Evaluating the optimism and pessimism of VEB community projections

For the last month, the Viva El Birdos community has gathered to project how several notable players will perform for the 2018 St. Louis Cardinals.

As the projection forms were hosted on a website dedicated to St. Louis Cardinals coverage, it should come as no surprise that projections were largely more optimistic than the projections of ZiPS, the system Dan Szymborski (Googles name...got it right the first try!) established to project players. But there were some exceptions, and while some of the variances seem influenced by narrative rather than substance, others were seemingly influenced by logical causes that the rigidly mathematical ZiPS equation (Szymborski himself acknowledges that while the projection system is largely solid, there are occasional blind spots) did not properly capture.

Below is a ranking of players from most to least optimistic projections. The players are ranked by VEB projection relative to ZiPS projection, using OPS as the key statistic for batters and a hybrid of ERA and FIP as the key statistic for pitchers. Is it a perfect way to rank them? No, but...actually, who cares, yes. Yes, it is a perfect way to rank them. Let’s move on to it.

1. Tyler Lyons—15% above ZiPS: On the part of the VEB community, this one makes a fair amount of sense. Lyons, recently a less than excellent starter, was a revelation last season in the bullpen and figures to continue to handle high-leverage outings for the Cardinals in 2018. While the ERA projection of 2.81 feels a bit optimistic, the 3.62 projected by ZiPS also feels a bit high—Lyons did manage a 2.83 ERA in 2017, and while there is something to expecting regression, I’m slightly more with VEB than ZiPS on this one.
2. Luke Weaver—13% above ZiPS: Weaver was a bit of a FIP monster in 2017, coupling a 3.88 ERA with a 3.17 FIP, and VEB expects a similar, if slightly less extreme, trend in 2018. Weaver’s ZiPS projected ERA-FIP gap is 0.04, a very modest one, but the gap for VEB is 0.31—Weaver’s VEB-projected FIP of 3.12 is the lowest of any community-projected pitcher, starter or reliever. I’m probably slightly more on the side of the 3.77/3.73 ERA/FIP of ZiPS than the 3.43/3.12 of VEB, though I do agree with the community that Weaver is a solid bet to outperform his projections.
3. Tommy Pham—10% above ZiPS: How does one evaluate Pham, a breakout star from 2017 who did something completely unexpected in his age 29 season? Pham has the highest WAR projection of any player on the Cardinals (5.4, a high number which is lower than his FanGraphs mark from 2017) and while his OPS projections aren’t quite as outlandish, a .896 OPS, which would fall somewhere between last year’s versions of Anthony Rizzo and George Springer, is quite excellent. The major factor with Pham has always been health and thus I find a 5.4 WAR projection, while possible, wildly optimistic, but considering Pham exceeded his ZiPS projected OPS of .811 not only last year but also in 2015, I can buy being bullish on his (rate-based) OPS.
4. Marcell Ozuna—10% above ZiPS: While VEB expects regression, it doesn’t expect much of it. At .907, Ozuna is projected as the team’s best hitter, with an OPS which would rank tied for 21st (with Marwin Gonzalez) in 2017, and just one spot below 2017 Marcell Ozuna. As with Pham, Ozuna’s projected 5 WAR seems like a big reach, and while I’d take the under on his 2018 VEB OPS projection, the community seems to be buying that Ozuna’s new approach has truly made him a great hitter rather than simply a good one.
5. Paul DeJong—8% above ZiPS: DeJong might be the toughest player to project on the team just because there is so much variability. He could be 2017 DeJong; he could be “not much of a prospect, just learned a new position, was expected to be a third baseman coming up through the minors” DeJong. VEB doesn’t quite buy him to be 2017 DeJong, losing 37 points off that mark to have an OPS which falls to .820, this clearly exceeds his .756 expectation from ZiPS. If I had to take one, I’d take ZiPS, but it also wouldn’t be totally absurd for DeJong to blow his 2018 VEB projection out of the water.
6. Michael Wacha—8% above ZiPS: Projected as 8% above ZiPS by both ERA and FIP, the community has Wacha with his best ERA since 2015 and his best FIP since 2014. Of course, the same also applies to Wacha’s ZiPS projections. It makes sense, though—Wacha has shown signs of stardom throughout his career and many are waiting for him to make a step towards that rather than continuing down a good-not-great path on the mound.
7. Jose Martinez—8% above ZiPS: The highest projected Cardinals WAR gain (329% of his ZiPS total) from VEB projections, readers are believers that Martinez, whom they projected for an .833 OPS (fourth on the team), is for real. And so am I.
8. Carlos Martinez—7% above ZiPS: VEB loves Carlos Martinez, and the only reason he doesn’t rank higher on this is that ZiPS does. Martinez appears to have moved on from the “they still think he’s a reliever” stage of his career and ZiPS has him as a 3.36/3.57 ERA/FIP pitcher. The community says 3.27/3.20. I think I’d go with ZiPS and still have Martinez as one of the better pitchers in the game. He’s very good. Love and cherish him.
9. Yadier Molina—4% above ZiPS: Appropriately, the longest tenured Cardinal is also the one with the median level of homerism. Fans know who Yadier Molina is, and ZiPS has plenty of data to draw from, and Cardinals fans see the ZiPS-projected .708 OPS hitter as a .738 OPS hitter. Molina also has a still-moderate WAR edge, with a 2.4 vWAR (hey, I’m going to start calling VEB-projected WAR this!) compared to his 2.2 zWAR.
10. Kolten Wong—4% above ZiPS: Wong had his best offensive season in 2017 despite a general lack of power. VEB projects very mild regression, going from a .788 OPS in 2017 to a .786 OPS in 2018. If his vWAR of 2.7, compared to 2.0 zWAR, is any indication, readers seem to believe that his defensive dip in 2017 was an aberration (as do I).
11. Dexter Fowler—4% above ZiPS: This is one where the readers project Fowler more optimistically than ZiPS and I don’t think they go far enough. The readers have Fowler at a .825 OPS following a .851 OPS while battling injuries throughout the season and with a perfectly reasonable, solidly below career totals .305 batting average on balls in play. Give me the over.
12. Jedd Gyorko—3% above ZiPS: That he started to cool down in the later months may have tempered expectations a bit (or at least relative to the rest of the projections). As with Wong, the community has Gyorko’s OPS declining by exactly two points, going from a 2017 mark of .813 to a 2018 mark of .811.
13. Matt Carpenter—2% above ZiPS: No player tracked closer to his ZiPS projection than Matt Carpenter, whose VEB projections have him with 14 more points of OPS, putting him offensively halfway between Dodgers Corey Seager and Chris Taylor. I’ll take it.
14. Adam Wainwright—3% below ZiPS: ZiPS cannot account for velocity drops, at least not as quickly as 2017 Wainwright had his. Whether this drop is permanent remains to be seen, but VEB is at least somewhat wary on the matter, tabbing Wainwright for a 4.40 ERA (ZiPS says 4.09), though interestingly enough a 3.88 FIP which is slightly more favorable than his 3.94 zFIP.
15. Miles Mikolas—4% below ZiPS: One of the trickier players to project is Miles Mikolas. Both ZiPS and VEB agree that he will improve from his pre-Japan form, though neither believes he will be quite as dominant as he was in Nippon Professional Baseball. But VEB is slightly more bearish, predicting a 4.37/4.33 ERA/FIP, where ZiPS is going with 4.13/4.26. He was also the only person in these projections with a projected WAR below that of ZiPS.
16. Luke Gregerson—5% below ZiPS: ZiPS is mostly forgiving Gregerson for his lackluster 2017, pegging him for a 3.34 ERA and 3.49 FIP, and while VEB views him as better than the replacement level pitcher he was, his twin 3.60 ERA and FIP are a step down. Note: These projections were made before it came out that Gregerson will begin the season on the Disabled List.
17. Brett Cecil—15% below ZiPS: I’ll just cut straight to the numbers on this one: ZiPS has Cecil with a 3.26 ERA and 3.02 FIP (both best on the team). VEB has him as essentially an equivalent pitcher to Luke Gregerson, with a 3.56 ERA and 3.64 FIP. If forced to pick one, I’d predict the ZiPS projections come true.

Overall, I think this is a mostly rational set of projections. Slightly optimistic, but none so outlandish that it ruins the sense of realism. This could be about right, and we shall reconvene in a little over six months to determine how we did.