http://blog.rogersbroadcasting.com/mikewilner/
about whether the Managerial Changes made last June 20 had an impact. This will not answer this to the satisfaction of Mr. Wilner, as his explanation is that the changes are not due to Cito, Gino, et al, but simply due to the fact that the hitters finally got there act together. Fair enough. There is no way to test that assertion statistically, unless you were prepared to look at the hitting situational statistics to see if the batters were being more aggressive and swinging earlier in the count. One thing I can say, which is not in these slides is that bases-on-balls were down by more that one per game under Cito (p-value of 0.0005), without changing the OBP significantly, which suggests to me that the hitters were being more aggressive.
Enjoy!
There is a typo in the above slide - it should be p = 0.0499 less than alpha = 5%. When I figure out how to replace the slide, I will do so. (I'm new to this blogging stuff.)
Great blog! Pleasure to read.
ReplyDeleteI am a firm believer that the Jays are a better team with Cito than they were with Gibbons and some of these statistics that you have calculated prove it. Statistics aside, managing a ball club isn’t just about penciling in names on a line-up card and throwing the players onto the field. I think the best example in baseball of just how important a manager is to a team is the New York Yankees. A team stacked with all-star caliber players (A-Rod, Jeter, Cano, Damon, Matsui, Posada, Rivera, Wang etc.), yet have not been able to get it together in the Joe Girardi era. Torre knew how to handle all the egos of the all-stars in his ballclub, Girardi has yet to do so. Now you look at what Torre has done in Los Angeles, he’s able to control the biggest personality in baseball in Manny Ramirez and make that team a legitimate threat in the National League.
I strongly disagree with Mike Wilner. I believe the Jays are a better team with Cito and that he was the difference the Jays took it up a notch offensively in the latter part of the 2008 season. I believe they recalled Adam Lind the day after Cito was hired, who had been demoted earlier in the season due to a slow start. For the rest of the 2008 season, Lind put up great numbers offensively after starting the season off terrible under Gibbons, and continues to put up big numbers so far into the 2009 campaign. If I’m not mistaken, the Jays as a team early into the 2009 campaign are in the top 5 of almost every offensive category and lead the MLB in Batting Average, Runs Scored, Hits, Doubles, Runs Batted In. I don’t believe this is just a coincidence. Cito and Tenace have great knowledge on hitting and what approaches to take at the plate, and have done a great job in bringing out the potential of this Jays line up.
Yes, the jays are well up there in most offensive categories. I waited until the first 20 games before posting my first statistical analysis because very little can be said based on that small sample of games. As the season wears on, and more "mean reversion" - the tendency to see less extreme averages - works through the system, I'll start posting more statistical analyses as time permits.
ReplyDeleteHey, seems like you have a good concept for your blog here and I'm going to add it to my roll.
ReplyDeleteMy question about this post comes from a different angle. First, let me say that I do NOT subscribe to the "the switch made no difference" premise. I get what Wilner is saying and both agree and disagree with elements of it.
But the premise is more complex than that. Wilner defends Gibbons vs Gaston on IN GAME management, for the most part, while openly admitting that Gaston is "the best manager in the game between the end of this game and the start of the next one" - I think he's right on both those things, but I am convinced (where maybe he's not) that Gaston's strength trumps Gibby's strength in a real way and not in a coincidental way.
Not, with that set-up in mind - here's my question.
Can you statistically determine - and did you factor into your analysis - the following premise:
that the Jays hitting under Cito is their "real" ability and that what they did under Gibby in 2008 was "bad luck"
in other words, if I understand your thesis, you were testing the probability that the Bad Hitting Gibby-Jays could have improved to the level of the Cito-Jays through chance or because of the influence of the new manager. But whichever the cause, you were looking at the probability of improvement.
What I'm trying to say is what is the probability that a good hitting team would somehow hit poorly for a third of the season for no other reason than bad luck?
Am I making sense? is such a question even testable?
Anyway, I'm always a fan of someone looking at actual factual data to reach conclusions instead of subjective emotional impressions. Keep up the good work.
WillRain
The problem with observational analysis is that it is impossible to determine exactly cause and effect. As I mentioned in an exchange with Mike on his blog, another explanation of the 2008 Jays hitting phenomenon was the disruptive influence that Gary Denbo's approach had on the hitters. I didn't post this as part of the analysis, but under Gibby/Denbo, the Jays had about 1.1 more walks/game than under Cito/Gino had a t-stat of 3.379 and a p-value of 0.00047, enormously significant.
ReplyDeleteOnly 5 Jays had more than 450 PA in 2008: Rios, Overbay, Scutero, Wells and Rolen. Overbay and Rolen were playing through injuries that had an impact on their swing: Overbay was still getting back on track from the hand injury and Rolen was nursing a bad shoulder. For these two, their numbers look about the same before and after the managerial change.
Consider the other three:
Wells:
Career upto 07 281/331/478 OPS 809
Denbo/Gibbons 277/325/436 OPS 761
Gaston/Tenace 316/355/537 OPS 892
Rios:
Career upto 07 288/338/453 OPS 791
Denbo/Gibbons 285/336/411 OPS 746
Gaston/Tenace 306/343/555 OPS 897
Scutaro:
Career upto 07 259/320/384 OPS 704
Denbo/Gibbons 248/343/317 OPS 660
Gaston/Tenace 276/338/376 OPS 714
All three were performing below their career numbers under Gibbons/Denbo. All improved markedly under Gaston/Tenace, to above their career numbers.
To me the numbers do not lie. SOMETHING was happening that was affecting the hitters' performance in the first half. It may have been the manager; it may have been the hitting coach. It IS possible that it was bad luck, and Mike Wilner suggests, but the stats suggest that the probability of it being due to luck or chance is very small.
By the way, I agree 100% with your third paragraph.