YEEK! Looks like half of my pitching staff will double-sucking through a Siebe Gorman Savox!
For the purposes of this post, consider ‘risky’ to be a “greater chance than average that they have a significant drop in their skills and/or miss over a 1/3 of the season.” So I’m not going to cherry pick ‘lucky’ 2009 starters like J.A. Happ whose ERA was significantly lower thanks to unsustainable luck in terms of batted balls finding fielder gloves and fly balls not finding the mitts of spectators.
#4– Joba Chamberlain
MLB Pitches 2009-2010: 1,711 -> 2,733 (+1,022)
2009 was first year > 2,500+ Pitches: Yes
Slider %: 22%
I know….Joba has more warning signs than a cigarette pack – 1.55 WHIP last year, declining fastball speed, starter vs. reliever status, his mom, his surname-inherited guilt for appeasing Hitler’s pre-WWII Central Europe land grabs.
As a nominal Yankee fan, I’d send him to the bullpen anyway. But the fact that his fastball was crushed last year (-21 wFB) while his slider was solid (+7.5) is just one more reason to do it.
Pass on him as a starter. Pick him up on waivers if he shows promise again as a set-up guy.
#20 – Josh Johnson
MLB Pitches 2009-2010: 1,412 -> 3,284 (+1,872)
2009 was first year > 2,500+ Pitches: No
Slider %: 25%
Like Tommy Hanson, Josh Johnson is a pitcher I really like, would draft, and sounds like an actor on the CW (b/w him and fellow Marlin John Vander Wal, their pitching staff reads like a Dawson’s Creek reunion). He’s got a great fastball (95.1 MPH). He’s got a decent changeup but just doesn’t use it that much.
But with that slider rate, I’m a little more hesitant to draft him than I otherwise would be. I know he’s built like a truck but so was Kerry Wood. Discount him a little bit and don’t cry to me in June if your Tommy Hanson and Josh Johnson-led staff has some injury troubles.
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
1. Danny Posted: March 10, 2010 at 10:21 PM (#3476893)Well, kind of fair point but not completely fair. Here was the preceding analysis: http://razzball.com/predicting-risky-pitchers-take-deux/
BTW, Verducci's success criteria of "no injuries, ERA doesn't go up" means that about 3/4 of pitchers don't succeed year-over-year. So he is good at predicting - it's just he's predicting the equivalent of a sunny day in San Diego based on that criteria...
Armando Galaragga: 178.2/120, 143.2/82
Ricky Nolasco: 212.2/123, 185/84
Gavin Floyd: 206.1/119, 193/114
Brett Myers: 190/96, 70.2/88
Ryan Dempster: 206.2/155, 200/123
Andy Sonnanstine: 193.1/101,99.2/67
Jonathan Sanchez: 158/88, 163.1/103
Todd Wellemeyer: 191.2/115, 122.1/70
Dana Eveland: 168/95, 44/61
Johnny Cueto: 174/92, 171.1/97
Zack Greinke: 202.1/126, 229.1/205 (oops)
Ervin Santana: 219/127, 139.2/90
Jesse Litsch: 178/118, 9/48
John Lester: 210.1/144, 203.1/138
Mike Pelfrey: 200.2/113, 184.1/82
AJ Burnett: 221.1/104, 207/106
Matt Garza: 184.2/119, 203/114
Javier Vazquez: 208.1/98, 219.1/143
Ted Lilly: 204.2/112, 177/145
Scott Baker: 172.1/121, 200/100
Better than I thought at first glance.
I see big improvements from Greinke and Vasquez, and decent improvement from Sanchez. Lester, Burnett, Garza, Lilly, Cueto, Floyd and Baker were about a push.
The rest of them fell off. So only half the guys on the list had serious deprovement.
In the preceding article, I mea culpa'd that my picks/analysis last season resulted in a wash - http://razzball.com/predicting-risky-pitchers-take-deux/.
Part of that was reading too much into breaking ball % alone. If I had focused on 2+ of the criteria, would've swapped out guys like Vazquez and Lilly for Volquez and Parra.
The 'success' rate is tough on something like this. I based the analysis on 2004-2008. Those that fit none of the three criteria dropoff about 20% of the time (Webb last year was an example) while 1+ dropped off 30%, 2+ dropped off 34% and, all three dropped off 41%. So if you can guess 50% out of 20, you're doing a great job (which I did not do last time).
It is pretty hard to fall from the bottom of a canyon.
EDIT: I mean, I guess the guy means Rick Vanden Hurk or something... but seriously, if you're going to make a comment like that on a guy's name, get the right guy. Geez.
First, the increase in pitches: this penalizes all guys who were rookie starters in 2009. Sure, minor-league pitches are presumably easier but suggesting that Brett Anderson going from "0" to 2800 is not necessarily a bigger risk than somebody who went from 2300 to 2800 if Anderson threw about 2300 pitches last year.
Second, the increase also "penalizes" previously injured starters. That might be fine in terms of predicting injury risk but it could clearly be a specious correlation. Josh Johnson had a big jump because he was hurt for at least half of 2008 (and nearly all of 2007). 2009 happened to be his first healthy year in a long while. He's not necessarily an injury risk because of the increase in the number of pitches ... he's probably an injury risk because he had such a serious injury that he missed 1.5 seasons. Guys with past injuries are, generally speaking, larger injury risks (although this, too, needs some controls added).
The same can pretty much be said for the "first season over 2500" criterion as well.
Of the guys on your list, Anderson, Wells, Niemann, Scherzer, Romero, and Hanson were rookie 2009 starters. At least Wainwright, Carpenter and Johnson were hurt for a good chunk of 2008 (and 2007 for the latter two).
Essentially, every rookie starter will tick at least one of your boxes for the next year and, if he's reasonably full-time, probably at least 2. But it's not clear they should. Those injured the previous year will tick at least one box -- they probably are injury risks but probably not due to an increase in workload from the previous year. Now, it could be that injured starters brought back into heavy work are a bigger injury risk than those who aren't worked so hard but since you don't control for injury, you can't tell from the piece.
Anyway, the question is to what extent have you picked up on anything besides "heavily used young starters are risky" (this has been known for ages) and "previously injured players are more likely to be injured again and/or fall off a cliff."
Also I think you want to compare their 2009 performance/FIP (or similar) to their projected 2009 performance or 2008 FIP rather than their 2008 actual performance. (Maybe the study did, but not the table earlier in the article.) Managers always ride the hot hand in starting pitching so having a fluke good season will result in lots of innings -- i.e. "overperformance" and "overuse" are related. Plus, Ricky Romero had a 101 ERA+ last year but is projected to an 86 this year (ZiPS), so it would hardly be a shock if he turned in a 75. That would be regression (and maybe some bad luck), not insight about inning increases and sliders. (If he pulls up lame after 20 innings, that's a different story.)
Also of the 88 pitchers who threw at least 162 IP in 2008, only 53 did so in 2009. A couple of those guys retired (Mussina and Rogers I noticed), but (based on that small sample) you expect maybe 35% of the guys who threw 162 IP in 2009 (and there were only 78 of them) not to repeat that feat.
Pitchers get hurt or decline in talent or get old or get unlucky or stop getting lucky all the damn time.
Ha! Great catch. Fixed it and left in the crossout for posterity's sake
Walt -
Great points. You're right that these criteria are particularly critical of 2nd year pitchers and previously injured pitchers. I imagine if one went with those two criteria - removing the same pitching seasons (pitchers > 37, FIP > 5.00, removed retirees) I did - that you'd index above-average in finding pitchers more likely than average to break down. I did look into 'Age' as a variable and there's no consistency there for < 38. The only variable I found that reduces the likelihood that one of these pitchers' risk is % of sliders thrown.
All my analysis has been based on actual xFIP vs. projected. I assumed that got rid of most/all of the flukiness that ERA/ERA+ would have.
I went lower than 162 IP for 'dropoff' for the analysis. I went with 2500+ pitches in Year 1 and < 2000 pitches (~130-140 IP) in Year 2 - removing retirees, pitchers < 38 and FIP > 5.00 in Year 1. For 2008-2009, I have 15 of 65 pitchers (23%) falling below 2,000 pitches.
So the analysis acknowledges that a little more than 1/4 (27%) of pitching seasons following relatively full, successful seasons (2500+ pitch, < 38, non-retired, < 5.00 FIP) will fall below 2,000 pitches or have a +0.75 FIP change YOY. The analysis focuses on the drivers that help target pitchers at greater risk to fall above this norm.
So pitchers do get hurt or decline in talent every year but some are more likely than others...
You must be Registered and Logged In to post comments.
<< Back to main