|Testy Copy Editors
|How good baseball analysis works
|Page 1 of 1|
|Author:||Wabberjocky [ Fri Dec 09, 2005 5:42 pm ]|
|Post subject:||How good baseball analysis works|
Here's a question: If you were the general manager of the Seattle Mariners and had the opportunity to acquire disgruntled Baltimore shortstop Miguel Tejada in a straight-up, one-on-one trade for Seattle rookie shortstop Yuniesky Betancourt, what would you do?
Most of us would said not only "Yes," but "HELL, yes!"
After all, in 2005, Tejada hit .304 with 26 home runs and 98 RBIs. He's a former league MVP and is still young (he'll be 30 next May). Betancourt, a Cuban refugee, took over as Seattle's shortstop more than midway through 2005, hit .256 with 1 home run and 15 RBis as a 26-year-old.
And yet, as a friend of mine pointed out, such a trade would be a very bad deal for Seattle. He argues quite convincigly â€” I would even say irrefutably â€” that Tejada isn't worth an unproven rookie for the next few years His reasoning is a classic example of new-school baseball analysis, and I thought I'd share it here for your edification.
Last night, news broke that Miguel Tejada is unhappy in Baltimore and has requested a trade.
Immediately, every fan in the Seattle area who is still pissed that the Mâ€™s didnâ€™t sign him two years ago began concocting trade scenarios in which we could rectify the mistake and bring Tejada to Safeco Field. Well, Iâ€™m not part of that group. Not only do I not want the Mariners to trade for Miguel Tejada, I wouldnâ€™t swap Yuniesky Betancourt for him in a one for one deal.
And no, Iâ€™m not insane. Put the straightjackets away.
Letâ€™s get past the labels here. Forget that Betancourt is â€œunprovenâ€ and Tejada is â€œa starâ€. Letâ€™s evaluate what we actually think theyâ€™ll do on the field the next four years.
Tejada hit .304/.349/.515 in 162 games last year. Depending on which runs conversion formula you want to use, that performance was worth about 100-110 runs of offense. Weâ€™ll use Runs Created just for consistency and give Tejada credit for the full 110 runs on offense.
Betancourt hit .256/.296/.370 in 60 games last year. His offensive value in his given playing time was 22 runs, which would project out to 60 runs over a full season.
Ignoring park effects, Tejada was worth about 50 runs more than Yuniesky Betancourt offensively in 2005, projected over a full seasonâ€™s playing time. Thatâ€™s substantial, no doubt.
This gets a little trickier. The defensive ratings for Tejada are mostly consistent, placing him right around average. Baseball Prospectus has him at 3 runs below average. From 2000-2003, UZR (the best system out there, in my opinion) had him as exactly average. Chris Dialâ€™s range system had him at 3 runs below average. The outlier in the metrics is David Gasskoâ€™s RANGE system, which has him at 25 runs over an average defensive shortstop in 2005. Iâ€™m not buying that itâ€™s the only system that picked up on his defensive greatness, especially considering the scouting reports on his glove have been generally mediocre. So, weâ€™ll accept that Tejada is something like an average defensive shortstop.
Due to sample size issues, we donâ€™t really have good data for what Betancourt is defensively. Even the most ardent supporters of defensive metrics would like two full years of numbers to draw any conclusions, and with Betancourt, we only have about 1/6 of that. The defensive data we have on him simply isnâ€™t reliable. However, I think most of us would agree that heâ€™s fairly good with the glove. To what degree, though, we canâ€™t really quantify, so Iâ€™m going to take the conservative route and give him credit for saving 10 runs over an average defensive shortstop in a full season. In reality, heâ€™s probably better than that, but Iâ€™m erring on the side of caution here.
Adding offense and defense together, and assuming no change in performance from either player, the difference between the two players is about 40 runs over the course of a full season.
40 runs. Trading Betancourt for Tejada, straight up, if neither player changed one bit from their 2005 season, would net the Mariners approximately 40 runs. This number is the absolute best case scenario for Tejada supporters. It assumes Tejada, in his age 30-33 seasons, wonâ€™t decline one bit, which as weâ€™ll see in a second, is quite the reach. It assumes that Yuniesky Betancourt wonâ€™t improve at all with the bat. It assumes that Betancourtâ€™s glove is only solid, and that heâ€™s not really an elite defensive player. And, lastly, it assumes that thereâ€™s no difference in run scoring environments between Safeco Field and Camden Yards.
We know those assumptions arenâ€™t true. Letâ€™s try to quantify how much weâ€™re overstating the case, point by point.
Miguel Tejada is heading into his age 30 season. Heâ€™s far from over the hill, but heâ€™s also undoubtedly on the downside of the career development arc. Not every player follows this arc, but as a general rule, trading for a player hoping heâ€™s the exception is folly. Tejadaâ€™s most comparable player, according to both Baseball Reference and PECOTA, is Vern Stephens.
Stephens, also a shortstop, hit .295/.361/.511 as an SS in his age 29 season, almost a dead ringer for Tejadaâ€™s 2005 season. The next year, at age 30, .300/.364/.501, basically an identical line, but he did so in 40 less games. It would be his last season of any impact. At age 31, he hit .254/.343/.383 in 92 games, and he was reduced to being a role player for the remainder of his career.
Itâ€™s not just Stephens, either. A look through most of Tejadaâ€™s comparable players shows a similar trend. Cal Ripken, Joe Torre, Travis Fryman, Buddy Bell. They all endured serious decline, to the point of losing most of their value, in their early thirties. There are a few success stories, such as Bobby Doerr, but they are overshadowed by the clear trend of decline for a player with Tejadaâ€™s profile.
Prior to the 2005 season, BPâ€™s PECOTA projection for the next 5 seasons had Tejada holding steady through the 2006 campaign, then losing 8.5 percent of his value in 2007, 8.5 percent more in 2008, and 8.2 percent more in 2009. The projection for Tejada essentially costs him about 7 runs per season in decline from 2007-2009.
Again, thanks to sample issues, we donâ€™t really have the same kind of data as we do with Tejada. However, Betancourt is almost a certainty to get better offensively, even if only marginally, as young players who reach the majors at his age almost always improve to some degree through their age 28 seasons. How much improvement we should expect is up for interpretation. Iâ€™ll go with a modest 3 percent improvement each year for the next four years, which is in the ballpark of what players often compared to Betancourt experienced. Itâ€™s not a huge improvement-2 runs in in 2006, 3 runs in 2007, 5 runs in 2008, and 7 runs in 2009-but itâ€™s still significant enough to mention.
I used the 10 runs above average marker for Betancourt in the intitial comparsion, even though I think thatâ€™s low. Elite defensive shortstops, by most metrics, are 25-30 runs above average defensively. Thatâ€™s more the range I tend to think Betancourt falls into, but without any real evidence, Iâ€™m reluctant to use those kind of numbers in the comparison. If you happen to think Betancourt is one of the best defensive shortstops in the game, however, feel free to add 10-15 runs to Yunieskyâ€™s side of the ledger to account for the difference in glovework.
Starting with Tejada, the myth that Camden Yards is some kind of bandbox has been perpetuated in the media, but it simply isnâ€™t true. Overall, Camden has skewed slightly towards pitchers historically. In 2004, Tejada hit better on the road than he did at home. This year, his numbers were better at Camden, but thereâ€™s no reason to think thatâ€™s evidence that he benefited from his home park, as much as it was likely just the way things shake out in smaller samples. We shouldnâ€™t penalize Tejada at all for his offensive performances coming in Baltimore. We should, however, note that Tejada is exactly the kind of hitter that Safeco can destroy; a right-handed fly ball pull hitter. Thereâ€™s almost no way he would be able to sustain his offensive performances while playing half his games in Seattle.
Betancourt, on the other hand, certainly deserves a boost to his offensive numbers to account for Safeco Field. Because of its low scoring environment, runs are simply worth more in Safeco than elsewhere. Betancourtâ€™s offensive line in Safeco helps his team win just as much as if he had hit ..275/.320/.400 in a neutral park. The run difference in his translated line is about 5 runs. Again, not a huge difference, but a significant enough one to mention, especially when combined with all the other false assumptions we had to make to get Tejada to the +40 mark in the first place.
Run Value Conclusion
I realize reasonable people can differ on some of these points. Iâ€™m sure people will quibble with PECOTAâ€™s projected decline or Betancourtâ€™s defensive evaluation. Thatâ€™s why I broke out each questionable piece. This way, you can make your own evaluations on how much value you think Tejada adds over Betancourt. The most optimistic scenario, for Tejada, is a 40 run improvement. Depending on your interpretation of the other factors, it could be as low as 15 runs next year.
Essentially, youâ€™re looking at a possible range of performances from both players. Extreme Tejada Fans should assume a 40 run improvement if the Mâ€™s made the swap straight up, one for one. Extreme Betancourt Fans should expect about a 15 run improvement. More normal folks will fall somewhere in the middle. Thatâ€™s where I am. Iâ€™d probably expect a total net of about 25 runs in 2006 if the Mariners swapped the two shortstops.
Now, we get to the critical point we have yet to mention: salary. Tejada has 4 years and $48 million left on his contract. Betancourt has 3 years and $3.3 million left on his contract. Assume he gets a decent size paycheck in 2009, weâ€™ll say that Betancourtâ€™s total cost for the next four years is in the $8 million range. So, on average, youâ€™re paying an extra $10 million per season for Tejada. $10 million buys you somewhere between 15-40 runs.
That, folks, is an awful deal. There are several formulas out there, including one published in this yearâ€™s 2006 Hardball Times Annual, that place the value of a win at about $2 million. Itâ€™s generally accepted that 10 runs is generally equal to 1 win. So, letâ€™s do the math.
40 Run Tejada: $2.25 million per win
A slight overpay, but not a significantly bad one.
30 run Tejada: $3.33 million per win
A definite overpay, and a problem contract.
20 run Tejada: $5 million per win
An albatross contract, one of the worst in baseball, and a player who keeps you from winning.
10 run Tejada: $10 million per win.
Russ Ortiz style debacle.
So, there you go. The range of possible outcomes of swapping Betancourt for Tejada, straight up, range from â€œnot the worst deal in the worldâ€ to â€œfranchise cripplingâ€.
Thereâ€™s no way to justify that trade from an on field performance stand point. None whatsoever. The difference in value Tejada is likely to add over the next four years is dwarfed by the difference in salaries. Yuniesky Betancourt, at $2 million per season, is worth more than Miguel Tejada at $12 million per season.
Championships arenâ€™t won by acquiring stars in their decline finishing out free agent contracts. Just say no to Miguel Tejada.
|Author:||Wayne Countryman [ Fri Dec 09, 2005 8:52 pm ]|
|Post subject:||Re: How good baseball analysis works|
Ah, statistics and projections.
Here's another way of looking at this:
1. I've never seen Yuniesky Betancourt play. Or if I did, it didn't register.
2. I live in Baltimore. I'm not an Orioles fan, but as a baseball fan I watch quite a few of their games on TV and get to the stadium a few times a year.
We're comparing a 26-year-old rookie to a 30-year-old All-Star and recent 2004 MVP who's seldom injured.
Did Betancourt play in Cuba? Where's he's been? Are we sure he's only 26? (I honestly don't the answers, and am too lazy to look them up; I'm more interested in this analytical process.)
Tejada had a weak second half of the season. He tried too hard to carry a team that for the second season got off a great start and then sank into mediocrity. Then, frustration set in. He made far more errors and had far more strikeouts because of this. Unable to carry the team, he sank toward its level. Still, he was among the best shortstops in the league.
Did Betancourt play every day after becoming the starter? Can he hold up for an entire season? How does he do the second and third time he goes up against a pitching staff?
In other words, do we trust projections made after 60 games?
Tejada will be one of the best all-around shortstops for another year or two. He's only one season from driving in 150 RBIs. He's a team leader, which is why Orioles fans are really stirred up over his desire to be traded: he's respected far more than the front office; his unhappiness is taken as proof of what Baltimore fans think.
In one of those half-full, half-empty perspectives: Tejada is either an All-Star still in his prime, or a very expensive player with a lot of mileage on him who's bound to slow in the next few years.
Betancourt is pretty good, and it's a reasonable guess that he'll get better.
If you have the money and want to make a run for the World Series in 2006 or 2007, take Tejada.
If you don't have the money, or you're thinking long-term, and you think Betancourt could be someone to build around, then you might be far better off with him.
And you probably need to take into consideration what the rightfielder wants.
|Author:||wordygurdy [ Sat Dec 10, 2005 12:34 pm ]|
I dearly want to say Wabber's friend's analysis is deeply flawed because it doesn't take into consideration the fact that Stephens, Torre, Bell and even Ripken were facing much better pitching than Tejada faces today (Fryman played most of his career in the expansion era). Except it's hard to make that argument with Tejada, whose numbers (at least those I could find from 2002 to 2005 at ESPN.com) are inexplicably average or worse than average against the expansion teams. Tejada seems to be one of those players who play better against the better teams--his numbers against perennial playoff contenders are much better than those against the dregs of both leagues.
I would be very surprised if Tejada's production drops off the table as Wabber's friend seems to be predicting. Wayne's assessment seems more in line with past star players' performance--very productive till age 36 or so, then beginning an inexorable downward slide.
Of course, if the steroid allegations that Rafael Palmeiro heaped on Tejada's reputation prove to be true, then all predictions about Tejada's production are off.
|Author:||jjmoney62 [ Mon Dec 12, 2005 10:32 am ]|
|Post subject:||Re: How good baseball analysis works|
And yet, as a friend of mine pointed out, such a trade would be a very bad deal for Seattle. [/i]
Let's identify this friend and provide a direct link to the friend's analysis if there is one.
|Author:||Wayne Countryman [ Mon Dec 12, 2005 12:26 pm ]|
|Post subject:||Re: How good baseball analysis works|
According to news reports in Baltimore, the Orioles scrambled to reassure Tejada this weekend that he's wanted and that the team will improve.
Through his agent, Tejeda's supposed reply was that he doesn't really want to leave the team. He does wish that it would go after better pitchers.
My guess is that he just wanted to push management to spend money on healthy arms--the team's major weakness last year even when B.J. Ryan was rested.
Also: Reports out of Boston say the Red Sox want Tejada for Manny Ramirez.
Might be good for Ramirez to play in a more laidback baseball town, but who knows. Tejada would solve Boston's biggest problem, but he wouldn't protect David Ortiz the way Manny does, batting after him.
Orioles need leaders--since the end of Eddie Murray's career, that's been a weakness. B.J. Surhoff apparently is gone, too. Tejada has filled that role for two years.
|Author:||Wabberjocky [ Wed Dec 14, 2005 8:18 pm ]|
Sorry about that, JJ.
The direct link is here.
Nobody's saying that Betancourt is a better player than Tejada. That would be ridiculous. What's being said is that in the specific context of both teams, both parks, both players, their salaries and their projected futures, this deal would not be a good one for Seattle. That's all.
The point is that most people would say, "Tejada for Betancourt? Hell, yes!" without examing the full context of such a deal and all the factors that could influence its outcome. Good new-school analysis does that.
|Author:||ADKbrown [ Thu Dec 15, 2005 6:19 pm ]|
Many ballplayers continue to perform well late in their career these days, for a variety of reasons (more exercise, better physical therapy, profit motive, steroids). So I'm skeptical of the comparison with Vern Stephens and other players from an earlier era. But I don't know if Tejada is an exercise junkie.
|Page 1 of 1||All times are UTC - 5 hours|
|Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group