Inspired by Dave’s approach to our pitching staffs — using our allocations to estimate playing time — I’ve generated new projections for our entire rosters for the 2015 season. (Where you hadn’t allocated your team yet, I sort of mocked up an allocation for you. You might do better with your own allocations, so take this with extra grains of salt.) I did it with four projections systems: Fangraphs Depth Charts, Zips, Baseball Prospectus, and Fan Projections from FanGraphs. Then I also averaged them to gin up a 5th set of projections.
Some caveats. BP’s numbers come in lowest, in general, especially with some of its insanely hostile projections of pitchers (Corey Kluber at only 0.8 WAR???). The Fans on FanGraphs come in the highest, especially for the stars who get lots of projections — these are more projections onto the players of fans’ hopes than projections about the players’ realistic talents.
I am not going to take you through all the results for all five systems, which are partly written in pencil in the margins of a piece of junk mail. Here are some highlights:
* On the average projection, all of us project to be at or above .500. At least one of us will win over 100 games. Which will, of course, prompt the Commissioner to look for ways to tweak things for 2016 to bring us back into line with MLB competition.
* The races look tight. Probably they won’t be as tight as various teams stray from these projections in various random directions.
* Given that these projections, if they’re any good, will predict the central tendencies of each team, which would over 1000 iterations produce a bell curve of outcomes around the central tendencies, you only have to project sort of close to the top (I don’t know — 10 games? 15?) to have a realistic chance of winning. (This site suggests teams starting the season projected within 13 games have a 10% chance of winning. I know, the math is different for 9 team-races than it is for 5-team races, but it’s the best I can do in the short time before I need to go to bed.)
I also noticed that each of us will have a favorite projection, and a least favorite.
Canberra:
Favorite projection — BP: 28.0 WAR, good for 6th place, only 0.1 games behind the Rosebuds.
Least favorite — Depth Charts: 34.2 WAR, good for 7th place, only 0.1 games ahead of the Tornados, and only 3 games out of 9th place.
Range of finishes: 6th place, 14 games out to 7th place, 18.6 games out.
Cottage:
Favorite projection — FanGraphs Fans: 57.6 WAR, good for 2d place, only 0.9 games out of first.
Least favorite — ZIPs: 47.7 WAR, good for 3rd place, 3.4 games behind the Alleghenys and only 0.5 games ahead of the Pears.
Range of finishes: 2d place, 0.9 games out to 3d place, 7.9 games out.
Flint Hill:
Favorite projection — Depth Charts: 34.1 WAR, good for 8th place, only 0.1 game behind Canberra.
Least favorite — ZIPs: 34.0 WAR, good for 9th place, 0.2 games behind Kaline.
Range of finishes: 8th place, 18.2 games out to 9th place, 21.6 games out.
Haviland:
Favorite projection — FG Fans: 50.5 WAR, good for 6th place, 1.5 games behind Pittsburgh.
Least favorite — BP: 23.9 WAR, good for 7th place, only 0.6 games ahead of Flint Hill
Range of finishes: 6th place, 12.9 games out to 7th place 18.1 games out.
Kaline:
Favorite projection — ZIPs: 34.2 WAR, good for 8th place, only 2.8 games behind the Kangaroos.
Least favorite — FG fans: 33.6 WAR, good for 9th place, 6.7 games behind the Tornados.
Range of finishes: 8th place, 21.4 games out (or 9th place, 20.2 games out) to 9th place 24.9 games out.
Old Detroit:
Favorite projection — ZIPs: 55.6 WAR, good for 1st place, 4.5 games ahead of the Alleghenys
Least favorite — FG Fans: 58.5 WAR, good for 1st place, only 0.9 games ahead of the Cheese, 5.1 ahead of the Pears, 5.6 ahead of the Rosebuds, 6.5 ahead of the Alleghenys, and 8.1 ahead of the Dragons. Well, actually, this would be a fun race, so maybe it’s my favorite after all.
Range of finishes: 1st place 4.5 games up to 1st place, 0.9 games up.
Peshastin:
Favorite projection — BP: 37.8 WAR, good for 2d place, only 4.2 games behind the Wolverines.
Least favorite — ZIPs: 47.1 WAR, good for 5th place, 0.1 games behind the Rosebuds.
Range of finishes: 2d place, 4.2 games out to 5th place, 8.5 games out.
Pittsburgh:
Favorite projection — ZIPs: 51.1 WAR, good for 2d place, only 4.5 games behind the Wolverines
Least favorite — BP: 35.2 WAR, good for 4th place, 0.5 games behind the Cheese.
Range of finishes: 2d place, 4.5 games out to 4th place, 6.8 games out.
Portland:
Favorite projection — FG Fans: 52.9 WAR, good for 4th place, only 0.5 games behind the Pears
Least favorite — BP: 28.1 WAR, good for 5th place, only 0.1 games ahead of the Kangaroos.
Range of finishes: 4th place, 5.6 games out to 5th place 13.9 games out.
Hmm. Looks like some people depend on projections when building their teams, and others don’t. Some of us (I’m including myself) seek to obtain players with little present value, looking ahead to future value. It’s a risky strategy, but it uses money more efficiently.
Of course the projection systems don’t know as much as they seem to know. How good are they, anyway? It would be interesting to compare your projections with the actual results in October. Can you upload the spreadsheet?