Tennis scoring differs from basketball scoring in two important ways. First, in tennis, points are grouped into games (and games into sets) and the object is to win games, not points. If this were the only difference, then it would be analogous to the difference between a popular vote and the electoral college in US Presidential elections.
The other difference is that in basketball the team with the highest score at the (pre-determined) end of the game wins, whereas in tennis winning a game requires a pre-specified number of points and you must win by two. The important difference here is that in tennis you know which are the decisive points whereas in basketball all points are perfect substitutes.
Then to assess statistically whether tennis’ unique scoring system favors the stronger or weaker player (relative to a cumulative system like basketball) we could do the following. Count the total number of points won by each player in decisive and non-decisive points separately (perhaps dividing the sample first according to who is serving.) First ask whether the score differential is different for these two scenarios. One would guess that it is and that the stronger player has a larger advantage in the decisive points. (According to my theory, the reason is that the stronger player can spend less effort on the non-decisive points and still be competitive, thus reserving more effort for the decisive points.) Call this difference-in-differential the decisiveness effect.
Then compare matches pitting two equal-strength players with matches pitting a stronger player against a weaker player. Ask whether the decisiveness effect is larger when the players are unequally matched. If so, then that would suggest that grouped scoring accentuates the advantage of the stronger player.
10 comments
Comments feed for this article
September 8, 2009 at 12:54 pm
moses
and here’s the paper:
Click to access Paserman_Tennis_February2008.pdf
September 8, 2009 at 1:39 pm
jeff
beautiful. thanks. (the paper has a different question in mind but the data are there.)
September 8, 2009 at 7:10 pm
Reader X
How does the analogy with the popular vs. electoral vote go? Is there any possible scenario where the winner of the tennis match is also the player who won less points? Because this seems to be the most crucial feature of the electoral system.
September 8, 2009 at 9:33 pm
jeff
absolutely it can happen and no doubt it has happened many times. the winner of the match could win his games by narrow margins while the loser of the match wins his games by large margins so that the loser of the match has more total points than the winner.
September 9, 2009 at 8:20 am
Shrik
Not just points – it’s even possible that the winner of the match has won less GAMES than the loser.
September 8, 2009 at 10:33 pm
ryan
what if we used the natural experiment offered by golf? my girlfriend and i were discussing this issue today, and realized that stroke play in golf, where every shot matters equally, is closely analogous to the basketball game, while match play, where you only count holes won, not total strokes, is closer to tennis.
obviously there’s issues with the fact that stroke play isn’t explicitly a mano-a-mano (or teamo-a-teamo) competition, and you lose some of the zero-sum nature of tennis points or basketball possessions. also, the holes in match play are all pretty much worth the same, but you do know which ones are decisive.
September 8, 2009 at 10:39 pm
ryan
for what it’s worth, tiger woods has generally been a great match play player, with the most match wins of anyone in the accenture match play championship. his record at the ryder cup is a bit mixed though.
September 8, 2009 at 10:53 pm
Twitted by playwright000
[…] This post was Twitted by playwright000 […]
September 9, 2009 at 8:19 am
Shrik
This is super. Sound scientific backing for the way (I think) Roger Federer is playing thus far in the US Open.
I always felt that this was what he does (and Rafa too) in matches against much weaker opponents – they rarely annihilate them. Instead seem to conserve their energy by turning up the heat only during key points.
Brilliant. Getting objective validation of one’s subjective instincts is always a great feeling 🙂
October 30, 2009 at 10:37 pm
Etl World News | How to improve basketball
[…] Here is Jeff Ely on related issues. […]