BORAS Blowout 2018/19: Intro and Catchers

For the past several Octobers (give or take an extended playoff appearance), the writers here at Camden Depot would put together organizational blueprints for the off season to suggest what the Orioles should do to improve the squad.  That will not happen this year.  However, a major component of that work was our contract projection model: BORAS.

Catcher | 1B | 2B | SS | 3B | LF | CF | RF | DH | SP

In case you forgot or never knew, BORAS takes into consideration several characteristics of a player (e.g., age, offensive performance, defensive performance, historical contract comparisons) and uses that information to project what the new contract will look like.  Each year, we add on another season of data and try to tinker with the model to improve upon it.  This year, the primary model improvement comes from using a batch age model to weight things differently.

Our first series of posts will consider BORAS(po), the position player contract model.  What we found was that younger players and older players were really being valued differently beyond their age.  For instance, our analysis suggests that older players are sought after for their offense with considerably less value placed on their defense.  If we look at encapsulated metrics, defense was valued about half as much on a per run basis as offense was in the 31 and older player set.  For 30 year olds and younger, the value gap shrank from 52% to 17%.  Differences like these can really impact the quality of a model.  By taking these into consideration, we improved the performance of our model internally by about 15%.

In the past, I would just drop the whole player list and then run away.  This year, we will try to be more comprehensive and cover a large swath of the player market as well as putting in some arbitration eligible Orioles because...eh, why not?  Again the caveats remain as they have in the past.  This model does not know about injuries or suspensions.  This model is based on a data set that largely ignores players who get bench level playing time.  Those can be two major blindspots at times.  Finally, our first post here deals with catchers and this model has always struggled with catchers.  In my opinion, this will probably be the worst performing projection group.

BORAS(po) 2018/19 Catcher contract projections

	Years	Total (MM)
Drew Butera		Invite
Robinson Chirinos	1	9.9
AJ Ellis	1	6
Yasmani Grandal	3	45.3
Nick Hundley	1	5.6
Jonathan Lucroy	1	4.4
Martin Maldonado	1	4.6
Jeff Mathis		Invite
Brian McCann	1	7.7
Devin Mesoraco	2	14.9
Wilson Ramos	3	26.3
Rene Rivera	1	7
Kurt Suzuki	2	20
Matt Wieters	1	4.1
Caleb Joseph	1	3.9

The adjustments made this year have resulted in a much more competent model at first look, but there are a couple peculiar projections.  I have a hard time seeing Suzuki making 2/20 even though he has experienced a bit of a late career bloom.  Second, I think the model might be a bit too hard on Wilson Ramos who has had to deal with some misfortune with respect to his health.  I also think Ramos gets dinged a bit due to his poor base running, which I think in general does not impact catchers all that much when it comes to signing a contract.

As this is an Orioles blog, I added arbitration eligible Caleb Joseph.  He should be looking at 1.5 to 2 MM in arbitration while BORAS projects Joseph to be worth slightly more in the open market.  With the Orioles organization in possession of a few decent catchers that would come in under the 1.5 MM mark, it may well make sense to non-tender Joseph even though his arbitration value might be a slight bargain.

Tomorrow, I will post first basemen.