The Padres Must Think That They’re Not Far Away
The last time the Padres won at least half of their games was 2010. Last year’s team finished with 71 wins and 91 losses, and, according to the underlying numbers, the club was actually even worse than that. Looking immediately ahead, the picture doesn’t look much better. Steamer thinks the Padres are the worst team in the NL West. PECOTA agrees. We don’t have everything we need from the ZiPS projections yet, but that system’s probably in agreement with the others. The 2018 Padres almost certainly aren’t going to make the playoffs. They’re just another organization that’s tried to rebuild.
The 2018 Padres are also going to play Eric Hosmer just about every day. News came out Saturday evening that Hosmer finally decided between the Padres and the Royals. The terms from San Diego, given to the 28-year-old first baseman: eight years, $144 million. There’s an opt-out after year five. Hosmer will get $105 million over the first five years, with the last three worth $39 million, in the event Hosmer sticks around. Reports recently had the Padres offering seven years, while Scott Boras wanted nine. These things so often end up with the obvious compromise.
Hosmer has been out there so long, and he’s been polarizing so long, that there’s hardly even anything new to say. If you’re a regular reader of FanGraphs, you know what Hosmer is, and what he isn’t. I wrote about the idea of Hosmer signing a big contract with the Padres back in the middle of December. Everything I said then still applies. This is an interesting deal, of course; Hosmer is admittedly fascinating. But what might be even more interesting is the signal this sends. The biggest contract of the offseason was given by a last-place team. That last-place team clearly has no intention of remaining there very much longer.
There’s no getting around the humor of this. It was just last month the Padres hired away our own Dave Cameron. It was an odd pair, because, while the Padres were enamored of Hosmer as a free agent, Dave wasn’t quite so enthusiastic.
Matt Adams got 1/$4M to be a backup. Eric Hosmer is probably getting $100M+. So I’ll just share these career numbers and walk away. https://t.co/Na60BaIsKU
— David Cameron (@OneDaveCameron) December 20, 2017
In November, Dave wrote about his biggest expected free-agent landmines, and Hosmer ranked at No. 1, with an expected deal of six years and $126 million. Hosmer is being guaranteed even more than that. He’s coming with an opt-out clause. It was interesting a month ago, because the Padres were hiring a guy who had publicly disagreed with their evaluations. It spoke well of their willingness to be challenged. Now, there’s a lesson here. I can’t imagine it’s that the Padres sold Dave on Eric Hosmer. Rather, the lesson is that the decision-makers are the decision-makers. Let it be remembered that one person has but one voice in a room, and in the Padres’ room, A.J. Preller’s voice is the loudest. At the end of the day, he gets to do what he wants.
I do think people have gotten carried away with the idea that the numbers hate Eric Hosmer. There’s this whole stats-versus-eye-test thing people love to bring up, but it’s overblown. Yes, there’s disagreement regarding the quality of Hosmer’s defense. Observers think he’s fantastic, while DRS and UZR think he’s literally one of the worst defensive first basemen around. Neither side can be entirely dismissed. But if you look at Hosmer’s last five seasons, he’s had three seasons in which he’s been worth at least three wins. The projections believe that Hosmer is solidly above-average. That’s just…it. And he’s followed a weird path, one that’s seen his WAR bounce around every single season.
In Hosmer’s three most recent odd-year seasons, he’s been good. In his three most recent even-year seasons, he’s been replacement-level. Hosmer broke in back in 2011, meaning he’s gone year-to-year six times since. Over that duration, Hosmer’s WAR has moved by a sum total of 22.1. That puts him in second place out of everyone who’s played since 2011, behind only Chris Davis‘ mark of 24.1. Hosmer has been a volatile player. It shows up in the regular numbers, just as it shows up in the advanced ones.
I don’t know what to do about volatility. I don’t think anyone does. When you think about Eric Hosmer moving ahead, you can’t expect that volatility to continue. Hosmer is assumed to be an above-average player. He’s getting a long contract because he’s only 28. As you might’ve gotten sick of reading about, Hosmer’s been a pretty good hitter, but he’s teased the ability to be a great one, if he can ever consistently get his batted balls off the ground. It seems so simple, but Hosmer’s been a ground-ball guy since he was a rookie, and he might not want to change. I’m sure the Padres aren’t unaware of the upside.
As far as that upside goes, I should make a note. There is one quirk to how this contract is structured. This deal is front-loaded, before the year-five opt-out. If it were front-loaded without an opt-out clause, that would make some sense, because Hosmer’s best seasons are probably his nearest ones, and the Padres have few other places to spend that money. But as the contract is written, Hosmer will get more money up front, and then it won’t take very much for him to opt out. The Padres have made it harder for themselves to get a premium return. Overall, it’s a somewhat minor point, but Hosmer essentially can’t be a bargain.
We should get back to the nature of this match. It was strange that Hosmer wound up having to pick between the Padres and the Royals to begin with. Neither team is competitive, and good free agents tend to sign with better teams. And given the choice, it would’ve been easy to see Hosmer sticking with what he knows. He’s always been with Kansas City. He’s won with Kansas City. The people there love him. He could be a franchise icon, the rare one-team player. I don’t know Eric Hosmer, so I don’t know why he picked the Padres, but maybe he just wanted a change. He’s already been a leader through one rebuild; maybe he simply thinks the Padres are closer to emerging from theirs. The Royals might be down for a while yet. Hosmer might see the Padres on the rise.
You can’t talk about Eric Hosmer without at least mentioning the perceived intangibles. He’s considered a leader, a heart-and-soul type, and the Padres are hoping that translates, to a different clubhouse full of different players, who’ve had different experiences. Every young team needs its positive role models, and in that sense, the Padres like Hosmer for the same reason they like Clayton Richard. I do imagine the Padres sold Hosmer on the idea that he could steer the ship. He’ll be the guy, for an up-and-coming ballclub. Hosmer’s someone who could command instant respect.
But to go all the way back to the original point, at last, I doubt the Padres wooed Hosmer by saying he could guide their rebuild another three or four seasons. And that wouldn’t even make much sense for the team, since Hosmer’s next few years will probably be the best ones he has left. I think there was a message delivered to Hosmer, the same message that’s now being more widely broadcast. This would be silly if the Padres didn’t think they were almost competitive. It’s hard to see how that could happen in 2018, but, come 2019, the Padres might attempt a real push.
This is a club that just ranked third in Baseball America’s organizational rankings. The quality and depth of the Padres’ farm system is no secret, as Preller has turned the organization around in a hurry. What’s missing, very obviously, is a sufficient number of talented major-league players, but sometimes teams like this can arrive sooner than expected. For all I know, Preller is growing somewhat impatient. He could be getting ready to cash some chips in, just as others start to graduate from Double- and Triple-A.
The risk is that, well, prospects are prospects, and they fail all the time. The Braves ranked No. 3 in the Baseball America list pre-2016, and they still don’t look competitive. Teams can’t always predict their timelines. But while the Padres can’t compete with, say, the Phillies’ resources, Wil Myers — who’s now an outfielder again — is the only other player with a major long-term guarantee. The Padres have financial flexibility to play with, since so many of their players over the coming seasons are going to be arbitration and pre-arb guys. It allows the Padres to take a chance like this, and even though Hosmer is being acquired early, the same was true of Jayson Werth. Werth joined a 69-win Nationals team. It won 80 games, before it won 98. The Diamondbacks, Rockies, and Giants are more set up to win now than they are to win in the future. The Padres have long been excited about the next wave, and this contract is like a commitment to trying to make the most of it, quickly.
It’s putting a lot of faith in very young players. If even just Fernando Tatis Jr. disappoints, I don’t know where the Padres will be. Many of you probably came to FanGraphs looking for us to rip this deal to shreds, and I won’t do that, because it’s not stupid. I think it’s just an overpay, by a team with nowhere else to spend. I certainly wouldn’t have wanted to give Eric Hosmer this contract, myself, if I were in charge of a team that’s not good. But now we get to see the Padres try to follow up. By signing Hosmer, the Padres are placing a bet on themselves. They must believe that they’re getting very close. It’s up to them to prove it. Baseball’s more fun when there’s pressure.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
Wow that is a strong commitment. 144m over 8 means they bet on him producing 2 war per year or 16 war and this is including decline years. Basically the Padres must somewhat bet on him improving and also think his bad 2016 was a fluke.
If you think steamer is right his projection for next year is 2.8. Applying “normal” aging the curve would be
28: 2.8
29:2.8
30:2.8
31:2.3
32:1.8
33:1.3
34:0.8
35:0.1
36:-0.6
Total:14.1
So the contract is under water around 18m. Not terrible but not good either and it could get worse if 2017 was the fluke and 2016 closer to normal. sure he could also learn to elevate and become a star but what is the incentive for a 28 year old with a big contract to change his swing?
The backloaded nature makes the last two years less bad but i still wonder why they had to give him the 8th season, opt out and ntc in this market. Big time win by boras.
I just want to point out that this is almost certainly not what the Padres are calculating. They are definitely counting on him producing more than 2 WAR per year, because they are almost certainly counting on him to produce more than 2.8 WAR next year.
There are two reasons why this is likely.
First, 2.8 WAR would make him somewhere between the 50th and 60th most valuable position player next year (I just looked it up on Steamer, and FWIW for some reason it has him projected for 2.4 WAR now anyway). That’s a ridiculous amount of money to pay for someone who wouldn’t sniff the all-star team.
Second, that $/WAR figure is taken from Matt Swartz’s work, which is the amount of WAR teams got for their money…not how much they expected to get. Swartz’s methodology takes the 3 year rolling average of contracts and actual WAR numbers and averages them out. What this means is if a Jason Heyward busts, the $/WAR figure goes up. If a Logan Morrison breaks out, the $/WAR figure goes down. This is mostly independent of the expectations teams have.
If you want to find out how much the Padres expect Hosmer to produce, just take whatever projection system you think is closest to the Padres’ line of thinking. What other teams paid other free agents to produce independent of what they actually did doesn’t help us much here.
To be clear, I do not disagree that this is a ludicrous move for the Padres. I just don’t think the Padres see it that way at all.
the bit about inferred preference is a piece of standard economics style. in this case tho, it is possible that teams are operating far below a sustainable market clearing budget, so there’s a lot of room for them to overpay before being forced to confront the cost.
This means that the observed market clearing price is in fact not a perfect reflection of team preferences, and there can exist systemic over-valuation of free agency for long periods of time. Considering that the portion of baseball WAR available on the free agent market vs the portion of WAR in controlled players is in steady decline since the late 90’s, an increase in $/war could certainly have contribution from teams behaving as fairly silo’d organizations, with imperfect tracking of efficiency to adjust for relative efficiency of various sources of talent.
the above described effect shouldn’t account for all of the rising $/war, but it does say that an adjustment is possible as teams become more aware of more efficient teambuilding strategy.
*revealed preference.
I more or less agree with what you have said, but one other note on the the “revealed preference” thing: Yes, it is a pretty standard economics approach and I think that this is the sort of market where it doesn’t work at all. Revealed preference only makes sense as a construct if people know what they are buying.
So let’s say I go to the grocery store, intending to buy a bag of 100 pretzels. I go and find a bag that looks like it has 100 pretzels, but lo and behold when I get home and open it up I only find 80 pretzels. Or worse, I find potato chips instead. What does the “revealed preference” tell us in this case? It doesn’t at all. What “revealed preference” do we figure out from Jason Heyward busting? It mostly tells us we suck trying to figure out how many pretzels are in the bag.
In this case, you really do need to figure out what the expectations are, or make some kind of a guess. Manski has some ideas on how to do this, although I’m not sure how realistic they are in an MLB environment: https://cepr.org/sites/default/files/4797-final.pdf
I don’t think anyone should be that shocked that Heyward is a bust, or that the highest bidder among 30 may be overly optimistic (or in that case, just plain nuts).
The revealed preference approach is not without merit, but one has to keep the timing the same, and this can be murky. Is the pitching market the same now that Darvish has signed? Yes and no. I would say mostly yes since he signed where he was expected to, but it is possibly that the uncertainty was large enough that this did change things for some teams.
Certainly, expectations are not the same as actual performance so retroactively applying the revealed preference logic is silly.
In this market, Scott Boras lands an eight year deal for a 1B with most of the money upfront and an opt-out after the upfront money is paid.
Boras is a stud among studs.
I’d say the MLBPA should make Boras the head of the union, but I doubt he’d be willing to take the pay cut.
If Boras thought he could do enough at the MLBPA to increase overall player compensation, he might figure that that could make up for the pay cut by increasing the size of his commissions when he went back to being an agent after leaving the MLBPA.
Well it’s really a five year deal at 21/yr with a three year insurance policy in case he is the bad hosmer instead of the good hosmer.
Padres get him for his 5 prime years.
Boras got an NTC too which is a free option if they want to deal him
This doesn’t seem like a “great” Boras deal. It seems ok. It signifies to me that the market is very soft.
Boras got Hosmer more guaranteed money than Dave or the crowd predicted with an opt out giving him the chance to get another roll through free agency at 33 if he produces. That sounds like a huge win. And there’s the no trade clause so the Padres can’t exactly move him if their rebuild goes downhill.
So, a 25 HR .318 hitting, 66 walk-drawing, team leader, who runs well, plays decent defense and is in his prime, got a good deal.
It looks like he just got what he is worth to me, not much more. I don’t see the “great deal by great Boras”. Then again, I’m not here to celebrate great agent-performance.
Based on the totality of his career, he’s a lot more likely to be .290/23/60. You know that.
Id say that a new peak at age 27 > earlier seasons. But, there is really only one way to find out.. see what happens going forward.
But he doesn’t play decent defense. And all of his good offensive years came with unsustainably high BABIPs. He’s had 3 good offensive years, and they each came on BABIPs above .335. When his BABIP has been normal or average (below .312), he’s been a below average hitter.
It’s not impossible for him to post some high BABIPs going forward, but it is virtually impossible for him to be above .335 every year.
The issue, of course, is that we don’t now that “he doesn’t play decent defense,” because UZR and DRS only measure 1B range, not how many throwing errors he saves (which is, of course, the main selling point on his defense).
If the Padres’ metrics evaluate Hosmer’s fielding ability as closer to his FSR numbers than his UZR/DRS, that’s nearly a full win added to his WAR.
It’s always critical to be aware of what we don’t know.
DSR includes scoops for 1Bmen.
The backloaded nature makes the present value of the deal greater. Time value of money is a thing.
I know what you’re saying, but it’s actually frontloaded.
That’s what I meant. I wish comments were editable after more than a few minutes!
Front loaded but even the front loaded portion was in line with projections and he may opt out before the back loaded if he is good for 5 years.
If he stays at 13 million a year, by the last year that is worth under 10 million in todays dollars plus he can be traded.
Actually, no, he can’t be traded without his permission. When he serves the five years, he becomes a 10/5 player and can block a trade. If he’s really bad, the Padres are stuck with him.
Whoops, that’s nine seasons. The total should be 14.7.
I think the total $/WAR figure should be higher than what you’re using, which would make this a good deal for the Padres. But I also don’t think he’s going to be nearly this good, which would make it a pretty big overpay.
Actually, Matt Schwartz’s work (which is the best public research on the price of a win) puts expectations at $11.1M per win, with 5.9% inflation going forward. While a lot of time has spent wondering if it would be lower than that this year, the deals made are largely in line with expectations.
However, the lack of decline you gave Hosmer for his age 29-30 seasons more than makes up for it. The average player declines about a 1/3 of a win per year over those years (link to my own research on this matter at the bottom).
My value calculator puts Hosmer as worth $113M over 8 years, though if you stop playing him when he reaches replacement level it rises to $123M. Even with giving the Padres that benefit, adding in the opt-out and the loss of a draft pick means this deal could project to be something like a $40M overpay. He needs to be considered something like a 3.3 win player right now to worth $144M plus the value of an opt-out and lost pick. It’s also how good he has to be to still project above-replacement-level (0.1 WAR) by the end of the deal in the mean outcome. If they buy into the intangibles and like his defense better than the public metrics than I guess they could reasonably believe that.
Average WAR/600 aging curve:
https://www.vivaelbirdos.com/2016/10/15/13283054/taking-stock-of-jedd-gyorko
Again, let’s remember: Matt Swartz’s work does not calculate expected value. That calculates total efficiency. It is a retrospective accounting of how much teams got, not how much teams expect to get. (The example I used up above is that when a Jason Heyward busts, $/WAR goes up, and when a Logan Morrison breaks out, $/WAR goes down.)
This might seem like kind of an arbitrary difference but what it means is: Many analyses using Swartz’s $/win number says something like “Because Jason Heyward and a bunch of other free agents totally busted, and we don’t project Eric Hosmer to totally bust, that means he’s a good value.” The part of that sentence that makes sense is not the part about Jason Heyward and his fellow free agent busts…it is the part about what we expect out of Eric Hosmer.
I think a more productive way of thinking about this is by thinking of opportunity costs. Say your goal is to improve your team by 5 wins this offseason. What is the most efficient way of getting there? How about if your goal is to improve your team by 3 wins three offseasons from now? And is Eric Hosmer really the most cost efficient way to get to whatever your goal is? I would think it is not, even at this late stage of the offseason, but someone on the Padres (I don’t know who) has some kind of an analysis where it is.
Don’t be ridiculous. Schwartz’s research looks at all players with more than 6 years service time, single players busting has an extremely little effect on the average.
That his work is accounting after the fact rather than the team’s internal estimate is true, and a valid criticism of using a general $/WAR figure to judge a deal. However, *we can’t know any individual team’s internal valuation of a win or even of any certain player*. The best tool we have is to use this along with noting other relevant factors like the team’s market/revenue, place on the win curve, etc. The observed average cost per win isn’t perfect, but pointing it out does nothing to change the fact that it’s the best metric we have to judge contracts. Obviously if there was credible public analysis that gave us a way to understand each team’s individual marginal benefit per win at each point of the win curve in terms of dollars that would be better. If that exists somewhere I’d love to see it.
“Say your goal is to improve your team by 5 wins this offseason. What is the most efficient way of getting there? How about if your goal is to improve your team by 3 wins three offseasons from now? And is Eric Hosmer really the most cost efficient way to get to whatever your goal is?”
These are all complex situations where knowing the average observed cost of a win in free agency helps us answer them in a quantitative manner. It also gives us an understanding of Surplus Value of both MLB players and prospects, so it gives us away to quantitatively measure trades/draft pick compensation. That way we can compare trades and free agent signings in an apples to apples fashion. Tell me, what public tool exists that allows us to better answer the questions you posed above?
A single player busting is not likely to have an effect, unless the player has an enormous contract. If several players with big contracts bust, it will dramatically inflate the $/WAR.
More generally, the “revealed preference” construct just isn’t valid here. I’ll use the analogy from above: Let’s say I go to the grocery store, intending to buy a bag of 100 pretzels. I go and find a bag that looks like it has 100 pretzels, but lo and behold when I get home and open it up I only find 80 pretzels. Or worse, I find potato chips instead. What does the “revealed preference” tell us in this case? It mostly tells us I suck trying to figure out how many pretzels are in the bag.
There is no credible public analysis telling us what each team thinks a player is worth–that is a closely guarded secret. Charles Manski argues we should measure expectations directly, but I just don’t think that is likely to work here. However, we do have pretty good systems for figuring out how much a player is likely to be worth. It seems much more likely that a prospective projection system like ZiPs, combined with the amount they actually pay them, will tell us how much a team values a player and its contributions than player outcomes. You could even aggregate it to get a similar $/win number as you get from retrospective data.
(Important note: If anything, the “revealed preference” we see here is not how much the player performs, but how much the team pays.)
I would then compare them to other opportunities on the free agent market, or even via trade (which I actually like the concept of the surplus value framework a lot for, more on that later).
I also actually think there’s a lot of analysis here about whether you’re on the “right place on the win curve” to determine whether an acquisition makes sense, how it fits into contention windows, etc–I think you mention doing that too, so I don’t know why you think it can’t be done.
Short version:
1) I would urge using prospective, not retrospective measurements to get at how a team values $/win
2) I would urge looking at opportunity costs available to a team to determine whether it was a good move
3) I would urge using Swartz’s work to look at how efficient spending was, either in aggregate or by team
4) I would be quite wary of combining #2 and #3. Retrospective and prospective judgments of worth are not really the same thing.
(I don’t really have a problem with using the “surplus value” framework to evaluate trades and prospects and whatnot, as long as it is all pegged to the same system. The irony is that the thing that it would seem to be best at–evaluating free agent contracts–is the part where it stumbles. To the extent that my post was seen as a critique of the whole surplus value framework, that was not my intention.)
This is silly. Is it impossible to know how much a team values a win or a certain player? Of course. Does the average observed cost of a win give us a better idea? Of course.
The average observed cost of a win is likely biased by two factors: The team’s individual value system and the discrepancy between performance and expectations.
There is nothing to do about the first one…you’re always going to have team-level heterogeneity because teams will guard that information carefully.
For the second one, it would be better to measure expectations more directly rather than inferring that how assets collectively perform after purchase. The expectation is what the value system is trying to explain. If you can model the value system (as opposed to efficiency of purchase) you can get a lot closer to an actual value framework.
This article gives some more theoretical background as to why expectations should not be judged as equivalent to outcomes.
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.9.5406&rep=rep1&type=pdf
I have serious doubts that Manski’s specific analytic suggestions are feasible here because valuation systems effectively constitute trade secrets. But if there has been a convergence of analytic methods across teams (through some sort of isomorphic process, and which has been written about here before) then ZiPS would be a good stand in (or if you like to be fancy, a factor score of ZiPS, PECOTA, and Steamer).
That’s a lot of words for not addressing anything i said. Guess this is a good spot to stop.
I’m sorry I wasn’t clear, but I’ll just leave with this:
1) The average observed win is better than nothing.
2) Measuring how much a team “values” something is actually a measurement of expectations of rewards, not what you actually get.
3) We have better measurements of expectations than what actually occurs in the following 12 months. We should use those instead.
4) I think the surplus value framework is really smart. I just think the way we calculate the $/WAR on the free agent market captures something other than expected value.
5) I think Matt Swartz’s work is excellent, but most people are using it to answer an unrelated question.
This is something I am happy to discuss any time, but if you still feel this doesn’t address anything you said, then you’re right, it is time to stop.
He is being paid through age 35 (8yrs) and not age 36 and is far more likely to be a 4 WAR player away from Kaufmann stadium his first 3 years than 2.8 WAR. Thats 17.1 WAR
Entering age 28 season coming off a great age 27 season and going to a stadium and division that may encourage him to elevate the ball more, we may be seeing the next David Ortiz, one who plays 1B .
WAR is stadium/league/position neutral…
Petco plays like Kauffman Stadium anyway. It’s a little smaller, but the sea humidity kills the high fly ball which doesn’t carry. Ryan Ludwick went nuts hitting moon shots that fell straight down into gloves, rather than carry into the seats.
Hosmer has already adjusted to that, and there’s still plenty of grass for line drives to grab. Unfortunately, the humidity prevents low liners from dropping, so a ball in the gap will hang up long enough for a fielder to catch it. Petco will suppress his doubles like everyone else’s. A high liner over the outfielder’s head is a triple though.
Well, it tries to be stadium neutral. I don’t know how successful it is.
I’m not sure what you mean by position neutral, since there is an adjustment for position. A good offensive season for a average fielding shortstop would be worth more WAR than the same offensive season for a first baseman, no matter how good his defense was.
Last summer, Matt Swartz looked at forecasted $/WAR on the free agent market through 2022 and projected an increase from $11m/WAR in 2018 to $14m/WAR in 2022. With $/WAR costs increasing steadily year by year, is there any value to locking in the cost for 2024 production in 2018? Would the 18m overpay partially or fully evaporate through rising inflation of $/WAR for free agent signings on the market at that time? I suppose this all hinges on the assumption that there would still be positive WAR production to value in 2024.
I think the contract is structured with the money up front. If so, doesn’t this deal have the characteristics of a dump and run?