NBA 2k12 Ratings Reasonable Explanation

[Ken North]

Re: NBA 2k12 Ratings Reasonable Explanation

Close isn't mid range. Close is close. There's no way you can call that area a mid range shot.

Just to bring the Dirk vs Kobe discussion to this thread;

Spot Up shooting for non-3PT for Dirk is at 47.6%, for Kobe it's at 43.1% (my previous number was wrong).

[chia51]

Re: NBA 2k12 Ratings Reasonable Explanation

This has been done already many times in the roster section. http://www.operationsports.com/forum...-update-6.html

Here is a picture of what 2k uses for their shooting zones.



And here is the same pic with ~distances

As you can see.. the red section is inside, the next is close, then the blue section is medium and beyond that is 3.
Hoopdata collects 5 locations.

At rim
3-9 ft (not counting at rim)
10-15 ft
16-23 ft
3pters

So with that and with 2k's four zone system something needs to be added together. Initially I thought as you did, to add (at rim and 3-9) needed to be together for inside. But after looking at 2k's zone system it doesn't quite work out too good as there are too many inbetweens in their zone locations.

In real life the paint is 8 ft wide from the center of the goal outwards. Which is pretty close to 9 ft. The free throw line is 19ft from the out of bounds line in real life, while 2k's system for close goes all the way to the ft line. So there are too many inconsistencies with the 2k zone system and real life data points.

Close would extend out to 19 ft as that is how far the FT line is from the out of bounds line in real life. Granted the close range is shorter on the sides in the game but this would still lead to inconsistencies with shots from the ft line.

Either way there is going to be a smudging point as the data from hoopdata does not exactly correspond with 2k's system.

What's important about the My Player map is that it perfectly fits the NBA Hotspots map. So, then you can see how the Hoopdata numbers compare to the NBA Hotspots numbers. Here's an example:

Andrew Bynum (09-10)

NBA Hotspot:
Inside: 341-546
Close: 48-126

Hoopdata:
At Rim: 258-370
3'-9': 96-207
10'-15': 29-71

My method gets you 354-577 for Inside. Pretty close to 341-546. Your method underestimates Inside attempts by 176 (mine overestimates it by 31). Your Close method overestimates by 81 (mine underestimates it by 55).


JJ Barea (09-10)

NBA Hotspots:
Inside: 110-220
Close: 28-68

Hoopdata:
At Rim: 94-171
3'-9': 18-61
10'-15': 21-43

My method gets 112-232 for Inside (again very close to the Hotspot). For Barea, your method does better for Close (underestimating by 7 vs. underestimating by 25). However, you're giving up a lot of accuracy on Inside (underestimating by 49 vs. overestimating by 12).

The best thing I can think to do is use the aggregate data for the entire league and figure out how to best combine the ranges. So, Inside might be 100% At Rim and 50% 3'-9', while Close might be 50% 3'-9' and 100% 10'-15'. That's just hypothetically speaking, but I think it could be done without too much difficulty.

EDIT: I started looking at a way to match Hoopdata to Hotspots. Based on some early analyses (using five teams from the 09-10 season), it looks like you can most closely match Inside by doing 100% At Rim + 75% 3'-9'. Close is best matched by 25% 3'-9' + 100% 10'-15' + 10% 16'-23'. Oddly, 90% of 16'-23' very closely matches the Mid range numbers.

With all of that... you can see that mid is alot farther than what most of you guys are 'thinking' (in the game). Kobe at 78 for mid is fine, if his close is >85 than he will be beast on the block. Also when you factor in SOD and SIT as well as Post Shot he is going to be Kobe.

Kobe is not known for consistently knocking down a 18-20 footer all the time. Please show me Kobe, pulling up consistently from 20 feet and dropping them like Dirk or Ray Ray. Please do.

The problem with the 2k ratings are not the ratings themselves but how they determine their ratings. Some years 2k has not even touched some players while at the same time some players get overhauled. That is the main problem.

When 2k actually does update a player, its usually fairly close to the data at hand.

[B-Ball Life]

Re: NBA 2k12 Ratings Reasonable Explanation

This is crazy. its bothering me not because im a Laker fan but because its just wrong.

2k says that the ratings are based on last years ratings. but somehow last year Lebron and Kobe were both a 97. this year kobe is a 94 and bron a 98 behind jordan. why? because Kobe had realistically had the same season from the year before, so i dont wanna hear that hes geting old thats not a stat. why not rate him the same then wait and see progression if there is a season. but his mid range is far better than 78, and with signs of no off the dribble shot stat how fare is that?

Then he has a generic dunk package? : he should have a sig dunk package and he still does dunk too.

http://www.youtube.com/watch?v=WBDOklrqxvs 2010-11 season.

smhh he had a career season. http://www.nba.com/home/playerfile/kobe_bryant/

78 midrange? he had 36 points vs dallas game 1 and only one shot in the paint. http://espn.go.com/nba/shotchart?gameId=310502013

SHOW SOME LOVE. all this hate killing me. LOL

[Ken North]

Re: NBA 2k12 Ratings Reasonable Explanation

With all of that... you can see that mid is alot farther than what most of you guys are 'thinking' (in the game). Kobe at 78 for mid is fine, if his close is >85 than he will be beast on the block. Also when you factor in SOD and SIT as well as Post Shot he is going to be Kobe.

I'm also not concerned about how Kobe will play. I'm sure he's going to be producing just fine in the game.

Kobe is not known for consistently knocking down a 18-20 footer all the time. Please show me Kobe, pulling up consistently from 20 feet and dropping them like Dirk or Ray Ray. Please do.

Like the two best in the game? No. Better than "slightly above average"? Yes, absolutely. Go get Synergy Sports and look through his spot up shooting.

The problem with the 2k ratings are not the ratings themselves but how they determine their ratings. Some years 2k has not even touched some players while at the same time some players get overhauled. That is the main problem.

The problem with 2Ks ratings are that they're merely based on FG%, which means it doesn't take context into consideration. Which gives us some role players who doesn't face defense with too high numbers, and contested shooters getting too low stats.

When 2k actually does update a player, its usually fairly close to the data at hand.

Yes, but the wrong data. This is why Rashidi's rosters play more realistic than 2K's do. 2K are absolutely addicted to HoopData. Rashidi uses Synergy Sports. His ratings are taken from the right context of the games.

[NY State of Mind]

Re: NBA 2k12 Ratings Reasonable Explanation

Some of you are so bent on your love for people. Comments like why is this guy better then this guy Overall is a useless argument. The higher range of skill sets will always be higher but on paper. Just cause a person or team is better on paper/stats doesn't mean they will win. Just look at these Finals this yr. I know last yrs Lebron had a 70 close and now Kobe 78 mid and I have no problem with either. Some of you are acting like Kobe's mid shot is like Jordans in his final yrs while I agree Kobe is clutch and can make harder shots better then those role players with higher mid which I'm sure is shown in his attributes which will make him still his dominant self...the fact is he shoots a lower percentage....if he improves during a season it will be updated so be happy he has a 99 close cause I don't remember that last yr and I'm sure thats where he was most accurate this past season

[iceberg0982]

Re: NBA 2k12 Ratings Reasonable Explanation

Has anyone thought about the concept of the ratings increasing in the players hot zones?

I hope the ratings increase if they are in their zones.

[chia51]

Re: NBA 2k12 Ratings Reasonable Explanation

I'm also not concerned about how Kobe will play. I'm sure he's going to be producing just fine in the game.

Agreed.

Like the two best in the game? No. Better than "slightly above average"? Yes, absolutely. Go get Synergy Sports and look through his spot up shooting.

That was directed more towards the fellas who think Kobe deserves a 85+ mid range. He doesn't shoot that shot often and he isn't great at it, realistically. Disclaimer: I did not look up his stats on it, but from watching him play he doesn't look to shoot from 20 often

The problem with 2Ks ratings are that they're merely based on FG%, which means it doesn't take context into consideration. Which gives us some role players who doesn't face defense with too high numbers, and contested shooters getting too low stats.

Understand though, that consistency plays a big part on how well the player will shoot even with defense. So someone like (for example) Pape Sy (first name thought of) has a mid of 95 (don't know, but the number is arbitrary) and Kobe has a 78, now everyone gets bent out of shape, but Kobe's consistency is >85 while Sy's is <60 (#'s are arbitrary and used for example) so Kobe will shoot more consistently at 78 with defense guarding while Pape will get a lower shot rating with defense.

[GLZ43]

Close isn't mid range. Close is close. There's no way you can call that area a mid range shot.

You get no argument from me there, Ken.

What I'm saying is that it appears 2k uses the term/verbiage "close" as a subcategory of what is considered "midrange". i.e., it's being used as label, not a definition, same way the word "inside" is being used for what would be considered close shots (by definition). No player in the game has a "Midrange" attribute (the label, not the definition). What they all have is a "close" and a "mid" rating, which, when combined, represent a player's overall midrange game. I think this helps further distinguish players skillsets from one another.

It's just like all the dribbling related attributes (which are labeled better to me). Before it was just "ballhandling", which didn't go deep enough, leading to "off hand" and ball security. Midrange, and shooting as a whole, has been handled the same way: by adding multiple attributes/categories.

[blackbird23]

Re: NBA 2k12 Ratings Reasonable Explanation

But again like in the other thread.... Kobe med range is 78 but he probably has the highest shoot off the dribble and shoot in traffic rating for a shooting guard. While Amare on the other probably has a very high Med range because he actually takes more spot up jumpers compared to Kobe who probably takes and make most of his med jumper off the dribble. I bet you Kobe can hit way more tougher jumpers than Amare in a game situation.

Med range is a base rating. The supporting ratings are SOD and SIT. And that's where Kobe is great at.

Last year in 2k11 I played with someone 2001-2002 rosters. But when I saw Iverson med range rating was 70.... This was after dropping 50 plus with severally times lol... But his SOD and SIT were 99...

this coming from a laker fan, you said it all man the shoot in traffic, shoot off dribble and consistency rating all matter.

[djsider2]

Re: NBA 2k12 Ratings Reasonable Explanation

this coming from a laker fan, you said it all man the shoot in traffic, shoot off dribble and consistency rating all matter.

shoot in traffic and shoot off dribble doesn't matter if you're shooting a wide open jumper. in this scenario, there's guys like Jeter and Gooden that would smoke kobe as a spot up shooter. THAT AINT RIGHT.

There's a reason kobe is never open for a wide open mid range jumper, it's because he'll kill you. a 78 does not reflect that. Same goes for brandon roy. what kind of SIM is it if the game is telling you that if you had a choice between roy and gooden shooting a wide open jumper, you'd choose gooden...

[houston911]

Re: NBA 2k12 Ratings Reasonable Explanation

2k has too many nerds that never played basketball developing the game

these stat geeks understand numbers but they dont know the game

kobe takes more CONTESTED mid range jump shots than just about anyone in the league...where as teams will encourage a guy like amare or lebron to shoot from mid range

you have to factor that stuff into the game. a wide open lebron or amare is not making more wide open mid range shots than kobe, but on the game they do

shoot off dribble and shoot in traffic are not enough to counter the flaw in their system

2k needs to add in "CONTESTED JUMP SHOT RATING" a contested shot is different than a shot off the dribble...shots in traffic refer to shots in the paint by the way

[Ken North]

Re: NBA 2k12 Ratings Reasonable Explanation

Teams encourage Amare and Lebron to shoot from mid range? That sounds rather unlikely.

[houston911]

Re: NBA 2k12 Ratings Reasonable Explanation

Teams encourage Amare and Lebron to shoot from mid range? That sounds rather unlikely.

compared to kobe they do

do you watch their games? those guys don't get defended the same way kobe does. priority number one when defending lebron is keeping him out of the paint

i would link film but nobody posts lebron jump shooting on youtube...you can see the type of shots kobe takes though.

lebron and amare having a higher midrange rating than STEVE NASH....lmao

the rating system is flawed, they need to mix statistics with game watching...their pulling straight stats with no context

[Ken North]

Re: NBA 2k12 Ratings Reasonable Explanation

do you watch their games?

More than you do.

those guys don't get defended the same way kobe does. priority number one when defending lebron is keeping him out of the paint

That's something completely different than what you said, though.

i would link film but nobody posts lebron jump shooting on youtube...

Luckily, they're all on Synergy Scout.

you can see the type of shots kobe takes though.

Yes. Bad ones.

lebron and amare having a higher midrange rating than STEVE NASH....lmao

Over the last season, Nash shot a grand total of 26 spot up jumpers from inside the 3PT line, and hit 16 of them. Distributed over 75 games, that's a fairly low data sample.

Lebron scores 1.01 points per play on spot up jumpers. Compared to his isolation number of 0.92, I think it's fair to say if defenders wants to encourage him to take that shot, he'll be happy to rain on them all night long.

[houston911]

Re: NBA 2k12 Ratings Reasonable Explanation

More than you do.

That's something completely different than what you said, though.

Luckily, they're all on Synergy Scout.
Yes. Bad ones.

Over the last season, Nash shot a grand total of 26 spot up jumpers from inside the 3PT line, and hit 16 of them. Distributed over 75 games, that's a fairly low data sample.

Lebron scores 1.01 points per play on spot up jumpers. Compared to his isolation number of 0.92, I think it's fair to say if defenders wants to encourage him to take that shot, he'll be happy to rain on them all night long.

ive had nba league pass for the past 6 years, the odds of you watching more games than me is very very low

you sound like a stat nerd with zero understanding of the game outside of numbers

wtf does the number of "spot up numbers nash took inside the 3 point line" have to do with anything? we're talking about midrange jump shooting as a whole. if you think nash attepmpted 26 midrange shots over the course of an entire season, then lmao

the fact that you acknowledged that kobe takes "bad shots" shows how stupid your argument is. kobe takes bad shots and still puts up a comparable mid range fg% to lebron, who is taking easier shots.....what does that tell you?

kobes offensive awareness or shot selection rating should suffer, not his midrange. like you said, his percentage is a result of shot selection, not his ability to make shots

anyone who believes that lebron is a better midrange jump shooter than kobe is ignorant