Friday, March 10, 2006

Searching for undervalued pitchers

The Reds need pitching. This won't come as a big surprise, as we lead the league last year in both runs scored and runs allowed. When a team is this unbalanced, the obvious thing to do is to deal from one's strength (in this case, our offense) and acquire some defense.

The "Moneyball" approach is to search for players who are undervalued. I've been learning more and more about defense-independent pitching statistics of late (see my stats quicksheet for links to learn more about this stuff). A major discovery stemming from this research over the past year is that fielding and simple dumb luck can have a big influence on the number of hits allowed by a pitcher after a ball is hit into play. In fact, there is very little correlation from year to year in most pitchers' BABIP – the opponent batting average on balls hit into play. In contrast, pitchers' strikeouts, walks, and home runs are all fairly consistent across years. What this means is that we can use strikeouts, walks, and home runs to predict pitcher performance, and treat hits allowed as essentially a factor due to chance.

So getting back to finding undervalued pitchers, one thing we might try to do is to find pitchers who suffered from very bad luck last year. It is possible that their teams will undervalue them due to their (apparent) poor performance, and we might be able to get a good pitcher for relatively cheap. To do this, I'm going to use FIP, which essentially is the ERA one would predict for a pitcher given his K's, BB's, and HR's. The mean FIP equals the mean ERA, but luck and team fielding will determine how closely each player's ERA matches their FIP.

Among pitchers who pitched at least 100 innings last year, here are the top-10 most unlucky players based on the difference between ERA and FIP.

Player Name, FIP, ERA-FIP
1. Hideo Nomo, 5.67, 1.57
2. Chan Ho Park, 4.33, 1.33
3. Sidney Ponson, 4.93, 1.28
4. Zach Greinke, 4.67, 1.13
5. Jose Lima, 5.89, 1.10
6. Mark Hendrickson, 4.81, 1.09
7. Joel Pineiro, 4.63, 0.99
8. Eric Milton, 5.62, 0.85
9. Seth McClung, 5.79, 0.80
10. Jeff Francis, 4.92, 0.76

The problem with this list is that even FIP predicts that most of these players will be pretty poor performers. We don’t need players who can put up a high 4's and 5's ERA's, we need players who can give us ERA's under 4.25 or so. Here's what happens if I filter out anyone with an FIP over 4.25.

Player Name, FIP, ERA-FIP
1. Cory Lidle, 3.90, 0.63
2. Jeremy Bonderman (det – 4.57 era), 4.08, 0.49
3. Glendon Rusch, 4.04, 0.48
4. David Wells, 4.01, 0.44
5. Odalis Perez, 4.17, 0.39
6. Matt Clement, 4.22, 0.35
7. Jake Westbrook (CLE – 4.49 era),
8. Brian Moehler, 4.20, 0.35
9. Jason Schmidt, 4.06, 0.34
10. Erik Bedard, 3.67, 0.33

Interesting to see Cory Lidle at the top of that list. Anyway, unfortunately we're not dealing with as much of a difference as I'd hoped we might see. A 0.4 undervalue of ERA might not be enough to drop a pitchers' perceived value significantly enough to help us make a steal. Nevertheless, the let's look at a few possibilities:

Erik Bedard – While he had a 4.0 ERA last year, his FIP was 3.67, best among the players on my list. Bedard had Tommy John surgery three years ago, and has had some non-elbow injury problems since then, so health might be a problem. He has decent k/9 numbers (7.95 career) and hr/9 numbers (0.74 career), though he is a bit more wild than I'd prefer (4.12 over his career, though he got it down to 3.62 last season). I don't know how high Baltimore is on the 27-year old lefthander, but he seems like a very good person to target as a pickup. Sure, he's still probably considered a prospect, but he's unproven and might come fairly cheap.

Jeremy Bonderman – Bonderman made good strides last year; even better than his 4.57 ERA indicates. He's a young guy who made the jump from single-A to the majors with Detroit in '03, and has been improving ever since. His strikeout numbers declined last season from 8.22 to 6.90, but his walk and HR numbers improved. He suffered from some bad luck, however, and might be had more cheaply if Detroit doesn't realize his performance was actually worth a 4.08 ERA. The only concern about him is that his HR allowed rates are a bit high (career 1.14, though he was down to 1.00 last year); otherwise, he seems like a stud.

Cory Lidle – Cory has pitched very well since we traded him to Philadelphia, even though his numbers last season didn't look so great. His BABIP last year was 0.316, which is horrendously high and is almost certainly bad luck. Cory doesn't strike a lot of guys out, but he walks very few and doesn't give up many home runs…even pitching in Philadelphia. Maybe he's worth going after?

Odalis Perez – Perez has a fairly strong big-league record and is highly touted, so he might not come cheap, even if his era was a little bit inflated last year. But it wouldn't hurt to at least inquire about him and see what they'd want for him. He's about the same age as Josh Hall, and he could be much better than he has been already. The only concern I'd have about him is that his hr/9 rate is rather high at 1.07, which might make for a bad pitcher*GABP interaction.

Matt Clement & Jake Westbrook – These guys just keep coming up, eh? Both players were discussed in a possible trade with Austin Kearns this past offseason, and both would seem to be good fits with our team. Clement is the high-k, low-hr type of pitcher that we all dream about, as he would probably fair well in Cincinnati. It's doubtful he'd come cheap though. Westbrook should be a less costly option, although he is a high groundball pitcher…which means fewer home runs, but also works best when there is a solid infield defense behind him. And the Reds can't offer him that.

The other guys on this list may be good pitchers, but they're all aging enough that they won't be effective for much longer. This was a fun exercise to run through, even if it might not practically have much of an implication. All I know is that Krivsky will have to start wheeling and dealing to improve this team. I understand not making many moves during spring training, as he needs to figure out what players he has right now. But I'd look for at least one big-time trade by the all-star break, and possibly more. There's no other way we can improve our club as it stands. -JinAZ

2 comments:

  1. "A major discovery stemming from this research over the past year is that fielding and simple dumb luck can have a big influence on the number of hits allowed by a pitcher after a ball is hit into play."

    Is this also a reason to favor hitters who don't strike out so much (unlike, well, the Reds)? That is, hitters who consistently put the ball in play are more likely to benefit from "simple dumb luck," too. Not that the Reds didn't score lots last year, but every little bit helps--especially when the pitching is so bad.

    ReplyDelete
  2. Absolutely true. I'm a fan of Dunn et al and I do think the to-do made over the strikeouts is a bit overdramatized. But strikeouts are the least productive kind of out that you can make, as any hit ball at least gives the chance that something good could happen. It's one of the reasons (along with the low OBA and poor defense) that I'm not wild about Wily Mo Pena.

    It's also the reason why strikeout pitchers would be very helpful to the Reds. If the pitcher gets the batters out himself, the Reds' fielding ineptitude doesn't come into play. :)
    -JinAZ

    ReplyDelete