Monday, March 15, 2010

Fantasy Baseball Analytics

Reactions: 
Aaron Crow, Royals
2010 Goose Joak Original
by Goose Joak

Nerd warning!

Ok, with that out of the way.

I have assembled a "VORP" system for fantasy baseball.  You basically:
  1. Set team targets for each statistic (i.e. 800 runs, 800 rbi, etc.)
  2. Input projections (I used Chone, tested and regarded as the best of its kind)
  3. Define replacement level at each position (i.e. the mean projections of 10th-16th ranked shortstops). I used ranks from Fantasy Baseball Hot Stove, regarded as the most thoughtful and thorough of all the player ranking systems.
  4. Boost the replacement level performance at each position, and each statistic, to a full slate of games played (I used 159 for hitters, 140 for catchers)
  5. Boost each player's individual forecast to a full season, using replacement players to make up the difference.  For instance, assume Chipper is projected to play 120 games.  His "total" value in my system would be 120 games of Chipper plus 40 games of replacement level (which varies, depending on if you play him at 3B, CI or UT).
  6. Figure out VORP or value over replacement player for each player.  That is, if Chipper's total value is 90 runs, and the replacement level 3B is 80 runs, Chipper gives you +10 runs.
  7. Create a theoretical team of replacement players and deduct that from the goals in each statistic. So say you need 800 runs, and the replacement team projects to 750.  That means you need a total team VORP of 50 runs to make up the difference.
  8. Divide each player's VORP into the total team VORP.  So say Chipper is +10 runs, and your team needs +50 runs.  That means that he nets you "20%" of your total team need in runs.
  9. Average the VORP for each player across all categories, and then rank order them
  10. Compare the net rank of each hitter to the Yahoo! prerank and the Fantasy Baseball 365 average round values, which are an invaluable asset.
Got it?  Granted, it's still highly dependent on projections, and I wouldn't use this tool blindly.  But it does reveal some fun findings.

Here are a few nuggets I can share right now, based on my keeper league format.

The top three players are:
1. Hanley Ramirez (52%)
2. Joe Mauer (38%)
3. Alex Rodriguez (33%)

There is another catcher in the top 15 overall players (I was surprised).

The system does a really good job showing how valuable top tier talent is.  For instance, the true value of Hanley Ramirez is HIGHER than Miguel Cabrera, Grady Sizemore, Mark Teixeira and Justin Upton...combined!  I believe this because my brother had a stars and scrubs style roster last year with his team, and almost won with it, despite having only three hitters who were really above replacement level.  This helps validate his approach and results.

This utility also comes into play when planning your team.  For instance, better to reach on a +20% OF now and save yourself from having to use three spots early on highly touted +7% guys to make up the difference.  This really becomes interesting when you see +7% guys in the early rounds (3-6) and +5% guys at the end of the draft.

Anyway...fun stuff if you ask me!  Now I've just got it convert it for my other draft next Tuesday.  Should be fun!

(And by the way, I don't do this for pitchers.  I pretty much exclusively look at K/9, BB/9 and GB%.  That's about it).

And speaking of which...I am BUYING on the early Strasburg hype.  He induces so many freaking ground balls in addition to the strikeouts this Spring.  I had no idea!

2 comments:

  1. Yeah, I drafted with the Cap'n... He didn't "get it.

    I assumed you would post this, so I drafted Hanley with the #2 pick in the Card Blog League draft.

    Unfortunately, the rest of my picks were a big pile of VORP.

    ReplyDelete