clock menu more-arrow no yes

Filed under:

PECOTA projects the Orioles for 73 wins. Here’s how the Orioles can beat it.

New, 16 comments

Projection systems like PECOTA were wrong about last year’s Orioles. PECOTA projects 73 wins this year. What do they have to do to be better than that?

Baltimore Orioles v New York Yankees
PECOTA was hilariously wrong about last year’s Orioles, which is why the above photo happened.
Photo by Mike Stobe/Getty Images

Close to death and taxes on the list of certain things in life is the idea that the Orioles will receive no respect from the myriad preseason projections. This was demonstrated yet again on Tuesday morning with the release of the 2017 PECOTA projections, which asserted, after a brief hiccup, that the Orioles will win 73 games in the upcoming season.

PECOTA does not hate the Orioles. It’s just continually wrong about the Orioles, missing by 10+ wins in four of the last five years and by 15+ in three of the last five, including last year, when a 72 win projection turned into an 89 win season.

Even if someone believes the 2016 Orioles overachieved to make the playoffs, which is a reasonable belief, they did not overachieve by 17 wins. The Orioles were just better than they were given credit for being in the preseason, much like they were in 2012 and 2014. O’s fans will be hoping this trend continues.

PECOTA’s 73 win projection is formed from the sum total of its projections about the value of individual players, expressed in WARP (Wins Above Replacement Player), a similar statistic to the now-more familiar bWAR and fWAR for Baseball Reference or Fangraphs Wins Above Replacement.

For WARP, a team of 0-win players will win 50 games. The Orioles being projected for 73 wins means that they are expected to get a combined 23 WARP (actually 23.5) from their whole team - 7.4 from pitchers and 16.1 from hitters. That is a dismal projection, the third-worst in all of MLB, worse than is projected for openly-tanking teams like the White Sox.

Operating under the assumption that the projection will turn out to be colossally wrong, where are the areas where PECOTA might have egg on its face about the O’s at season’s end? You know, if it had a face.

Manny Machado

Here’s a simple question for you. Do you think Machado will be as good as, or better than, he was in 2016 in this season? If you answered yes, you’ve already added at least one win to the Orioles. PECOTA hits 2017 Machado with a 4.8 projected WARP, worse than the 5.7 he achieved last year and notably lower than the 7.3 WARP he earned in 2015.

One specific way he could, and should, beat his projection is with power. PECOTA only projects 25 home runs with a .476 slugging percentage. Machado, in 2016, hit 37 home runs and slugged .533. Only 25 projected home runs for a guy who’s combined for 72 in the last two years? That’s funny.

Chris Davis

Davis, with what we now know was a dislocated left thumb, hit 38 home runs last year. PECOTA pegs him for 35 in 2017. Even with his 2016 struggles, he was still worth 2.0 WARP. A 2.1 WARP projection for 2017 is one that it feels like he could exceed without much difficulty. His 47-homer campaign in 2015 was worth 3.8 WARP.

PECOTA’s algorithm gives Davis only 85% of the playing time at first base and does not give him any time at DH. That’s only about 138 games. So another way Davis can beat his PECOTA projection is simply by playing 155+ games, as he has done in three of the last four seasons, and doing good things in the 15-20 games it doesn’t think he’ll play.

Chris Tillman

Before we get into talking about a pitcher, a note about WARP. Tillman’s 2012 season, where he pitched to a 2.93 ERA over 86 innings once he was called up, is valued at -0.7 WARP. That is a negative number for a very positive half-season. For Tillman at a minimum (and probably all pitchers) WARP is, at best, suspect.

Yet even if we accept it, Tillman was worth 1.2 WARP last year. Tillman is projected to have a 0.2 WARP this year with a 5.03 ERA. Could it happen? It could. It did in 2015. But if the O’s blow by PECOTA yet again, they’ll probably do it with Tillman exceeding his projection significantly.

None of these are super optimistic, pie-in-the-sky guesses. That’s something like four extra wins just from looking at three players and believing they will be close to as good as they have been very recently.

Miscellaneous

We can play this game with more of the roster. I don’t know why Welington Castillo is projected for a -0.1 WARP - his defense, perhaps. But he was worth a 0.8 WARP last year in Arizona, so if he can do at least that much again (maybe he can’t, or maybe he’ll hit even better and be worth more) that’s a win there.

Stroll through the bullpen and several numbers stand out: Zach Britton with a 1.5 WARP, Darren O’Day at 0.5, Brad Brach at 0.6, Mychal Givens at 0.5. The projected ERAs that go with those WARP projections, which I wrote about yesterday, are actually laugh-out-loud funny.

Again just skipping right over a possible argument about whether WARP undercuts certain pitchers’ values - Britton’s 2016 was worth 4.3 bWAR and only 2.4 WARP - we can nab some wins just by supposing, “These guys will keep doing what they’ve been doing.”

Now, Britton’s not going to have a 0.54 ERA again. That would be unprecedented. So we’ll just give him a 2 WARP. And even an injured O’Day was worth 0.8 last year. I like his chances to do better than that. We’ll give him a half-win as well.

Givens was worth 1.4 WARP last year. I think he could well exceed even that this year, but this isn’t the optimistic scenario. This is merely the “Come on, he’s not THAT bad” scenario. Half a win more? Half a win more!

If we suppose that Brach, who had a 2.1 WARP in 2016, is good enough for just half a win more than his projection as well, then we’ve added another two wins just from believing the back end of the bullpen, based on its recent track record, is a little better than PECOTA thinks.

Over the course of one silly little article, we’ve “found” nine wins and the Orioles suddenly look like an 82-win team. And an 82-win team that gets lucky, like having a closer go 47-for-47 in save chances and a set-up man who parlays an injury to the regular eighth inning guy into an All-Star appearance, just for one hypothetical example, can end up in the postseason, or at least in the chase.

It’s not Orioles Magic that makes them beat their PECOTA projections. A lot of it is just their individual players being better than PECOTA gives them credit for being.

The people who came up with PECOTA are a lot smarter than me. The people who maintain it today are a lot smarter than me. That superior intellect has not imbued their creation with a good history of predicting the fortunes of Orioles seasons from 2012 onward.

Perhaps this will be the season where PECOTA is finally right, and the Orioles will, due to strokes of misfortune, injuries to key players, or players sinking down to true talent levels they’ve been (according to PECOTA) overachieving for years, stink once again.

If so, this blog post, more than most that I write, will be a tale told by an idiot, full of sound and fury, signifying nothing. For now, I like my chances to avoid that... at least until the first time the Orioles get swept in a series and everyone panics because everything is going to be terrible again forever.