Thoughts on the George Springer signing

George Springer is officially in the fold for $150-million, the biggest contract in Blue Jays history, surpassing at least in nominal terms that awarded to Vernon Wells in the halcyon days of December 2006 for his age 29 to 35 seasons (he was already under contract for 2007). Wells sold at the top of his earnings power, and though he had some good seasons left, he was done as a star. Hopefully what’s past is not prologue with Springer.

As with last year’s signing of Hyun-Jin Ryu, I’m somewhat taken back by the sheer exuberance to the Jays simply winning an auction and landing a major free agent, as if it were an end in and of itself. I view total resources more or less as a given, and the challenge is translating that to maximizing on-field results over the medium term. In that respect, my initial reaction was the commitment was an extra year at the $25-million annual value beyond what was comfortable. But then that generally the nature of the game; it’s what it took to land Ryu and Russell Martin once upon a time.

That said, given how little salary was committed in the near term, it’s not going to crowd out anything. The real danger is what it may do in a few years, just as with the Wells deal there were reports the Bautista extension was possible only once it was moved. It would be terrible if in three or four years we’re in a situation of not being able to extend someone or land a needed piece due to having the backend of a collapsed Springer deal on the books.

Otherwise, my thinking is largely along two tracks. First, where Springer fits in the roster, and what was the best way to upgrade the roster for 2021 and the medium term given the available resources and opportunity set. Second, the value that Springer represents on a $150-million contract specifically.

---

As to the first part, while many seemed to think a major upgrade in the outfield was critical was winter, in my view, there were far more pressing areas to allocate resources. The Jays already had three players who project as at least fringe regulars there, and glaring holes elsewhere. Granted, I also think Randal Grichuk is at least passable defensively in centre, or at least not a black hole out there.

To the extent that the front office prioritized landing an impact, star level player this winter and there was a fairly limited opportunity set, then you take it where you can get it and work from there accordingly. But whereas an average or better infield or starting pitcher acquisition would essentially be a 100% upgrade on replacement level production at the margin, Springer is replacing at least a projected low end regular.

To be more specific, if he’s a 4-5 win player in 2021, but replacing close to a 2 win player, that’s only a 2-3 win net upgrade for the money spent. Perhaps one can squeeze out another win given the existing roster) if Teoscar Hernandez with significant time at DH, perhaps as the weak side of a platoon with lefty Rowdy Tellez.

The caveat to this is Vladimir Guerrero Jr.’s role. From the outside its pure guesswork, but if the front office has enough basis to think he can hold down the fort at third base on a team that now not only as legitimate aspirations but the expectation of contending, then it’s a much better fit. Then instead of a spot open somewhere on the infield, Teoscar can take the lion’s share of the DH at-bats with Grichuk and Lourdes Gurriel on the corners. It seems like a heck of a gamble though.

And of course, we still need to see how things get rounded out and what dominoes are still to fall. If there’s plenty of money left in the budget and those other areas can be addressed, then getting Springer makes even more sense. One of the outfielders could be moved, and maybe that is the plan all along. But in any event, while I didn’t really have Springer on the radar at the beginning of the winter in terms of needs, on reflection even in isolation it’s less of a suboptimal fit than I was thinking even recently.

---

What about $150-million specifically for Springer? The best acid test starting point is the standard framework of projecting a performance level and applying aging and then translating that to dollars.

For that, we can turn to Dan Szymborski’s ZIPS model, and it’s relatively optimistic with a base case 2021 projection of 4.7 WAR and 17.5 WAR across the contract years. How that translates to value depends on a number of assumptions, but if that projection came to pass, particularly Springer averaging ~4 WAR over the next three years, I think the Jays would be plenty happy with the bargain and we can say it worked out quite well even if the last few years are marginal.

My issue with relying entirely on forecast models is they’re really sensitive to collapse points. As an expectation across groups of players, player aging is a yearly curve as reflected in that ZiPS forecast for Springer (4.7/4.2/3.4/2.5, etc). But at an individual level, aging tends to be more about falling off a cliff.

If the player’s production collapses early in the contract, the team doesn’t get much of the near-term impact that contending buying near-term wins prize, and then get saddled with the backend. If the Jays only get one impact-type season from Springer before a major decline and then maybe a couple of decent years after that, they won’t have gained much above their existing baseline, while tying up a bunch of spending capacity. That fundamentally the big risk here, the risk of an early collapse.

That’s exactly what gave me pause about the Jays going after Springer: a poor history of free agent outfielders living up to large contracts, seemingly littered with early collapses shortly after signing. I keep thinking to four years ago when 31-year old Dexter Fowler seemed like a great fit with a six-year track record as a solid or better regular, the Jays severely deficient in outfielders and otherwise slim pickings on the free agent overall. The Cardinals stretched to five years, and while Fowler had a good 2017, off the cliff he went.

One data point does not a trend make, but $100-million contracts to free agent outfielders in recent years that came to mind seemed: Justin Upton, Yoenis Cespedes, Jason Heyward, Jacoby Ellsbury, Shin-Soo Choo, Jayson Werth, Carl Crawford. Granted, my personal bias may be to remembering the disasters, so I looked at all major free agent contracts (4+ guaranteed years, AAV at least three times MLB average) in the last 15 years:

Unsurprisingly. these were really good players, averaging almost 600 PA, 122 wRC+, and 4 WAR in the three years before free agency. After they signed their contracts...it was more quantity than quality, with the bats really declining. One thing really sticks out: of the 19 contracts in the last decade, only Curtis Granderson has even matched his WAR from the previous three years over at least four years of the ensuing contract. Even adjusting Springer’s total for 2020 being shortened, that implies well under on that ZiPS projected total.

In fairness, many of those players aren’t great comparison points for Springer, whose value is really tied to his elite bat whereas many others on that list had good bats with value in other facets. So I took out all players with less than a 120 wRC+ over the three years running in:

Though this subset has almost the same PA and WAR, it looks a lot more like springer with a 130 wRC+ leading up to free agency. Likewise, their bats held up better, and that quality drove a little more overall production (9 WAR vs. 8) in similar playing time. Still, I don’t think this is particularly encouraging as a prognosis for Springer.

Looking at individual players, Cespedes is probably the closest recent comp, though probably of limited inferential value. Shin-Soo Choo presents a more relevant cautionary tale, as he had been a longstanding offensive force (135 wRC+ in 3,477 PA in the six years leading in)...and then promptly collapsed. Josh Hamilton might be even closer to Springer’s offensive prowess, but again there are unique circumstances. On the other hand, Matt Holliday aged really well, giving the Cardinals five impact years. That would basically be the best case scenario for how this signing workout. Carlos Beltran wasn’t quite the same offensive force, but a CF would age well (the second entry adjusts him to be the same age/contract as Springer, though not included in averages).

---

The takeaway above is that big free contracts don’t turn out very well, and almost uniformly worse than forecasting models based on the individual’s previous performance would suggest at the time. This reflects a broader problem of what’s called asymmetric information made famous in a Nobel-winning economics paper by George Akerlof, “The Market for Lemons” about used cars.

As applied to free agent baseball players, the idea is that the player’s previous team has more complete information about the player to make a valuation. They will tend to keep players they think will hold up and let those they don’t (lemons) go to the market, and thus the players on the market will on average underperform. It’s at least interesting in this regard that Houston didn’t really seem to make any serious effort to keep Springer.

But beyond that general observation, there’s so few good comparisons among free agents for Springer that inferences about future expected production are necessarily limited. While he plays a premium defensive position competently, the majority of his value is generated from his bat, and that will particularly the case in his 30s. Hence, yesterday I looked at how similar players have aged offensively over their age 31 to 36 years.

Briefly, I came up with two sets of comparable hitters, the first based off Springer’s age 27-30 performance, and the second based off his power breakout and even higher 2019-20 production (chart below). The age 27-30 group averaged about 2,400 PA from ages 31-1 with a 114 wRC+ and about 10 WAR; the more elite 29-30 group averaged about 2,700 PA, a 117 wRC+ and 12 WAR.

By comparison, the ZiPS projection adds up to about 2,950 PA with a 122 OPS+, so the offensive side is more optimistic.

One note in terms of the WAR averages is that players selected based on hitting will tend to be more limited defensively/positionally than Springer is and reasonably projects. The average player in the 29-30 group averaged 10 defensive/position runs below average in those two seasons whereas Springer is close to neutral, so he’s about a half win ahead. Assuming that continued, one could add back 2-3 wins, for a total of 14-15 WAR.

---

On the positive side, there is a solid incidence of players who were really good, with more than 25% of both samples exceeding Springer’s ZiPS baseline; generally players who were offensive forces and/or retained defensive value. Scott Rolen, Chase Utley, and Carlos Beltran show up in both sets and posted about 20 WAR. If Springer approaches that level as a Blue Jay, the signing will simply be a smashing success. That’s a much higher rate than I’d have expected beforehand.

On the flip side, depending on how you want to define it, up to about 50% of the players in the samples who represent somewhere between disastrous to really bad outcomes. There’s a few complete collapses (Hanley Ramirez), but otherwise, almost everyone turned in at least 2,000 PA of above average offensive production. It’s a quality contribution, just not at a $150-million price tag.

In the middle is a mixed bag of profiles. Paul Konerko only posted 12 WAR, but was an offensive force. If Springer’s bat held up close to that, it’s a win. Albert Belle posted three impact offensive seasons before injuries ended his career. That would be a disappointing outcome for Springer, but probably better than a quick decline to 2-3 WAR averageness for most of the contract. Ken Griffey Jr. was pretty good with the bat but struggled to stay on the field, that would be ore disappointing.

There’s one final data point that’s encouraging. For both samples, instead of matching for similar hitting profiles, I matched based on overall offensive production (all players with 10 points wRC+):

In both cases, the age 31-36 outcomes are significantly better. If one buys into Springers as a truly elite bat based on the past couple years, similarly productive offensive players age very well into the contract horizon.