No replies to this topic

[Sir Thursday]

I've been following some work done by a poster on Cavs: The Blog in which he uses regression analysis to determine which if any of the physical measurements of draftees available before a draft correlate with success at an NBA level. Here is an introduction to his methodology, and he put up the first post in the series yesterday.

Now, I find this interesting because it seems like the sort of thing the Rockets would be all over. There are clearly some strong correlations between certain physical tools and NBA success, but they are not necessarily the ones that GMs would have traditionally slathered over. Somehow I imagine that it would be much harder to obtain the sorts of proprietary statistics that Morey and co. favour from College ball given how fragmented the scene is, and the alternative is this sort of thing.

His post about point guards is one that we had a stake in over this past offseason with the signing of Scott Machado (admittedly, we didn't actually draft him, but given that we've picked him up as a rookie I feel like it's justified to include him in this conversation). According to draftexpress.com, Machado posted a 3/4 court sprint time of 3.23 seconds, which places him just outside the core group of <3.2s upperclassmen PGs that this analysis predicts will do well in the NBA. If Machado's performance matches up with the regression, you would expect him to contribute around 1 OWS next year. For comparison, that's somewhere between John Lucas and Beno Udrih, both of whom are competent but unspectacular. I'd take that over Douglas based on how he's played so far though!

Anyway, I thought this post and the series as a whole seems like a really insightful piece of analysis and that it was worthy of a wider audience. What do the rest of you think?

ST