Re(4): New Hall of Famer for Craig G re: Game Point winners Posted on February 11, 2018 at 07:36:00 PM by Craig G
does not seem like rocket science, and i have been a rocket scientist in my career
No disrespect intended, but it appears that the rocket scientist has overlooked the fact that there are elements of near symmetry involved.
You are looking at it from only the server side. What about the reverse?
server had to be at 5 or 6 points (spectacular 7 is assumption of course) receiving team could have 0,1,2,3,4,5,or 6 points
fine, but if R had won the point and game, we could reason...
receiver had to be at 5 or 6 points, serving post could have 2,3,4,5, or 6 points
the receiving team would only have had a chance to win if they had 5 or 6 points
true, but the flip side is...
the serving team would only have had a chance to win if they had 5 or 6 points
So we aren't making much headway with that line of thought. It isn't as simple as you depict it.
Which is why I was impressed that Bennett worked out 80% plus, presumably without doing any computer analysis.
tell me how this matters and how you use it to any advantage in betting?
Well, it's a curiosity, like retrograde analysis in chess.
But I think that to look at it as a scientist, once you have solid data, you then try to work out the theoretical implications.
As I said in another post, if nothing else, it illustrates how unlikely it is to receive your way to victory, especially in the worst posts. When we see a player winning from high posts in doubles, we strongly suspect that he is not struggling to win points when serving.
If in reality there is an ebb and flow to a player's ability to win doubles points on serve, it might work out that a 5-6-7-8 results index could reveal it to some extent.
If true, for someone like Tiger or anyone using a Skiena-based model, who use data mining techniques from the published results, that kind of correlation might be intriguing.