Varitek’s Value
Scott Boras is known for his ridiculous assertions in an effort to boost his client’s worth on the free agent market. We’ve already talked about his hilarious Oliver Perez as Sandy Koufax argument, so today, we’ll look at his claims that Jason Varitek’s defense is so valuable behind the plate that he’s worth “Posada money”, or about $13 million a year.
The heart of Boras’ claim is that Varitek has a significant, positive influence on Boston pitchers, far more so than any other catcher, and this influence directly translates to wins on the field. He presents the ’08 Red Sox .608 Winning Percentage with Varitek behind the plate versus just a .524 Winning Percentage when he didn’t start as evidence. It doesn’t just stop there, however – if you look at the pitching performances in those respective games, there’s a massive difference.
In games where Varitek started behind the plate, the Red Sox pitching staff had a 3.94 FIP. In games where anyone else started behind the plate, the Red Sox pitching staff had a 4.80 FIP. That’s a pretty striking difference, and FIP is obviously a better measure of the team’s pitching than winning percentage. So, is Boras on to something in regards to ‘Tek?
Maybe, but the macro view of the two statistics just presented don’t help us find out. There’s all kinds of problems with using the data just presented as evidence that Varitek helps his pitchers, starting with the fact that games started by Varitek isn’t a representative random sample. He caught all of Josh Beckett’s 27 starts, 30 of 33 Jon Lester starts, and 27 of 29 Daisuke Matszaka starts. Of the 89 starts made by the Red Sox top three starters, Varitek caught 84 of them. Of the 73 starts made by the various #4 and #5 starters that Boston cycled through, Varitek only caught 36 of those.
If the Red Sox hadn’t performed better with ‘Tek behind the plate, it would have been a massive upset. When 70% of your starts come with one of the big three on the mound, it’s a virtual lock that the team will have a lower ERA with you behind the plate than when you’re not, simply due to the talent level of the pitchers that you’re catching.
So, instead of looking at the macro approach, let’s look at how each individual pitcher fared when ‘Tek was catching compared to when Kevin Cash was behind the plate. There were some huge dropoffs among some pitchers when Varitek wasn’t catching.
David Aardsma: 3.33 FIP w/Tek, 6.71 FIP w/Cash
Hideki Okajima: 3.13 FIP w/Tek, 6.06 FIP w/Cash
Julian Tavarez: 3.06 FIP w/Tek, 5.62 FIP w/Cash
Craig Hansen: 3.76 FIP w/Tek, 6.08 FIP w/Cash
Jon Lester: 3.64 FIP w/Tek, 4.60 FIP w/Cash
On the other side of the coin, the swings weren’t as large, but perhaps more interesting.
Mike Timlin: 6.37 FIP w/Tek, 3.48 FIP w/Cash
Bartolo Colon: 4.54 FIP w/Tek, 3.37 FIP w/Cash
Clay Buchholz: 4.86 FIP w/Tek, 3.90 FIP w/Cash
Daisuke Matsuzaka: 4.12 FIP w/Tek, 3.80 FIP w/Cash
Three starting pitchers did better with Cash than with Varitek, while only one did better with Varitek than Cash. If Varitek really was significantly better at calling pitches and helping his pitchers improve, wouldn’t that be manifest most strongly with the guys he prepares with ahead of time, rather than the ones that come into a game without much notice?
Still, though, the sum difference of the FIP by the pitchers that Varitek and Cash had in common was +.252 when Varitek was behind the plate. That’s a huge difference, worth about 28 runs over a full 1000 inning catcher season. If we could actually prove causation, and not just correlation, Boras’ argument for Varitek wouldn’t be all that crazy after all.
Unfortunately, we just can’t. We just don’t have precise enough tools to judge whether the performance fluctuation can be credited to Varitek, or if it was just random variation. So, while there is some evidence that the Red Sox pitchers did quite a bit better when he was behind the plate, we just can’t say with any kind of certainty that it’s a sustainable skill that will go with him wherever he plays in 2009.
Buyer beware – if you pay a significant amount of money for Jason Varitek’s catching skills, hoping that your pitching staff will magically improve, you’re betting on a hope.
Dave is the Managing Editor of FanGraphs.
There seems to be about 68MM Americans willing to bet on hope and change, I would say it is likely there will be GMs out there willing to do the same.
Use, or rather abuse, of statistics as Boras does, is why many people seem to mock the value of statistics and the statistical community.
Either way, it is clear that opinions formulated on the basis of incomplete and/or inaccurate data lead to poor opinions, often resulting in very poor decision making.
This only works if the team looking to sign Varitek has the equivalent of Cash behind the plate. There’s really no way to compare catching defense across teams though, so who’s to say if Cash is replacement, average or good behind the plate relative to other MLB catchers.
Also, these stats should probably be adjusted for opponent. Cash started a small enough sample that there might be a significant difference between his opposition and Varitek’s. Probably the adjustment will lean in the direction of helping Varitek (he probably started against better opposition) but it would be worth checking.
Just wondering why you left John Beckett’s numbers out of you individual pitchers comparisions? Where his numbers nutural showing virtually no advantage of one catcher over the other?
Varitek caught all of Beckett’s starts.
And yes, Larry, that’s another important caveat – how do we know if the margin is Varitek’s bonus or Cash’s flaw?
Measuring catcher defense is just very hard with the tools we currently have. A team banking on Boras’ claims that Varitek will improve their pitching staff significantly is simply overestimating the reliability of the information we have.
Let’s try to break Varitek down into three parts, batting, defense, and “pitcher handling”. He is clearly replacement level at the plate, though he should regress up slightly in 09 and if he were to catch less often and be healthier through the season he might get closer to league average. We could say Tek is 0-0.5 WAR for batting.
His defense has not declined like his batting. He is about 5% worse than average in CS%, but is 0.130 points better in pass balls and wild pitches. He only committed 4 errors last year in over 1000 innings. I would say he is at least average defensively, maybe even 0.5 wins above average.
As far as handling pitchers goes we don’t have a good way to put a number to it. It appears as though he does a good job of handling them. He is at worst average in this department.
We should average the defense and pitching metrics into one grand defensive wins number. Lets call it average (2 WAR).
Based on the value of a win I can see paying him about 10MM/ year (150 games). A few business issues here. 1000 innings is only 111 games for starters, so that 10MM should be a lot closer to 7.5MM/year on his contract.
As a Sox fan I could see and be happy with 2 years for 16MM.
Bill James projects a .325 wOBA, Marcell a .312.
That is certainly not replacement level hitting.
Some fans complain about Tek’s relentless penchant for fast-ball calling.
I’ve always thought it was bit too much anecdotal; however, those two groups kinda supports such criticism, I guess.
pitchers with 1. diminished fastball-life (Colon, Timlin): 2. off-speed outpitches (Dice-K, Buchholz) performs worse while Tek is behind the plate.. mm..
My mistake. I just looked up the average wOBA and I saw .338. Given the projections Tek is a little below average, which is probably also a little above replacement, 1 win.
Sounds like Boras isn’t too far off in evaluating Tek, just he is going about it the wrong way.
What is FIP?
Fielding Independent Pitching, it only takes the three true outcomes (strikeouts, walks, and home runs) into account when evaluating a pitcher. It is represented just like ERA and can be looked upon as such (anything less than 3 is really good, above 4.5 not very good, etc.)
Some think tRA is better. Homers to flyballs can be random as well. And some pitchers do suppress babip better than others on average. Ask Carlos Zambrano. if you say he doesn’t, he’ll go all crazy on you and throw a ball into the outfield.
Scrappy,
I want to challenge the notion that Varitek is 1.5 WAR in “pitcher handling.” I don’t know how you can just assert “at worst average”. The anecdotal evidence to me is he’s nothing special there. He calls for so many fastballs that I wonder sometimes if his non-index fingers are broken. Of course my experience watching the Red Sox is primarily when Beckett or Papelbon is pitching and with those fastballs, you’re definitely going to call the pitch a lot more. That being said, I think I saw Papelbon use a non-fastball 4 times when I watched him (which was maybe 7 or 8 times this year). There are many examples on the web out there that basically say that Varitek called for the fastball too much this year. If his “pitcher handling” was really 1.5 WAR, then I don’t see how there should be ANY doubt about his calling.
Sorry that should be Scappy.
I know part of Papelbon and all of the fastballs is that for some odd reason management wants to try and get his pitches per plate appearance down. To do that they have shelved the splitter in favor of throwing more fastballs and trying to keep them low in the zone. Papelbon’s slider kind of sucks too, so he has become a one trick pony.
Beckett is one of the few two pitch starters (fastball / curve ball). With that kind of choice selection there really isn’t much choice.
The evidence I can supply for Tek calling good games, are the four no hitters he has been a part of. I know a vast majority of the credit needs to go to the pitcher, ~95%. Some credit needs to be given to the guy telling the pitcher what to throw. You could probably assign more or less credit based on how many times the pitcher shook him off. I still remember Schilling shaking him off in 07. I just don’t think the dollar amount is that unreasonable, the number of years are though.
You would also have to control for the strength of the opposing teams during those starts if you wanted to do it right. If Daisuke is starting against an anemic offense like Toronto with Cash behind the plate and Texas when Varitek is catching, you’re going to get quite a big split in how their performances look. Or vice versa. You could also look at multiple seasons to see what (if any) Varitek seems to have as an impact on pitchers.
Does Varitek call more fastballs? My first instinct is to doubt assertions like these. Shouldn’t it be easy to check (controlling for the pitcher).
Because I have no better way to occupy my time, I calculated the average opponent’s wOBA, EqA and OPS+ of the games each man started and found no difference:
Kevin Cash — 42 starts, .330 wOBA, .259 EqA, 99 OPS+
Jason Varitek — 120 starts, .329 wOBA, .259 EqA, 99 OPS+
Not the best method, but yeah.
LarryinLA,
There are numerous reports if you just search for “varitek too many fastballs” or some combinations. Those are all subjective, but I don’t really have the database access to check how much % wise and in relation to the league.
Scappy,
There are complaints when he catches the other pitchers too. All I’m saying is that I don’t think we can really say he’s above average in any capacity when there are doubters out there. If all we have to go on is anecdotal evidence, then I’d expect it to 99% overwhelmingly point to him being a good pitch caller.
what is variteks throw out % vs other starting catchers??