Opt-outs are the new hot thing. This winter, four players have signed contracts worth more than $120 million (David Price, Zack Greinke, Jason Heyward, and Johnny Cueto), but only one of those doesn't have an opt-out that might change the overall value and length substantially. The player signing that opt-out-less contract was Greinke, himself coming off an opt-out, leaving $72 million over three years with the Dodgers on the table to sign his long-term deal with Arizona.
As a result, opt-outs have been on everyone's mind. There's been a lot of great stuff written on the topic over the last month or so, here and elsewhere, but I wanted to touch on a few more things while taking a slightly different approach than most of the discussions I've seen. I'm calling it a "case study" method, and it involves identifying players who had a similar track record to a free agent at the same age, and looking at how they performed over the next several years.
For the most part, this is not a novel idea, at all – this Dave Cameron article is an example of the genre from after the David Price signing – but what is just slightly original is applying the concept to opt-outs. I'm going to focus on the comparable players' performance before and after the age when the opt-outs will trigger in all these contracts. This will give a sense of the likelihood an opt-out gets exercised or not, and a sense of the risk to the team, as well as illustrating a few broader concepts about opt-outs along the way.
5-Year WAR: 26.3
Contract: Three years, $90mm; opt-out; four years, $127mm
Price was the first to sign, so he gets written up first. Taking a cue from the Dave Cameron article linked to above, each former free agent's player comps will be the players in the last 20 years who totaled a similar amount of WAR over the same age range as the free agent did in his last five years. For Price, I set the window at +/- 4 WAR, to pick up a reasonable number of players, and it's a pretty impressive list.
|Name||25-29 WAR||30-32 WAR||33-36 WAR|
Eleven other players in the last twenty years have pitched roughly as well as Price between their age-25 and age-29 seasons, and on the whole it's a pretty good list; unsurprising, since Price has been very good. What if they had all signed the Price contract? Of the nine who have completed their age-32 season, when Price's opt-out will trigger, four would likely not have used it: Johan Santana (two decent seasons, the third totally lost to injury), Brandon Webb (too upsetting to talk about), Ben Sheets (coming off Tommy John), and Dan Haren (one very good season followed by two average ones). Greinke just finished his age-31 season, but barring a serious catastrophe, it seems highly likely he'll make it into the "would have opted-out" group. So five of nine or six of ten would have opted out, which suggests that, while Price exercising the opt-out is likely, it's not exactly a given.
Only six of these players have made it through their age-36 season, the final year of Price's contract, three of whom would have opted out and three of whom wouldn't have in this hypothetical. The three who wouldn't have – Santana, Webb, and Sheets – would have been serious drags on any team, the sort of bad player/expensive contract combo that can weigh heavily on any franchise, even a rich one. Two of the three who would have opted out, Roy Halladay and Andy Pettite, continued to be excellent, with only Roy Oswalt declining noticeably. Haren and Sabathia are only two seasons into the 33–36 portion of their careers, but without a serious renaissance, both of them would fall into the "bad contract" camp, Sabathia on the opt-out side of the line and Haren on the opt-in.
While I don't suggest drawing strict conclusions from this (e.g., Price has precisely a 50% chance of opting out), the general trends basically confirm what one would expect from this contract. There is a substantial chance that a great pitcher quickly turns into an average or below-average one, turning the massive contract into an anchor around the team's neck, and the opt-out does little to help with that. The chance that a player opts out and remains good, and the team loses out on the still-productive later years of a player's career is also fairly high.
5-Year WAR: 17.0
Contract: Two years, $46mm; opt-out; four years, $84mm
For Cueto, I narrowed the WAR window to 2, mainly because many more players have about 17 WAR over their age-25 to -29 seasons. I cut two players who otherwise would have made the cut, Adam Wainwright and Corey Kluber, because they made the threshold with only four years of WAR instead of five, and are overly optimistic comps as a result. I also cut Josh Johnson, who did fit the criteria but had Tommy John in his age-29 season, and the hypothetical of him signing the Cueto contract before his age-30 season breaks down pretty severely as a result.
|Name||25-29 WAR||30-31 WAR||32-35 WAR|
After those adjustments, we're left with twenty other players who had between 15 and 19 WAR over five seasons from age 25 to 29. Looking at this list with the benefit of hindsight, there are some terrifying names next to the very good pitchers, which reinforces the perception of Cueto as a player with his fair share of risk.
Eighteen of the twenty have completed their age-31 season, when Cueto will get the chance to opt out. Five of those eighteen seem very likely to have opted out – Lester, Shields, Radke, Millwood, and Burnett – and eight of eighteen seem just as likely to not have opted out – Jimenez, Penny, Bedard, Fister, Zambrano, Floyd, Lincecum, and Garcia. The remaining five comps had between four and five WAR over that two-year span, and my guess is that they too would stay in. Mike Leake, a roughly average pitcher, just signed a five-year, $80mm contract, and even with inflation, I'd guess most players of that caliber would prefer four years/$84mm, the amount that will remain on Cueto's deal after the chance to opt-out. The first takeaway of this analysis is therefore that there's seemingly a much higher likelihood of Cueto choosing not to opt-out than Price, even with his smaller contract. Somewhere around thirteen of Cueto's comps would have taken the guaranteed money, or more than 70%.
Ten of the comps are either retired or through their age-35 season, and can give a sense for the total value over the life of Cueto's contract. Seven of ten wouldn't have opted out in this hypothetical, and were uniformly quite bad for the next four years, averaging less than 1 WAR per year as a group, and the best (Freddy Garcia) still averaging less than 1.5 WAR per year. The Giants seemingly have taken on a significant risk with this contract. Of course, that's true of every contract for a free-agent pitcher, but the difference is that the presence of an opt-out cuts into any chance of upside substantially. Three of these ten would have opted out, and two would've been roughly average (2.1 WAR per season) and one below-average (about 1 WAR per season). As upside goes, that's pretty marginal, but there's surplus value to a Cueto-like pitcher staying with the team through 35 and providing decent performance under this contract, and that scenario becomes very unlikely with an opt-out.
Certain comps of Cueto and Price bring up a counterargument I've seen a lot in opt-out discussions, which is that an opt-out can be valuable to a team. Situations like Radke's or Oswalt's provide some precedent for the idea that Cueto or Price could opt-out, then decline sharply, leaving the Giants or Red Sox, respectively, happy that the player has exercised the opt-out.
But there are a few problems with that idea. Players, in the abstract, produce value, and once they're free agents, they can sell that value to the highest bidder by signing a contract with a team. Once a team has a player under contract, they control the value and can either keep it for themselves or sell it elsewhere via trade. An opt-out gives the player the chance to take that control back, which he will do so if and only if there's a higher bid on the open market than the remaining contract. In other words, if the contract is providing value to the team, the player will opt out, and if it's a drag on the team, the player won't.
The only way to determine whether a contract is valuable or not is by projecting future performance. In hindsight, some players' careers unfolded such that an opt-out would have saved their team from an albatross contract, but at the time the contract is being signed, there's no way to predict whether the pitcher will be Roy Halladay or Roy Oswalt after the opt-out. A team might be retrospectively glad, but at the time, the opt-out is always costly.
This is true even if the original team values the player less than other teams. Perhaps they have secret knowledge about the player's health or are a rebuilding team without any need for expensive players over 30. They might prefer the player opting out to having to pay him to play for them, but that's not truly the choice in such a situation, since they can always trade the player while they continue to have him under contract. If the player is opting out, it means the market as a whole thinks he's more than worth his current contract, and him opting out means he gets to sell those years of performance, instead of bringing them to the team.
As a result, opt-outs can simply never, ever be a positive thing when compared to the same contract without the opt-out. If you want a more quantitative perspective, Sky Kalkman (formerly of Beyond the Box Score, and on Twitter @Sky_Kalkman) put together a spreadsheet after the David Price signing that used some basic assumptions about Price's possible futures and $/WAR inflation to come up with an estimated total value for his contract, opt-out included. You can use the same method with any level of performance, contract structure, and year of opt-out, but no matter the method, the takeaway is the same: opt-outs are a benefit to the player and a cost to the team.
Finally, there are two more opt-outs in contracts signed this winter, both with the same player: Jason Heyward.
5-Year WAR: 23.1
Contract: Three years, $78mm; opt-out; one year, $20mm; opt-out; four years, $86mm
Heyward's contract is a strange one and difficult to evaluate without some serious thought. Multiple opt-outs are unintuitive, and while the second opt-out is conditional on reaching 550 plate appearances in the prior season, that's basically irrelevant: if Heyward is too hurt to play regularly and reach the plate appearance threshold, he almost certainly isn't going to opt out and walk away from a four year/$86mm guarantee. I went back to +/- 4 WAR for Heyward, and as above, required players who took fully five years to get to that point (eliminating Evan Longoria, Andrew McCutchen, Hanley Ramirez, and Nomar Garciaparra, all of whom would have qualified otherwise).
|Player||21-25 WAR||26-28 WAR||29 WAR||30-33 WAR|
This yields nine comps, eight of whom have completed their age-28 season. Immediately striking is that either eight or nine of these players would likely have opted out under Heyward's contract structure. Reyes has one of the lower WAR totals, but was coming off a season featuring his best wRC+ to date by almost 25 points, so it seems safe to assume he would exercise the opt-out. The only player I think would stay with the contract is Ryan Zimmerman, who averaged less than 3 WAR over that period. It seems unlikely that he'd walk away from five years/$106mm with an opt-out after the first year, but even there it's plausible. It's worth noting, however, that had Carl Crawford's collapse come a year earlier, he would make this contract look a lot riskier. That said, these comps should make one thing clear: a good player is almost guaranteed to take an opt-out after his age-28 season.
This brings up another important concept: opt-outs are more valuable to inconsistent or unpredictable players, or alternately, less costly to teams when they can predict with more certainty whether the player will use it or not. This can be demonstrated mathematically with relative ease. Using the same method as Sky Kalkman did, we can calculate the value of an opt-out to a contract.
Consider two players, both signed to two-year contracts worth $20mm with an opt-out after year one. Player One has a 50% chance of being worth 0 WAR both years and a 50% chance of being worth 2 WAR both years, while Player Two has a 5% chance of being worth 0 WAR both years and a 95% chance of being worth 1.05 WAR both years. Both players will opt out if they're worth more than 0 WAR in year one. These players have the same expected value, but the opt-out is worth $5 million in expected value to the inconsistent Player One and only $0.5 million to the extremely predictable Player Two. The more consistent the player, the less the existence of an opt-out changes the value of the contract.
Intuitively, this makes a lot of sense. Heyward is a good example; at 26, he's unlikely to decline sharply any time soon and, as the above analysis showed, can be almost guaranteed to take the opt-out. That has advantages in planning, but it also has tangible financial advantages, as the chance that he stays for the full eight years and fails to produce is very small. An opt-out to Heyward, therefore, is much less expensive than an opt-out to, say, Johnny Cueto, who is both older, seemingly more inconsistent, and a pitcher, and therefore inherently unreliable.
Hopefully this has cleared up some of the confusion around opt-outs and provided a roadmap for what to expect out of each of the three above contracts. It's easy to see why opt-outs provide value to players, as they're a form of insurance in case things go badly while still preserving the possibility of hitting the market again in just a few years. Unlike third-party insurance, however, the underwriter is the team, and opt-outs come at a real cost in lost control over the player and his value. They're seemingly becoming extremely common, if this offseason is any indication, which makes understanding them crucial to evaluating contracts.
. . .
Henry Druschel is a Contributor at Beyond the Box Score. You can follow him on Twitter at @henrydruschel.