"Success come from better data, not better analysis" D Morey

[@docoolstuff]

Daryl Morey recently made this argument in an article for Business Insider: http://www.businessinsider.com/success- ... sis-2011-8

Agree or disagree?

I see where he's coming from, but would argue that the bottleneck in terms of better measuring player productivity is hardly with the data gathering aspect, but rather figuring out the best ways to apply the abundant data we already have. Moreover, I'd argue that more data is not always better, especially when you begin counting all data equally.

Thoughts?

[PD123]

In his defense, that was written three years ago (i.e. before teams actually had all the data they have now).

Now that teams have the SportsVU data, it's probably slanted back in the direction towards having the right analysts who can figure out what to do with it.

[JimiHendrix]

Now that teams have the SportsVU data, it's probably slanted back in the direction towards having the right analysts who can figure out what to do with it.

Ditto.

Semi off-topic, but it's related: I'd like to see more statistics used for things like this. How does a player perform when they use technique A versus technique B in situation X or Y? I think these types of statistics, since they incorporate basketball strategy/technique, are valuable since it's something that can be analyzed by a "basketball mind" just as much as it can be a statistician, and from a player/coach perspective, it could tailor programs for a certain players. In the case of that article, the stats suggest that Olynyk should focus on incorporating "the hop" on his perimeter jumpers more than "the one-two."

[@docoolstuff]

Awesome links, thanks for sharing.

[Mike G]

http://www.slate.com/articles/arts/book ... f_nba.html

The trick wasn’t just to build a better model. It was to listen both to it and to the scouts at the same time. “You have to figure out what the model is good and bad at, and what humans are good and bad at,”

[Nate]

It's always going to be a 3-legged stool: data, analysis, and implementation.

The best data and analysis in the world won't help if you can't change what you're doing.

[mscbuck]

In response to Morey saying "you have to figure out what the model is good and bad at", I remember talking to Philip Maymin about draft models and he said "If you had a scout that was wrong 100% of the time, you'd pay him millions of dollars because you could just do the exact opposite every time".

There's immense value in knowing what the model is good and bad at. It's all about having a systematic and reliable decision making process year in and year out. You can get a ton of value from a draft-model CONSISTENTLY over-valuing or under-valuing a certain type of player