I expect the Sox to stay in the race most of the year, barring injuries, and they have a chance to win it all if their rotation matches its potential.
High fives all around! Can I call ‘em or what? OK, even the recently retired Al Leiter hit the ball sometimes, and I know I’ll probably never call one like that again. So it is with a mix of arrogance and humility that I review this year’s White Sox in order to answer the only question that really matters: Can they do it again?
1. Which players played over their heads last year?
Pretty obviously, no one had a breakout year at bat, but things were different on the mound. A simple way to gauge pitching effectiveness is to compare a team’s actual ERA to its Fielding Independent Pitching (FIP), which focuses on the most essential components of a pitcher’s talent (strikeouts, walks and home runs). In general, a team’s ERA will resemble its FIP, and major differences between the two will decrease over time.
The White Sox had an ERA of 3.61 but a FIP of 4.11. The difference (0.50) was the second-largest in the majors last year, behind only the Cardinals (0.63). The White Sox were one of the best fielding teams too, which accounts for some of that difference. But only some. And if you look at the difference between the team’s starters and relievers, you see that fortune truly smiled on the bullpen:
IP ERA FIP xFIP FIP-ERA xFIP-ERA Starters 1084 3.74 4.18 4.13 -0.44 -0.39 Relievers 387 3.25 3.93 4.06 -0.68 -0.81
xFIP is just like FIP, except that it “normalizes” the average number of home runs allowed per outfield fly. You can skip all the stats if you want and just take away one thing: the Sox’s bullpen was very “lucky” last year (though a couple of other bullpens had even larger xFIP variances. Watch out, Indians!). Here are some key bullpen stats for each Sox reliever:
WXRL LEV IP ERA FIP DER HR/F xFIP LOB% Dustin Hermanson 3.9 2.13 57.3 2.04 3.74 .757 6% 4.23 80% Cliff Politte 3.8 1.35 67.3 2.00 3.77 .799 8% 4.13 91% Neal Cotts 2.0 1.10 60.3 1.94 2.98 .763 2% 3.67 79% Bobby Jenks 1.4 1.79 39.3 2.75 2.71 .687 10% 2.54 76% Damaso Marte 1.2 1.74 44.7 3.77 4.48 .661 10% 4.25 81% Shingo Takatsu 0.8 1.42 28.7 5.97 6.57 .712 24% 4.15 81% Luis Vizcaino 0.0 1.18 68.0 3.73 4.63 .704 11% 4.34 80%
WXRL is a measure of each reliever’s Win Probability Added and is reported by our friends at Baseball Prospectus. LEV stands for Leverage, a measure of how important each reliever’s innings were (same source). LOB% represents the percentage of baserunners left on base, and relievers typically have higher LOB%’s than average because they don’t always start at the beginning of an inning.
As you can (hopefully) see, the Sox’s three top relievers had very advantageous contextual stats.
- Dustin Hermanson‘s DER was .757 (league average was .696), only 6% of his outfield flies were home runs (league average was 11%) and 80% of his baserunners allowed didn’t score (league average was 71%).
- Cliff Politte was even luckier in DER (.799) and LOB% (91%!).
- Neal Cotts experienced the same trends, and only 2% of his flyballs passed the outfield fence.
After the season ended, Kenny Williams traded away some of his bullpen depth (Luis Vizcaino to the Diamondbacks, Damaso Marte to the Pirates), and the news from spring training hasn’t been positive. Hermanson’s back is hurting (prompting rumors of retirement) and Bobby Jenks has reportedly lost some of the zip from his fastball.
In last year’s White Sox preview, I talked about their “luck” batting with runners in scoring position (RISP) in 2004 (.291 vs. .268 overall) and made a not-very-impressive prediction that they wouldn’t do as well in 2005. Lo and behold, they batted only .259 with RISP. So let’s make a similarly unimpressive prediction right now: the bullpen won’t be as good this year. In fact, it may be no better than average.
2. Why did they trade for Jim Thome?
Because they had to. The Sox’s offense wasn’t only not lucky last year. It wasn’t terribly good. They scored only 4.6 runs a game, slightly lower than average (4.8), despite the fact that they play in a batter-friendly park. When Frank Thomas was healthy, the Sox offense was fine. When he wasn’t, it wasn’t. Letting Thomas go was obviously the right thing to do, but so was finding another big bat to replace him.
His health is the issue, of course. Nate Silver’s PECOTA system estimates that Thome has a 38% chance of attrition (major loss of playing time) and a 26% chance of having a breakout season (much better than expected) in 2006. That is the most extreme top/bottom split of any established major league slugger (unless you think Michael Tucker is an established major league slugger), so the Sox can’t really be sure what they’ll get from Thome in 2005.
Other than Thomas, the Sox were lucky in the injury department last year, too. They need to remain lucky with Thome this year; I think 30 home runs is a good benchmark for the guy. It would indicate that he’s adapted to playing in the American League again and he’s stayed healthy. Of course, Sox fans should hope for more.
Lots of knowledgable White Sox fans are Aaron Rowand fans, but he only hit .270/.308/.407 last year. Baseball Prospectus and Ron Shandler both predict better stats from Brian Anderson. Anderson almost certainly won’t match Rowand in the field but the overall impact of inserting him in centerfield should be relatively small.
3. Was the Vazquez deal really necessary?
Kenny Williams also improved the starting rotation in the offseason, picking up Javier Vazquez from the Diamondbacks. I initially liked this deal for the Sox, but that’s because I didn’t realize how good Chris Young is. He’s one of the ten best prospects in baseball and he’ll probably roam Arizona’s center field for many years, starting soon. Trading away both Rowand and Young seems risky to me, and the Sox didn’t need Vazquez that badly. They have a fine young starter in Brandon McCarthy.
If McCarthy can help stabilize the bullpen this year, or if one of their starters loses it, this deal may help the Sox. But they paid a high price. At some point during the year, I expect Williams will make a deal to fortify the bullpen. Maybe he thinks that his rotation depth will give him more flexibility to make the right deal. Time will tell.
4. Won’t Ozzie take care of things?
Two years ago, I wondered what kind of manager Ozzie Guillen would be. I think we can now say that Ozzie’s biggest strength as an in-game manager is his effective use of his pitchers. He tends to stick with his starters, though it helps that most of them are efficient with their pitch counts. He’s also an imaginative and flexible bullpen manager, as he showed by moving the closer role from Takatsu to Hermanson to Jenks when necessary. He’s taking a good calculated risk with Boone Logan this spring. But Ozzie isn’t a miracle worker.
The 2006 versions of the Hardball Times Annual, Bill James Handbook and Baseball Prospectus all agree that the Chicago White Sox were statistically “lucky” last year. In particular, the Sox were 35-19 in one-run games, about six more than you’d expect based on their runs scored and allowed totals, and they were about seven games better than expected overall. This is the pythagorean variance baseball analysts talk about.
Virtually every championship team is “lucky” in some way. That doesn’t take anything away from their accomplishment. But if we want to understand how they might do this year, we need to understand what went their way last year. In general, strong bullpens help teams win close games, but that’s not always the case. For instance, Boston was 37-27 in close games last year, but that was due more to David Ortiz’s bat than anyone in their bullpen. I wondered how much Chicago’s bullpen helped them win the close ones.
To figure that out, we analyzed all close games (won by one or two runs) last year and determined the Win Probability Added that could be attributed to each team’s bullpens from those games. Bullpens tend to have positive WPA in close games as a matter of course, about .05 WPA points per game. So we calculated the average amount of WPA accumulated by each team per close game and compared them to each other. By this measure, the Sox had the fourth-highest contribution rate in the majors last year.
Team Relief WPA Games Per Game MIN 8.37 77 0.11 SDN 8.92 82 0.11 WAS 8.40 87 0.10 CHA 9.11 95 0.10 ANA 8.17 87 0.09 ARI 6.21 76 0.08 SLN 6.63 83 0.08 CLE 6.40 90 0.07 FLO 4.55 66 0.07 OAK 4.86 72 0.07 HOU 4.91 78 0.06 TOR 4.41 73 0.06 SFN 4.72 82 0.06 CHN 3.99 76 0.05 DET 3.99 83 0.05 SEA 3.69 80 0.05 PHI 3.06 69 0.04 CIN 2.49 65 0.04 LAN 2.94 84 0.04 NYA 1.95 64 0.03 NYN 2.17 73 0.03 TBA 1.71 79 0.02 COL 1.63 86 0.02 PIT 1.20 74 0.02 TEX 1.21 80 0.02 KCA 1.06 76 0.01 MIL 1.11 82 0.01 BAL 0.43 64 0.01 BOS -0.10 63 0.00 ATL -2.61 70 -0.04
I plan to write a full article about these findings soon. But the bottom line is that the Sox’s bullpen was a significant contributor to their record in close games last year, and their bullpen is not going to be as good this year. Expect a worse record in close games, regardless of Ozzie’s bullpen mastery.
5. So? Can they do it again?
Of course they can. The offense should be better, the starting rotation could be just as good. Their fielding may decline a bit, but not a lot. Their problems will most likely be in the bullpen and (not coincidentally) matching their great record in close games.
But no team is strong enough to run away with this division. The Indians appear to be awesome, but they had some great years from some of their young players, and their pitching is questionable. I expect the Twins to be in the thick of things again; their pitching is too strong to be denied. The Tigers could surprise. The White Sox should win 85-90 games and stay in the pennant race until the end. Injuries could hurt them. Conversely, a breakout year by one or two players (Uribe? Crede?) could help them surge to the top.
Yet I have a feeling that management decisions will be most critical this year. Can Kenny Williams find some solutions to the bullpen problems, or can Ozzie manage his way through them? If injuries hit, where will the two of them turn?
Oops. I just answered a question with a question, didn’t I?
References & Resources
Special thanks to John Walsh for his WPA programming.