Sunday, April 3, 2016

How Will These Yankees Defy Their Projections (For Better or for Worse)?

Last season during spring training, I posted two separate stories about which Yankees would outperform their preseason projections and which ones would underperform. This year, let's tackle both in one long post.

So what qualifies you to do this again? you might ask. Were your predictions better than those of the projection systems last year?

Well, yes and no -- quite literally! I correctly predicted the performance of both Yankees pitchers (Masahiro Tanaka and Dellin Betances) I examined while incorrectly forecasting the performances of a pair of position players (Mark Teixeira and Chase Headley). I correctly guessed one strong season relative to projections (Betances) and one weak season (Tanaka). On the other hand, I totally messed up in terms of Teixeira (much better/healthier than I expected) and Headley (much worse). So I went two-for-four -- like I said, yes and no.

I probably would have done just as well last year if I had just flipped a coin for a few Yankees. Still, my predictions were a little better than many others, including anyone who prophesied the Marlins to win the World Series and everyone who thought the Nationals would treat the National League like General Sherman treated Georgia.

Which is all a long way of saying that you can't stop me from doing this again. So, which Yankees will preform better than their projections and which will do worse? On to the predictions!

All projections based on the Steamer system, which is neatly displayed on Fangraphs. Yankees Steamer projections are found here.

Players Who Will Outperform Their Projections

Player: Didi Gregorius

Projected WAR: 2.3
Why Steamer is Pessimistic: Steamer's actually not that pessimistic. The system has Gregorius pegged as the tenth-best shortstop in baseball, which is much higher than it would've been after his rough first few months in New York. However, Steamer does still doubt Didi's offense, projecting him to cost the Yankees almost a full win at the plate. His projected .255/.313/.375 line matches his career norms.

Why You Shouldn't Trust the Damn Projection: Gregorius turned a corner in the second half last season. From June 20 until the end of the season, he slashed .284/.337/.401. That's hardly a small sample size, and the production was real: Gregorius hit that well without any sort of crazy batted-ball luck. He also looked much more comfortable on defense after a horrid start in the field last year. He should put up positive value at the plate and in the field, and I'd expect at least three WAR from him.


Player: Dellin Betances

Projected WAR: 1.4
Why Steamer is Pessimistic: The computer has a much longer memory than humans, and Steamer (and other projection systems) factors in Betances' struggles at a bunch of minor league stops for his first several years of pro ball. Steamer also projects just 65 innings for Betances, even though he logged at least 84 frames in each of the past two seasons. The Yankees will probably cut the righty's workload a bit, but Steamer's innings prediction seems very conservative, especially with Aroldis Chapman's early-season suspension and Andrew Miller's recent health questions. Steamer thinks Betances will have stellar rate stats, so that lower innings forecast looks like the primary reason for the reliever's modest projected WAR.

Why You Shouldn't Trust the Damn Projection: Just like last season, Betances is extremely undervalued by the projection systems. The gigantic righty has been one of the most dominant relievers in baseball over the past two seasons. As his Fangraphs player page says, "At this point, he's more than put to rest any concerns about his past struggles as a starter in the minors, where he got shelled due to his spotty command." So even though Steamer can't forget about Betances' growing pains of a few years ago, we humans should. Enjoy Betances this season, and expect him to eclipse his Steamer WAR projection before Labor Day.

Players Who Will Underperform Their Projections

Player: Brian McCann

Projected WAR: 2.9
Why Steamer is Optimistic: Based on solid career production, Steamer thinks McCann will be above average at the plate and behind it. The 32-year-old is projected to match last season's WAR output, when he hit five percent better than league average and added about a win on defense.

Why You Shouldn't Trust the Damn Projection: McCann is projected to be the Yankees' most productive everyday player, a fact that doesn't bode well because his WAR projection is actually rather modest. Still, I'll bet that age, catcher-related injuries, and negative baserunning value cause the McCann-imal to fall just short of Steamer's forecast.


Player: CC Sabathia

Projected WAR: 1.6
Why Steamer is Optimistic: This is pretty much the inverse of the Betances situation. Whereas Steamer still remembers Betances' struggles in the minors, the projection system also remembers that CC was one of the most dominant pitchers of his generation. So while Steamer projects continued decline by Sabathia, it also forecasts that he'll log 132 innings at about league-average performance. For that reason, Steamer credits Sabathia with a higher WAR projection than Betances.

Why You Shouldn't Trust the Damn Projection: Spring training performances usually tell us very little about what to expect from a player in the regular season. With that being said, Sabathia has looked atrocious down in Florida. While he has traditionally heated up as spring turned to summer, I don't think that'll happen this year. The lefty's velocity is down, his knee is more rickety than the Cool Runnings bobsled, and he's had a rough go of it off the field lately. He's not long for the Yankees' rotation, with injuries and the bullpen looming as the two most likely outcomes. The Steamer WAR projection is modest, but CC won't be much better than a replacement player this year. I hope I'm wrong on this one.

Why I'll Probably Be Wrong about Both Predictions: Because computers are smart (and humans -- even ones named Ken Jennings -- are not).

Anyways, Happy Opening Day, everyone!

If you want to subscribe to How Blank, just type in your email address on the right side of the page. You'll get a notification every time we post new content.

Follow Francis Tolan on Twitter @frantweet

No comments:

Post a Comment