I have just finished my first round of data collection after watching 30 table tennis matches from the 2010 World Table Tennis Championship. The great news as that the Table Tennis players are PLAYING ACCORDING TO MINIMAX, or at least it seems! After observing approximately 1800 serves, it seems as though the winning probabilities for serving to the right and the left side are the same, exactly what minimax strategy predicts. The fact as to if the serve is from deuce court or ad court doesn't change the winning probabilities (which are about 53% for left serves, and 54% for right serves [from the server's direction]). What is interesting is that I observed that servers tend to serve more to the opposite side (server's left to returner's left, server's right to returners right) more often than to the same side. Table tennis strategy predicts that there is advantage to serving this way because the returner is limited in returning the serve if returning with the dominant hand. However, even table tennis strategy has to take a back seat to minimax strategy, because I was able to observe equal winning probabilities for each serving side.
However, I haven't done any confirmation statistics, such as P-values etc., so I can't declare the experiment a success just yet. I also have to observe every match and record data point-by-point. I have simply only done match summaries to this point because I wanted to see if the data would work. The point-by-point summaries of these matches might take quite a while, perhaps another three weeks! However, I know that this experiment is definitely heading in the right direction.
BTW, below is a picture of the Chinese Men's Table Tennis Team that won the 2010 WTTC. The players, from left to right, are Xu Xin, Zhang Jike, Ma Lin, Ma Long, and Wang Hao.
-- Jayanth (not Juanito6561 anymore)
No comments:
Post a Comment