Sabermagicians Discover How to Compare Similar Pitchers, Claim Cliff Lee is Randy Wolf

Well, not really, but kinda.

If you’re not familiar with BrooksBaseball.net, it’s an incredibly useful pitch f/x website where you can find data on just about anything your nerdy heart desires. The site recently rolled out it’s new Similarity Scores (a brief explanation can be found here). Essentially, the goal is to quantify how similar pitcher X is to pitcher Y, given information such as the pitchers’ release points, pitch selection, velocity, and so on. While Brooks Baseball understands that this brand new methodology is still very far from perfect (and probably still far from even being useful), I was curious to see who is similar to the Phils’ rotation according to this algorithm.  Below are the top five comparisons for each starter. Keep in mind that a lower distance value indicates a more similar relationship.

So, I suppose there are a few things to take away from this. Remember, it’s clearly not a performance-based comparison model. The pitches they throw may have similar spin, but Cliff Lee is pretty damn different than Randy Wolf. It seems like Worley and Blanton have some pretty close comparisons. This is consistent with what I expected—they possess pretty average, “normal” stuff, so it seems likely that there would be a higher number of back-of-the-rotation guys that they compare to. As for Lee and Hamels and especially Halladay, the distances become higher. In fact, for Halladay, the similarity scores are so off the charts that the guys behind this thing feel safe concluding that there are no true comparison for Doc. We know this to be true, of course, because medical science has repeatedly proven that Halladay is actually a robot.

Does this provide any real insight? Not really. At this stage, it’s probably no more than an interesting experiment. Even the website itself acknowledges this fact, calling it, “a very beta feature.” Regardless, it’s a neat idea, and we now know that Cliff Lee is Randy Wolf. It’s science.