Toying Around

Some More Projections

Tuesday's column over on ESPN.com on the Favorite Toy produced quite a reader response about omitted players, such as this note:

I enjoyed your ESPN column today--it is fun to speculate about these things.
One thing I noticed was the absence of the player who, prior to 2000, usually led all of these discussions--Ken Griffey. Has his 2001 season knocked him out of contention, or were you just focusing on other players? I'm not surprised that Alex Rodriguez has surpassed him, but I'm somewhat surprised that folks like Derek Jeter and Vladimir Guerrero have done so already.
On a more whimsical note, I liked your comment that A-Rod has already had a career's worth of numbers. Indeed, his numbers surpass the career marks of just about every 1970s shortstop (the era when I began watching the game), including Mark Belanger, Bucky Dent, Rick Burleson, Duane Kuiper, Bill Russell, Freddie Patek, Chris Speier and such, all of whom played 12-18 years.
Thanks,
Charles Abell

Griffey was the most popular name in my inbox, and with good reason. The mass media has spent the last two years telling us that Griffey was a lock to break Hank Aaron's home run record, just in time for Junior to enter a decline phase that now has some writing off his days as a superstar. Both reactions are too extreme, but Griffey has undoubtedly damaged his chances to make history. Here's a run through a few players who, like The Kid, didn't make the cut for the ESPN piece.

Ken Griffey, Jr.: Griffey was indeed on a pace to shatter Hank Aaron's vulnerable career home-run mark until this year, although his drop to 40 last year didn't help matters. Griffey is still a better-than-even (68%) bet to reach 600, but just a 13% bet to reach 700 and has just a 1% probability of passing Aaron. Next season will tell us a lot about his chances in this department; a 60-homer campaign would increase his shot at Aaron to 12%.
More interesting is the damage done to Griffey's RBI projection. He's currently below zero for Aaron's RBI record, and has just a 17% chance to reach 2,000. Again, one good year could turn these numbers around, because Griffey has strong career totals at a young age, but one more bad year could doom him entirely.
Juan Gonzalez: Gonzalez, like Griffey, has an awful recent season that skews his percentages downward. Unlike Griffey, however, Gonzalez has other holes in his game that make analysts question his staying power, including a terrible walk rate, a career spent largely in favorable hitting environments (good hitters' ballparks, behind high-OBP guys), and a series of nagging injuries. Even with his resurgence this year, Gonzalez has just a 24% probability to reach 600 homers and an 11% chance to reach 3,000 hits--both of which probably presuppose the existence of the Tampa Bay Devil Rays in their current close-to-the-record-veterans-acquiring form.
Gonzalez's best shot to do something newsworthy in any category is probably in the RBI department; the Toy gives him a 4% chance to break Aaron's record, but if he signs somewhere this offseason that puts him behind more high-OBP hitters, that estimate would be conservative.
Mike Piazza: Piazza got a bit of a late start, debuting at age 24 even though he'd crushed the high minors for two years. So while he's a lock to pass Johnny Bench's 386 homers, the most by any player who was primarily a catcher, the Toy gives Piazza just a 34% chance to reach 500 homers and a 5% chance to reach 600, with no chance for any other major milestone. A move to first base may or may not be the right move for the Mets, but for Piazza, it could give him the longevity he'd need to reach, say, 3,000 hits.
Jeff Bagwell: Bagwell was 22 when he first tasted the good coffee, but wasn't much of a "production" hitter until 1994, when he topped 20 homers for the first time en route to a total of 39 when the players struck. He does have twin 11% probabilities to reach 2,000 runs scored and 2,000 RBI, and a 3% chance to reach 3,000 hits. His recent home-run production gives him a 42% chance at 500 homers, but the Toy rates his chance of passing Willie Mays (660) at an even zero. Of course, the Toy doesn't know that Enron Field could help Bagwell keep his production totals up even as his age advances.
The big surprise in Bagwell's line is that the Toy gives him no chance at even 2,000 walks, surprising for a guy with 100+ walks in each of the last six years and nearly 1,100 to his credit on his career.
The Pitchers: The Favorite Toy isn't a great tool for predicting pitcher stats. The two cumulative numbers of interest are wins and strikeouts; wins are non-deterministic, with wild fluctuations from year to year, and strikeouts aren't that much better. (All those who predicted in 1997 that Randy Johnson would challenge Nolan Ryan's strikeout record are now dismissed.) But looking at what the Toy has to say about those two records might give us more insight at the aggregate level than the individual level--that is, how likely is it that we'll see another 300-win pitcher, or that we'll see a 5,000-K pitcher?
Greg Maddux and Roger Clemens are close enough to 300 wins that each gets a strong mark from the Toy in that category, but the slow pace of no more than 20 wins per year holds each to a 75% mark (76% for Maddux, to be precise). Beyond those two, however, it's a long way down: Tim Hudson is next, with a 7% mark that is slightly low because he didn't pitch a full season in the majors in 1999.
As for strikeouts, the 5,000 mark appears nearly unassailable, at least given the talent currently in the majors. The Toy doesn't give a positive probability to either Johnson or Clemens for reaching 4,500 strikeouts, and the majors' other current 300-strikeout pitcher, Curt Schilling, just passed 2,000 for his career this summer.
One name is, of course, conspicuous in its absence from the above discussion. If Pedro Martinez had been healthy enough to make 30 starts this year, he'd be a viable candidate in both categories, but his 10-week season has pushed all of his probabilities below zero. However, since the Favorite Toy is just that, a toy, we can play a little with his numbers and hope that he'll come back next spring at full strength.
If we use start with current career totals, but use his 1998-2000 seasons to construct our baseline, Martinez does have the right combination of age and career totals to have positive probabilities where they count. The Toy gives him a 21% chance of reaching 300 wins and a 36% chance of reaching 4,000 strikeouts, both of which would be remarkable enough. Amazingly, the Toy gives him a 5% chance of reaching 350 wins (behind Maddux's 8% chance) and a 7% chance of reaching 5,000 whiffs. Alas, even our cheat can't raise his shot at Ryan's 5714 strikeouts above zero. It'll take a remarkable combination of stuff and health to threaten that one.

Keith Law is an author of Baseball Prospectus. You can contact him by clicking here.