Back in late March, as Spring Training was wrapping up, I collated various pre-season projections of the upcoming Cardinal season and formulated my own positional-based projection here. For fun, I’ll walk back through that and compare what was projected against what actually happened.
We will start with the original projection table published in March and add an Actual WAR column next to the Projected WAR column. Note the actual WAR values are through the SF series, so won’t include the last 3 games of the year.
Recall that the Projected WAR was not an extract of individual WAR projections. Instead, I projected a WAR for each position (except for starting pitchers), expecting that we’d see a lot of plug-and-play variations to give young guys runway and the usual assortment of injuries and variation in performance. Even the runway assumption got severely tested as the season went on. If you click the link above, you can navigate to what each individuals’ projections were at the time, buried in Doc Howl’s excellent article coming out of spring.
PositionProjected WARActual WAROver/UnderExpected PrimaryExpected ContributorCatcher2.52.70.2PagesHerreraFirst34.31.3ContrerasBurlesonSecond31.7-1.3DonovanSaggeseShort352Winn??Third30.7-2.3ArenadoGormanLeft3-0.2-3.2NootbaarKoperniakCenter11.40.4SianiNootbaarRight1.50.1-1.4WalkerKoperniakDH21.6-0.4BurlesonHerreraStarter13.53.60.1GrayStarter22.50-2.5FeddeStarter31.50.7-0.8PallanteStarter41.50.6-0.9MikolasStarter511.80.8MatzFill-in Starts11.40.4McGreevyMatthewsRelievers45.31.3HelsleyEveryone elseTotal WAR3730.7Expected Wins8579
First base – The projected WAR of 3.0 was under-stated with Burleson and Contreras (primarily) accumulating 4.3 WAR. The projection anticipated a bit of a time share with Contreras on the heavy end, which is how it turned out. Contreras turned out to be superior defensively, 4th best in MLB as measured by both OAA and FRV. Burleson was -2 OAA in far fewer innings. Below average, but not horrid. Overall, I’d say the first base transition from Goldy to Contreras/Burleson went about as well as anyone could have hoped for.
Shortstop – SS was projected to accumulate 3.0 WAR, with Masyn Winn doing the heavy lifting. As seen in the graph, it was unclear who would back him up. Barrero got first shot. Saggese took the reins late and impressed. In spite of some lingering injury issues, Winn exceeded that with a 3.5 WAR. His replacements, mostly Saggese, added another 1.5 WAR. Sags accumulated 1.1 WAR primarily in September, covering SS when Winn finally succumbed to his ailments. Sags partially benefitted from improved offensive production in September, but his +3 OAA was a bit of a surprise (pleasantly). Small sample size alert here, but it appeared that he is at least adequate to handle SS. All this combined to make shortstop greatest producing position this year for St. Louis.
Starting Pitcher Mathew Liberatore – The original starting five coming out of spring looked to be Gray, Pallante, Fedde, Mikolas and Matz. Liberatore came out of nowhere to lock up a starting spot, so his starter projection really was zero and he delivered 1.8 fWAR. Matz was projected to accumulate 1.0 WAR in that spot, so Liberatore is measured as both a success for delivering more than the projected value, and for being a pleasant surprise.
Fill-in Starters – The projection here was based on typical 6th starter contributions, not really any individual expectation. Impeccable starter health resulted in an unusually low number of fill-in starter opportunities. McGreevy and Matz made all the fill-in starts, accumulating 1.4 WAR in a lower IP count than expected.
The relief corps – Taken as a whole, the bullpen was projected to produce 4.0 WAR. Note that this was my personal projection, which was well over what other systems projected. For instance, Baseball Prospectus had the Cardinals relief corp as full-on replacement level. As the season unfolded, this group has accumulated 5.3 WAR, sixth best in all of baseball. More telling, they’ve done this on considerably less innings than the teams above them (WAR is a counting stat). For example, the Cardinals relievers accrued 5.3 WAR in 565 IP. The next closest team, the Dodgers, accrued 5.3 WAR but took 637 IP to do it. Even more interestingly, this group was almost completely made over from the start to the end of the season. Maton, Helsley and Matz were dealt at the headline. But overall, a very good group.
Left field – This position takes the prize as the worst outcome, as in greatest under-run of actual below projected. Surprised? All the leftfielders combined accumulate -.2 WAR. The best was Nootbaar at .3 WAR, offset by Burleson’s -.4 WAR. Everyone else was replacement level.
Second base – This may also come as a surprise to some folks. Originally projected to be some combination of Gorman, Donovan and Saggese producing 3.0 WAR, this group combined to produce 1.7 WAR. Donovan did his part at 2.6 WAR in 400 PA, but Saggese contributed -1 WAR. It’s really hard to be -1 WAR in 125 PAs, but he managed it. Gorman was replacement level production, which was actually better than the actual replacement. I am scratching my head that Saggese was quite good at Short and approaching horrid at 2B this year. I’m going to chalk that up to him struggling when he first came up (playing second) and then settling in while he was playing Short. Someone will have to talk me into the notion that Saggese is a better SS than 2B.
Third base – As bad as second base turned out, third was even worse. A 3.0 WAR with Arenado carrying the bulk of the load was projected, with Gorman backing him. The average of projection systems had Arenado at 2.8 WAR. I was a bit more pessimistic, figuring 2.5 WAR for him, with another .5 WAR for Gorman in runway mode. Instead, Nolan produced .9 WAR and Saggese added .4 WAR in limited playing time. Gorman was -.6 for a group effort at third base of .7 WAR (against projected of 3.0). Arenado’s defense was rated slightly above average (+2 OAA, 5 DRS) whereas Gorman was -9 OAA. So, the defense at third was negative, net of all events over there.
Right field – This will surprise no one. Low-end expectation was 1.5 WAR and a group led by Jordan Walker combined to accumulate .1 WAR. Replacement level. In practice, the rest of the group (Burleson, Scott, Church, Nootbaar) combined to offset Walkers -1.4 WAR, while Walker got the bulk of the playing time. At the outset, I thought 1.5 WAR was an embarrassing projection. Boy, was I wrong. No, I was right. 1.5 WAR is an embarrassing projection, but indeed, it managed to be worse.
Erick Fedde’s spot in the rotation – A clearly overly optimistic projection of 2.5 WAR for this spot resulted in 0 WAR in 2/3rds of a season. Michael McGreevy helped salvage a bit of the expected value by producing .9 WAR filling in for Fedde late in the season. It does leave one wondering how the season might have gone if management had been quicker to react to Fedde’s performance implosion.
The designated hitter – Originally projected to produce 2 WAR (all on offense, of course), the group produced less than that at 1.6 WAR. This outcome belies the notion that the Cardinals have too many DH-only types. As it turns out, those DH-only types don’t produce in the DH spot. Herrera produced all the positive value with 1.8 WAR in just over half a season at DH (86 games). The remaining cast of characters produced -.2 WAR for the other half. To be fair, Gorman and Burleson produced limited positive WAR in very limited PAs at DH. Pozo, Baker, Donovan, Nootbaar, Contreras and Arenado all came in below replacement level in a hitting only position, again in limited PAs. I think this might call into question the Cardinals approach to allocating DH at-bats to players on a once-in-a-while basis – the proverbial “half day off”.
What went about as expected?
Sonny Gray – Projected at 3.5 WAR, finishing at 3.6 WAR. It probably doesn’t seem that way, since his ERA is higher than desired and he had a bit of a home-run problem, which always makes things seem worse. But the projection was right on.
Mikolas and Pallante – Both were expected to be 1.5 WAR pitchers, and they under-performed by a bit, producing a total of .9 WAR less than projected between them. These guys were never expected to carry a rotation and if they’d pitched to expectation, the Cardinals would have won but one more game. Not a difference maker, although it was frustrating to watch at times.
Centerfield – Expectations, seen as projections, were not high for this position coming in. This position was expected to accumulate 1 WAR, with playing time split between Scott, Siani and Nootbaar. With Scott surprisingly carrying the bulk of playing time, total accumulated WAR has been 1.4 WAR, a little bit better than expected, but improvement is needed.
Catcher – Actual WAR was a bit over the projection (2.7 over 2.5), although how they got there was not at all how it was expected. The projection included some sort of time share with Herrera and Pages. Herrera only played 14 games at C. Totally unexpected. In those 14 games, he accumulated 1.1 of the total 2.8 fWAR. A guy named Pozo accumulated .4 and Pages did the bulk of the catching and accrued 1.6 WAR. Crooks brought the total down with a disappointing -.5 in ten MLB games in September. Rookie debuts can go like this. Ask Masyn Winn.
Some observations about defense
There really aren’t any defense specific projections that I know much about. All we know is what happened. The Cardinals finished tied with Houston for first at +33 OAA. DRS was much less kind at +5, good for 19th in the MLB. You’ll see this 19th number repeat.
Winn and Scott combined to produce +38 OAA. If there was a projection, I’d venture they out-performed what it would have been. Gorman was -9 OAA at third base. I’d say that was a surprise to the negative. Contreras at +6 OAA at first base would seem to be over what a projection would have been.
Pretty much everyone else was average or a bit below, plus or minus. It should be noted that the Cardinals were ~30 OAA at mid-season, so their second half defense measured out markedly worse than their first half.
I note that Cardinal pitchers BABIP for the year was .299 (23rd in MLB). With such a volatile stat, any projection would likely be around league average (.286 this year). So, actual would be a bit worse. Makes one wonder. Was the pitching a bit worse, or was it the defense (outside of shortstop and CF) that made the pitching worse? By FIP (and WAR, of course), the Cardinal staff was 19th (there is that 19th number again). So, by projection one might conclude overall defense contributed to a decline in already poor pitching.
Not shockingly, some things went better than expected, some the same, and some went worse. Such is life in the major leagues. All the projections were provided with the caveat “given health”. It’s easy to see that playing through on-going injuries to Donovan and Nootbaar (and perhaps Arenado, too?) in particular contributed to reduce their WAR, and in too many cases, the “runway” guys weren’t able to improve their performance enough to make up for these injuries. So, depth was exposed as more of a mirage than a strength. An interesting topic for the boards … at what point does playing through injury hurt or hinder the team more than it helps?
One thing about the projections and actuals. The outfield projections themselves were really, really low. And the actuals were even worse. At some point, the Cardinals are going to have to address this issue and acquire players that are expected to produce greater value and then actually deliver. The same thing could be said of the starting rotation. And, of course, this story has been fundamentally the same for how many years now?