OK, obviously, I’m way behind in the blogging department, so apologies for that first off. In any case, a bit of news came out today that really floored me and I think has obvious implications for my work. Roger Clemens released a 45 page report today that supposedly uses statistics to show that his sustained greatness over time is not due to steroids but rather due to him being as good as Nolan Ryan and Randy Johnson. OK, those aren’t the words he used, but the clear intent of this document is to exonerate himself from the recent steroid charges from the Mitchell Report.

Now I haven’t had time to go through the report in depth, but upon flipping through it, it looks like the writers of the report use some basic sabermetric ideas. It talks about the effect of run support on Clemens’ won-loss record, it talks about the effects of injuries on his number, and from what I can tell it attempts to use statistics to draw comparisons to Ryan and Johnson. Like I said, I haven’t had time to read it in depth and determine how good the statistical conclusions are (not that I’d be able to anyhow), but more than anything, I think this could be seen as an example of the transition of sabermetrics into legitimate knowledge.

There are a couple of interesting things to think about I think in relation to my work. Read the rest of this entry »