APBRmetrics

What up everybody,

I am in love with the 'Scouting Report' feature on the https://fbref.com/en/players/01c3aff5/Weston-McKennie site and was in the process of making something similar for my site (Crafted NBA)https://twitter.com/craftedNBA/status/1 ... 43744?s=20.

I am wondering what stats you think would be relevant for the offensive and defensive portions of the scouting report. Here is what I have ->

Offense:

• Crafted O Plus Minus

• Passer Rating

• Box Creation

• True Shooting %

• Three Point Attempt Rate

• Load

• Portability

• Free Throw Rate

• Turnover %

• Rebound Ability

Defense:

• Crafted D Plus Minus

• Deflections

• Rim Defense Frequency

• Rim Defense FG%

• Versatility

• Rebound Ability

What stats would you add? What would you take away?


Thanks!

I'd want all the RAPM factors and at least one or more raw stat related to each. Allowing comparison of direct boxscore tangibles and tracking indicators to estimated / so-called black box impact estimates.

Your list gets to raw stats related to each factor except defensive fouling.

How long a list to make is debatable. Depends how hardcore you and the users are.

Very, very good post!!!

But, I have questions due to my lack of knowledge about "basketball language". These are my questions ...

What does it mean "Box Creation"?
What does it mean "Load"?
What does it mean "Portability"?
What does it mean "Deflections"?
Waht does it mean "Versatility"?

About which formula can you calculate some stats, I have some questions too ...

How can you calculate "Load"? "Portability"? "Deflections"? "Rim Defense Frequency"? "Rim Defense FG%"? "Versatility"?

I have arrived to this post becase I was looking information about "How to calculate who is the MVP"? Now, the selection of the MVP is a votation between journalists or fans. But, I would like to create a formula which can give me the MVP, of the regular season, for example.

I have been looking information in basketball reference and there is a prediction about who would be the next MVP https://www.basketball-reference.com/friv/mvp.html or this other article who try to make a prediction of who would be the next MVP https://medium.com/analytics-vidhya/nba ... c6cfe727ee

But my goal is not to make a predicction but to get the MVP of the season through his stats. And here is my doubt. What is the stats to get in consideration and what would be the weight of these stats to get the MVP.

My goal is to get a set of stats and assign a weight of this stats. For example,

USG% * 0.35 + TS% * 0.35 + ORTG + 0.15 + DRTG * 0.15

It's only a proposal to see what I would like to get.

Note: I can use only stats that you can get through the boxscore

I don't want to deviate the original post so if you want that my last question could be better in another post. I can create it without any problem.

Thanks in advace for your appreacited help.

The navigation to the glossary from the menu is messed up but here is the direct link:
https://craftednba.com/glossary


If you use something like
USG% * 0.35 + TS% * 0.35 + ORTG + 0.15 + DRTG * 0.15
you should standardize the values in some fashion (z-scores?).

Crow wrote: ↑Wed Mar 24, 2021 1:50 am The navigation to the glossary from the menu is messed up but here is the direct link:
https://craftednba.com/glossary


If you use something like
USG% * 0.35 + TS% * 0.35 + ORTG + 0.15 + DRTG * 0.15
you should standardize the values in some fashion (z-scores?).

Thanks for the glossary!!!

What do you mean with z-scores?

Look up full explanation of z scores on Google. Compare values in distributions and values in different distributions by variance measured in terms of standard deviations.

Crow wrote: ↑Wed Mar 24, 2021 8:36 pm Look up full explanation of z scores on Google. Compare values in distributions and values in different distributions by variance measured in terms of standard deviations.

Ok!!! Thank you very much for your appreciated help!!!