Yesterday, I wrote a lengthy comment comparing Mark Reynolds and Chris Davis, noting that Davis could be a very useful player if he made several small improvements in areas such as walk rate, strikeout rate and power production. Based on the data, I guessed that Davis may be the kind of player who shows flashes of excellent talent but never actually puts everything together.
I think most people would agree that that's a reasonable guess. But it seems like that kind of prediction lacks any value without some kind of statistical backing. What if we could take Chris Davis's skill set, find a list of comparable players - call them the Chris Davis Family of Hitters - and see how those hitters' careers played out? How many of those players improved their strikeout rates? How did they perform in their age-26 seasons? Every player is different, but we might be able to learn more about Chris Davis by looking at similar players.
This isn't a new idea, not in the least. Tom Tango used this method extensively in The Book. Bill James wrote an essay titled "The 96 Families of Hitters" in his 2009 Gold Mine (page 23). Dave Cameron recently used a similar approach in assessing the value of Adam Jones's upside.
So, this is probably a good way to think about Chris Davis's future. But I don't have the time, resources and skills necessary for doing this kind of analysis. Why not open up the ideas to the people who can analyze the data properly?
I figured I would start this thread as a place for everyone to ask the baseball questions that they've never been able to answer definitively. Some questions will be able to be answered by linking to someone else's work; others will have to be answered by combing through the numbers.
Note: It's going to be pretty important to give credit if you borrow somebody's work. If you follow through on someone else's idea, link to the source, be it a blog post, CC comment, tweet, etc. Fangraphs and Baseball-Reference have a ton of data, and Stacey can help anyone who needs data from B-R's premium content.