Table of Contents

Friday, January 04, 2008

Data for the masses, by the masses ii

Several folks have posted valuable stat resources around the 'net lately, and I thought I'd quickly link to them tonight:
  • Tom Tango has continued his investigations of positional adjustments using 5 years of UZR data. It's largely confirming the spectrum he's been using up until now, which I profiled here. This spectrum was the source of a lot of debate when I talked about it, so if you're interested in this research, this is your chance...
  • Chone Smith has a great little site here that has (currently) two primary features.
    • First are his TotalZone fielding ratings for all players, 2005-2007. TotalZone is a very interesting fielding system based strictly on Retrosheet data. If you've been reading Dan Fox's recent forays into creating a fielding stat, Chone followed much the same approach about a year ago when he put together his system. At some point, I'll probably run some comparisons between TotalZone, Dan Fox's SFR, and the other various systems like I did here.
    • Second are his CHONE projections for 2007. CHONE is a high quality projection system that is probably on par with ZiPS and perhaps PECOTA. Last year, it was roughly as good as PECOTA at projecting pitchers.
  • Speaking of ZiPS, Dan Szymborski posted his '08 Reds projections last week.
  • And Tom Tango posted '08 Marcels last month. Besides having the most clever name of any projection system, they are also the simplest projection system available. The system uses a weighted 3-year mean of MLB data, a simple aging curve, and regression to the mean. And that's it. The amazing thing is that this extremely simple system routinely performs almost as well as the vastly more sophisticated PECOTA, and often beats out "lesser" systems annual roundups. Anyway, at this point, all we need are the '08 PECOTAs to have the full gambit of projections, and those can't be far off.
  • Dave Studeman has updated his historical win shares database with 2007 data on his personal site. This is a tremendous resource containing all MLB players, past and present, and it seems to have flown under a lot of folks' radar screens. Win Shares aren't perfect, but you can do a lot worse in terms of comprehensive measures of player value.
  • Speaking of Dave, his recent article at The Hardball Times used WPA stats and Tom Tango's With or Without You defensive numbers to demonstrate that Davey Concepcion does clearly belong in the Hall of Fame. But you guys knew that already.
  • Retrosheet released 1956 and 2007 play by play data last month. Still missing 1999 play by play data, but aside from that, their database is incredible.. or will be, once I figure out how to handle it.
  • Finally, this isn't stat related, but Jay Bruce has a blog! Hat tip to Red Hot Mama.
Looks like the Reds acquired some minor talent today. I'll try to do profiles on these guys (welcome back, Jim Brower), plus some analysis of Edinson Volquez and Danny Herrera sometime this weekend. But my blog motivations have been fighting with Guitar Hero III and MLB Power Pros for my evening attention these last few nights, so we'll see. :)

No comments:

Post a Comment