Replacement Level Yankees Weblog

Disclaimer: If you think this is the official website of the New York Yankees, you're an idiot. Go away.

by Larry Mahnken

Good things happen when you hit the ball in the air

I'd like to apologize for my behavior tonight on the Game Chatter, allowing my superstitions to get the better of me, and trying that "reverse jinx" stuff all game. I'll stop that now.

Today was a fun day. Not only did the Yankees win and Red Sox lose, but Godzilla got three more hits. The past five games Matsui has been on fire, and while it's only five games, a tiny, tiny sample size, it's been one hell of a five games. Godzilla's batted .706/.762/1.353--a 2.115 OPS! I really think he's turning into the slugger the Yankees thought they were getting. I actually picked up on this after Thursday's game, when I wrote:

Godzilla had a great game today, four hits, three doubles and a homer. I think that he's going to go off on a tear, something like .706/.762/1.353, but that's just a guess.

I got lucky I guess.

Of course, I didn't really predict anything, but I did point out that Matsui might have turned it around, and have been repeating that belief all weekend. When you're a former 50 HR hitter, and you're hitting everything on the ground, there's definitely something wrong with your mechanics, not necessarily your talent. I hope he's really turned it around.

Earlier, the Yankees released Juan Acevedo's sorry ass. There was a great deal of discussion the past couple of days about Acevedo over on Primer. The point of contention was Acevedo's quality, and whether he should really have been in that game Saturday in that situation. Acevedo's "supporters" pointed out that statistically, Acevedo has been an average or above average pitcher in the past, and well above average the past few years. They further pointed out that 23 innings is a very small sample size, and not enough to draw a conclusion about his pitching ability going forward this season. They do have a point here, but I still don't think he should have been put on the mound in that situation.

For those that don't have much understanding of statistics...well, I probably won't be much help, because I don't understand them completely, either. To the best of my understanding, here's the relevant idea: A player has a definite level of talent, and if you took an infinite number of at bats or innings, you'll get a true, accurate picture of their talent from the statistics. Since you can't take an infinite number of anything, you have to take a smaller number--a sample. The smaller the size of the sample taken, the less accurate the statistics are, the larger the sample, the more accurate. For example, as you flip a coin an infinite number of times, it is likely land on heads 50% of the time, and on tails 50% of the time. But if you flip the coin ten times, it can land heads ten times--but the likelihood of it landing on tails is still 50%, and as you approach infinity, the rate at which it lands heads or tails will approach 50/50. That is the basis for regression to the mean. (Thanks to eric, who does not own a television, for correcting me).

So, what this means is that Acevedo's true talent is far better than how the statistics show he's pitched this season. He will pitch better, and if given the ball enough times, he will pitch at the level he has pitched in past seasons, which is above average. Nobody knows when he'll do it, but eventually he is bound to. Looking at the problem in this manner would indicate that putting Acevedo in the game was not a bad idea, and also a better idea than putting in a pitcher who has performed well this season but not as well in the long term.

But I don't think that in a real world situation it's a good idea. Juan Acevedo is not a coin, he's a human being. There may be some specific reason that he has performed poorly that we cannot perceive, we cannot simply assume that it has to be a result of small sample size. If there's nothing wrong with him, or if he corrects the problem, he will regress to the mean, I won't dispute that. But he should demonstrate that he is capable of pitching well again before he is placed in a high-leverage situation when there are other acceptable options. Should he have been released? He refused an assignment to Columbus, and I don't think they should have kept him in the bullpen, so I think yes, they should have released him.

I think regression to the mean should be taken heed of much more when looking at a player who has performed at well above their established level, and understand that they're likely to regress to their previous level--I'm looking at you Chris Hammond. There are many more reasons for a player to suddenly decline than there are for a player to suddenly improve, and as Branch Rickey said, it's better to give up on a player one year too soon than one year too late.

The highlight of my day was when I got an unexpected email from Keith Law, a consultant on baseball operations for Toronto GM J.P. Riccardi, responding to a comment I made on Primer. I took the opportunity to ask the former Baseball Prospectus writer and Riccardi's right hand man what he thought about Alfonso Soriano, who has continued to excel (well not the past month and a half, but you know, sample size) despite lousy plate discipline. Is he a Freak of Nature, as Aaron Gleeman said, or what? He was kind enough to reply:

I figured the same thing you did - that his performance would slip given enough reps. But I do think that he's a freak of nature (Aaron's not the first to say that, I believe JP was quoted as saying it last year in a Gammons column), kind of like Vlad was early in his career. What I do know about Soriano is that he has tremendous plate coverage - there are few spots where it's safe to pitch to him, because he has both great bat speed (allowing late decisions) and long arms (allowing him to successfully reach many pitches out of the strike zone). IOW, you can get away with swinging at pitches out of the zone if you can hit them consistently, but rare is the animal who can do so.

Did that make sense? Re-reading it now, I'm not sure that it does.

It makes sense to me, but the one big difference between Soriano and Vlad is that Vlad never struck out as much as Soriano. It is an inspiring thought, though, that one of the smartest guys in baseball thinks that Soriano will remain productive despite his poor approach. I'd still like to see him be more selective, though.

Speaking of Soriano, Jay Jaffe commented about the Soriano Wager Watch, which tracks some wagers for lunches I made based on Soriano not producing as well as last year: "Larry, if you win that bet and the Yanks miss the playoffs because Sori didn't produce, I'm gonna find a new way for you to take lunch."

Gulp. Frankly, I want Soriano to bat .299 with 39 homers and a BB Rate of higher than .062. If he's batting .299 with 39 HRs and one non-BB PA away from dropping below .062, and the Yanks need a homer from him to make the playoffs, I'll gladly spend $30 on lunches. If the Yanks miss and I win, those meals are going to taste very, very bitter.

--posted at 10:58 PM by Larry Mahnken / |

Larry Mahnken and SG's
	Replacement Level Yankees Weblog
"Hey, it's free!"