We’ve made what are likely (hopefully) our final methodological tweaks of the 2014 pre-season.
First, we started accounting for pitcher defense (including how pitchers affect the running game) which helps Clayton Kershaw and Mark Buehrle among others. We also tweaked how we project HR’s allowed for pitchers and shifted how we build our pitcher priors (regression means) so that they line up with today’s offensive levels.
We are now projecting a league ERA of 3.95 (the league had a collective ERA of 3.87 last year and 4.01 the year before) and a 3.95 FIP based on 7.56 K/9 (actuals of 7.57 and 7.56 the last two years), 3.01 BB/9 (3.02 and 3.05) and 1.02 HR/9 (0.96 and 1.02). We’re not forcing our averages to match a specific expected league average but it’s certainly reassuring if we’re quite close to what you’d reasonably expect.
The starter who benefitted the most from our tweaks was Clayton Kershaw’s whose ERA fell from 3.11 to 2.93 — thanks mostly to our inclusion of pitcher defense and, to a lesser degree, from our new HR allowed formula (note that while Kershaw’s actual HR projection ticked up slightly his HR/9 relative to the league actually went down). Zach McAllister moves in the other direction, from 4.40 to 4.62. Aroldis Chapman is the winner among relievers falling from 2.11 to 1.93.
As of this morning, we’ve updated our fielding projections. There are two major changes in the way we’re cooking this year’s fielding projections. First, these are now based on more data: 5 past years and the current season (in-season, that is) of UZR data instead of merely 2 past years of UZR’s and the current season. Second, they are regressed towards Tom Tango’s Fan Scouting Reports instead of towards zero. This ends up creating somewhat more aggressive projections. Andrelton Simmons leads the way and is projected to be 21 runs better than the average shortstop. The biggest gainer from the old system to the new system is Brett Gardner who goes from a projection of merely +1 all the way up to +14 (the 4th best UZR projection after Simmons, Machado and Arenado). Check out the table below to see how fielders fared under each system.
Our previous regression levels for hitters were too high and didn’t fully reflect the lower offensive levels we’ve been seeing in the last couple of years. Our new lower levels knock hitters down by roughly 3 wOBA points on average (less for more established players but more for players with less history and thus more regression).
One nice aspect of new lower our projected league offensive level is that our projected offensive level is now essentially identical to both ZiPS and Oliver making all three projection systems directly comparable.
Steamer has officially adopted Fangraphs
Parks Factors, the differences are considerably larger than I might have
expected and I think we can all feel good about this change. Park factors were
never something we had great enthusiasm for, some of our park factors were out
of date and, personally, I am quite pleased to pass the buck and rely on Fangraphs here. Go here to
read up on how Fangraphs creates its park factors.
So, how big are the changes and whose projections changed the
We’ve updated our save projections so that now they take advantage of information from the Fangraph Fan Ballots in addition to the Fangraphs Depth Charts.
Before the update, our system was quite simple: all pitchers who were rated as the leading closer candidate for their team were assigned 28 saves, all #2 guys were given 6 saves and all #3 guys were given 3 saves.
Now, Joe Nathan, Craig Kimbrel and David Robertson lead the way with 35 projected saves (our new maximum) and John Axford, despite being the #1 guy on the Indians depth chart falls all the way to 20 saves, with Cody Allen and Vinnie Pestano each taking 8. Danny Farquhar, the #2 guys for the Mariners, goes from only 6 projected saves up to 15, with Fernando Rodney taking 23. Pedro Strop is projected for 9 saves despite ranking 5th in the Cubs pen.
Let me know if anything looks amiss.