5 ms·

The stats look really damning. Notice the p-scores : 4e-19 for worlds 2013 and 5e-8 for olympics 2016. And you can see it so clearly in the graphs too. No qu

by dzdt 10y ago

The stats look really damning. Notice the p-scores : 4e-19 for worlds 2013 and 5e-8 for olympics 2016. And you can see it so clearly in the graphs too. No question, there was a big difference between one side of the pool and the other!

Alex3917 10y ago
Those aren't the stats that really matter though. At the end of the day, the question isn't whether the pool is fair or not, but whether the people watching the event on TV still go out and eat McDonalds or whatever afterwards. Most sports aren't really 'fair' per se. Making them fair wouldn't be cost effective, and wouldn't make for good TV. E.g. everyone knows that some lanes are faster than others in rowing due to the lack of wind screens around the venues. And that's not even to mention the people who spend ten years training only to get violently ill and unable to compete due to the water quality. The reality is that even at the highest levels of competition there is still a lot of luck.
- 0xdeadbeefbabe 10y ago
  I'm annoyed by NBC's approach to streaming online where you have to login. I suppose they don't care about me, because I don't eat at McDonalds. Edit: Also you are right.
- yodsanklai 10y ago
  > The reality is that even at the highest levels of competition there is still a lot of luck. It really depends on the sport. For instance, football (soccer) is very random, and this is amplified by the tournament format. But in many other sports, there is no suspens at all. Actually, it would be interesting to somehow evaluate the random factor for each sport.
  - pesfandiar 10y ago
    Even in sports with less variance in performance (e.g. weightlifting), we're only looking at one or very few data points. There are definitely random factors in the final outcome, but it may even add excitement to the events. That said, I still think fairness or at least perception of reasonable fairness is very important in high-level sports. No one is going to watch the games (or the ads as the GP pointed out) if they think the game's rigged.
  - mamon 10y ago
    Some even argue that the randomness of football (which Americans know by riddiculous name "soccer") is the main factor that makes it the single most popular sport in the world. Some unlucky events end up being remembered for generations. For example people in Poland are still discussing over the beer whether we would have been world champions if only it hasn't been raining so heaviliy during Poland - West Germany semi-final in 1974. P.S English is not my language and I'm pretty sure I messed up with conditionals, hope you still get what I meant in the last sentence.
    - evincarofautumn 10y ago
      The name “soccer” is just an old British slang abbreviation of “Association Football”, to differentiate it from other kinds of football (that is, games played on foot, as opposed to horseback—compare “footrace”). You can find old references to “Rugby football” and “rugger” as well. Americans kept the term, Brits didn’t. http://www.todayifoundout.com/index.php/2010/06/the-origin-of-the-word-soccer/ http://www.todayifoundout.com/index.php/2010/06/the-origin-o...
      - jacinda 10y ago
        Exactly. And America isn't the only country that kept the abbreviation. Australia, Ireland, Canada and some countries in Southern Africa use the same term. http://deadspin.com/map-what-every-country-calls-the-thing-we-call-soccer-1491638984 http://deadspin.com/map-what-every-country-calls-the-thing-w...
      - gsnedders 10y ago
        Soccer and rugger survive in parts of upper-class society in the UK, FWIW.
    - thaumasiotes 10y ago
      > P.S English is not my language and I'm pretty sure I messed up with conditionals, hope you still get what I meant in the last sentence. Change "hasn't" to "hadn't" and your conditional is perfect. (If you're curious, you've made what I would consider a definiteness error three times -- I would say 'the ridiculous name "soccer"', 'discussing over beer', and 'the Poland - West Germany semi-final in 1974'. Those won't affect understanding and you give the impression of being very fluent; there are no other errors.)
  - aidenn0 10y ago
    For sports are head-to-head match-ups with scores, and have leagues with sufficient games per season, you can evaluate the amount that luck is involved by comparing the points-scored/points-allowed to win/loss over a season. Merely empirically determine the appropriate exponent to input for the pythagorean win percentage[1]. Baseball uses an exponent slightly under 2, while basketball uses an exponent slightly under 14. The higher the exponent, the less likely a poor team is to defeat a good team. In fact this can be generalized to sports without points if you have some other way of estimating a team/player's "quality" normalized to 1 being "average" which is all the points-scored/points-allowed ratio does. [edit] Note that the Jamesian (points-scored/points-allowed) quality metric doesn't work as well in games where draws are common (e.g. soccer/football), and different quality metrics will yield different exponents, so do not allow apples-to-apples comparisons. For what it's worth, applying James's quality metric to league football yields an exponent of about 1.3[2], which is in line with the intuition of ranking these 3 sports from most to least luck per win: Soccer, Baseball, Basketball 1: https://en.wikipedia.org/wiki/Pythagorean_expectation https://en.wikipedia.org/wiki/Pythagorean_expectation 2: http://www.soccermetrics.net/soccer-pythagorean-theory/why-the-baseball-pythagorean-doesnt-work-for-soccer http://www.soccermetrics.net/soccer-pythagorean-theory/why-t...
imh 10y ago
The p values don't necessarily imply what you think they do. The p value you end up with is a function of the modeling assumptions you make. For example, if your errors are correlated (which I'm sure they are in these swim meets since many data points can correspond to the same swimmer and many are from the same race), then you need to account for that. It's not uncommon to have a p value like that, then you account for something you forgot and it goes away.

by dzdt 10y ago

Alex3917 10y ago

Those aren't the stats that really matter though. At the end of the day, the question isn't whether the pool is fair or not, but whether the people watching the event on TV still go out and eat McDonalds or whatever afterwards. Most sports aren't really 'fair' per se. Making them fair wouldn't be cost effective, and wouldn't make for good TV. E.g. everyone knows that some lanes are faster than others in rowing due to the lack of wind screens around the venues. And that's not even to mention the people who spend ten years training only to get violently ill and unable to compete due to the water quality. The reality is that even at the highest levels of competition there is still a lot of luck.

0xdeadbeefbabe 10y ago
I'm annoyed by NBC's approach to streaming online where you have to login. I suppose they don't care about me, because I don't eat at McDonalds. Edit: Also you are right.
yodsanklai 10y ago
> The reality is that even at the highest levels of competition there is still a lot of luck. It really depends on the sport. For instance, football (soccer) is very random, and this is amplified by the tournament format. But in many other sports, there is no suspens at all. Actually, it would be interesting to somehow evaluate the random factor for each sport.
- pesfandiar 10y ago
  Even in sports with less variance in performance (e.g. weightlifting), we're only looking at one or very few data points. There are definitely random factors in the final outcome, but it may even add excitement to the events. That said, I still think fairness or at least perception of reasonable fairness is very important in high-level sports. No one is going to watch the games (or the ads as the GP pointed out) if they think the game's rigged.
- mamon 10y ago
  Some even argue that the randomness of football (which Americans know by riddiculous name "soccer") is the main factor that makes it the single most popular sport in the world. Some unlucky events end up being remembered for generations. For example people in Poland are still discussing over the beer whether we would have been world champions if only it hasn't been raining so heaviliy during Poland - West Germany semi-final in 1974. P.S English is not my language and I'm pretty sure I messed up with conditionals, hope you still get what I meant in the last sentence.
  - evincarofautumn 10y ago
    The name “soccer” is just an old British slang abbreviation of “Association Football”, to differentiate it from other kinds of football (that is, games played on foot, as opposed to horseback—compare “footrace”). You can find old references to “Rugby football” and “rugger” as well. Americans kept the term, Brits didn’t. http://www.todayifoundout.com/index.php/2010/06/the-origin-of-the-word-soccer/ http://www.todayifoundout.com/index.php/2010/06/the-origin-o...
    - jacinda 10y ago
      Exactly. And America isn't the only country that kept the abbreviation. Australia, Ireland, Canada and some countries in Southern Africa use the same term. http://deadspin.com/map-what-every-country-calls-the-thing-we-call-soccer-1491638984 http://deadspin.com/map-what-every-country-calls-the-thing-w...
    - gsnedders 10y ago
      Soccer and rugger survive in parts of upper-class society in the UK, FWIW.
  - thaumasiotes 10y ago
    > P.S English is not my language and I'm pretty sure I messed up with conditionals, hope you still get what I meant in the last sentence. Change "hasn't" to "hadn't" and your conditional is perfect. (If you're curious, you've made what I would consider a definiteness error three times -- I would say 'the ridiculous name "soccer"', 'discussing over beer', and 'the Poland - West Germany semi-final in 1974'. Those won't affect understanding and you give the impression of being very fluent; there are no other errors.)
- aidenn0 10y ago
  For sports are head-to-head match-ups with scores, and have leagues with sufficient games per season, you can evaluate the amount that luck is involved by comparing the points-scored/points-allowed to win/loss over a season. Merely empirically determine the appropriate exponent to input for the pythagorean win percentage[1]. Baseball uses an exponent slightly under 2, while basketball uses an exponent slightly under 14. The higher the exponent, the less likely a poor team is to defeat a good team. In fact this can be generalized to sports without points if you have some other way of estimating a team/player's "quality" normalized to 1 being "average" which is all the points-scored/points-allowed ratio does. [edit] Note that the Jamesian (points-scored/points-allowed) quality metric doesn't work as well in games where draws are common (e.g. soccer/football), and different quality metrics will yield different exponents, so do not allow apples-to-apples comparisons. For what it's worth, applying James's quality metric to league football yields an exponent of about 1.3[2], which is in line with the intuition of ranking these 3 sports from most to least luck per win: Soccer, Baseball, Basketball 1: https://en.wikipedia.org/wiki/Pythagorean_expectation https://en.wikipedia.org/wiki/Pythagorean_expectation 2: http://www.soccermetrics.net/soccer-pythagorean-theory/why-the-baseball-pythagorean-doesnt-work-for-soccer http://www.soccermetrics.net/soccer-pythagorean-theory/why-t...

imh 10y ago

The p values don't necessarily imply what you think they do. The p value you end up with is a function of the modeling assumptions you make. For example, if your errors are correlated (which I'm sure they are in these swim meets since many data points can correspond to the same swimmer and many are from the same race), then you need to account for that. It's not uncommon to have a p value like that, then you account for something you forgot and it goes away.