, 2005


It sounds odd, but I think there's something to it. Look at football teams that share a city - invariably, the team in red does better. Lisbon (Benfica/Sporting), Liverpool (Liverpool/Everton), Rome (Roma/Lazio), London (Arsenal/Tottenham), Munich (Bayern/1860) or Manchester (United/City).

But what about Sheffield...
posted by ZippityBuddha at 2:47 PM on May 20, 2005


My math-fu is pretty weak, but isn't there about a 13.6% chance that 29 coin flips will result in at least 19 heads or 19 tails?

Add that to a less-than ideal sample (perhaps larger countries are likelier to use red uniforms), and this article is just fucking stupid.
posted by Kwantsar at 3:53 PM on May 20, 2005


London (Arsenal/Tottenham)

Charlton? the Addicks are in London, wearing red... :)
posted by First Post at 4:51 PM on May 20, 2005


I've already heard it spun into politics--GOP=Red, redstates vs. bluestates, etc.
posted by amberglow at 4:55 PM on May 20, 2005


My math-fu is pretty weak, but isn't there about a 13.6% chance that 29 coin flips will result in at least 19 heads or 19 tails?

Yes, except that's not the whole thing--it's not 19 matches, it's 19 weight classes with several matches within each weight class.

perhaps larger countries are likelier to use red uniforms

The researchers included only data from four sports (boxing, tae kwon do, Greco-Roman wrestling, or freestyle wrestling) where the competitors are randomly assigned red or blue uniforms for each match. So a) the color of uniforms are not dependent on what country the competitor is from; and b) the same competitor may not even have the same color across several matches.

However, I'm not defending the researchers' conclusions. In fact, I've found something very interesting in the raw data. The raw data is available neither in goatdog's link nor in the brief Nature article itself. It is available in the "supplementary material" for the article available from Nature's website, if you're a subscriber. Fortunately, I'm at an institution that does subscribe. The authors don't break down the results by sport in their article. I did, based on their raw data, and these are the results, expressed as the number of red victories out of the total number of matches.

Boxing: 150/272
Tae kwon do: 45/80
Greco-Roman wrestling: 25/51
Freestyle wrestling: 28/54

Of the four sports, the effect seems to exist only for boxing! The distribution for the other three sports are well within what's expected by chance. And the effect is seen for the overall data (248/457) only because there happen to be so many more boxing matches than tae kwon do and both forms of wrestling combined. Even looking at boxing alone, the data, while suggestive, fall a bit short of statistical significance. At best, the researchers have shown evidence that is suggestive that a red uniform may give a relative advantage over a blue uniform in boxing. That's another point: all the matches considered were red vs. blue, so it's overstating it to say "red outfits give an advantage" without noting that that is only compared to blue uniforms. There's no information on red vs. white, or yellow vs. blue, etc.
posted by DevilsAdvocate at 5:02 PM on May 20, 2005


The authors don't break down the results by sport in their article.

Whoops, now I'm overstating my own case--they do provide a graph of the breakdown by sport in the Nature article. However, the graph is at best misleading, because the graph shows the percentage of red and blue winners in each sport, rather than the raw numbers, which would more clearly show that for three sports the effect easily lies within what would be expected by chance. Also, their by-sport numbers seem to be slightly different by mine, apparently because they've excluded a few of the matches for various reasons (as described in their methods in the supplementary information) whereas I counted all matches listed in the raw data.
posted by DevilsAdvocate at 5:21 PM on May 20, 2005


Boxing: 150/272

And thats still only 55%. Not very strange, at all, really
posted by ZippityBuddha at 5:40 PM on May 20, 2005


Suddenly, Orks seem less dumb.
posted by john at 5:49 PM on May 20, 2005


Great breakdown, DevilsAdvocate. That's why I love Metafilter.
posted by goatdog at 6:08 PM on May 20, 2005


Manchester United are playing Arsenal in the FA Cup today (Saturday) and both teams play in red (albeit ManU's strip is more red). Each are as good as each other and hate the others with passion so it's evenly balanced. However I believe that ManU are playing in black tomorrow so if the rules are adhered to then Arsenal should win. Let's wait and see.
posted by movilla at 6:54 PM on May 20, 2005


Perhaps athletes have a subconscious reaction to red that puts them at a disadvantage, the researchers contend.

I'd be more suspicious of bias by the referees and judges, they may subconciously associate red with a "winner" more.
posted by bobo123 at 8:20 PM on May 20, 2005


As a Kansas City Chiefs fan I can assure you... it's bullshit.
posted by Hugh2d2 at 8:24 PM on May 20, 2005


Manchester United are playing Arsenal in the FA Cup today (Saturday) and both teams play in red (albeit ManU's strip is more red). Each are as good as each other and hate the others with passion so it's evenly balanced. However I believe that ManU are playing in black tomorrow so if the rules are adhered to then Arsenal should win. Let's wait and see.

Spooky! (Arsenal won 5-4 on penalties, despite being outplayed for the whole game).

As regards London: the rise of Chelsea (playing in blue) may contradict this theory. And what about Madrid? Real play in white, and Athletico in red and white.
posted by Infinite Jest at 1:37 AM on May 22, 2005


Boxing: 150/272 (=55.15%)
Tae kwon do: 45/80 (=56.25%)

Of the four sports, the effect seems to exist only for boxing!

Your stats seems to suggest Tae kwon do also.
posted by biffa at 10:06 AM on May 23, 2005


Thanks, biffa, for demonstrating precisely why you can't judge these things on percentage alone.

If you flip a coin four times and get three heads, do you conclude there is a strong bias towards heads? After all, it came up heads 75% of the time!

Which is why statisticians use the concept of statistical significance. 3 heads out of 4 flips is not statistically significant. Now, if you flipped the coin 400 times and got 300 heads, that would be statistically significant, despite the fact that the percentage of heads is exactly the same.

45/80 is not statistically significant. 150/272 is, barely.
posted by DevilsAdvocate at 8:48 AM on June 3, 2005


« Older Yes, they deserved to die! I hope they burn in...   |   Here piggy piggy...... Newer »


This thread has been archived and is closed to new comments



Post