College Hoops 2k8 Ratings Calculator Project

[bhrangerfan0809]

I've just started working on a ratings calculator that uses Per 40 Minutes and Advanced Statistics from sports-reference to generate attribute ratings for players in College Hoops 2k8. I'd like some feedback since this is very much a work in progress. If anyone would like to collaborate on this, please let me know. Below is a link to what I have so far.

https://docs.google.com/spreadsheets...it?usp=sharing

Player ratings for teams in the Big Ten, Big XII, ACC, SEC, Big East, PAC 12, and AAC are included. I will add more as the offseason progresses. You can find the ratings in the Per 40 Player Ratings and Advanced Player Ratings tabs. Final ratings are in Green.

Here are the attribute/statistic correlations I'm using:

O AWR: OBPM
D AWR: DPBM
Inside Shot: 2P%
Mid-Range Shot: 2P%
3PT Shot: 3P%
Free Throw: FT%
Ballhandling: TOV/USG
Passing: AST%
Blocking: BLK%
Stealing: STL%
O REB: ORB%
D REB: DRB%
Inside Shot Attempts: 2PA/40, adjusted to position
Mid-Range Shot Attempts: 2PA/40, adjusted to position
3PT Shot Attempts: 3PA/40
Foul Tendency: PF/40


Issues:

Ballhandling needs to be adjusted
Inside and Mid-Range shot ratings are still to low for guards

[itzbub117]

Re: College Hoops 2k8 Ratings Calculator Project

This looks great! I have also been working on updating this year's rosters using TarheelFan's (I think that is his username) ratings spreadsheet. There are some flaws in his system but for the most part, I feel it has given realistic results.

I do have some questions/concerns about the formula you use as well. I will use Kentucky for an example because I am most familiar with them. I notice PJ Washington has a 96 3pt shot rating which seems way too high. What stats/math factor the shot ratings? While PJ does have the highest 3pt % on Kentucky's team, he probably isn't the best shooter on the team and far from the best shooter in the country.

I'm also curious of your thought process on determining the Off. & Def. Awareness rating. The formula's I have been using I feel is very flawed for determining those ratings. It basically gives players who get more MPG a much higher rating than players that don't. IMO (using UK again here) Ashton Hagans and Immanuel Quickley are very close when it comes to an overall rating and both are much better than say a small conference/mid major PG. But because of the Awareness rating formula's I have been using, Hagans ends up being much higher rated than Quickley and Quickley has the overall rating of a small conference PG.

With that being said, I really don't know what Off. & Def. awareness affects in-game or why it has the most influence on a player's overall rating.

I've been meaning to make a forum post with a few questions before I got too far into these rosters so instead of making a new topic I will just ask in here:
- There are some attributes I have no clue how to rate (on-ball def, low post off & def, hustle, consistency, stamina, vertical, shot in traffic, shoot off dribble) and I was wondering if anyone had a good way to come up with those attributes without knowing every player and team.
- The are some attributes that I have no clue what they do or how they affect the game/player (hustle, consistency, quickness, awareness)
- What influences a team's unity. I have completed a few conferences and so far Duke has the highest rating in the game at 87 and also have the highest Unity at 63 and the next highest is below 40. IMO, they shouldn't be one of the highest and I have no clue what attributes contribute to a team's Unity.
- What all attributes influence a team's overall Defense rating? Duke also has the highest (which they should be up there) but Virginia's defense is 78 with an 86 Offense. Virginia has one of the highest Def. Ratings in the country and one could argue is the best defensive team in the country so the rating seems low to me.

I'm not very good when it comes to making rosters and determining ratings and I really don't understand the ratings well enough to come up with my own which is why I've been looking for some advice or a good formula to use for each rating lol.

Sorry for the wall and being nitpicky

[bhrangerfan0809]

Re: College Hoops 2k8 Ratings Calculator Project

Thanks for the reply!

3PT Shooting: Yeah, that is an issue that needs to be addressed. I'm just using 3PT% for that rating, but, like you mentioned, with players who have only shot a few 3PA and made most or all of them, they'll have an undue high 3PT rating. I have an idea for that. I could create a formula that would, if the player's 3PA per 40 fall below the average 3PA/40 for all players, it could regress toward the average 3P% rating depending on how far off they are from the 3PA average. That way Washington would only be very close to the average 3P rating rather than having it be an elite attribute.

Virginia's Defense: I think this is where using OBPM and DPBM for O AWR and D AWR comes in handy. Here are the defensive awareness ratings for all of Virginia's qualified players (minimum of 200 minutes played):

Hunter: 85
Jerome: 87
Guy: 70
Diakite: 98
Key: 98
Clark: 85
Salt: 91
Huff: 98

And Gonzaga's qualified players' O AWR ratings:

Clarke: 100
Hachimura: 93
Norvell, Jr.: 99
Perkins: 94
Kispert: 91
Jones: 92
Petrusev: 88
Crandall: 69
Tillie: 93

So Virginia would have a very high team defense rating and Gonzaga a very high offensive rating. Using OBPM and DPBM seems to do pretty well with having strong offensive and defensive teams actually excel on those respective ends of the court in 2k8.

[bhrangerfan0809]

Re: College Hoops 2k8 Ratings Calculator Project

I've updated the Ballhandling metric. You can find it as Ballhandling 2.0 in the Advanced Player Ratings tab.


Old Formula: (TOV/USG Percentile - Inverted)*65+35 = Ballhandling rating

New Formula: (TOV/(USG+AST) Percentile - Inverted)*65+35 = Ballhandling rating


EDIT:

I've now updated the 2P% and 3P% ratings as well. They are now adjusted by the amount of 2P and 3P attempts the player has taken. Players below the 50th percentile have their 3P rating weighted toward the average 3P%, and it is weighted more heavily the lower their 3PA percentile is. For example, Nick Ward of Michigan State had a 99 3P Rating because he was 2/2 on 3PA this year using the old formula. His rating is now 70 with the adjusted metric. These can be found in the L and M columns of the Per 40 Player Ratings tab.

I have also completed twelve conferences so far, so just shy of 2000 players.

[R1zzo23]

Re: College Hoops 2k8 Ratings Calculator Project

Wow, this is very impressive stuff! Love to see people continuing to work on this game! Please god someone give us a roster editor!

[dmx_133]

Re: College Hoops 2k8 Ratings Calculator Project

Great project ;-)

[BaconSunshine]

Re: College Hoops 2k8 Ratings Calculator Project

Was trying to do something similar with current 2k ratings instead of stats. Does anyone have any advice for what ratings to use that don't line up one to one with CH2k8?