This Trade in History: Josh Donaldson to the Blue Jays
Editor’s note: We’ve had people ask us if we could look back at major trades in history and see how they matched up with our model. This is the second installment of this series. (See the first one.)
Also note: Since this site (and our model) didn’t exist when these trades occurred, we may not have access to all the data we currently use, so some of the inputs are not as accurate as they would be today. But let’s try it anyway, just for fun.
On November 28. 2014, the Oakland Athletics traded 3B Josh Donaldson to the Toronto Blue Jays for IF Brett Lawrie and three prospects: LHP Sean Nolin, RHP Kendall Graveman, and SS Franklin Barreto.
This one came as a shock. The A’s had just come off three seasons of playoff contention. But after a strong first half where they had put up the best record in baseball, the team started falling apart. A’s Executive Vice President Billy Beane had tried to stem the downturn with July trades for Jon Lester and Jeff Samardzija to shore up the pitching, but it didn’t quite get them over the hump. Their season ended with a brutal loss to the Royals in the 2014 Wild Card game.
- Because the team had given no indication it would embark on a rebuild/retool after that, when news of this trade broke, it was quite a surprise — on a number of fronts:
- *Donaldson was in his prime, with four years of control remaining. Even Beane, who was known to trade early, hadn’t traded a star player that early in their control years.
- *There was no evidence that Beane had shopped Donaldson. This trade just… happened. (Later, he admitted as much, but said he was “blown away” by this offer.)
- *The return seemed underwhelming to many at the time. (We’ll soon see why.)
- The Blue Jays, meanwhile, were ascendant, coming off a third-place finish in the AL East with a record of 83-79. They had some good pieces — Jose Bautista, Edwin Encarnacion, and Melky Cabrera — but could use another impact player.
So what would our numbers show if we’d had the model then?
Imagine it’s November, 2014, and we’re looking ahead.
Donaldson has four years of control remaining, and his projections are sky-high, thanks to having put up 7.3 fWAR in 2013 and 5.7 fWAR in 2014. He’s a bit of a late bloomer, at age 28, so we’re looking at his age 29-32 seasons going forward (dollar values below in $Ms):
Donaldson has clearly hit his prime years with aplomb. He’s projected for 5.6 fWAR in 2015, then (applying aging curves), 5.1 WAR in 2016, 4.6 WAR in 2017 and 4.0 WAR in 2018. That’s a total of 19.3 WAR, which (discounted for inflation back to 2014), results in a projected field value of $154.4M.
Meanwhile, he’s just entering his first of four arbitration years as a Super-2, the total of which we estimate at $48.7M. (Rough breakdown of expected salary by year: 4.3, 8.6, 13.8, 22.) That’s very affordable for most teams.
The difference between the field value and salary numbers gives us a surplus of $105.7M. (Rough breakdown of expected surplus value by year: 38.9, 31.5, 23.3, 12.)
And given market dynamics — star players generally go higher than plain-vanilla field value due to high demand — we’ll skew his range slightly higher, giving us an estimated median value of $116.3M.
So it would take quite a haul to get Donaldson, right?
Hmmm. Let’s take a close look at the return package.
We’ll lead it with 3B/2B Brett Lawrie, who at this point has been in the majors for 3+ years. After a very promising rookie debut in 2011, during which he produced 2.6 fWAR in 43 games, he seems to have settled into a modestly useful, albeit underwhelming, career (relative to the excitement of his debut), producing 1.9 fWAR in 2012, 1.6 in 2013, and 1.6 again in 2014.
He’s also a bit injury-prone, as the amount of games in which he’s appeared has been steadily declining, from 125 in 2012, to 107 in 2013, to just 70 in 2014. He’s now entering his arb years. Even after we adjust his fWAR numbers to a full-season basis (but then account for the injury risk in a separate column), the projections are relatively modest:
The package also includes three prospects, whose valuation estimates at this time are (in $Ms; we’ve also adjusted all prospect numbers to 2014-era inflation to match the veterans):
|SS Franklin Barreto||24.6|
|LHP Sean Nolin||9.2|
|RHP Kendall Graveman||3.4|
So, looking at the total package:
- Oakland gets: 18.9 + 37.2 = 56.1
- Toronto gets: 116.3
That can’t be right, can it? I mean, we’ve seen a few lopsided deals once in a while, but nothing this far off. So let’s double-check to see if we’ve missed something.
Are we too high on Donaldson?
- *He has four years of control.
- *He’s produced 5.7 fWAR in 2014, after 7.3 fWAR in 2013. His WRC+ was 147 in 2013 and 130 in 2014.
- *Our projections, starting at 5.6 WAR for 2015 and declining steadily from there, seem reasonable (even a touch conservative).
- *We’ve even baked in more injury risk than usual, knowing that Donaldson has a tendency to play very hard at the hot corner.
- *So yeah, he’s older, and may decline a bit, but he’s playing at such a high level that even if he does, it would be down from an elite level, gradually, to an above-average level.
All that checks. So we feel confident about Donaldson’s numbers.
Are we too low on the Blue Jays’ package?
- *He is still young at this point (24), so presumably he still has his best years ahead of him.
- *However, we now have three years of data (fWARs around 2 if you adjust for full-seasons) to suggest he is what he is by now, and he comes with significant injury risk to boot.
- *As a hitter, he appears to be decidedly average — over his last three years his WRC+ numbers have been 97, 95, and 103.
- *So chances are, he’s just a league-average player. He’s meh. Maybe he rises slightly above that mark in his peak-age years to come, but it’s unlikely to be much higher than it is now.
What about the prospects?
- *At this point, he’s an 18-year-old kid with a few tools — bat speed, power, sprint speed.
- *He’s got upside, sure, but given his youth, he’s far away from the majors. And we know a lot can happen in the ensuing years.
- *Given that, he’s ranked in the back half of Top 100 lists — Baseball America has him at #86; MLB Pipeline at #85; and Baseball Prospectus at #74.
- *So there’s consensus on his profile of a talented teenager who has yet to prove himself. Those rankings validate our estimate.
- *He was a 6th-round draft pick (#186 overall), so it’s not like he has pedigree. Most 6th-rounders never sniff the majors.
- *But he’s a lefty with some good numbers as a starter. In 2013 in AA, over 17 starts, he put up a 3.01 ERA and a 2.82 FIP, with a K/9 of 10.0 and a BB/9 of 2.43. That’ll get you noticed.
- *So by 2013 he rose from obscurity (most 6th-rounders don’t make prospect lists) to something around a 40 on the FG scale.
- *In 2014, starting 17 games at AAA, his numbers went down a little — 3.50 ERA, 3.86 FIP; more concerning was that his K/9 went down to 7.63, and although his BB/9 improved to 3.61, the smaller separation between those two raises some eyebrows.
- *His GB% rose from 34.8% to 41%. So he’s pitching to contact more.
- *That’s typically not the profile of an ace. He’s more of a back-end starter type (or, at best, a Kyle Hendricks MOR control type, but he hasn’t quite proven that yet).
- *But given that he’s now at AAA, and giving him the benefit of the doubt, we figure he’s a 45+. That translates to his surplus value number above.
- *He’s more of a lunch-pail guy. He was an 8th-round pick (#235 overall), and figured to be org filler.
- *But he started defying that expectation by rising through the minors quickly. At the High-A level in early 2014, over 16 starts, he put up a 2.23 ERA and 2.88 FIP.
- *On the other hand, he was old for his level (at age 23, pitching against a lot of teenagers), so while that was impressive, it didn’t mean all that much.
- *Further, he’s not a strikeout guy — he put up a 5.96 K/9 at that level, against a 1.68 BB/9, with a 57.1% GB rate. So he’s been pitching to contact as well, with even less stuff than Nolin, and somehow getting by with it against younger minor-leaguers.
- *He continued that pattern in a smaller sample size after getting promoted to AAA, though, which tells you that he can probably be a back-end innings-eater in MLB.
- *Given his modest pedigree and his smoke-and-mirrors approach, we estimate he’s somewhere around a 40+ or a 45 at best.
So have we underestimated the return? It doesn’t look like it, to be honest.
So what have we got?
We’ve got a heist, is what we’ve got. Billy Beane appears to have been railroaded by the Blue Jays’ then-GM Alex Anthopolous.
Unlike other trades, such as this one, which looked fair on paper at the time but turned out disastrously, this one was lopsided from the get-go.
Why would that be? Was Beane really “blown away” by the offer? For a package headlined by a mediocre infielder, an unproven teenager, and two relatively unheralded pitching prospects?
Well, maybe he knew something we didn’t. Surely, if that were true, the results would prove him right in the end, no?
So let’s see what happened.
Josh Donaldson, fWAR and value produced (in $Ms, per Fangraphs), 2015-2018:
|Josh Donaldson||fWAR||Value||Salary||Surplus Value|
*Donaldson avoided arbitration by signing a two-year contract ahead of the 2016 season, then a 1-year contract ahead of the 2018 season. So the salary numbers here ended up about $7M higher than our model assumed, had he just gone through arbitration.
So Donaldson outperformed his projections across the board during the next four years of control, even with his down 2018. We estimated his field value at $154.4; he produced $182.0. We estimated his surplus at $105.7M (with a market adjustment up to $116.3M); he produced a surplus of $126M (which would have been $133M if he’d gone through arbitration).
One thing that’s clear: our model was not too low on him. Billy Beane might have been, but the model wasn’t. Neither was Alex Anthopolous, apparently. Beane was right to be concerned about injury risk, but Donaldson’s calf strain didn’t materialize as an issue until 2018, long after three elite seasons that included an MVP year.
So let’s see what Beane ended up with (A’s fans: look only if you dare):
|Brett Lawrie||fWAR||Value||Salary||Surplus Value|
So the stats didn’t lie. Our projection of Lawrie as a mediocre-at-best 1-2-ish player was, if anything, too high. He was close to replacement-level. In fact, the A’s traded him to the White Sox after 2015, for two relief prospects (who produced little to nothing, so that’s a wash). After 2016, he was out of major-league baseball. So in the end, he produced just under $10M in surplus value, compared to our estimate of $18.9M — a gap of $8.9M.
|Franklin Barreto||fWAR||Value||Salary||Surplus Value|
Barreto tantalized A’s fans who hoped he would become a star. He did not. He hit the occasional home run, but was mostly victimized by poor plate discipline (especially on low-outside sliders). The A’s traded him to the Angels for a rental in late 2020, after which he had elbow surgery. He was DFA’d recently, and is now a free agent. Couple all this with his lack of results over four years, and it’s not looking good. So the teenager whose surplus value estimate at the time of the trade was 24.6 produced -8.8 (and is now zero), for a gap of -33.4.
|Sean Nolin||fWAR||Value||Salary||Surplus Value|
Nolin made six starts for the A’s in 2015, with a 5.28 ERA, a 5.13 FIP, a 4.66 K/9, and 3.72 BB/9. So the pattern we saw in the minors played out — only worse. The K-BB gap closed even tighter, and overall, he was below replacement-level. Making matters worse, he got injured and did not pitch again until 2018 (after the A’s had released him; he’s since bounced around the minors, resurfacing as a reliever with the Nationals in 2021 and producing 0.1 fWAR for them). So this former 6th-rounder has essentially delivered like one — with little to no results.
Our value estimate of 9.2 for him as a prospect now looks overly optimistic. Given that he produced zero surplus value during his tenure with the A’s, we have a gap of -9.2.
|Kendall Graveman||fWAR||Value||Salary||Surplus Value|
Aha! We have a winner! Well, on a modest scale, anyway. Graveman was the throw-in; the one you’d least figure to make an impact. Yet in the four years the A’s had him, he produced $22.2M in surplus value (even with a lost year in 2018). Compared to the original estimate of 3.4, he exceeded expectations to the tune of $18.8M.
And to Graveman’s credit, he has gone on to accomplish even more. After wandering in the homeless wilderness of injuries, he transformed himself into a successful reliever for both Seattle and Houston, and recently signed a three-year, $24M contract with the White Sox.
Now let’s look at the big picture:
Blue Jays got:
|Player||Expected surplus value||Actual surplus value delivered||Gap|
|Player||Expected surplus value||Actual surplus value delivered||Gap|
As bad as this looks for the A’s, it could have been worse. At least they got some value out of it (mostly from Graveman) — just nowhere near what they gave up.
On a field-value basis, Donaldson delivered $182M. The four players the A’s received delivered $36.1M. So on that basis, Toronto “won” the trade by about $146M.
But the A’s, as it turned out, were retooling (they also traded Samardzija and Brandon Moss after this, and suffered through three losing years afterwards), so that’s not quite the right measure.
On a surplus-value basis, Donaldson delivered $126M; the four the A’s received delivered $23.4M, so Toronto really “won” the trade by $102.6M.
That’s right. Billy Beane — master trader, rock star GM, once played by Brad Pitt in a hit movie — threw away over $100M in value.
There’s an Occam’s Razor feel to this trade, in that the simplest answer is usually the right one. An elite player will most likely stay elite. A mediocre player will most likely stay mediocre. Two late-round pitchers will most likely not make an impact. Teenagers will most likely bust.
I think the main point here is pretty obvious: Beane blew this one. The question is: how? How could a GM, best known for steering the sport into the age of analytics, ignore the analytics? Or did he?
I see three plausible explanations:
- Beane did not have valuation metrics. Things have changed in the last few years, so maybe in 2014 it wasn’t as possible to value players like we do today. This is possible, but I have doubts. It’s not that hard to do even back-of-the-envelope $/WAR, surplus-value calculations, so surely he and his team could do that much.
- Beane had bad analytics. This would mean his valuation department produced flawed numbers, or at least numbers different than what we use today. Perhaps they overestimated the return, focusing on the best-case scenarios (say, 80th-percentile outcome) for the four players they received, and underestimated Donaldson, focusing on the worst-case scenario (say, 20th-percentile outcome). This seems far-fetched to me.
- Beane had the right analytics, but ignored them. Why?
One can only speculate on that last question. Maybe his ego got the best of him. Maybe he thought, having made many bold, successful trades in the past, he had the Midas touch. Maybe he believed that Donaldson’s body would fall apart and, having made a few good 4-for-1 deals in the past (including one of an oft-injured pitcher, which netted him Donaldson), he was on a roll, and could just lather/rinse/repeat the pattern.
Or maybe he was blinded by upside. This happens a lot with fans, and it’s possible with some GMs as well. To illustrate the point, imagine this hypothetical trade in today’s world:
- Yankees get: Austin Riley
- Braves get: Jasson Dominguez, Gio Urshela, Clark Schmidt, and Ken Waldichuk
Before checking the numbers, what does your gut tell you? Does this look fair? Do you make this deal if you’re the Braves? (Ignoring, for the moment, the team contexts, such as that the Braves plan to contend again.)
The structure of the deal is similar — the Braves give up a 3B in his prime with multiple years of control; they get back a serviceable replacement 3B, a mid-level pitching prospect, another pitching prospect with interesting numbers who may end up at the back of the rotation, and — the key to the deal — a teenager with tools and high upside.
Dominguez’s upside is such that many Yankee fans — and perhaps many front-office executives — would think he’s untouchable. But one look at his latest prospect evaluations, and his 2021 stats, would cast some doubt on that.
Many of us have a bias towards thinking the sky’s the limit with toolsy teenagers — so much so that we may overlook the reality that there is very high bust-risk there. But maybe we don’t think about that as much as we think about that upside, which, with some young prospects, seems infinite.
It’s not infinite.
Recently, I learned that most baseball GMs have read the popular behavioral economics book Thinking, Fast and Slow by Daniel Kahneman, and that many swear by it. One of the key points of that book is that everyone has biases, and most of us don’t see them. The best GMs are now careful to curb those biases by inviting input from other diverse voices on their staffs. It’s possible that Beane didn’t do that, couldn’t see his own biases here, fell in love with Barreto’s upside, and fell victim to it.
A friend of mine, who works with analytics, once gave me this advice: trust the numbers. In other words, don’t let your gut, or your biases, get the best of you if the numbers are telling you something else.
Alex Anthopolous trusted the numbers. Billy Beane did not. And he paid a high price for that mistake.