Documentary "The Other Boys of Summer” Tells Untold Story of Negro Leagues, Mamie "Peanut" Johnson
Old Dusty Baseball Book Vault: Adair's 'The Physics of Baseball' Delves into the Outer Limits of Baseball

Blue Jays Improved? An in-depth Analysis

By Tammy Rainey
Now that the Blue Jays roster is coming into clear focus, apart from a couple of middle relief jobs, it's time to analyze the extent to which the team has been improved.  Fans of any given baseball team are, by nature, prone to view the potential of the team in question through the lens of their own preconceptions. Some fans are given to negativity and are thus inclined to see dark clouds on the horizon and feel the need to blame management for poor choices that will surely bring about the impending doom. Likewise, the positivity inclined fan can be given to minimizing reasonable caution flags and presuming "best case" outcomes are easy to come by. This reality is what creates the need for more objective, data driven projection systems. While certainly not perfect, such calculations work to filter out the subjective and the emotional to provide a model that at least treats all players the same.
One of my spring rituals is comparing the previous year's roster to the one expected to break camp and seeing what the models project, as opposed to what I, with my acknowledged subjective bias, would project, and then using the former to restrain the excesses of the latter. In that spirit, let me share my results with you.  For the sake of simplicity, I'm going to confine my work to the numbers posted by Fangraphs. And from among those, I prefer the Depth Chart projections since it already serves as an amalgamation of two different projection models (Steamer and ZIPS) albeit adjusted by the staff estimations of likely playing time. 
The only way that seemed reasonable to me to construct a one-to-one comparison between the 2017 Blue Jays and the likely 2018 squad was to align players according to their role on the team. For example, Devon Travis did not get the majority of at bats among the team's second basemen last year, but he was obviously the team's starting second baseman. So I'll compare Travis '17 to Travis '18, rather than using Goins or Barney. Hopefully this will become clear as I go through the roster.
Catchers: The projections were kind to Russell Martin, figuring him for 2.6 WAR as opposed to the 1.8 he accumulated last season. Even the optimist that I am, I wouldn't have expected that and I'll note this again later. The real difference in the position though is the reserves. Five different reserve catchers amounted to a negative 1.5 fWAR last year, and the system projects Luke Malie for an even zero (perhaps generous) and rookie Danny Jansen for 0.5 which, if that held up would be a net swing of 2 WAR just among the relievers. For the position overall, +2.6 WAR ahead.

Justin Smoak - the system sees a decline from 3.4 to 2.2 WAR. That's understandable but possibly a bit severe. For now I'll note the drop of 1.2 WAR and move along
Devon Travis - In a mere 50 games, Travis was credited with 0.6 last season, and the projection is 1.7 for 2018 in twice as many games. This is one of the more volatile spots to project  due to the complete inability to  be confident in any playing time estimate. Still, in 2016 playing 101 games his WAR was 2.6 so this seems light to me. Still, there's a gan of 1.1 WAR per the projections.

TroyTulowitski - Despite totaling 0 fWAR in 66 games last year, DC projected 2.2 this year in just about twice the playing time. That may seem to be a lot but he accumulated 2.9 in a similar number of games in 2016.Obviously that's a net gain of 2.2 at this spot.
Josh Donaldson - In 113 games last year, JD accumulated 5.0 fWAR. Depth Charts expects him to play over 1/3 more games but only projects 6.2 WAR for 2018. I think this is low but note that here's another +1.2 WAR.
Reserve infielders - Here we compare the rather disastrous results of Ryan Goins and Darwin Barney (a combined -1.1 fWAR) with the projections for Aldemys Diaz and Yangervis Solarte (0.4 ad 0.9 respectively) for a net gain of a whopping 2.4 WAR
Kendrys Morales - Swings from a negative 0.6 to a positive 0.6 which, if it played out that way, is a +1.2 WAR net gain.
Right Field - Jose Bautista was scored at -0.5 last year, his replacement Randal Grickuk is projected at 2.0 (and that may well be conservative) for a net gain swing of +2.5 WAR.

Kevin Pillar - Depth Charts projections expect Pillar to rebound somewhat from the 1.9 fWAR he registered in 2017. Though somewhat more optimistic that I would be, the model has him at 2.7 for 2018 which is a net gain of +0.8
Left Field - Last year's pairing in LF was, however unintentionally, Steve Pearce and Ezequiel Carrera. While it was not perhaps your most traditional platoon the bulk of the playing time went to this pair and they combined for 0.7 fWAR. Carrera is gone now, replaced by veteran Curtis Granderson who is expected to be a considerable upgrade in a more traditional platoon role with the lefty-mashing Pearce. The projections have them combining for 2.1 WAR. 

Others - the next three most appearances last season went to Teoscar Hernandez (+0.7), Chris Coghlan (-0.5), and Richard Urena (-0.2) who combined to basically cancel each other out; the projection model identifies Hernandez, Dalton Pompey, and Gift Ngope as the "next three" and there's no gain or loss in this regard.

Overall, on the position player side of the equation, the Blue Jays project to be dramatically better. There's an accumulated +14.2 WAR here and while it's true that it's not as simple as taking last year's win total and adding another 14 wins, it's certainly a portent of a much better season. Moreover, I would suggest that some of these are off a bit. I personally think they are Martin, Morales, and Pillar (by a total of around one and a half wins) but also low on Donaldson, Travis, Girchuk and possibly Granderson (by a combined 3 WAR or so) but everyone will have their own opinions on individual projections.
Turning to the pitchers, the models are in some cases less satisfying but they are not a source of bad news. I'll order them by the role they are expected to play, more so than by the highest WAR contributors, for ease of direct comparison. 

Marcus Stroman - Accumulated 3.4 fWAR in 17, and projected for 4.3 in the upcoming season.
Aaron Sanchez - came in at 0 fWAR last year and is projected for a (far too low) 1.9 in 2018
JA Happ - 2.9 last year and projected to match that figure this year
Marco Estrada - He registered 2.6 fWAR last year in a season that went sideways for a couple of months mid-season, yet is only projected for 1.6 for 2018. Estrada is one of those pitchers that the projection models struggle with and even the folks who run these systems will acknowledge that some pitchers "break" the model.
Fifth starter - the season opened with Francisco Liriano in this role last year, and he's credited with 0.8 fWAR as a blue Jays before being traded. Jamie Garcia steps into this slot for 2018 and the model projects him for 1.5 WAR
Sixth starter - This was and likely will remain Joe Biagini. He was at 1.6 last year (some of which was accumulated in the bullpen) and is projected for 1.1 in the new season (mainly a function of playing time).
Others - Brett  Anderson was the most prolific of an otherwise motley crew that combined for - 0.7 fWAR (that's a total of everyone who made at least two starts though several of them pitched some in relief as well) and DC projects only one other option to acquire positive value as a SP in 2018, that being Ryan Borucki at 0.2 WAR.
As a unit, all these SP combined for 10.6 fWAR last year, and projects to 13.5 in 2018. One could, with some objectivity, argue that both Garcia and Estrada are a half-win too low, if not more, and that a full healthy season of Aaron Sanchez would be worth twice what the model projects (i.e. something similar to what he was worth in 2016) and there's another 3 WAR more or less.
The bullpen is a place where WAR doesn't work very well for middling guys. It's not common really for a reliever to even reach 1.0 fWAR even when he had a relatively good season.
Roberto Osuna - was at 3.0 last year, and though there's no objective reason to suppose a decline the model still projects 1.6 which may be as much as 1.5 too low in my opinion.
Dominc Leone/Seung-Hwan Oh - Since traded Leone accumulated 1.5 fWAR last year, his obvious replacement Oh is projected to only 0.2
Ryan Tepera -  1.0 last year, the model says 0.6 this year. 
Aaron Loup - 0.6 in 2017, projected for 0.5
Danny Barnes - 0.2 last year, 0.3 for 2018.
Joe Smith/John Axford(?) - it's not certain Axford wins this job but he seems to be leading the pack. He is one of a half dozen "other" relievers in competition for the last 2-3 bullpen jobs and all of them have a 00.0 projection. Smith, before the deadline trade, gave the Jays 1.0 fWAR. This contribution, though, is canceled out by Jason Grilli and Matt Dermody combining for a -1.0 total. 

Collectively, the bullpen that is credited with 6.3 fWAR for last year is projected for 3.2 this year. This loss essentially wipes out the uptick in projections for the starters. But here, too, it's easy to quibble. particularly with Osuna. All totaled, using the Depth Charts number you still have a team projected to total 14 more WAR than the 2017 team did. And that, in my subjective view, is conservative by something around 3 WAR. Such a level of improvement, if it played out on the field, would make the team very much a wild card contender and one that's within striking distance of the  Yankees and Red Sox if good fortune is on their side this time. 


Feed You can follow this conversation by subscribing to the comment feed for this post.

The comments to this entry are closed.