There will be a very short planned maintenance outage of the site tonight (7/22) at 11 PM ET
January 31, 2005
The Sabermetric Challenge
A few weeks back, I was doing my weekly "HotList" segment when Brian Kenny hit me with his "Sabermetric Challenge." This was the day after Carlos Beltran had signed his seven-year, $119-million contract with the Mets, so Kenny threw out a bunch of choices, asking me to decide which of two players I'd rather have, money notwithstanding. "Beltran or Adam Dunn?" is one I remember, but I've forgotten most of them. In almost every case, though, I chose Beltran.
The one that sticks in my head three weeks later? "Beltran or Bernie Williams at his peak?" I haven't wanted to call a time out on television too often, but if I could have paused the segment there to do some research, I would have. I ended up choosing Beltran at the time, for his defensive edge and the runs he creates on the bases, but it was a weak call. I just wasn't sure, and I guess that makes it a great question.
The two players are awfully similar in type. At their peaks, both are or were complete players, packages of power, speed and defense who made contributions in every phase of the game. Williams' weak spots were his arm and his base-stealing--for a fast guy he was an awful percentage stealer. Beltran doesn't have those flaws, but he hasn't demonstrated Williams' ability to post a very high OBP.
Through age 27--Beltran's age in 2004--Beltran had been a more productive player, 36.9 wins better than a replacement player (WARP). Williams had just 28.5 WARP. The difference is largely attributable to the Yankees' questionable handling of Williams in his first few seasons, where they refused to grant him a job he clearly deserved. Beltran was Rookie of the Year in his first season (1999), and even after a difficult sophomore year, lowlighted by a suspension, was able to produce more value than Williams through age 27.
Of course, we're a bit more concerned with how the two players will compare over the next seven seasons, the life of Beltran's contract. To help me answer this, I asked Nate Silver to project Beltran out through 2011, when he'll be 34 years old. Nate did so, while offering the caveat that the method he used here may produce a more favorable result than others that he employs.
It should be noted that in these comparisons, Beltran is about seven months younger than Williams in every season.
Age Yr AB 2B 3B HR BB SB CS AVG OBP SLG WARP EqA Df CB 28 05 527 31 6 26 80 30 6 .280 .377 .508 7.4 .312 +1 BW 28 97 509 35 6 21 73 15 8 .328 .408 .544 6.9 .314 -8The reason I preferred Beltran to Williams was the defense and speed he provides. That edge is clear in comparing their age-28 seasons: Williams out-hits Beltran, while still not having as valuable a season because of Beltran's glove.
Consider run environment in looking at these raw lines. In the late 1990s and 2000, baseball was at an offensive peak not seen in decades. Williams' superior OBP and SLG produce just a slightly better EqA than Beltran's do (Beltran has a big edge in net steals as well).
Age Yr AB 2B 3B HR BB SB CS AVG OBP SLG WARP EqA Df CB 29 06 513 28 6 26 82 26 6 .283 .383 .513 7.5 .314 +1 BW 29 98 499 30 5 26 74 15 9 .339 .422 .575 8.1 .324 -1Williams' improved defense is the key for him at age 29. He continues to out-hit Beltran, with a 56-point advantage in BA that shows up in the rest of his rate stats. His average defense makes up the age-28 gap.
Age Yr AB 2B 3B HR BB SB CS AVG OBP SLG WARP EqA Df CB 30 07 505 31 5 25 82 23 6 .286 .387 .516 7.3 .316 0 BW 30 99 591 28 6 25 100 9 10 .342 .435 .536 9.5 .318 4Williams' best season destroys another good Beltran performance. One of the weaknesses of the method is becoming evident: PECOTA is projecting a very flat performance curve for Beltran, with slowly diminishing playing time. These are very good seasons for Beltran, but it's fair to say that if he never again has an eight-win season, it will be a mild surprise.
Age Yr AB 2B 3B HR BB SB CS AVG OBP SLG WARP EqA Df CB 31 08 494 29 4 25 82 21 5 .287 .390 .515 7.3 .317 0 BW 31 00 537 37 6 30 71 13 5 .307 .391 .566 6.9 .305 -5Williams' defense takes an ugly, and permanent, turn for the worse. Beltran outplays him, but the continued degradation in his projected playing time minimizes his edge.
Age Yr AB 2B 3B HR BB SB CS AVG OBP SLG WARP EqA Df CB 32 09 459 27 3 22 76 18 5 .285 .389 .499 6.7 .313 0 BW 32 01 540 38 0 26 78 11 5 .307 .395 .522 7.1 .307 -5Beltran is hurt as PECOTA's steady drop in his projected playing time eats away at his value. He's better than Williams in every aspect of the game (the OBP and SLG difference are entirely run environment).
Age Yr AB 2B 3B HR BB SB CS AVG OBP SLG WARP EqA Df CB 33 10 442 25 3 21 74 17 4 .282 .387 .493 6.4 .311 +1 BW 33 02 612 37 2 19 83 8 4 .333 .415 .493 6.7 .311-17Beltran outplays Williams, dragging their contributions nearly even despite Williams having 30% more playing time. Williams playing center field in a lawn chair helps, but I think that's been mentioned in this space before.
Age Yr AB 2B 3B HR BB SB CS AVG OBP SLG WARP EqA Df CB 34 11 440 25 4 20 73 14 4 .284 .388 .491 6.0 .311 -2 BW 34 03 445 19 1 15 71 5 0 .263 .367 .411 3.3 .275 -9Williams hits the wall in an injury-riddled season, driving down his playing time. Beltran is a whopping three wins better at 34.
Over the seven seasons, the two players are as even as they can be: 48.6 WARP for Beltran, 48.5 WARP for Williams. Beltran trails in more seasons, but makes up for it in the last.
I'm certainly not comfortable with the Beltran projection. Even though Nate says it's favorable to Beltran, it strikes me as far too clean; it's the rare player who has seven seasons of gradual decline from 7.5 wins to 6.0 wins, and I'm even less comfortable with the loss of playing time that PECOTA projects. That he'll never hit 30 home runs in a season again seems out of place, and would certainly render him a disappointment to Mets fans.
However, this exercise served as a reminder of just how great Bernie Williams has been. I admit that my fandom has affected my view of him over the past couple of years. He was once one of my favorite players, an underrated superstar who exuded class even when he was being mishandled as a young player, and who never received his due at his peak. Watching him misplay center field the past few seasons, while the Yankees do everything in their power to make sure he can't be moved to a corner, has been inordinately frustrating. I remain convinced that, relieved from center-field duty, he can have a late-career offensive surge. He doesn't need much of one to cement his Hall of Fame case.
Look at that 1999 season: .342/.435/.536 from a center fielder playing 158 games with positive defense value? Derek Jeter's best year drew more attention, but Williams was more valuable. His ninth-place finish in that year's Internet Baseball Awards voting is inexplicable.
So after all that, who would I take? As much as I like Beltran, I don't think he'll have a year as good as Williams' 1999. However, the Beltran projection does look very low, low enough that if he's just a little bit better and a little bit more durable, he'll be the much better player. Then again, there's the certainty of knowing what Williams did versus the fact that we're dealing with a projection for Beltran. If he misses a half-season along the way, or becomes a .260 hitter in Shea Stadium, Williams wins handily.
I guess I'm no closer to a decision now than I was staring into that camera three weeks ago. Time out!