THE BOOK--Playing The Percentages In Baseball

I would not draw ANY conclusions from the data until a control group is also looked at, as Eric himself suggests.

I wrote this on his web site:

I agree that because Marcels are not nearly perfect and there may in fact be a bias in them (e.g., maybe they under-project home run rate for ALL players who have high HR rates in their most recent sample), using a control group of similar players who were not in the Derby is essential.  I am glad that Eric picked this up.  When I started reading the article, that is the first thing that came to my mind.

Yeah, we need a control group of players with similar AB/HR rates to compare this sample to in order to see if the effect is the derby itself or something else.

This is just step one, so nothing concrete at all should be drawn.

We might see that, when compared to a control group not in the derby, the AB/HR usually decreases while the other production increases.  Or perhaps the opposite will result where the slash line and OPS will decrease but AB/HR will increase.

For now, though, we know that players from 2000-2007 saw their 2nd half-actual numbers exceed projections in BA/OBP/SLG and OPS, while their 2nd half-actual AB/HR was less than projected.

I’m going to revisit it at season’s end so I can incorporate players from 2008 in the sample, extend the sample backwards through, say, 1989 or so to give us a 20-yr period.

I’m currently working to get a database that automates the projection process so I don’t have to manually enter data pre-2004 all the way back for 1986… to get the Marcel projections using Sal’s spreadsheet I had to manually enter all the numbers which took a heck of a lot longer than the actual tests.