Nate Silver on the Zito signing
In an “Unfiltered” blog post at Baseball Prospectus (“Barry Zito and “Smart” versus “Dumb” Projections“), Silver makes a really interesting point: he argues that Zito’s contract is entirely reasonable if you think ERA is awesome and predictive of future performance. Over the life of his contract, he should be good enough to earn that $100m (if you assume that the current rate/win isn’t a fluke, etc).
If you don’t – if you’re willing to pay more attention to his component stats, his stuff, look at comperable players, etc, you find that that’s way, way off. $57m off, according to a PECOTA long-term look (which also assumes that paying the current rate/win, etc etc).
Yup. The game is changing, and front offices are getting much smarter, but it’s not changing that fast, really, and teams aren’t getting smarter at the same rate.
Comments
23 Responses to “Nate Silver on the Zito signing”
Leave a Reply
You must be logged in to post a comment.
I’d be interested to know how Boras and other smart agents use stats (and which stats they use) to argue for their clients.
Do you think Boras speaks new-school metrics to old-school GM’s?
Apparently not, in this case.
I’m sure Boras is smart enough to use whatever makes his client look best. He probably played up Zito’s ERA, wins, IP, and awards.
If he had a client with sub par traditional stats, but good peripherals, I’m sure he’d play up those.
Y’know, baseball blogs focus on on-field performance, but front office types have other metrics based on a player’s impact on revenue. A big free agent signing may allow the team to negotiate better local television/radio/cable contracts; justify higher ticket prices; and boost local revenue. It shows the fans, advertisers, marketing partners, and the like that the team will spend money to compete.
Remember, to an owner, “park effect” has to do with revenue generated from luxury suites, ticket revenue, concessions, stadium advertising, government funding, etc. And that’s what makes Boras so good. ERA? PECOTA? Boras projects a player’s ROI. That’s a number owners understand when they evaluate the risk of a big-dollar contract.
I would LOVE to see the complete financial analysis for Zito — the spreadsheets Peter McGowan sees from both his own staff and from Boras.
Exactly — Boras is amphibious.
One of the key requirements for any person involved in sales – and Boras is essentially a sales person – is to find out how the prospect makes buying decisions and present the information to the prospect in that format. One big reason why Boras is successful is because he does exactly that.
When Boras was representing A-Rod in 2000, Hicks had dreams of making the Texas Rangers into a far bigger brand than just the baseball team. Hicks was focused on building a local broadcast network and doing additonal land development around the stadium. Hicks probably had other projects in mind as well.
Hicks had already identified A-Rod as the ideal player for the Rangers, seeing A-Rod as both a future Hall-of-Famer on the field, and the perfect face around which Hicks could build his brand. I’m sure that Boras presented Hicks with additional information on how A-Rod could tie into those plans and help bring in those additional revenues. Thus, Hicks was willing to give A-Rod a contract that reflected revenues beyond what A-Rod brought on the field.
I have heard that Hicks is himself a good salesman, and he also had to sell A-Rod on the concept of playing for the Rangers. I recall reading some news stories that Hicks used those same development plans to woo A-Rod, describing to A-Rod how no other team in baseball could give him a chance to be a key player in building something that was bigger than just a baseball team.
A-Rod and Boras came to Texas mostly as a courtesy call that winter. Hicks met with them and laid out his whole pitch about what he was trying to build in Texas and how A-Rod would fit in. That pitch played right into A-Rod’s desire to be uniquely challenged and recognized, and once Hicks got that notion planted in A-Rods head, Texas went from being a courtesy call to having the inside track in signing A-Rod.
Considering how bad they are at evaluating (relatively simple) on-field production, one could only venture a guess as to how catastrophically terrible these supposed revenue-prediction metrics will end up.
so the rumored alternate offers for Zito were:
Mets 5/$75M
Rangers 6/$88M with an option to 7/$99M
and a supposed Boras proposal to the Yanks of 7/$119M (declined)
oh, and have I mentioned lately how different it is reading Geoff Baker than Finny in the offseason?
I’m so glad to see posts like those by Bellacarmella and Steve Nelson.
While I agree that on-the-field measurement of a player is the true value of his worth to his teammates and the team’s winning percentage, there is a lot more going on in these free agent decisions than just those things.
Just simply Zito sells some more tickets for San Fran. And he takes a little bit of spotlight off of Bonds. Maybe he becomes the focus of the team’s marketing and identity after Bonds leaves.
I think the Royals $11 million to Meche is similar. They wanted to prove to fans they could outbid other teams for a player. They are currently working on revamping Kauffman Stadium and need some evidence they are committed.
Of course they probably should be committed for giving a player of such low quality so much money.
But a lot of these decisions are made by men who are thinking about things from a whole different perspective.
Yes, nice piece by Baker. It is refreshing to see intelligent writing in the Times. However, it is also good to see that Chuck Armstrong remains as clueless today as he was when George Argyros called the shots.
Some things never change.
Not exactly a sly insult. Funny but direct.
Has anyone gone back 10 or 15 years and done historical PECOTA analyses on free agent signings and then looked at subsequent year performances against those projections? Seems like that would be pretty interesting to look at.
PECOTA has only been in existence for 3.5 years. I’m betting Silver wouldn’t want to do that much retrojecting. But you could take the projections for the 2003 through 2006 seasons, it seems to me, at least the ones in the BP book. Not sure the historical spreadsheets are readily available online.
I suspect that GM’s and agents, for the old-school ones, talk about classic stats, and new school ones wouldn’t get more “out-there” than peripherals, and of course, marketing effects. Like I can see a player being sold on potential, because his K rate, BB rate, and his great stuff and reputation, and OPS being used. I can’t see PECOTA or VORP or anything like that being brought up, just the classical stuff. I’ve also noticed that since its a very public and media-saturated market, that reputation and things that influence reputation DO affect GM”s whether they they know it or not. Things like Cy Young’s, Gold Gloves, no-hitters, and memorable catches all have inflated their players salaries automatically in future years when they become free agents. In that order, would Barry Zito’s contract been as big if he didn’t win a Cy Young in 2002? Would AJ Burnett have such a big contract based on “potential” if he didn’t have a no hitter, in which he walked 8 guys by the way? What about Gary Mathews and his amazing catch? I’m sure there are gold glove guys out there (suggestions?), none come to my mind.
One side point on Barry Zito is I read somewhere that the Giants see him as a Tom Glavine type, who when he is 35, will still be having 200 IP with a decent ERA. But Barry Zito in his prime, even a great player, was not even close to Tom Glavine or Greg Maddux in their primes. Those guys were both some of the best pitcher’s of all time in the 90’s, relative to league and what not.
Just simply Zito sells some more tickets for San Fran.
The Giants home games were at 93% capacity last year. Exactly how many more fans could Zito really put in the seats? And how many fans could/would the Giants draw in 2011 if they had $18M to spend to up your win total by 5 instead of getting 1-2 wins from Zito.
There is some merit to a NAME player drawing fans (though I don’t think Zito is that name). But usually this is overblown. I have a friend who insists that in a big market Dontelle Willis would put 10,000 people in the stands everytime he pitches, ignoring the fact that Florida’s has 40,000 empty seats to fill and so while it might be true there, the seats don’t exist in most cities.
Long term, the best way to draw fans is to win consistently. And dumb contracts are the #1 obstacle to that goal.
DMZ (or Dave), your parenthetical comment suggests you don’t even believe the $57M PECOTA/MORP valuation. What do you think Zito is worth in today’s dollars, or what would kind of contract would have been required for you to say it was an OK deal?
No one would ever do such a thing, because results matter less than the process to these people. The satisfaction is in applying the formula and the prediction itself, not in actually being correct. Player performance is irrelevant. It would be like asking them to trade in a Natalie Portman poster for an actual female.
The Giants are in a bind. They have gone from a have-not to a have, albeit one with a big mortgage payment. So far, due to a variety of circumstances, they’ve be able to avoid the big attendance drop-off that occurs for most teams once the newness of the fancy ballpark begins to wear off. The Giants can’t absorb a big dip in attendance quite as easily as the M’s. So they are motivated somewhat differently than most other clubs. While the Zito contract seems reckless, one can see that the Giants ownership may have needed to send a strong message to its money-spending fan base that they need to keep coming to the ballpark. Knowing that Bonds’ big contract (and drawing power) will disappear soon, the Giants must be hoping to stoke interest for the next 2-3 years and then hope like heck that the team finds a way to compete after that and that Zito’s salary on the back end doesn’t sink the team later. In some ways, I admire the Giants strategy, if only because it differs from the oh-so-frustrating one adopted by the M’s.
And to add to that Jon, I hate the smug snickers of most baseball writers about the amount of money spent. Jerry Crasnick’s caption to his article was something like “Giants signed Barry Zito, but here’s the joke: for $126 million!” (exclamation mark present). Seeing the inflation in the market this was no surprise, and to think you could get Zito for a reasonable deal is idiocy…if any of these guys were GM’s, they’d be sitting in a corner muttering “I’ll pay his fair price…” over and over whlie landing zero free agents and watching their team become crap. As much as I hate JP Riccardi, he brings up a good point, and that is, the fans only see the tip of the ice berg of what goes on for a GM…which is true for even writers too. There are so many factors involved, that to think you wouldn’t have to pay up for Zito is just stupidity. It’s called reality, not some fancy Microsoft Word dream land.
A little harsh, I think and I wasn’t suggestng validating (or challenging) anyone’s methodologies. Rather, I was thinking about using a standardized player rating system such as PECOTA to get a long term look at how baseball’s been paying for free agents — adjusted for inflation, how have prices changed for measurable skill sets? What’s the shape of the growth curve? What’s the rate of change and is it constant, accelerating, or variable? Is, in fact, management getting smarter about how much to pay for x number of incremental runs scored or saved per season? What differentials exist among the teams on these issues over time? How strongly correlated are prices paid for various player metrics over time and how are these changing?
w/r/t
and related comments:
That’s ridiculous, for one — it’s been fascinating to see PECOTA evolve over the years as evaluations of the results have led to new factors being introduced and others discarded or given less weight.
You can also follow the debates and comparisons on how different systems do, and see the fruitful competition of the various approaches people are using.
It’s taught us a lot about what’s important to player success, for one thing. We’re richer fans for the work Nate and others have done on projections.
Further, the whole “stathead = joyless tool” thing is namecalling at its worse, and I’ll ask you to please knock it off. No one’s less a fan for being geeky.
“namecalling at its worst”
I agree with your point. Nate Silver works hard to test pecota
Personally, I find the churning out of predictions pretty boring until I get to the end of the season and find out I’ve nailed Adrian Beltre’s stat line. Only then is it pretty cool.