Opt-outs have been around since Alex Rodriguez's groundbreaking $252M/10 year deal in the 2000-2001 off-season. That deal allowed him to opt-out after the 2007 season, and the Yankees obliged by giving him a new $275M deal over 10 years. C.C. Sabathia also received one as part of his 2008-2009 deal for $161M over seven years. Like Arod, C.C. opted out when he could, after the 2011 season. Sabathia passed on $92M remaining over the last four years, and instead landed a new deal for $122M over 5 years.
Zack Greinke negotiated one into his deal prior to 2012 and exercised it this year. After making $76M in the first three years of his deal, he opted-out and scored a $200M+ guarantee from Arizona, albeit with some significant cash defferred until after the deal is over. A lot of people point to Greinke's deal as an example of an opt-out not working out for the team. I think it's the exact opposite, it's a great example of what's great about them.
Most reactions to Greinke's opt-out involve the narrative that if Greinke couldn't have opted-out, the team would have had the ace on a very team-friendly contract. The Dodgers don't get this now, and that's true. But, they had him on a friendly team contract already! Fangraphs has a way of tabulating value for each player's season, based on his production and the calculated cost of a win in the previous offseason. Based on that metric, Greinke was worth just over $105M from 2012 to 2014, nearly $30M more than he was paid over those years.
And that to me, is the silver lining to these opt-out deals. If the player opts-out, Then the team in essence had the player on a short-term, surplus-laden deal. If things go wrong then it's a long-term deal, but the players do indeed appear to be taking discounts in order to get these deals.
Opt-outs have exploded this season. Eight players have received at least one in their contracts, with Jason Heyward managing to get two in his deal. Of those eight deals, Ian Kennedy's deal with the Royals is the only one where the costs before and after the opt-out haven't been reported. So we'll exclude him for now. Here are the other seven:
The projections use an average of Steamer and Zips on a rate basis, while using Fangraphs' depth charts for playing time projections. ZiPS creator Dan Szymborski made two projections for Cespedes, one as a corner outfielder and one as a centerfielder, and I averaged those two together. Some common assumptions: decline each player a half a win each year, and use $8M for the cost of a win in 2016 with 5% inflation for the years after. Here's a value estimate for each player's pre-opt-out years:
A lot of surplus value to look at there. Cespedes is the only player that offers a negative surplus value over his pre-opt-year(s), but that's to be expected. Cespedes was thought by many to end up getting overpaid, due to a walk season way better than previous seasons as well as the fact that home run hitters tend to get paid better than WAR/$ would predict. Cespedes is also a guy whose name value has generally exceeded his numbers. Even counting Yoenis, on average the players' pre-opt-out years come with nearly a 25% surplus value. And free agent contracts are generally supposed to be 0% surplus value, by the nature of how surplus value is calculated.
The pre-opt-out portion is just part of the deal if the player doesn't opt-out though. The other main consideration is how good the player must be to opt-out. I calculated Break-even points for each player's post-opt-out deals in order to get a sense of how good each player needed to be in order to provide the same level of value they would be getting paid for. Theoretically, if they could be expected to be better than that break-even point, they would opt-out.
So, based on the half a win decline method, most of the players would be worth less than they're paid over the post-opt-out portion of the deal. But in general, the total pales in comparison to the surplus value gained in the pre-opt-out portion. Generally, if all these players aged linearly by the half a win a year model, only a few would opt-out, but overall they would average a good chunk of Surplus Value. That speaks to the discounts the players are giving in order to get the opt-out in the first place.
Cespedes is predicted to run a deficit over his three year deal, as we covered that was a fairly likely scenario no matter how his deal was constructed. Justin Upton also projects to produce less than he'll get paid, but he's also only 28. He likely declines at a slower rate than the non-Jason Heyward free agents in this discussion. Decline him just a quarter-win in his first two seasons, and he earns a couple million more in surplus value pre-opt-out, and the post-opt-out portion of the contract reaches the break-even point.
Cueto projects as a very slight deficit, though of course there is an ongoing debate on how to value his fielding-dependent numbers. The Cardinals villain has consistently posted an ERA better than his FIP. While ZiPS' WAR is calculated based on projected ERA, I can't find out how Steamer values pitcher's WAR on their end.
But, the half a win decline isn't set in stone. Fangraphs writers, for instance, have been using several more contract calculators than just the standard half a win decline method. One characteristic in many other estimators is to decline players in their early 30's just a quarter of a win rather than a half win. While this is fairly small when considering the pre-opt-out years, it can add up to a pretty large total over the whole guarantee. If you age each player just a quarter of a win during their pre-opt-out years then they would almost all project better than the break-even points presented above.
The important thing to me to remember is all of the players project to be around the value of the post-opt-out portion of their deal by the time the opt-out comes around, which is the goal.
Betting on the short-term rather than the long-term
The teams are essentially making a bet. And that bet goes "I believe you're going to be so good over the next 1-3 years that I'm willing to give you a long-term deal with an opt-out because I don't think you'll end up preferring this deal compared to the one you'll be able to get later. And in return, I want this deal to be cheaper than it would usually be." That's a mouthful, but the point is that it's a fundamentally different bet than the one a team makes when it gives a player a long-term deal without an opt-out.
When a team signs a player to a typical long-term deal, it's making a bet that the player will be worth it over the whole deal, through some combination of strong performance in the early years and aging well. The opt-out deal centers the bet on the early years which are, by baseball's nature, easier to project than the later years. The later years are still important as if the player does opt-in you don't want him to fall off a cliff, but you avoid the situation altogether if he opts out.
Recouping a draft pick
The Cardinals have avoided the opt-out in a similar nature that they've avoided signing players with compensation attached. But while draft picks are very scarce resources (and the same resource that helped build the Cardinals' run of success) the Cards shouldn't feel similarly about opt-outs. In fact, opt-outs make signing a player with compensation attached more palatable.
If the player is playing well, he most likely opts out. Also, if he's playing well enough to opt-out of a significant guarantee, he's going to turn down a Qualifying Offer, netting the team a draft pick once he signs with someone else. So in that sense, the Cardinals would essentially be pushing the pick down the road a few years. The pick isn't a sure thing and there's Net Present Value considerations, but this means the average cost in terms of draft picks is only a fraction of the usual penalty.
The upcoming new CBA and two years of a weak free agent market
However, there will be a new CBA in place by next year's off-season, and its been reported that the Qualifying Offer compensation system will be on the agenda. While I believe that there will still be some way for teams to extract compensation for departing players, the loss to the signing team and gains for the former team will probably be lessened.
The new CBA also provides a reason for why the players are pursuing so many opt-out clauses. The uncertainty about what the rules will be like afterwards likely provides motivation for the players to give themselves the ability to get a long-term deal before the new CBA, while still having the chance to get a new deal afterwards. Perhaps this is a one year outlier.
Another reason for players to push for opt-outs this year is that the market looks to be weaker the next two years, followed by an all-time great free agent class in the 2018-2019 off-season. The only player who can only opt out after 2018 is Price, and if he's still close to as valuable as he is now he'll have no problems still getting a good deal. Heyward could opt-out after the 2018 season, but unless he has a particularly strong season he might just take another $20M and opt-out after the 2019 season.
Team Strategy in Opt-Out deals
We see a variety of ways to structure a deal with an opt out. Price and Chen's deals are back-loaded like a typical free agent deal, with the player taking a lower AAV in the pre-opt-out portion than the post-opt-out portion. Upton and Kazmir's deals are even-loaded over both portions. Cespedes, Cueto, and Heyward's deals are front-loaded so that they receive a higher AAV before the opt-out than after.
I can see the strategy in both back-loading and front-loading deals. Back-loaded deals allow the team more surplus value during pre-opt-out years but decreases the chances of the player opting out. The strategy here is paying as little as you can, and putting just enough money on the other side until you get just short of an amount the player will pass up free agency for. Front-loading on the other hand minimizes the surplus value the team will get prior to the player opting out, but helps make it more likely the player will opt-out
Comparing the Chen deal to the Leake deal
The Marlins signed Wei-Yin Chen to $80M over five years, same as the deal the Cardinals gave Mike Leake. Chen is the far better pitcher, but his deal came with both draft pick compensation and an opt-out. So these two contract terms presumably make the two deals look more or less equal. But I'm not so sure that's true. Yes, Chen cost a draft pick, but the Marlins have a decent chance to get that draft pick back (or some other form of compensation worked out in the next CBA) in two years if Chen continues to perform well.
Chen's deal is also one of the most preferable deals I see, as he offers more than $20M surplus value over the pre-opt-out portion and projects as break-even after the opt-out. So the Marlins put themselves in a spot where they are way more likely to get Surplus Value than the typical free agent deal. About 50% of the time Chen will opt-out, and the Marlins will end up only paying just $28M for more than six wins over a two year period. The other about 50% of the time he won't opt out, but a good chunk of the time he still ends up delivering surplus value over the entirety of the deal. Leake's deal is fine and all, but I think the Cardinals could have done better by guaranteeing Chen $80M instead.
The opt-out clause adds quite the interesting wrinkle to free agent deals. While the negatives for the team are very apparent on first inspection, looking at the deals more in depth allows one to come away with the notion that these deals are benefiting the team as well. They're generally getting deals with smaller commitments than you would otherwise expect, and if things go well its a huge success. If thing go wrong, they go a little less wrong because of the discount the player is taking.
For the last few off-seasons, long-term deals have ruled the day. Players were generally sacrificing some AAV in the name of long-term security. The opt-out seems to be a counter to that model. Players can have their cake and eat it too, but the team has a chance as well to both have and eat cake. Want to sign David Price to a three year deal? Impossible without an opt-out. But with an opt-out you can have him for three years, and for much less than what he would command if he exclusively pursued a 3-year deal. You just have to give him most of the long-term guarantee as well, and bet that he ages well enough over the next three years to not even want it anymore.