clock menu more-arrow no yes mobile

Filed under:

Steamer Projections: Underrating the D-backs offense?

The Steamer projections are out, and on first glance appear not to be too optimistic about the chance our offense repeating.

Joe Camporeale-USA TODAY Sports

Here are the OPSs projected by Steamer for the top men in the offense, along with the expect WAR . It presumes 600 plate appearances for everyone except catchers (450 PA), which I kinda like, because it makes it easier to see at a glance which players are going to be most "valuable". Along side are the actual OPS, PA and fWAR put up by the players on the team this year.

2016 2015
Paul Goldschmidt .931 4.8 695 1.005 7.4
A.J. Pollock .768 3.4 673 .865 6.6
Jake Lamb .739 2.4 390 .716 2.0
Welington Castillo .717 1.9 303 .813 1.7
David Peralta .782 1.7 517 .893 3.7
Ender Inciarte .702 1.6 561 .747 3.3
Jarrod Saltalamaccha .698 1.2 194 .805 1.1
Peter O'Brien .744 1.0 12 1.300 0.2
Nick Ahmed .635 0.8 459 .634 1.7
Phil Gosselin .682 0.7 76 .927 0.5
Chris Owings .655 0.3 426 .707 -1.3
Yasmany Tomas .728 0.1 552 .587 -1.4
Aaron Hill .688 -0.1 353 .640 0.1
Brandon Drury .666 -0.6 59 .629 -0.2

I suppose it is in the nature of projections to expect regression, but it seems a one-way track, with almost everyone putting up WAR at a lower rate than they did this season. The two exceptions are Tomas and Owings, who are expected simply to be replacement level, rather than significantly below that, as they were measured by fWAR this year. Certainly if Goldschmidt puts up anything below five wins in 2016, that is going to seem like a disappointment. But considering Steamer inexplicably value him in negative territory on defense, I'm not sure how credible this system should be considered. Wish I'd kept the 2015 projections to hand to see how accurate they were(n't).

Welp. In the 24 hours since I wrote the above, they appear to have updated the stats to included projected playing-time. No, I am not going to go back and copy all the revised figures into the table above. Somehow, civilization will cope, I feel. On the other hand, from shoe's link, I found the 2015 Steamer projections, but they're in CSV format, so will require more analysis.


The boot seems on the other foot with regard to pitching. This was a major issue for the Diamondbacks, who came 27th in the majors for fWAR by their pitchers, managing a total of only 8.1 fWAR. The top five starters and seven relievers listed by Steamer total 11.0 WAR, so about three wins better.  And that discounts entirely any new additions, or contribution from Aaron Blair who is a) inexplicably listed as a reliever, and b) is outside the top dozen, at -0.1 WAR. I probably won't be the most disappointed here if that turns out to be true, but I expect to be among the top ten... Here are the numbers for those top 12. Now with projected inniings!

2016 2015
Patrick Corbin
85.0 3.60 1.5
Rubby De La Rosa
188.2 4.67 0.3
Chase Anderson
152.2 4.30 1.6
Archie Bradley
35.2 5.80 0.0
Jeremy Hellickson
93 4.09
146.0 4.62 0.8
Robbie Ray
63 3.90
127.2 3.52 2.1
Daniel Hudson
65 3.16
67.2 3.86 0.7
Zack Godley
64 4.42
36.2 3.19 0.1
Brad Ziegler
65 3.40
68.0 1.85 0.6
David Hernandez
55 3.52
33.2 4.28 -0.2
Enrique Burgos
35 3.55
27.0 4.67 0.4
Silvino Bracho
20 3.00
12.1 1.46 0.1

Some of the projected innings seem a little off; though I imagine Hellickson's innings are affected by the injury which took him out for much of 2014, and pretty sure Ray will end up with more than 63 innings, so you can likely push that projection to 150-odd innings and around two WAR. Anderson and De La Rosa are both projected to have a better ERA than they did this year, and Hudson is projected to be the most valuable of the relievers. Between him, Burgos and Bracho, I think we are more than adequately covered for potential closers in 2015 from internal candidates, which is one of the reasons why we should not be shopping for one on the trade or free-agent market.

Of course, these are all only projections, and should be treated as such, not gospel. As shoe mentioned, the MARCEL projections are also out, though I haven't found a location for those on a single page. However, a study on the 2014 data showed Steamer to be the most accurate in terms of predicting younger players; given where this team is coming from, that should perhaps be borne in mind.