Standardizing Advanced Statistics

[Joe Gerrity]

Mason Ginsberg, Hornets247.com

In order to provide a point of reference for the many advanced statistical categories, we have created a chart that differentiates, by position, the good from the bad.

Below is a chart that attempts to standardize some of the advanced statistical categories to allow people to easily distinguish, in a nutshell, what is good from what is not. If an advanced statistic is referenced with which you are unfamiliar, refer to this post to help determine how to interpret it.



Standardizing Advanced Statistics

When utilizing this guide, remember to look at all statistics in the proper context. For example, the listed levels of assist rate are given with their respective positions in mind; while a 55% AR for a PG would be considered elite, an 18% AR for a center would be almost equally impressive (Marc Gasol had an assist rate right around 20% last season). Similarly, while a DRR of 12% would be considered very bad for a big man, a similar DRR for a guard would actually be viewed as a strong mark by most people (James Harden had a defensive rebound rate of just over 12% last season).

As always, leave any questions in the comments section, and we will be sure to address them as soon as possible.


More...

[The JNR]

Looks like you meant to put 2.0 in the WSP48 column and put 0.2 instead.

Otherwise, really awesome chart. Thanks for putting this together guys!

[Pelicans78]

I really like this chart. Can it be stickied?

[Joe Gerrity]

I really like this chart. Can it be stickied?

Done.

Mason actually made it after we had a conversation about how people know off-hand that averaging 20 rebounds or 10 points or 10 assists or whatever is good, but when it comes to the more advanced and arguably meaningful statistics, they're largely clueless as to what is excellent and what is terrible. Hopefully this proves to be a nice reference going forward.

[GeauxHornets88]

Glad people are finding this helpful. Joe, not sure if you can edit the first post, but I corrected the chart to fix the win shares/48 minutes box.

[Joe Gerrity]

Working on it.

[Joe Gerrity]

Done

[FlyGirl]

For those interested, here's another site for some more advanced stats plus/minus stuff.

http://godismyjudgeok.com/DStats/asp...orp/2013-aspm/

[JOE in NO]

Is this basketball advanced stuff in any recommended book form? I like to learn from a book then poke around websites, at least that strategy worked for me with baseball and sites like fangraphs.

[AD23forMVP]

What's better, efg% or ts%?

Looking at Durant last night, he posted a efg% of .425 which would be "poor" but a ts% of .633 which would be "elite."

[42]

TS% includes FT%.

It's not a matter of better. The two stats say that he did really well from the line and got there a ton, but his actually shooting while the clock was running was not good, even giving `extra credit' for the threes.

Now to the data . . . 20/21 for 1, 4/10 for 2, 3/10 for 3.

The stats together give you the rough outline of that picture.