Archive - clutchiness RSS Feed

Cameron isn’t clutch

November 28, 2007 in clutchiness, player evaluation, San Diego Padres with 4 Comments

5 reasons to keep Mike Cameron; 5 reasons to let him go - Good piece by the folks over at the the UT assessing whether or not to keep Mike Cameron. It’s nice to see an approach that looks at both the negatives and the positives of a prospective move.

I’ll take issue with the second reason to keep him, though:

2. He’s clutch. Since 2005, no major league center fielder can match Cameron’s batting average (.302) or slugging percentage (.554) with runners in scoring position.

When talking about clutchness, you can’t just pick certain stats to make a case (at least, imo). They mention his good numbers with runners in scoring position, which is certainly, in general, “clutch.” However, how clutch is a double with men on second and third in a 10 run ball game?

If you look at Cameron’s numbers in, say, late and close situations, they are the total opposite:

2007: .226/.337/.398
2006: 176/.324/.275
2005: .249/.314/.391

Yea, not so clutch anymore.

The best way, at least that I know of, of measuring a players clutchness is to look at their “clucthiness” measure over at fangraphs. That’s the case because it looks at how many wins a player adds due to clutch or timely hits (by leverage index) over their normal batting line (in OPS wins). So if there’s a “clutch hit” to be had in the 5th inning, and Cameron picks it up, he’ll get credit for it. At the same time, if he gets a single in the 9th with bases loaded in a 8 run game, he’s not going to get a lot of credit, if any, in his clutch score (and in that situation, the UT article is essentially saying it’s a clutch hit). So here is his clutch score since 05:

07: +.06
06: -.77
05: .09

Since 2002, he’s actually -4.02 wins in the clutchiness measure. If anything, he’s not clutch! With knowledge in hand that clutch hitter’s don’t really exist, I’d definitely hesitate to use clutch performance at all in deciding whether or not to bring back Cameron. But I certainly wouldn’t use it as an argument in his favor.

Cameron’s 20 most costly outs

November 21, 2007 in blogging, clutchiness, San Diego Padres with 2 Comments

Courtesy of the fantastic Fan Graphs (by WPA).

  • grounded out to third (-.141)
  • struck out swinging (-.141)
  • grounded into dp (-.132)
  • grounded into dp (-.118)
  • flied out to center (-.110)
  • struck out (-.110)
  • flied out to center (-.110)
  • stuck out looking (-.110)
  • struck out looking (-.109)
  • caught stealing (-.096)
  • fouled out to first (-.093)
  • flied out to center (-.092)
  • flied out to center (-.092)
  • flied out to center (-.092)
  • flied out to first (-.091)
  • grounded into dp (-.084)
  • struck out swinging (-.082)
  • fouled out to first (-.081)
  • struck out swinging (-.078)
  • caught stealing (-.076)

Strike out variety: -.52
In play: -1.236
On the bases: -.172

Total (these 20 outs): -1.928
Remaining negative plays:  -10.192

This is more of Tim Wakefield fastball, get-me-over fastball type post (more on that later). I’m not sure there’s much here to work with, as it is more just for fun.  First off, you’re working with the WPA (and it’s assumptions), which uses historical averages, rather than the exact, specific situations (i.e., Livan Hernandez on the mound and Adrian Gonzalez at 3b). So, rather than actual win probability added (or subtracted in this case), it’s more like “generic” WPA, based on historical averages. On that note, a costly out may be followed by a home run, and for the most part, be erased. Finally, selecting the bottom 20 was just an arbitrary cutoff and holds no real significance whatsoever.

Anyway, the next time someone says, “those Cammy k’s were killer in the clutch,” you can come back with “those fly outs to center didn’t help either!”

******

Just a few random, blog related notes. I can’t use firefox at school, but I’m using it now. I notice some things that I don’t with IE, like the graphs extending into the sidebar. If you ever notice anything with any browser, please let me know and I’ll check it out.

Also, I don’t have a lot of time now, so I may not post much until later next week (I know, seriously, like maybe two people really care).  I’m home now for the week and then I have a huge project due thursday of next week. I’m interested in looking into Cameron’s k’s more in depth and of course my exclusive, cutting edge, critically reviewed top 30 prospect list is going to be making its annual appearance ; ) I may wait on running that until late next week to get a little more research in and not rush it. No need to really, with the whole winter in front of us. So expect posting to be a little lighter than usual over the next few days. Then again, unexpected time may present itself and I may make a flurry of posts … kind of reminds me of this:

Shoeless Joe Jackson: The first two were high and tight, so where do you think the next one’s gonna be?
Archie Graham: Well, either low and away, or in my ear.
Shoeless Joe Jackson: He’s not gonna wanna load the bases, so look low and away.
Archie Graham: Right.
Shoeless Joe Jackson: But watch out for in your ear.

More importantly, have a great Thanksgiving!

A garbage time special: Kevin Cameron

October 14, 2007 in clutchiness, San Diego Padres with Comments Off

Before making my post on the bullpen, I was poking around through baseball reference. This was the first time I noticed the innings/score appearance matrix (check at the bottom of a player’s game logs page … here is Cameron’s). Now you can easily see when a reliever entered the game (i.e., the run differential and inning). Cameron entered 4 times with a tie. He never once entered the game with a 1 or 2 run lead. 27 of his 48 appearances came when the Padres were either ahead or behind by 3 runs.

Looking at fangraphs’ leverage index, Cameron was used in the least important situations of a Padre with at least 15 innings pitched (.60 LI). We already knew that he wasn’t used in many clutch spots and this just drives home the point.

Now, if you didn’t know anything about Kevin Cameron you would probably expect that he had a high ERA last year. Somebody has to eat up these innings, right? That’s not the case. His era was 2.79 and below the Padre bullpen average of 3.06 (actually higher if you take out Cameron). So why wasn’t he used in big spots? Well, obviously, Bud Black, management, etc. didn’t think he was truly a 2.80 era pitcher. If there’s one thing we know about relievers, it’s not to put too much emphasis on one year of performance. Also, when attempting to assess their performance, we shouldn’t put too much emphasis on plain old earned run average.

I’ll leave it up to Kevin Towers and co. to determine Cameron’s true talent level going forward. We know that Towers does a great job of picking up underrated relievers on the cheap. What does he do when he may have an overrated one on his club?

Is a trade looming? As always, we’ll have to wait and see.

Another look at Buddy’s bullpen

July 29, 2007 in clutchiness, Sabermetrics, San Diego Padres with Comments Off

I’ll have a post up on all of the recent roster moves soon, but I wanted to check in on the pen again. The last time we looked was July 6th, and here’s how the pen stacked up by leverage index back then (with their LI from June 13th in parenthesis):

Hoffman: 2.33 (2.39)
Linebrink: 1.66 (1.58)
Bell: 1.42 (1.35)
Meredith: 1.37 (1.43)
Hampson: .87 (.80)
Cameron: .65 (.82)
Brocail: .57 (.66)

Now, it’s been almost a month, so let’s see if anything interesting pops up.

Hoffman: 2.33
Linebrink: 1.53
Bell: 1.44
Meredith: 1.17
Ring: .86
Hampson: .83
Brocail: .62
Cameron: .54
Thompson: .52

So there were really no huge changes over the past 20 something days, as we’d probably expect. It looks liked some of Linebrink’s crucial spots went elsewhere. Meredith has been a case study in the effect that balls in play can have on a relievers yearly numbers. Last year his BABiP was .203, while this year it sits at .362. He went from the extremely lucky side of the spectrum to the extremely unlucky side in one year. His peripherals are really quite similar to last year. Check it out:

k rate (percentage of total batter faced)
06: 20%
07: 16%

bb rate
06: 3%
07: 5%

hr rate
06: 1.6%
07: 1.8%

gb rate (percentage of balls in play)
06: 69%
07: 71%

He’s been relatively similar pitcher with his peripherals,  yet his ERA is about 3 runs higher. There is a lot of luck and randomness involved in both last year and this year (and you’d expect more of that in smaller 50 inning samples). Meredith’s true talent probably, almost definitely,  lies somewhere between this year and last. He’s a perfect example of why relievers are so tough to predict.

I always feel compelled to talk about Cameron in these posts. Make no mistake, he’s been pretty darn good. He has a great k rate, he keeps the ball down, and he’s yet to allow a homer. He also has 19 unintentional walks in 156 total batters faced (12%). That’s just too high. At some point, I think you do have to start letting pitch some more critical innings, but I generally like the way Black has handled him so far. Cameron isn’t close to a 1 era pitcher when he steps out onto the mound, and Bud knows it. And he doesn’t treat him like one.

Now, who takes Linebrink’s innings? Afterall, he was the second most leveraged reliever on the club. Well, I guess the better question may be … if you could create a pecking order in the pen, by leverage index, what would it be? Here’s mine:

Hoffman
Bell
Meredith
Thatcher
Cameron
Hampson
Brocail

We’ll wait and see with Hensely. The list isn’t concrete or anything. Of course you’d play match-ups and worry about rest and what have you. It also doesn’t mean I think Bell should be the “8th inning guy”. I think, optimally, he should be used in the second most important situations, overall. Sometimes, that may come in the 6th, 7th or 9th inning…

The clutchness of Gonzalez

July 25, 2007 in clutchiness, Sabermetrics, San Diego Padres with Comments Off

Adrian Gonzalez has been more valuable this year than last. If I told you that, you wouldn’t believe me, right?

This year he’s hit .261/.335/.470 and last year, in the same ballpark, he hit .301/.362/.500. There is no way he could be more valuable this year, right? Well, wrong, at least kinda. This year his WPA (win probability added, calculated by fangraphs) is 2.04. Last year it was a much lower 1.02. You might wonder how the hell that happened. It’s all about clutch performance. Gonzalez’s “clutchiness” this year is 1.31 (4th in all of baseball!) which is a complete turn around from the -.64 that he put up last year. Before we go any further, let’s look at how these things are calculated.

WPA is just win advancement minus loss advancement. A player gets credited for positive events and debited for negative ones. Say, a 2 run single to in the game in the 9th is worth .4 wins. The player is credited with .4 win advancements. An out in the first inning trailing 5-0 is worth -.02 loss advancements. The player is “credited” with -.2. Add all of these up at the end of the year to get WPA. Just as a note, I made those numbers up.

Clutchiness is basically how many wins the player added over what would be expected. The formula is WPA – (OPS Wins * pLI). OPS Wins is .025 * (1.7 * obp + slg – 1) * PA. The numbers are also park adjusted. pLI is leverage index per plate appearance. Usually for hitters it’s around 1.

Here are Adrian’s PA’s sorted by leverage index. This is what he has done in his 5 most crucial PA’s:

Double, fly out, double, double, sac fly. Pretty clutch, ey? In those 5 PA’s his WPA is .795. Last year in his 5 most critical PA’s he singled, walked, struck out, singled, and GIDP for a total WPA a .292 … a pretty large difference when we’re just talking about 5 times at the plate. Anyway, this is just cherry picking, as you obviously look at all PA to determine this stuff.

The merits of WPA (and “clutchiness”) are something that certainly could be discussed, as could the implications of this on any evaluation of Gonzalez. I don’t really think he’s been “better” this year. He’s just leveraged his performance so that, in a sense, he has provided more value, by WPA’s standards, to the Padres. Going forward, you’d probably regress his clutch performance a long way toward average and evaluate him like that. Afterall, how does a player go from -.82 clutchiness from 04-06 to +1.31 clutchiness in 07. There’s just a lot of random flucuation involved.

Page 2 of 3«123»