Why the projection systems disagree about the Cubs

Every year, there is a fan base raging about projection systems. Over the last few years that would have been the Royals’ fan base, who argued that they were consistently undervalued by the projections as below-.500 teams. The projection systems were right about them maybe not being a one-hundred win team, but getting to the World Series twice in a row required a level of dare-I-say-intangibles that the systems might have missed.

This year, that honor falls upon the Cubs, who have a fan base that I am sure is taking it rationally as well. After winning “only” 95 games last year and getting bumped in the Wild Card game, you could say fans were both disappointed and even underwhelmed by a team that looked better on paper than in person.

Unfortunately the paper is starting to change its mind. By FanGraphs’ projections, they have a modest and understandable projection of 88 wins, good for first in the NL Central. Via PECOTA, they are projected to win just 80 games, good for last in the division. How could two systems project two opposite results, even if they differ by just eight games?

One of them is the strength of the division, obviously. The Pirates were always going to be a near-.500 team that would be on either side of both of those projections. The Cardinals added Paul Goldschmidt and Andrew Miller. The Reds added Yasiel Puig, Alex Wood, Sonny Gray, Matt Kemp, retained most of their other talent, and will add the likes of Nick Senzel from the minors. And the Brewers just won the pennant and added Yasmani Grandal to the fold.

The other reason is just plain disagreement in player performance and how these metrics are calculated. Take, say, Kyle Hendricks for example. He has been worth nearly 20 RA9-WAR as a Cub and has thrown 180+ innings in three of his last four seasons.

Yet, PECOTA projects 165 innings and 1.8 WARP (3.95 ERA) while FanGraphs projects 196 innings and 2.6 fWAR. Because he has consistently had a higher FIP than ERA it is naturally assumed that there is a regression, even on FanGraphs’ side, but reasonable observers could also make the argument that he is a FIP-beater as a rule. In this case, I may side with Hendricks.

Similar to Hendricks are the rest of the starters, who are both regressed to fewer innings and also closer to a 4.00 ERA. Call that a benefit or a curse of the mixed-model approach of DRA, but the assumption is always going to be to regress it to career and league norms, as well as adjusted for defense and park.

Another example would be Wilson Contreras. By FanGraphs he is something like a three-win player, while PECOTA thinks he is worth just a shade over one. The divergence here is obvious: framing. By Baseball Prospectus’ more complete metrics Contreras is projected to be worth -10.5 FRAA as opposed to the other’s nearly-league average approach, and here I side with BP. Just last year alone, he was dinged for over 17 runs (or nearly two wins) of framing; that makes a big difference standings-wise.

Going back to DRA projections just one more time, and this time for the bullpen. BP projects not a single reliever to put up more than a half-win, while FanGraphs has Brandon Morrow, Carl Edwards Jr. and Pedro Strop at least adding a modicum of value. This again could be due to PECOTA/DRA’s pitching projections; it assumes almost exactly an ERA of 4.00 for all three despite them not having the mark either in their careers or over the last 3-5 seasons. All of this could also relate to the Cubs’ defense dinging them by DRA as well, but that’s hard to tell with all of the FRAA caveats that come with that.

But by and large the rest agree on Kris Bryant, Anthony Rizzo, and Javier Baez that they will be... very good. Yet the success or failure of a season likely comes down to a weak underbelly of a bullpen and getting length out of the rotation, two things they have struggled mightily with over the last couple of years.

Call that ignoring their weaknesses in the offseason or trust in their own internal projections, but suffice it to say that this club could be great, or could be meh, and the difference in opinion is enough to warrant this being called a Jenga-tower roster that Theo Epstein and company just hope holds up over the long haul.