Better pitchers leave more runners on base.
Last week, I reviewed a stat we produce at THT called LOB% (or percent of baserunners left on base). It’s a stat that will help you understand how individual pitchers did what they did last year, and how they might do next year.
In the article, I noted that LOB% is a repeatable stat to a degree—to about the same degree as home run rates (home runs per game). I should have also mentioned that it’s relatively easy to figure out which pitchers will maintain high LOB% rates. Good ones.
The reason is simple: baserunners accumulate. If you allow only a couple of baserunners per game, chances are very good your LOB% will be 100%. However, as you allow more runners on base, chances are better that they will get on base in the same inning, making it more likely they will score. As you allow more runners on base, your LOB% will fall at an ever-faster rate. So good pitchers—who allow fewer baserunners—will have better LOB% rates.
In fact, you can estimate a pitcher’s expected LOB% by using our xFIP stat (expected FIP), which is a measure of the basic skill components of pitching: strikeout, walk and fly ball rates. Here’s a graph of all pitchers who appeared in at least 100 innings last year, with their xFIP on the horizontal axis and their LOB% on the vertical one:
I inverted the xFIP axis to show you that, in general, a pitcher’s LOB% increases when he’s a better pitcher. The graph may look like a “scatterplot” to you, but there is a relationship between the two variables, as the fitted line shows. The “R squared” of these two variables is .20, which means that 20% of the difference in LOB% between pitchers can be explained by their difference in xFIP. The other 80% is probably due to things like “clutch pitching” or pure randomness.
For you math geeks, the equation to calculate “expected LOB%” is 86%-(.033*xFIP).
This graph allows you to more easily identify the true outliers, several of whom I’ve labeled. The more a pitcher deviates from the line, the more likely it is he will regress to the line in 2006. Note the two recent free agent signings, Jarrod Washburn (due for a big correction downward) and A.J. Burnett (upward correction in the forecast).
What Javier Vazquez’s home run rate is likely to be.
When Javier Vazquez was traded to the White Sox a couple of weeks ago, I participated in a South Side Sox discussion about Vazquez’s home run rate. Home run rate is the key factor for predicting Vazquez’s future performance with the Sox, particularly considering that U.S. Cellular is a home run park.
Here are my comments:
FWIW, here are Vazquez’s numbers for HR-given-up-per-outfield-fly over the last four years:
The trend is up, but he’ll almost certainly come down from that 2005 number.
Chase Field increases HR/OF rates about 3%. The Joan (my pet name for U.S. Cellular) increases rates about 14%.
Let’s say, based on the above data, that Vazquez is a home run pitcher, and his “innate” level is around 13% (two points higher than the league average). That indicates his rate will be 15% next year, which is pretty high.
Last year he gave up 1.5 HR per game. At 15%, that would decrease to 1.25. The net impact is an ERA decrease of 0.1. Not as much as you might imagine.
At first, I liked this deal for the Sox a lot. But Young is a great prospect and, unless there is a terrific market for Garland or Contreras (which there might be), I’m not so sure now. I’ll wait and see what happens next.
The key question is whether home run rates (per outfield fly) are totally random from year to year, or whether there is some pitcher tendency that shows up over time. I tend to think that specific pitcher tendencies crop up over four years’ data (though you still have to regress them to the mean a lot), and I hope to dig into this later in the offseason.
For more about home run (and other) rates, I highly recommend the article by J.C. Bradbury and David Gassko in the THT 2006 Annual.
Why Not Will?
I’m a little late to the Hall of Fame debates. Ballots have been submitted, winners will be announced January 10. Cases have been made for Bert Blyleven, Albert Belle, Jim Rice, Alan Trammell, Andre Dawson, Bruce Sutter and others. But I wonder, why not Will?
Will Clark is eligible for the Hall of Fame for the first time this year, and he has a solid case. Michael Schell, in his book Baseball’s All-Time Best Sluggers ranks him the 14th-best hitting firstbaseman of all time. Plus, Clark was an excellent fielder and a clutch hitter (batting .293/.358/.486 with no one on and .314/.420/.538 with runners in scoring position). Bill James ranks Clark the 14th-best first baseman of all time in his Historical Abstract.
His 1989 season, when he batted .333/.407/.546, was the best season of the 1980s, according to Win Shares. He racked up 44 Win Shares that year, which prompted Bill James to write a special column about it in his Win Shares book. In addition to piling up some great stats, Clark did all the “other” things very well that year. He didn’t hit into many double plays, he stole some bases, hit in the clutch and played a stellar first base. Teammate Kevin Mitchell, with 47 home runs and 38 Win Shares, won the Most Valuable Player award.
Altogether, Clark amassed 331 Win Shares in his 15-year career, the 114th highest total in baseball history. (There are 258 players in the Hall.) I won’t go into all the details of Clark’s Win Shares, because this column already did a pretty good job.
Clark was almost certainly the best first baseman in the majors from 1988 to 1991, sort of post-Mattingly/Murray and pre-Thomas/Bagwell, and he wasn’t too shabby in many of his other years. Personally, I would vote for Clark ahead of his college teammate, Rafael Palmeiro (regardless of the steroid controversy).
Still, Clark is a borderline call at this point. He didn’t dominate at his position for a long enough time, and he didn’t win an MVP award. First base is a deep position; If you vote for Clark, you should strongly consider McGriff too. More importantly, you’ve got surefire Hall of Famers McGwire (er, except for the steroid thing), Thomas and Bagwell coming up in the next five or six years, which means Clark’s chances will probably fade with time.
If I had a ballot? Blyleven, Trammell and Gossage.
Kim Allen never got the chance he deserved.
While I’m calling out another blog, let me refer you to this story about Kim Allen, a former short-term major leaguer. Now, I have to admit that I don’t know if this story is true or not, but the background statistics check out. And it’s a great story about a talented player who should have gotten more of a chance. It reveals as much about baseball’s past as anything else.
Johnny Damon is a very rich man.
Johnny Damon signed a big contract with the Yankees last week, for $52 million over four years. That’s $13 million a year. This contract has generated lots of discussion on the web, so I’ll keep my comments brief.
The Yankees overpaid, but not by a lot.
Both the Bill James Handbook and Ron Shandler’s Baseball Forecaster project that Damon will hit around .295/.360/.430 next year and Shandler gives him a Reliability Score of 87, which is dang reliable (100 is max). He’s a pretty good center fielder with a relatively weak arm; David Gassko’s Range system puts him 12 runs saved above average. Based on those numbers, he projects to have about 22 Win Shares next year, 11 above a bench player.
Given Damon’s career, 22 is about right. As you can see from the sparkline, Damon’s Win Shares the past two years have been among the highest of his career:
Based on what free agents were paid last year, that would make him worth $12-13 million. And that’s for just the first year of the contract, when he’ll be 32. My guess is that his performance will decline faster than the inflation rate for free agents increases.
Damon’s agent Scott Boras tried to convince major league executives that Damon is among the best leadoff batters of all time. U.S. Patriot took a look at his claim in a recent blog entry.
So is Kevin Millwood.
Kevin Millwood reportedly just signed a four-year deal with Texas, including an option year that vests with playing time, for $11 million to $12 million a year. Several commentators have drawn comparisons to the Chan Ho Park contract (five years, $13 million a year starting in 2002), or they have assumed that the Rangers overpaid for Millwood’s “spike” year in 2005. Personally, I don’t think either is the case.
Millwood did have a flukey good year in 2005, with an ERA of 2.86 vs. a FIP of 3.77. But that FIP was seventh in the league, and it’s comparable to his 2004 FIP of 3.82. A pitcher who can consistently produce a FIP between 3.50 and 4.00 is worth a lot of money these days.
Ron Shandler’s expected ERA for Millwood has been between 3.43 and 3.82 each of the past four years. For 2006, Shandler projects a 3.64 ERA in 203 innings for Millwoood, a performance that would have been worth $18 million last year, according to my Net Win Shares Value calculation.
If he continues to perform at that level for the next four years and the market for starting pitchers remains robust, this will be a fine deal for the Rangers. Of course, it’s a huge risk. All of these pitcher contracts are. But one of the keys is that the fifth year only vests if Millwood pitches enough innings in the previous years. Compare that to the Burnett contract, which the Blue Jays guaranteed for five years, and it includes a clause that lets Burnett walk after three years.
It’s interesting to me how often people assume major league executives don’t know what they’re doing. But a recent study by J.C. Bradbury (link is to a PDF document) has shown that, since 1985, the best pitching contracts have gone to free agent pitchers with the best strikeout and walk ratios and not necessarily the best ERAs. In other words, major league execs typically understand what’s important. Kevin Millwood’s deal is no exception. We’ll talk about Jarrod Washburn’s deal another day.
Julio Franco signed for two more years of baseball.
This isn’t news anymore, but Julio Franco signed a two-year deal with the Mets for over $1 million a year. He’ll be 49 years old when the contract ends. I mean, can you believe he actually got a two-year deal? How crazy is that? I’ll bet Giants’ GM Brian Sabean is jealous.
The Cubs are emphasizing groundball hitters for next year.
The Chicago Cubs are trying hard to upgrade their team this year. They signed Ryan Dempster, Scott Eyre and Bobby Howry to relatively rich contracts in an effort to bolster the bullpen. They weren’t able to sign Rafael Furcal, but they did trade for Juan Pierre and signed Jacque Jones. And they’re talking with the Orioles about a trade involving Mark Prior and Miguel Tejada.
The Pierre and Jones acquisitions were interesting. I expect Pierre to bounce back from his 2005 season, though he will probably never have another season as good as 2004. Jones should only play against right-handers, regardless of how much he’s paid.
I’m guessing that you don’t think of these two guys as similar offensive players. After all, Jones hits home runs and strikes out a lot. Pierre hits singles and steals bases.
But they are similar in an important way. They’ve been among the most prolific groundball hitters of the past four years. Last year, for instance, 83% happy, 9% disgusted, 6% fearful and 2% angry. Must have been an Astros fan.
Something I never wonder about.
A recent poll found that over 50% of recently-married Japanese women fart in front of their husbands within the first year of marriage. I’m guessing the figure is much higher for the men.
Just to reprove my nerdiness, check out this addictive spatial relations game called Planarity. Try not to get discouraged too quickly.
Happy New Year, everybody. I’m glad I’m not on this list.
References & Resources
A nice review of Will Clark’s career is available at the Baseball Library.