On Monday, FanGraphs released their 2020 Steamer projections, which I used to cobble together this bad contract trade post this morning. It took me most of the day to realize that I was using numbers that warranted their own article.
Last year, Kenny broke down the projections in an adversarial nature, acknowledging the Steamer algorithm’s sentience in a sort of man vs. machine social compact while remaining skeptical of the output. Before I make a similar pact concerning this year’s numbers, let’s review those projections and see just how well the machine really did . . .
Steamer 2019: 551 PA, .286/.363/.425, 12 HR, 4.4 WAR
Actual 2019: 445 PA, .257/.320/.368, 7 HR, 1.8 WAR
Steamer 2019: 622 PA, .252/.358/.435, 20 HR, 2.8 WAR
Actual 2019: 616 PA, .234/.339/.403, 17 HR, 0.5 WAR
Steamer 2019: 626 PA, .254/.306/.428, 21 HR, 2.1 WAR
Actual 2019: 508 PA, .254/.325/.437, 20 HR, 2.0
Steamer 2019: 552 PA, .275/.342/.399, 9 HR, 2.2 WAR
Actual 2019: 491 PA, .244/.315/.336, 5 HR, 0.2 WAR
Steamer 2019: 604 PA, .254/.322/.401, 14 HR, 2.7 WAR
Actual 2019: 560 PA, .228/.304/.350, 11 HR, 0.4 WAR
Steamer 2019: 555 PA, .237/.306/.347, 7 HR, 0.9 WAR
Actual 2019: 281 PA, .234/.278/.31, -0.5 WAR
Steamer 2019: 399 PA, .256/.323/.376, 7 HR, 0.4 WAR
Actual 2019: 192 PA, .238/.333/.417, 5 HR, 0.7 WAR
Steamer 2019: 476 PA, .226/.292/.386, 16 HR, -0.1 WAR
Actual 2019: 144 PA, .156/.250/.258, 4 HR, -0.5 WAR
Steamer 2019: 173 PA, .240/.295/.390, 5 HR, -0.2 WAR
Actual 2019: 296 PA, .268/.313/.507, 14 HR, 1.0 WAR
So, yeah, being generous, that’s 3 out 9 that wound up in the ballpark of accurate (Longoria, Slater, and Williamson). Feel free to take these projections with a grain of salt. The computer has yet to replace . . . everything . . . or, at least, in this case, the public-facing algorithm of a fan community has yet to close the gap between humanity and software or even bloggerdom and front offices. Context is still important.
The computer didn’t know the Giants were finishing off an emotional fadeaway or saying goodbye to Bruce Bochy or undergoing a complete re-thinking of the way they approach the game of baseball, both on the field and in the executive offices. It didn’t know that Buster Posey was coming back from hip labrum surgery or that Brandon Belt’s knee is made of a bag of rubber bands. Sure, it understands age-based regression and park factors, but it’s not all knowing.
That said, please enjoy these early offseason Steamer projections.
518 PA, .274/.374/.409, 11 HR, 3.4 WAR
Posey is coming off the worst season in which his power all but vanished. He hit two more home runs in 2019 (7) than he did in 2018 (5), but just one at Oracle park. He slugged just .368 on the season. Digging a little deeper into Steamer’s line, the thinking is that he’ll be about a league average bat (102 wRC+) but provide all of his value on defense.
It’s safe to assume that’s what the Giants were expecting from him, anyway, though this exact triple slash and league-average hitting seems extremely unlikely. He might move down in the order and maybe even get fewer at bats over the season, but he got plenty of rest last year and for the third straight year simply turned to dust over the final two months of the season. He’s still a master defender, just go get a snack when he comes up to hit.
560 PA, .253/.357/.439, 18 HR, 2.0 WAR
FanGraphs has always hated Belt’s defense (career -46.2 Defensive Runs Above Average), but he has consistently been one of the top defenders in the National League over the last five years. I like this projection a lot more if he’s not playing home games at Oracle Park; otherwise, figure he’ll have about the same season in 2020 as he had in 2019. But, you know, with actually decent defense, FanGraphs be damned.
620 PA, .260/.299/.415, 17 HR, 1.4 WAR
This is probably the best projection of the lot, as Steamer thinks he’ll basically repeat his 2019 season. That wouldn’t be bad for the Giants, even if the defense slips just another tic.
556 PA, .266/.305/.402, 14 HR, 1.0 WAR
A slightly below average hitter with a little pop and a slightly above average defender with some speed. That’s how the system sees him and that’d be a good bar for him. Steamer also projects 26 doubles from our beautiful boy.
564 PA, .252/.310/.439, 21 HR, 1.6 WAR
Despite doubling his walk rate from 2018 to 2019 and posting gains in slugging and, especially, his defensive runs above average, Steamer really doesn’t like a 34-year old third baseman. The system suggests he’ll have a season closer to 2018.
572 PA, .247/.317/.393, 14 HR 1.7 WAR
Dropping a 33-year old player by a full win year over year is pretty staggering, but that’s how bad Brandon Crawford was in 2019: he cratered his own projection. And yet, Steamer still can’t believe that he’d be that far away from league average (87 wRC+), even though his 2019 line (74 wRC+ — tied for the 2nd-worst hitter in baseball) and our 2019 eyes tell us that he’s nowhere close to league average these days. And Steamer thinks he’ll be close to a +9 Defensive Runs Above Average, even though his two-year average is 5.7. Steamer just can’t believe he’d be as bad as he was in 2019, but it can’t help but acknowledge that he won’t be all that good.
What’s all this then?
432 PA, .271/.336/.446, 13 HR, 0.8 WAR
Not complaining about the line, just stunned by the playing time projection.
522 PA, .242/.315/.413, 16 HR, 0.6 WAR
Steamer rightfully sees his high strikeout rate (26% in 2019) as a concern and considers a breakout performance from a 28-year old outfielder more of an outlier, which is reasonable, but he was able to keep that strikeout rate closer to 21% across seven minor league seasons (22% from 2017-2018), so maybe a slight reduction in that department could keep him in the category of perfectly cromulent starting outfielder on a rebuilding team versus bench and platoon option.
Anyway, these are just projections and aren’t better than our dreams. Dream big!