Now Providing Even More Player Info
Posted by Brett on 09/14/2009
Drum roll please...Opponent stats are now available on player pages (go through the Stats Page). And expected stats. Basically, posts like "Chandler and Okafor Comparison" just got a lot easier - for me and for you. An example:
The stats are broken down by position, along with cumulative numbers for anyone who played more than one position. Same caveat as always applies: These are based off lineups - based on the players who were on the court for both teams, my program puts players in their likely positions and grabs the opponent from the same. It does not account for the handful of zone possessions a team might play in a game and it does not factor in when assignments get switched up because of a fast break. But it gives a general idea of whether a man was more effective than average in slowing down the opposition.
Just so we're all clear: Everyone in the league is available. This is exciting to me because previously it took anywhere between 1 and 4 hours to run the numbers for a single player - after some changes to the code, I can run the entire Bobcats team in 90 minutes. Look it over now and get used to it, because it will continue to be available once this season starts, updated on a daily basis. Now I am going to go do my nerd dance in celebration.
Click here to comment/view comments for this post.
| Crow Says: | |
Good job.I like that you provided the data per 40 minutes and per 100 team possessions.I checked a few players and the detail was consistent with 82 games after accounting for its display of the data per 48 minutes. Exception on PER. Theirs are higher than yours. Not sure who is right but I mention it.There have been some unresolved concerns about the displayed PER at 82 games before. |
|
| Mon Sep 14, 2009 10:02 pm |
| Crow Says: | |
Minutes that you assigned a player to each of their positions would be a desirable enhancement. Difference between Actual and Expected opponent production rows would be handy too, as would a minutes weighted player roll-up for all positions. If you are willing to use the space. |
|
| Mon Sep 14, 2009 10:18 pm |
| Crow Says: | |
Err... I guess you already have the roll-up. I assume minutes weighted. |
|
| Mon Sep 14, 2009 10:20 pm |
| Crow Says: | |
Any interest in calculating and displaying the average stats allowed by the opponent of the displayed player when on offense to even up with the detail you now have on their defense? That would help in calculating net counterpart matchup results actual vs expected which is something of an |
|
| Mon Sep 14, 2009 10:32 pm |
| Brett Says: | |
Crow, Thanks for the feedback. I do plan to include player offensive stats by position as well as team efficiencies with the player at each position, including a time breakdown. |
|
| Tue Sep 15, 2009 11:13 am |
| Brett Says: | |
Ask and receive. I have updated the stats pages to include:
Time spent at each position - on the efficiencies tab.
Player stats by position - the position tab.
Net stats by position on the...I think you get the idea. |
|
| Tue Sep 15, 2009 8:39 pm |
| Crow Says: | |
Thanks I plan to use this stuff |
|
| Wed Sep 16, 2009 4:30 pm |
| Crow Says: | |
This was actually the thing I most wanted. I don't see it yet.
Any interest in calculating and displaying the average stats allowed by the minutes weighted / season set of opponents of the displayed player when on offense to even up with the detail you now have on their defense?
To see who faces the toughest defense or easier. Useful in looking at a player and his sub especially in relatively even minute job shares but also at division and conference differences. |
|
| Thu Sep 17, 2009 3:00 pm |
| Crow Says: | |
If the depth of stats never ends I'd think dividing defenders into 2, 3 or 4 quality bands and seeing how offensive players fare against each would be quite helpful, better than relying on overall team quality splits. And do the same with player defensive perform against bands of offensive players. Would help see the highest level that players should be used against and would tip off who might fare best in the playoffs against tougher opponents and whose regular season is too fluffy based on what they did against the average and the weak. |
|
| Thu Sep 17, 2009 3:15 pm |
| Crow Says: | |
The Difference between Actual and Expected opponent production row idea didn't happen yet either. I assumed without checking they were done. Your choice of course. |
|
| Thu Sep 17, 2009 3:25 pm |
| Crow Says: | |
Alright, thanks for the actual=expected difference row.
Also the player pace info and by position . Your database is the only place that is available anywhere.
I'll mention the innovations at APBRmetrics. |
|
| Fri Sep 18, 2009 1:23 pm |
Leave Your Comment
