A deeper look into simulated statistics in Association Mode....

[CWSapp757]

Well I have always been a serious stats guy when it comes to my video games. NBA 2k normally does a decent job when it comes to simulated statistics. I have peeped at a few threads and see that a lot you guys are planning a roster project for this game. So I decided to take my first night with the game to simulate 8 years ahead to see what issues may arise so we know what it is that we are dealing with. I recorded some data that I feel may highlight some problems that we will have to deal with. Unfortunately, I don't know if we will be able to manual fix one of these things. I hope that I am wrong:

One of the biggest changes that I noticed in this years Association mode with the increase in unhappy players. I initially thought it was cool, but then I really started analyzing. Almost half the freakin league has a morale below 50%. Because of this, many overall ratings are decreased and lots and lots of players are being traded away. I noticed this throughout my entire Association (and really hope its not as bad as it seems at first glance) but I didn't decide to document it until 2017 (the year I decided to take all my stats from ultimately. Here is the Denver Nuggets 2017 roster.

Jonny Flynn 85(-1) morale(100),
Mitch Graves 82(1) morale(100),
Monty Ervin 79(-1) morale(100),
Randy Foye 75(-2) morale(90),
Oleksandr Gornev 74(-10) morale(0),
Donnell Reed 73(-1) morale(100)
Glen Davis 70(-2) morale(75)
Donnie Marion 64(-9) morale(10)
Brandon Jennings 63(-11) morale(0)
Ty Lawson 63(-11) morale (0)
Christophe Girbaud 62(-11) morale(0)
Renaldo Balkman 55(-9) morale(0)
Kyle Chambers 52(-8) morale(0)

And this was just the first team I came across. Unfortunately their were teams who had up to 7 players with morale under 50. Interestingly enough, I counted and only two teams had the entire team over 75% morale. I am hoping this can be patched or maybe you guys can do something about this? I'm not sure.


Another thing I noticed was that no generated rookie reached a rating of 90 by 2017 although an average of 3 to 7 rookies came out of each draft with an 80+ overall (This is more than likely due to the huge amount of unhappy players which lead to lower overall ratings). You can take this for what it is worth. Could be a good thing or bad thing depending on how you look at it. Regardless, we can tweak the draft classes to progress appropriately and maintain realistic statistics for each respective category. Which brings me to my next batch of data........

Points leaders at end of 2017
D. Wade 37.3, T. Parker 35.9, C.Bosh 35.1, L.James 31.6, K.Durant 29.5, D. Williams 28.6, A.Jefferson 27.3, C.Paul 26.7, D.Harris 26.4, C.Anthony 26.3, D.Howard 25.9, J.Johnson 25.6, R.Gay 25.5, O.Mayo 24.9, B.Roy 24.9, D.Rose 24.8, D.Granger 24.0, R.Westbrook 23.8, S.Curry 23.7, K.Martin 23.2, B.Gordon 22.0, M.Ellis 21.9, G.Arenas 21.2, A.Iguodala 20.5, L.Aldridge 20.2, A.Stoudemire 20.1, C.Butler 20.0

One thing I noticed about this that bothered me is that all of the individuals in the top 25 (or 30 or however many I listed) are real players. None of the generated players were in the top 25 in scoring even after 8 years of computer generated players entering the game. No biggie though. We'll just have to adjust the tendencies of these incoming rookies.

Rebound Leaders at end of 2017
D.Howard 11.0, A.Jefferson 10.9, C.Bosh 10.8, G.West 10.6, L.Flowers 9.7, B.Griffin 9.6, A.Horford 9.4, A.Randolph 9.4, W.Harvey 9.4, C.Page 9.2, A.Booker 9.1, A.Bynum 9.1, CBoozer 8.9, D.Cassell 8.8

I'm happy to see that some computer generated players are on the list. However, I noticed a steady decline throughout the years as the leader at the end of 2009 had 13.9 rebounds and there were about 12 players over 10 rebounds. Not sure what to change there but ill leave that to you guys much more about this than me.

Assist Leaders at end of 2017
D.Rose 9.3, R.Rondo 9.1, R.Westbrook 8.2, C.Terrell 8.2, C.Paul 7.9, D.Williams 7.8, D.Harris 7.8, A. Iguodala 7.8, B.Roy 7.7, J.Knox 7.6, L.Buckner 7.5, Lebron James 7.1, T.Parker 7.0, D.Toole 6.9, S.Curry 6.5, J.Flynn 6.4, D.Wade 6.3, C.Aldridge 6.0

I feel the exact same about assists as I do about rebounds.

Well guys this is all I have for now. Ill be happy to go more in depth but I'm not sure if anyone is even interested in this info. I suppose ill make the decision to do more tests based on the response I get from this post. Hope someone found this informative.

[fjccommish]

Even with the problems those are overall good stats. They're reasonable for an NBA league, maybe too low for some, but better than 24 PPG from the league scoring leader.

[CWSapp757]

Oh yeah, believe me commish, I think they are wayyyyy better than the "other" game. My boy who I saw at Gamestop tonight was talking about how the stats on that game were so ridiculous that he was taking it back first thing tomorrow. I don't have a huge problem with the simulated statistics. Like I said, I was just hoping to help shed some light on things ya know. I know that this roster project is gonna take a lot of work and man power so I just want to help out in anyway possible ya know.

[emo_nerd73]

more unhappy players= more realistic. just because a team is winning, doesnt mean that every player on that team is going to have 100 morale. thats exactly what happened in 2k9 and it tainted the overall realism of the game. anyone read jordan rules?

question with the stats in association; is the win-loss spread fixed up. cos you notice as you play through in 2k9 that the highest win total for a cpu controlled team was the mid fifties and that was good enough for 1 seed in either conference. then there was like almost 20 teams at the .500 mark. in reality youve got at least one or two teams that hit 65+ wins.

[emo_nerd73]

hold on a second

Quote: Originally Posted by CWSapp757

...Points leaders at end of 2017 D. Wade 37.3, T. Parker 35.9...

im sorry, 37.3 for d-wade is mildly acceptable but like thats bordering on the second coming of wilt.

but tony parker when hes like what 40 dropping nearly 36! wow. i think ill start training for the 2017 training camp cos i probably could make a cut in the squad.

[CWSapp757]

Yeah, I feel where you're coming from emo. I just imagine that it'll get to a point where an overabundance of players are refusing to sign and what not. But then again, I don't know. I mean, its 4:30 in the morning and my brain juices are flowing like they should be so it may not be as bad as I'm thinking.

As far as the win-loss thing, I can give you some simulated stats from 2009 to whenever if you'd like. It'll be interesting to see if they fixed this.

[Whoman]

so i guess it is safe to say that the shot tendencies for created players are way too low again ? Ah well, nothing i can´t fix but still a tad bit disapointing.
Will just go throught the teams and edit the shot tendencies of all rookies again every year and adjust a couple cpu one as well.

This will also help lower the PPG for the league leaders as the generated players take shots and points away from them when you raise their shot tendency (and maybe lower the rela players´ as well) .

[CWSapp757]

Lol, yeah you peeped that tony parker stat. Caught me off guard too when I saw it the first time emo, lol. Don't forget deron williams dropping 28.6 in his early thirties. Somewhat acceptable, but I seriously doubt it.