premium blogad:
blogads:
advertisement:
sitemeter:
|
|
|
May 5, 2008
How Inconsistent Is Oliver Perez?
[Editor’s note: From time-to-time, the MetsGeek editors may come across a particularly poignant or informative user journal. When that happens, the journal will be highlighted via a promotion to the site’s main page. Today, we’d like to present this journal written by user Danny. Enjoy!] The enigmatic Oliver Perez. We assume this to be true, no? The most popular criticism of Oliver Perez last year is that he was woefully inconsistent. I think this is pretty common among young pitchers lacking the experience to fight through the days where they don’t have it. The Mets have another young starter named John Maine who does not face the same level of criticism in terms of his consistency. I happen to think that both players are roughly equivalent, but my opinion seems to be in the minority. Now, I want to make it clear that this exercise is not intended to denigrate John Maine in any way. I love the dode. But he is the most logical candidate to compare to Perez to obtain some measure of perspective on how inconsistent Oliver Perez really is because they are roughly at the same developmental slope in their careers. I think the best metric for this little exercise is Game Score. Game Score is a metric devised by Bill James to determine the strength of a pitcher in any particular baseball game. Here’s how it is calculated.
My plan is to take the game scores from each start in 2007 and distribute them among buckets, and see if Ollie’s beta is really higher than that of Maine’s. I will pick some arbitrary buckets and apply each game score to it, and let’s see what we get. A quality start is considered any game score above 50. Oliver Perez - 29 starts - Average Game Score - 53.9 - Median Game Score - 57 >70 - 5 games (79, 72, 71, 72, 72) 69-60 - 7 games (69, 62, 64, 63, 67, 60, 64) 59-50 - 7 games - (57, 54, 59, 58, 55, 52, 55) 49-40 - 3 games - (40, 43, 40) <39 - 7 games - (39, 33, 32, 28, 39, 38, 27) John Maine - 32 starts - Average Game Score - 53.5 - Median Game Score - 59.5 >70 - 6 games (79, 76, 74, 70, 72, 89) 69-60 - 10 games (68, 64, 62, 60, 61, 68, 65, 66, 65, 64) 59-50 - 3 games (59, 50, 50) 49-40 - 5 games (44, 48, 42, 45, 45) <39 - 8 games (32, 38, 30, 24, 31, 38, 20, 13) They are essentially the same pitcher. The distribution of scores looks pretty similar to me. If anything, Maine is slightly more volatile. Both guys had 19 qualiy starts by the Game Score measure. Maine had a higher percentage of very high-end and very low-end games, but not significantly. They are both young pitchers still trying to find their way. Let’s not be so judgmental with the bumps in the road that each will face. Really, their development is paramount to the Mets success this year. Edit: Jessica wrote a particularly informative piece on Oliver Perez and all of the unearned runs that he allowed last year. Please take a gander before referencing those unearned runs and how it is applied in Game Score. 25 Responses to “How Inconsistent Is Oliver Perez?”Leave a ReplyYou must be logged in to post a comment. |
|
It’s funny. You don’t remember Maine disintegrating that much but I guess it’s true. Maybe, for some reason, his disintegrations aren’t as spectacular? Or is it this (and maybe some research can find out for us):
When Maine has a bad game, is it spread out over the whole of, say, 4-5 innings? Because it seems that when Ollie has a bad game, he’ll throw 4 scoreless and then implode in the 5th (or something like that). Maybe that’s what makes it more memorable.
Maine actually has 3 Game Scores worse than Ollie’s worst effort, which surprised me.
Ha, i like your tag-line. i thought it said bunnies at first, i was like HEY! lol
Very nice analysis Danny.
very clever and smaht of you to choose game scores which takes unearned runs into account so that’s covered already. very smaht.
nice work. Interesting it came out this close. I had Ollie’s back too but I didnt think it would turn it out as close as it did.
You can do some standard deviations on the 2 also to see who’s variance is higher and all that. 32 points isnt great, but its enough to do it!
THANK YOU
now make a journal about how zod’s step dad looks like steve trachsel in that turtleneck.
Dep, here are the standard deviations for each (I think I calculated this right):
Maine: 18.35
Perez: 14.69
Koo koo, thanks danny. that makes sense since Maine has that 13,20 and 24 and both maine n ollie have the same high-end type scores. those low scores require the larger distribution that the higher SD indicates.
The StDevs I got were 14.96 for Ollie and 18.64 for Maine.
I was close!
Oliver Perez has a sick sigma
DSL
Billy Wagner still says Ollie sucks ;-)
Great piece Danny
Good peice, it leaves food for thought.
user Danny got promoted
LOL!
Dannys a user? If you need help im here for ya buddy. Maybe we can do a Geek Intervention
I will go first. Your one of us, guy. When your a user, we all suffer. Look at poor Wilbur. He changed teams again. Ben doesn’t come here anymore. He says the User on the site keeps using and such. As for me, we love ya Danny and do not want to see you be a user.
FYI, good article
Very nice work, Danny. In fact, one might call it downright Geeky!
I wonder what it would look like to see these scores graphed against time? Maybe Ollie seems more volatile because his weaker starts are interspersed among the great ones? That would explain the impression that “you never know which Ollie you’re going to get.”
But last season, Maine was terrific heading into the All-Star Break and weakened dramatically as the season progressed. So we remember “All-Star Maine” and tend to repress August/September Maine.
I’m guessing that Perez’s line for 2007 looks spiky and Maine’s is a declining trend. It doesn’t change the outcome of Danny’s great analysis, but it might explain their very different reputations.
Interesting idea, Athena. Another approach would be to run a regression, with game score for the xth start as the response and game score for the x-1th start as the predictor. I’m going to be spending too much time in Minitab this weekend anyway, might as well have some fun with it.
That would be better, Jessica. And much more informative. But since I haven’t used Minitab in a thousand years (and forgot that it existed), I didn’t think of it. ;-)
Oh. And good luck with your exams. I hope you kick ass. Happy Graduation.
Athena, Jessica’s smarter than me, so let’s have her do the fancy work!
Haha, Danny. It think it’s fair to say that Jessica is smarter than all of us. Possibly combined. Sigh.
Besides she’s the baby of the family. (If we don’t count Izzy. And I don’t.) Let’s haze the rookie!
I like both pitchers, but I probably like Maine’s stuff better. Perez used to be a hard thrower, but he seems to be trending towards more of a finesse pitcher as he doesn’t typically throw harder than 90-91. He needs to hit his spots. When his control isn’t spot on, he struggles. I think his looming free agency is affecting him. I’d like to see the Mets sign him now while he’s sucking.
Maine’s fastball lets him be a lot more aggressive. In his last 2 starts he has consistently hit 95-97. Personally, I think Maine has one of the best fastballs in baseball. Not only is it hard, but it moves a ton and it moves in a very unique way that makes it uniquely hard to hit. His fastballs tail sharply up and to the right in a way which batters just aren’t accustomed to. Even though they know it’s coming, they can’t hit it, because their brains and muscles aren’t wired to hit a ball that moves like that. The typical rule of keeping the ball down does not apply to him…he’s most effective when he keeps the ball up. He’s gotten very good at throwing his fastball so it hits the upper right quadrant of the strike zone, and guys just flail at it. When he learns to throw his fastball at the hips of lefty batters that tail back over the inner half (upper left quadrant), he will get tons of taken strikes. Lately he’s been sporting a decent 85 mph change up which has me quite excited. That pitch, when combined with his 95 mph fastball and decent curve, makes him really tough to hit. With that changeup, I think he can become a legitimate # 2, and if he further masters his control of his fastball can be a legitimate ace.
Because Perez throws softer and relies more on control, I think his ceiling is lower. He needs Santana or Pedro like control to become a consistent #1 or 2.
Without checking, it seems that variance for performance from Ollie is greater from game to game but greater from Maine from month to Month. That is Ollie may have a great game followed by a horrible one followed by a great game followed by a horrible one. Maine (last year at least) might have 6 good to excellent starts followed by 6 mediocre to poor starts. I guess neither temporal scale is better than the other. I’ll take consistently good!
They are definitely similar pitchers: flyball guys with good K rates and questionable control. But I like Maine slightly better.
So far this year, Ollie is allowing 5.18 runs per game, while Maine is allowing 3.46. That is not a trivial difference. Maine has stranded 78.6% of his runners, while Ollie has stranded 69.1%. It’s a small sample size, but Maine is pretty easily outperforming Ollie so far in 2008. Neither guy has had mindblowing K/BB rates, but Maine is doing a better job of minimizing the damage from his mistakes.
Of course, Maine faded down the stretch last year, so it remains to be seen if he can keep it up. But even with that second-half fade, Maine’s xFIP was nearly a half a run better than Ollie’s in 2007, and I get the feeling that he’s slightly pulling away from the flaky left-hander.