The Commissioner Spins the Record

You may recall how a few days ago I posted WAR projections using three different projection systems.  The one I liked best was fWAR, as embodied in FanGraphs depths charts. This is the system I use most, because I think it is more reliable than PECOTA (WARP) or ZIPs.  It has nothing to do with the Wolverines coming closest to winning in that projection.

Here is how that projection looks, as a reminder — right side up this time:

FanGraphs (fWAR)

Cottage           104 – 58     (22.8 + 35.9 = 58.7)
Old Detroit    100 – 62     (25.9 + 29.6 = 55.5)
Peshastin          97 – 65     (24.8 + 27.5 = 52.3)
Portland            97 – 65     (22.7 + 28.9 = 51.6)
Flint Hill           85 – 77     (12.2 + 28.2 = 40.4)
Haviland           85 – 77     (17.5 + 22.5 = 40.0)
Canberra           85 – 77     (18.1 + 21.5 = 39.6)
Pittsburgh         83 – 79    (17.2 + 21.3 = 38.5)
Kaline                 83 – 79    (16.0 + 22.3 = 38.3)
DC:                      75 – 87    (15.1 pitching + 14.8 hitting = 29.9 total)

When I posted this I forgot to issue the standard caveats:  in addition to their normal inaccuracies, these projections do not reflect our allocations.  The projections assume we’ll use players the same way they are used in MLB.  We don’t. Most of our teams are better than average, meaning we are more likely to use full-time players in part-time roles than the other way around.  Some of our WAR is wasted on players sitting on benches part-time for us when they are playing for their MLB teams.

Now that everyone has posted their allocations (good job, everybody!  We were all on time!), I thought it would be fun to see how we look just considering our starting players — the 9 starting hitters, 5 starting pitchers, and the “closer.”  Since most of us didn’t do neat clean allocations, I had to use some judgment:

• If the player was the only one allocated 100% to a position, he was the starter.
• If the player was the only one rated as a relief pitcher you had allocated at 100%, he was the closer.
• If more than one player was allocated 100% in a slot, I took the one with the highest fWAR.
• If your allocations were a complex web of part-time allocations to multiple positions (I’m looking at you, Haviland), I did the best I could.  Which includes trying to decipher which of the seven players you allocated 100% to the outfield would actually play, since 5 of them are in the minors. (I decided Sano was an outfielder. Oh, and Soto was your OH, since all your other OH’s were busy elsewhere.)

Here is how our starters — our “core” players getting maximum use — rank as fWAR makers:

Portland	43.3	( 17.3 SP + -0.1 RP + 26.1 H)
Old Detroit	42.9	(17.2 SP + 1.3 RP + 24.4 H)
Peshastin	42.4	(17.9 SP + 1.0 RP + 23.5 H)
Cottage	41.9	(11.8 SP + 1.5 RP + 28.6 H)
Canberra	32.4	(10.0 SP + 2.4 RP + 20.0 H)
Pittsburgh	32.1	(13.9 SP + 0.0 RP + 18.2 H)
Haviland	32.1	(11.1 SP + 0.3 RP + 20.7 H)
Flint Hill	30.9	(7.8 SP + 1.1 RP + 22.0 H)
Kaline	29.1	(12.0 SP + 1.3 RP + 15.8 H)
DC	25.3	(10.7 SP +1.2 RP + 13.4 H)

OK, confession time:  I did this first without the “closer” since some of us don’t think that way about our relief pitching.  But then the Rosebuds were 1.5 WAR ahead of the Wolverines.  That was intolerable.  So I added in the closer position, since the W’s have an actual closer-quality relief pitcher and the Rosebuds’ best reliever is projected for negative fWAR.

One could defend the inclusion of a closer — it’s a “colorable” argument, as we say in the law.  And with that we rest our case.