World Series time! Enjoy Premium-level access to most features through the end of the Series!
March 17, 2003
Spring Training Stats Mean Nothing
Stathead Rejects Stats. Film at 11.
Before last Monday's Padres/Angels game in Phoenix, I was perusing the package of information distributed by the media relations people. While eyeballing the stuff, which included league leaders and the complete stats for the two teams, the thought struck me:
Is there anything as meaningless as a stat sheet on March 10?
Sample size is the obvious reason to be dismissive of exhibition stats, and it's the one we've been pounding on for years. No one gets enough playing time in preseason for the performance to mean anything. Of course, teams make rash decisions based on limited evidence in all months, so there's no reason to expect March to be any different.
Watching Francisco Cordova last week really brought the whole thing into sharp focus for me. Cordova, trying to become the Padres' fifth starter after a couple of years out with elbow problems, came into the game with an ERA of 16.20 in four innings. He proceeded to strike out the side in his first inning, and tossed three good frames overall.
But it wasn't the Angels he was facing. Oh, the uniforms were red and Adam Kennedy and Troy Glaus were in the lineup and someone somewhere in the stands had a stuffed monkey, but it wasn't the Angels. It was the typical hodgepodge of irregulars and Triple-A players that makes up the traveling team for most any Arizona road game; Eric Owens and Julio Ramirez and Benji Gil and guys like that.
I'm not trying to specifically pick on Cordova, but the experience was illuminating. In the regular season, we can use stats to evaluate performance because the numbers are compiled in games between major-league teams and MLB players in the process of trying to win games. The stats aren't the goal, they're the byproduct, which is why we can feel comfortable using them to ascertain who's doing well and who isn't. No one is gaming the system, no one is compiling numbers against a substantially different subset of the player pool.
Over the course of a season, we expect any anomalies to even out, If they don't, well, they get close enough for us to be comfortable with our tools. In spring, however, we know that's not the case. I can't look at a pitcher's 1.65 ERA and say, "hey, he's been pretty good," because the ERA doesn't give me enough information. It's possible he's put up that number by facing cobbled-together lineups loaded with guys without enough pull to get out of the road trip. Even within a game, the caliber of competition can change considerably; a pitcher getting the bulk of his innings in the sixth or later may not face a player with MLB service time until St. Patrick's Day. Pitchers sometimes absorb beatings they would never take in a real game, because they have to get their work in.
Winning is not the goal in spring training, and without that at the core, the competition and the numbers produced in it are meaningless. Add that to the small sample size and the wide variation in the caliber of competition, and I'm left to reach the conclusion that spring training statistics are completely meaningless. There are just too many problems with them to make them useful as tools for analyzing performance.
If spring stats are meaningless, how do we tell what players are doing in the spring? Well, performance can still be important, but instead of cumulative stats--skewed by those first few weeks when everyone is in major-league camp--you want to look at how players do in the last two weeks, when games are a lot closer to MLB caliber. Scouting information becomes paramount; it's almost as if you're evaluating everyone the way you would amateur players and low minors, where stats aren't very meaningful. Pay attention to when guys are playing, if you can get that information. Are they starting and getting three at-bats, or coming off the bench? Are they playing mostly in home or road games?
If you're dedicated, you might want to dig past the stat sheet and go into the box scores to see if a particular player has played representative competition. This is particularly important with pitchers, who often make a team with 10 innings of shutout ball in the spring; you'll want to know if those ten shutout innings were for real, or the baseball equivalent of garbage-time stats.
The A's Steve Schott announced that the team would not be re-signing shortstop Miguel Tejada. This isn't a huge surprise, as the A's have committed the bulk of their payroll to their big three starters over the next few years, and also have long-term commitments to Ramon Hernandez, Jermaine Dye and Terrence Long, with Eric Chavez coming up on a payday in two years.
Here's the funny thing: While Schott's whining about "the system" is the kind of self-serving crap that we've supposedly left behind in the New Era of Good Feelings, the decision to not sign Tejada is completely defensible from a baseball standpoint, and making that call public right now might help avoid an in-season circus.
The bizarre thought process of the BBWAA aside, Tejada isn't an elite player, but rather a very good one who leveraged quality teammates and an uptick in batting average into a major award. His 2002 season was the best of his career by far, and at that, wasn't anything historic. There were also some nasty indicators mixed in with the highlight-reel three-run jacks. Tejada's walk rate trended down for the second straight season, and his isolated power and speed indicators actually slipped. His strikeout rate continued to fall, and the extra balls he put in play helped him get his average up to a career-high .308.
Tejada is likely to be one of the league's best shortstops in 2003, but you can interpret that statement any way you like. He's supposed to be 27 this year, but there's been rampant speculation--admittedly, there's no proof--that he's actually due to turn 30. Even if his listed age is correct, Tejada doesn't really deserve to be listed with the league's trinity. The following chart lists the EqAs posted by the four in their full seasons, from best to worst:
Tejada Rodriguez Garciaparra Jeter .297 .342 .341 .337 .281 .339 .337 .314 .276 .339 .311 .307 .259 .334 .303 .306 .240 .312 .294 .293 .311 .280 .294 .279At his peak, Tejada had a year that would rank as one of the lesser ones in the careers of the Trinity. Give him points for defense over Derek Jeter, certainly, and perhaps even Nomar Garciaparra. That still doesn't get him to their level of accomplishment.
Tejada is a good player, not a great one. He's not a franchise player; he has exactly one line on his resume that suggests he could be the best player--the highest-paid player--on a championship team. If he does it again in 2003, then you have to think about an eight-figure salary, but until then, there's no reason to commit to him into his 30s for 15-20% of a payroll.
By washing their hands right now, the A's should be able to avoid the circus of negotiations during the season. While comparisons to the Jason Giambi situation have surfaced, I don't believe the issue of trading Tejada will come up. Unlike in 2001, when they wrestled with trading Giambi as they chased down a playoff spot, the A's aren't likely to have to come from behind to make the playoffs this year. They should win the AL West going away, and may well be the best team in the AL, which will render any discussion of trading their shortstop moot.
The A's are scary good, so much so that they can enjoy Tejada for one more season, then allow him to leave. The two draft picks they get for him--what, you're not waiting for the recommendations of the Committee to Pretend to Make Changes to the Draft, are you?--will be used wisely, on college talent that can get to the major leagues quickly. The A's can move Mark Ellis back to shortstop in 2004 if need be, and let him fill the hole for a year while Bobby Crosby gets ready to take over, with Esteban German filling the second-base hole. The short-term hit they take would be minimal, and would be more than offset by saving, even in an optimistic scenario, $50 million over five years. Crosby might be a better player than Tejada by 2005, and he won't make $4 million, total, through 2007.
Reading the story that described Steve Schott's decision, I immediately thought of John Schuerholz. Just as the Braves' GM did in the wake of dumping Kevin Millwood to a division rival, Schott blamed "the system" for his team's inability to compete for Tejada's services. It was a load then, and it's a load now. Schott's true motivations lay in a paraphrase buried in the wire story:
"Until the A's can generate more revenue--Schott said it will take a new ballpark to do that--they won't be able to keep all of their star players."
Steve Schott wants a new mallpark, one paid for by someone other than Steve Schott. Until he gets one, he'll keep bashing the current park and crying about the unfairness of it all. This time, though, he's the blind squirrel stumbling upon a meaty acorn. What's best for his ballclub and best for his estate are probably one and the same.