Today I’m introducing a new projections source, which I’m calling RV Current. The idea is to use the same basic projection model I use in the preseason, but also include current year data in the model. The goal is a current “true talent” estimate for each player.
So in addition to taking up to 3 years past data, I also add in current year’s stats. I give the current year weight based on the portion of the season we’ve played, such that after 81 games, I’d count this year’s stats as heavily as last year’s, and by the end of the season this year would count twice as much.
After doing the raw projections, I still make playing time adjustments based on depth charts and injury status. When a player is on an injury report with an estimated return timeframe, I’ll pick a date in the middle as a projected return date. When showing future statistics, I’ll use those dates to adjust projected playing time for players. The depth chart checks try to ensure starters get some minimum projected playing time, while bench players are capped (unless they’re on an injury report). Finally, I check a team’s projected totals to ensure they’re reasonable, both at the overall team level and at a per-position level.
While I have recently improved how I prorate current year statistics, prorated numbers are quite noisy, especially early in the year, due to small sample size issues. Charlie Blackmon is off to a great start, batting .348 with 7 HR and 8 SB, but he’s not likely to keep that pace up; on the other hand, Prince Fielder is off to a slow start, but will likely improve markedly. Simply prorating 2014 numbers assumes both will keep playing at their current rates.
So here are the projection systems I have for Blackmon:
Now Blackmon is playing almost every day, and his rate statistics project better under RV Current than they did before. I’m trying to strike a balance between new information and older data. If anything, I’m still probably too conservative on Blackmon, but in rate terms I’ve gone from having one of the more pessimistic forecasts for him to one of the more optimistic ones. Why is the new model projecting Blackmon with just 8 HR and 9 SB, when he nearly has that many only 6 weeks into the season? Two reasons.
First, what I’m aiming for with RV Current is a “true talent” estimate. To be blunt, so far my model thinks Blackmon has been playing over his head, and he’s likely to regress significantly. He’s not yet a proven .300 hitter, nor has he historically shown all that much power.
Second, the model does see him now as a starter, but it still sees a crowded Rockies outfield, and so it’s not projecting him for all that much playing time. The hot start and early success have moved him up, to just under 300 AB, but with Carlos Gonzalez playing all regularly, and other players like Drew Stubbs, Brandon Barnes, Michael Cuddyer, and Corey Dickerson all on the roster, it’s not clear to the algorithm that Blackmon will still play every day. Sure, he’s currently batting .348, but Barnes is at .351, and Stubbs, Cuddyer and Dickerson are all over .300 right now.
By contrast, Fielder’s projections have hardly budged:
Why? Because Prince Fielder has a long track record of success, and so a 6 week slump really shouldn’t alter our perceptions of his value that much. Oh, and happy birthday, Prince! Today, May 9th, Fielder turns 30.
A player with a bigger boost from preseason projections is Dee Gordon, who won the Dodgers’ second base job this spring:
The newer model gives Gordon more playing time, and bumps up his rate stats and also SBs. Like Blackmon’s projections, these numbers may still be conservative, but as in Blackmon’s case, Gordon has never had played a full season as a regular, so the playing time remains muted.
The player whose numbers improved the most compared to my preseason forecast might be Yankees’ ace Masahiro Tanaka
:
Since Tanaka had never played in the majors before, I only used spring training stats in my preseason projections, and regressed quite highly to the MLB mean. Now, however, Tanaka has shown success against big league lineups in regular season games, and the model takes that performance more seriously.
Here’s a link to a page showing the RV Current projections, scaled to the rest of the season.
I will be updating RV Current daily for the rest of the season, and will certainly be looking at it it when I’m considering trades or free agent pickups.