Tuesday, December 20, 2011
Projecting how Yu Darvish will do in Texas
By Richard Durrett
Don't we all love projections? Dan Szymborski of the Baseball Think Factory looked into some computer models on how Darvish would do in Texas the next five years (and what he might be worth). Here's a look:
Using his ZiPS projections, Szymborski has Darvish's numbers for 2011-2016:
2012: 13-7, 3.62 ERA with 46 walks, 169 strikeouts and a 4.5 WAR
2013: 13-7, 3.55 ERA with 44 walks, 167 strikeouts and a 4.7 WAR
2014: 13-7, 3.52 ERA with 42 walks, 163 strikeouts and a 4.6 WAR
2015: 12-6, 3.54 ERA with 39 walks, 153 strikeouts and a 4.3 WAR
2016: 11-6, 3.46 ERA with 38 walks, 150 strikeouts and a 4.3 WAR
As for what Darvish might be worth, here's part of Szymborski's column (you can read the entire version here as long as you have an ESPN Insider account):
If we consider that a win costs about $4.7 million on the open market and we assume 3 percent yearly growth, Darvish's performance over the next five years will run you roughly $100 million on the free-agent market. And while Darvish will likely cost more than that to Texas when you factor in the posting fee, he's unlikely to get a contract that large for himself since he can only negotiate with one team.
What makes this particular signing more risky than the average pitcher is simply the amount of uncertainty we have as to projecting Darvish going forward. Even with a dependable and long record of major league stats and full staffs of various trainers, coaches, scouts, and statheads, teams frequently whiff on guessing what a pitcher will do in the future.
Japan's professional league is a very high level of play, somewhere between Triple-A and the majors, but there's still a lot we don't know about how Japanese pitchers will fare in MLB. Going back through history, there are only 21 pitchers that had full-time jobs in Japan who went on to pitch 50 innings in the majors. That's the extent of our knowledge of how Japanese pitchers fare in the majors.
What can conclusions can we draw from 21 pitchers? Unfortunately, not a whole lot, given what a mixed bag the results have been. Some pitchers have been extraordinarily successful in the MLB, such as Hiroki Kuroda or Takashi Saito. Some have been flops, like notorious New York Yankees hurler Kei Igawa. Some are in the middle.
Using the simplest regression technique, linear regression, comes up with an r-squared of 0.08. What this means, in layman's terms, is that only 8 percent of the differences between ERAs in Japan explains the differences between ERAs in the U.S. The rest of the difference in performance are things that we can't really quantify with such a small sample size.
In terms of Darvish, it means that there is a sizeable amount of risk. While it's unlikely that Darvish would pitch as poorly in the U.S. as Igawa has, Daisuke Matsuzaka, generally believed to be a disappointment, is a better pitcher than Darvish's downside.
But the news isn't all gloomy. Even with the risk associated with signing a Japanese pitcher, the chances are very good that the Rangers will be getting a star pitcher for years to come. In Darvish's favor is that he's been more successful in Japan than any other migrated hurler. Despite only being 25 and pitching in NPB since age 19, Darvish has a career 1.99 ERA in Japan at ages when most pitchers are still learning their craft. Matsuzaka was a star in Japan as well, with the lowest Japanese ERA of any starter to come over, but even his 2.95 ERA in Japan pales in comparison.
That's, obviously, just one man's opinion. But I found it interesting and thought I'd pass it along. We'll file this away and see how close these projections are in a few years.