How’s our model doing? Plus: updating Bryant’s valuation; Starling Marte analysis
How is our model doing vs. real-life trades?
As of January 31:
Total trades: 39
Total trades accepted by our model: 33
Acceptance rate: 84.6%
Variance: +/- 4.0
While we think that’s pretty good, we actually think the model is doing even better. That’s because the publicly available sources we rely on for data — both for projections and prospect ratings — release their updates sporadically throughout the offseason. Sometimes real-life trades occur before our model has been updated with the latest data. So we took another pass and asked, what if we had plugged in those numbers before those trades? The results would be:
Total trades: 39
Total trades accepted by our model: 36
Acceptance rate: 92.3%
Variance: +/- 2.0
Trade analysis: Starling Marte to the Diamondbacks
On Jan. 27, Starling Marte [22.5] was traded by Pittsburgh to Arizona for two 19-year-old prospects, SS Liover Peguero [16.0] and RHP Brennan Malone [7.5] . Pittsburgh also sent $1.5M to Arizona, while the Diamondacks sent $250K in IFA slot allowance.
This one worked out with a variance of 3 in our model, which is right in line with our averages. Those who were expecting a larger overpay for Marte might be disappointed, but we think this is a good return. Marte is now on the wrong side of 30, which is never good for a CF (quick: how many CFs do you know over 30?), as Joshua Iversen wrote about in November: aging CFs lose their value quicker than other positions, because either they lose a step and can’t cover the ground as well, have too much wear and tear by that age from banging into walls, and/or get moved to a corner. Injury risk is also an underrated factor with the over-30 crowd. So even though Marte is still an above-average player, we didn’t think he had a huge amount of surplus. He’s a declining asset.
The two prospects going to Pittsburgh both have high upsides: Peguero’s performance in rookie ball, combined with his tools and age, earned him a 50 rating by Fangraphs (BA was not quite as high on him); Malone was a first-round draft pick with a 45+ on FG (slightly lower on BA, again). The two represent the kind of prospects who seem to get traded more often: younger, with some projection; as opposed to middling-ish ones higher up in the system. For comparison, Simeon Woods Richardson was traded to the Blue Jays by the Mets in the Marcus Stroman deal; he wasn’t high on the Mets’ list, but his youth and performance now have him high on Top 100 lists.
Our model says it’s a fair deal (minor overpay, just barely), and we agree.
Kris Bryant valuation update
Now that Kris Bryant [38.7] has lost his grievance case, we’ve updated his valuation higher. We had been factoring in a probability that he could win, but now that probability has been removed, and his number reflects the certainty of his two years of remaining control. We would not be surprised by a trade at some point soon.
Note, however, that Bryant’s salary numbers are quite high. He will make $18.6M in 2020 and, if patterns hold, around $27M in 2021. That doesn’t leave a huge amount of surplus. There are also industry concerns about his defense, as his numbers at 3B have been in decline, and if he moves to, say, LF, that could hurt his WAR contributions. Nonetheless, we project him at this point for 8.8 WAR over the next two years, but after adjusting for standard injury risk, it’s 7.3. That translates to a bit over $70M in field value over two years.
On paper, that gives him roughly $24M in surplus. However, we also need to add in the value of the draft pick his team would receive assuming he is offered, and rejects, a QO after 2021. Finally, we adjust up due to market demand. The latter two modifications elevate our estimate of his median value to roughly $38M at this point.
Note, however, that we are still waiting on two remaining data sources to update 2020 projections, so this number could change again.