Re(1): Who do you think is going to Win the Dania Doubles Championship? Posted on February 21, 2024 at 02:59:13 PM by Craig G
Stats are cool, and Mo's presentation is beyond amazing.
However, for me what is more important than stats are the ideas and context behind them.
I view the stats as similar to setting up a position on the chessboard. A hack will look at it and see very little. A master, much, much more. A grandmaster will be off the charts.
Of course Jai-alai analysis isn't chess, but it can be more complex than simply taking things at face value.
Clueless: Mr ITM, can you tell me which record is better, 37: 11-6-3 or 37: 3-6-11?
Mr ITM: Not a problem, they are dead equal.
I would bet that if you took the 4 most skilled players from any American fronton of seasons past, they would all have records with a descending WPS pattern. The team record of Urreisti-Oyhenard is a quintessential example.
I would also bet that as a predictor, a metric that addressed the paramount importance of winning would do much better than ITM. Eg: campeonato points per game.
(FWIW, computing show % (finishing third) might even produce a negative correlation with winning.)
An example of context would be the realization that the first couple of weeks of this meet would be maximum discombobulation for the players. Big Dave discussed what they would be going thru when they first arrived, so no need to reiterate. But the question arises, how important are the results from that period? Wouldn't we rather see how the players are doing once they are fully acclimated?
If we disregard the first 10 perfs, we get:
B-A 21: 4-4-3
U-O 33: 9-6-3
Another consideration is that with these sample sizes, there is no guarantee that the post positions were equally distributed, or that the average strength of their immediate neighbors were approx equal. So factors like recency, post position, strength of field and immediate neighbors add to the fun.
