David Appelman is the creator of FanGraphs.

16 Comments
Oldest
Newest Most Voted
Bikozu
17 years ago

But still no wOBA?

bikozu
17 years ago
Reply to  David Appelman

Awesome, thanks. No hurry, just curious… My other question would be if you are going to have a park-neutral wOBA number at some point. Inflated/deflated wOBA from extreme parks puts a bit of a limit on the usefulness of the WAR calculations.

Off topic, but I just want to say that this site is amazing. I’ve been coming here for the win probability graphs since 2007 and I never thought it would have exploded into the best single site for baseball statistics as well as a great place to read baseball-related articles. Keep it up.

David
17 years ago
Reply to  bikozu

Is the player wOBA and wRAA calculated using park factors? If it is not, why isn’t it?

David Appelman
17 years ago
Reply to  bikozu

The player wOBA and wRAA is not park adjusted because sometimes it’s nice to see what a player actually did as opposed to what he would have done in a neutral park.

“Batting” under the Value section is park adjusted wRAA, so you have that going back to 2002. We just haven’t gotten around to having a separate park adjusted wOBA stat yet.

bikozu
17 years ago
Reply to  bikozu

Is a park-adjusted wOBA stat in the works?

VoiceOfUnreason
17 years ago

Who did the work to get the bisID’s in place? I wasn’t able to find a download that included them when I looked.

TangoTiger
17 years ago
Columbo
17 years ago

I have a question about how the Chone projections are calculated. I know they look at several years of data and the most recent years having more weight , but how is a player who has been injured such as Michael Cuddyer calculated? Beyond other factors, such as, age and ballpark, is the injured year included in the forecast model?

Rally
17 years ago

The injured year is treated the same as every other year. The CHONE system does not recognize injuries. If I had full injury data, it would help in cases like removing Ben Sheets from the list. For Cuddyer, I don’t think it would make much of a difference.

I could produce one forecast and say what he should do healthy, and another if he’s hurt. Then weight the probabilities of each and come to a weighted mean. I don’t think it would be significantly different from the current output.

The R150 column is park and league adjusted linear weights runs. The forecasts themselves are not adjusted, but context dependent. You can really see the effect if you compare the OPS and R150 for David Ortiz and Ryan Howard, or Mark Teixiera and Ryan Braun, or Prince Fielder vs Miguel Cabrera.

Leagues and parks make big differences.

Columbo
17 years ago

Thanks for the info Rally. Cuddyer was probably a bad example.

Rally
17 years ago

The kind of injury that would “break” a projection system is the type where it affected a player’s performance negatively for most of a season, but has a relatively little chance of recurring. Off hand I can’t think of an example.

Something like Casey Kotchman getting mono in 2006 – it had a major effect on his stats, low chance of recurrence, but then he barely played that year so his 2006 stats have little weight in his projection. At least for rate stats, the missed time might hurt his playing time projeciton depending on the method for those. My method minimizes the impact of catostrophic injuries to playing time – a guy who plays 150, 155, 160, and 2 games will show a larger PT forecast than a guy who plays 110 games every year.

Greg
17 years ago

I’m not familiar with the history of CHONE Projection releases. Will the be any additional updates prior to opening day, or is the update of 2/23 considered the “final” version prior to the season beginning?

Zach
17 years ago

Greg, usually the new releases of projections just mean that recent free agent signings have been included in the projections, so that players’ stats can be adjusted based on their new park and league. I can’t speak for Rally, but I’m fairly certain that players who are not free agents do not have their projections changed at any time after the first release.