"Over Reyes' 9 years with the Mets the team has a total W-L record of 720-737 for a .489 winning percentage. In the 9 years when Jose Reyes has STARTED for the Mets they have a W-L record of 538-496 for a .520 winning percentage. This means that when Reyes does not start the Mets are 182-241 for a .430 winning percentage.
Thus the Mets win 9% more games with Reyes starting than without him starting. In a 162 game season 9% more wins turns a 70-92 team into an 84-79 team."
That's a GREAT statistic...
...or is it?
What is Ruben Tejada's Mets W-L record when he starts, versus when he does not start?
What is David Wright's Mets W-L record when he starts, versus when he does not start?
What is Angel Pagan's Mets W-L record when he starts, versus when he does not start?
What was Luis Castillo's Mets W-L record when he started, versus when he did not start?
There's no basis for comparison, here. Jose Reyes's numbers might be exactly the same, they might be markedly better, or they might actually be below other Mets players numbers. We have no way of knowing if this is really a valid statistic, unless we have something to compare it to, do we?
There was a statistic cited on a thread here, earlier this year, which shows the fallacy of citing statistics without any basis for comparison. The discussion began with the citation of the statistic of the Mets winning about 70% of the time when Reyes scored a run. Here are the relevant comparitive statistics, which were later cited in that discussion:
"Reyes: scored at least a run in 335 games, Met record of 237-98, .707 win%
Wright: scored at least a run in 318 games, Met record of 226-92, .711 win%
Beltran: scored at least a run in 319 games, Met record of 223-96, .699 win%
From 05-08, the Mets played .697 ball when Ramon Castro scored a run"
Catcher Ramon Castro has a strikingly similar Mets winning percentage when scoring a run, as Jose Reyes does. What does that tell you about that statistic? Maybe it's a little deceptive? Maybe it's not a terribly meaningful statistic?
By comparing Jose's Mets W-L percentage statistic to other Mets players, that will clearly show whether or not Reyes is as valuable to the team as he SEEMS to be, statistically speaking.
I was never too good in math, so please don't look to me to calculate those other players stats. Is there anyone here who has a program to calculate this? Is there some site on the web which would have this info? If we can't compare it to other Mets stats, then this statistic is just a number in a vacuum, in my opinion. It may actually show how valuable Reyes is to the Mets winning percentage...or, it may show that his value is about the same as Ruben Tejada's or Luis Castillio's, statistically speaking, just like Ramon Castro's run-scoring/Mets win % numbers were compared to Reyes's. Without that comparison, we just don't know.
Edited 10/15/11 by BigtimeFan