A look at Association Player Development, problems and possible ideas..

[Madden's Jowels]

Re: A look at Association Player Development, problems and possible ideas..

Agree with everything in here. And the worst part about it for me is that it absolutely destroys stats in the future. The league gets so flooded with talent that nobody stands out as special anymore.

After the 1st or 2nd year of an Association I almost never see anybody scoring 27+ points anymore, let alone a 30 point scorer. I've seen people win scoring titles averaging less than 24ppg. People win rebounding titles with like 10 rebs. It makes it soooo boring because you don't see any special players. Instead what you see is like 50 players scoring between 18 and 24 points a game which is incredibly unrealistic. It kills the entire mode for me. I want there to be guys drafted into the league who are true STARS and put up the appropriate numbers.

After 5 years you have a league with like 100 good players, but no great ones. Their ratings say they're great (because you'll have like 40 players rated 94 or higher) but their actual numbers are run of the mill.

[funnyfiggy]

Re: A look at Association Player Development, problems and possible ideas..

I once saw a team with 3 point guards rated 90+, and 2 of them were under 25. It made me LOL and decide to stop that association.

On an entirely unrelated note, 2k gets plenty of complaints about ratings of their players because they will never be perfect. Can you imagine all the ruckus it would create if players started getting their potentials lowered, especially if it appears to be for no apparent reason?

2k ultimately has to appeal to the casual gamer, and while this may not seem true because most people on this forum are hardcore, it is the truth, and the casual gamer wants to see his favorite team flourish. 2k is really in a situation where they cannot win; either they please the majority of their audience or they please the hardcore gamer too lazy to download rosters that they give you the resources to.

Look at pretty much every popular user-created 2013 draft class. They all have a curve considerably weighted towards the top. Yet these are the ones that most people download and use. I saw one that had more than 10 different players with 5 sig skills in the top 10. I stopped using it because I saw what it would do to my chise, but the majority of 2k's audience wants to see their favorite players become stars.

[mango_prom]

Re: A look at Association Player Development, problems and possible ideas..

2k ultimately has to appeal to the casual gamer, and while this may not seem true because most people on this forum are hardcore, it is the truth, and the casual gamer wants to see his favorite team flourish. 2k is really in a situation where they cannot win; either they please the majority of their audience or they please the hardcore gamer too lazy to download rosters that they give you the resources to.

You have a point there, still there are man issues you could fix without much effort needed and include an arcade alternative for casuals.

Even if they don't improve tendencies/ratings/player types for association (which would be pretty easy by the way), simply by including a "talent level" setting with maybe 2 or 3 options you could keep anybody happy.
I've said it before, generate random draft classes and only use the ones with 0-5 players >70 OVR or manually remove a few top prospects before every season, and even decades into the game you'll get rosters varying from Lakers-like depth all the way to teams with nobody >80 OVR.

That's as simple as it gets, the same with overall realism. Let's say casuals want to see superstars put up real-life stats in 5min qtr games.
That's fine, and to achieve that, tendencies and playcalling pretty much have to be "broken".
You'll notice that playing 48 min games on default settings/rosters. Just saw an asscociation vid of someone playing the Kings with CPU-Thornton taking like 20 3s in a game...
So I suppose the current game mechanics are built for short arcade games, and there's nothing wrong with that. 2k could keep this the same for 2k14, but there could be simple optional fixes you can turn on/off if you like realism.

Let's even keep the current rookie generating logic, but why not include an alternative setting for better ratings distribution? The same way, 2k could leave rosters/tendencies unchanged, but at least include a simulation setting where shooting tendencies and usage settings let us completely decide how CPU teams run their offense instead of hardcoding playcalling decisions and so on.

So I agree with your post, I mean 2ks main objective is to keep EA out of business.
Still, when it's as simple as coding the game to generate fewer >70 OVR rookies to get a realistic league picture in the long run, I just don't see why this hasn't been fixed in 2003 already.
Maybe the bball nerds are on OS/NLSC, and 2ks devs are simply making a game they like that way? Maybe it's a game from casuals for casuals?

[sword1986]

Re: A look at Association Player Development, problems and possible ideas..

I liked the way NBA 2K12 rated players. Not many players were rated in the 80s, a lot of players had 50, 60, 70 overalls and they still played like their real life self. A lot of people complained and now everyone is rated high to please everyone.

[mango_prom]

Re: A look at Association Player Development, problems and possible ideas..

I liked the way NBA 2K12 rated players. Not many players were rated in the 80s, a lot of players had 50, 60, 70 overalls and they still played like their real life self. A lot of people complained and now everyone is rated high to please everyone.

Maybe I'm wrong, but isn't that because 2k changed the formulas for overall ratings a bit? I don't know if individual ratings are that much different.

[thegiantworm]

Re: A look at Association Player Development, problems and possible ideas..

Great thread

My idea is for a flexible ratings system where players, instead of having a set potential, have a somewhat randomized rating, for example, for every association you start Davis's potential is randomized between 80 and 90, or henson's is randomized between 70 and 83, or favors's is randomized between 83 and 88, and make it so the ratings actually matter to players' stats. That way you won't see the same exact guys becoming great in every single assoc you play.

[threattonature]

Re: A look at Association Player Development, problems and possible ideas..

Yeah there should not be a bell curve of any sort for player skills. There should be plenty of randomness from association to association.

I think there should be behind the scenes ratings for something such as work ethic and longevity. Have it so that some players peak very early in their careers and then start to fall off. Some remain the same player throughout the course of their careers and others peak late. And make it random per association.

Have personality play more of a part. Have the players motivation determine their ratings. So that if a player is underpaid he pouts or doesn't give his all so his ratings drop. Or have players whose ratings drop based on if they are on a winning team or not.

Also add in other variables such as players receiving a one year bump during a contract year, or players receiving a ratings drop after receiving a big contract.

Add occassional sleepers in the draft.

The overall goal IMO of all games should have the same ratings spread as there was when the game is shipped. So if theres 10 90+ rated players and 50 80+ rated players than make it the same spread in year 5 and in year 10.

I would also love to see the magic behind the creation of draft classes. Is it all 100% random? IMO they should have just a database or pool of players created by 2K that the different classes feed off of. Just generate random names for each one. That way you'd see more ratings that make a lot more sense compared to if they are just generating completely random ratings for each player.

[mango_prom]

Re: A look at Association Player Development, problems and possible ideas..

I would also love to see the magic behind the creation of draft classes. Is it all 100% random? IMO they should have just a database or pool of players created by 2K that the different classes feed off of. Just generate random names for each one. That way you'd see more ratings that make a lot more sense compared to if they are just generating completely random ratings for each player.

It's not completely random right now, there are a few basic player types, which is better imo than a fixed database. Only the formulas need a few tweaks, but random is great for variety. The last time I've played EAs NHL, I think back with the 2002 version, there were pre-created rookies and it became stale pretty quick.

[alburst]

Re: A look at Association Player Development, problems and possible ideas..

You'll notice that playing 48 min games on default settings/rosters. Just saw an asscociation vid of someone playing the Kings with CPU-Thornton taking like 20 3s in a game...

Yeah it'd crazy I played a game in my hornets association and monta ellis scored 55 pts on 51 FGA's

[mango_prom]

Re: A look at Association Player Development, problems and possible ideas..

Yeah it'd crazy I played a game in my hornets association and monta ellis scored 55 pts on 51 FGA's

Ellis is one of the guys really screwed up with default rosters. Do you play on PC or console and what roster do you use?

[Colts18]

Re: A look at Association Player Development, problems and possible ideas..

Great thread

My idea is for a flexible ratings system where players, instead of having a set potential, have a somewhat randomized rating, for example, for every association you start Davis's potential is randomized between 80 and 90, or henson's is randomized between 70 and 83, or favors's is randomized between 83 and 88, and make it so the ratings actually matter to players' stats. That way you won't see the same exact guys becoming great in every single assoc you play.

Ooooooooo I like that.

[NYR LundqvistSTEPANu]

Re: A look at Association Player Development, problems and possible ideas..

It's funny because I made a HUGE post of this topic a few years back on 2K forums, during 2k11 suggestions for 2k12.

After 3 seasons the entire Asso is already ruined so the best thing you can do is tone down all ratings and potential ratings. I did this and it delayed the process by 2-3 years. But even an 80 Pot rating on someone like Waiters is too much. But somewhere in the 60's & he may not improve at all. So I have to find a middle ground there, which would take some testing.....

True about the drafts though. The post I referred to, my section about the Draft was HUGEEE. It was practically game-breaking to have some goofy, alien looking mental patient with a wild name as the #1 overall pick! Now they appear to follow some kind of formula, but still the drafts still inject too much talent into the league.

I've actually made a couple drafts with these concerns in mind. Names that work. No one gets TOO good unless there is a legit star. A top 10 pick has the ability to get to 84 OVR, average a double-double, and stay roughly at that rating. But only a handful of guys have the chance to reach 80+ OVR star status. Then I throw in a fair amount of role players. Your rebounding/ Post Presence PF/C, either big, or a Rodman/Faried/Bridges type. Your spot up shooter at the 2, 3 or stretch 4. Either straight up a Korver/Novak, or a player with a more all around game ala Ray Allen or (I guess) Peja, (not as lights out of course). Then good wing defenders and athletic "glue" guys with Scrapper, Bruiser, Charge Card and intangible sig skills like that. Also sometimes, I throw in volume scoring 6th Man types in the low-mid 60's with potential to get to mid-high 70's in time.

Also I absolutely mess with their signature animations, hot spots & tendencies, so they play like the roles players I make them to be. So no 2 guards who shoot like a big man, or even worse, never shoot at all & ect. crap like that.

I combine the toned down roster with my drafts, and it's a far more enjoyable experience. But it could still use some editing. I could use 2-3 more drafts . Have only made 3. If anyone would like to try them out or anything just say so and I'll PM you the name to search. Because they honestly work much better if you're looking for a more 'sim' approach to Asso. But ultimately you have to tone down OVR ratings and potential rating on basically everyone who gets too good....which is almost everybody.....Then Asso plays far, far better.

[charley farley]

Re: A look at Association Player Development, problems and possible ideas..

After 3 seasons the entire Asso is already ruined so the best thing you can do is tone down all ratings and potential ratings. I did this and it delayed the process by 2-3 years. But even an 80 Pot rating on someone like Waiters is too much. But somewhere in the 60's & he may not improve at all. So I have to find a middle ground there, which would take some testing.....

I combine the toned down roster with my drafts, and it's a far more enjoyable experience. But it could still use some editing. I could use 2-3 more drafts . Have only made 3. If anyone would like to try them out or anything just say so and I'll PM you the name to search. Because they honestly work much better if you're looking for a more 'sim' approach to Asso. But ultimately you have to tone down OVR ratings and potential rating on basically everyone who gets too good....which is almost everybody.....Then Asso plays far, far better.

Interesting, I've just started testing my own toned down rosters and draft classes. I've taken Slimm44's rosters that already have much more realistic player ratings and Egodynastys draft classes for 2013 and 2014, then using the default generated draft classes as a comparison from then on.

I have massively scaled back the average POT of the player pool. My aim is to manipulate as best I can the broken 2K player progression system.

Broadly, in real life, most players improve up to the age of 24-25, then remain at a consistent level to 30 when they begin to decline and then they fall off a cliff at some stage thereafter. In 2K players seem to continually improve at a fairly consistent rate until they are in their early 30's and then maintain that level untill they retire.

Of the 438 players currently assigned to teams in my association I have allocated the following POT to the following no. of players

POT No. Of Players
95 None
90 4
85 9
80 12
75 47
70 57
65 231
60 10
55 46
50 14
45 7

From my testing it appears that POT of 65-70 results in a static or a small increase, 75-80 means +1-2 OVR per season, 85-90 means +2-3 per season, sometimes 3+.

Some examples:

There is obviously some variation between individual players. Jeremy Lamb and Perry Jones both begin with OVR 70 and POT 70. After 4 seasons PJ is OVR71 and Lamb is OVR 75.

Reducing Kevin Martin and Dwayne Wade to POT 65 results in them beginning to decline slightly more realistically. KM has gone from a 78 to a 67 after 4 seasons at age 33. Wade held his higher rating till he's 32 then began to decline, but much more slowly, reflecting his higher OVR, but also that he's a much more all-round player. Once KM's 3PT shooting goes, well, he has little else.

Reducing Lebron James to a 75 POT stops his becoming an Atomic Powered Superman by 2014 and staying that way till he's 37. After 4 seasons he's still the best player in the league, but at a 95 OVR, not the comical 99 that he had previously.

I have also done the same for the Draft classes. Players with starting OVR of 67, POT 85 progress very well and can end up as 80+ OVR players in 4-5 years. Players who begin OVR 72 POT 75 end up as high 70s/low 80s. The fact is most players from any draft end up as scrubs, role players and bench warmers. The stars should be few and far between and the generational players like James, Duncan, Durant (?) should only come along every few years. The edited 2013 draft class has 1 player with POT 85, 2 with 80, 3 with 75.

Nerlens Noel, begins OVR 73. His original POT of 87 meant that he improved to OVR93 within 3 seasons. With his POT set to 75 he has now become an OVR85 after 5 seasons and OVR 87 after 8 seasons.

All this means that after 4 years, the association has gone from 5 players with 90+ (Paul, James, Durant, Westbrook, Howard) to 6 with 90+ (Paul, James, Durant, Howard, Drummond, Rose). Honourable mentions to Harden (88), Irving, (89) Davis (89) - Note Davis is actually more productive than Drummond, I just think Drummonds stats reflect better on the OVR calculation.

Once the default 2K draft classes kick in it all goes off the scale. Within 2 seasons we have 16 (!) 90+ players in the association. This seems to equate to there being minimum 3 players per 2K draft class reaching 90+ OVR within 3 seasons (often in 2 seasons).

Also, please note I have taken some small liberties with exceptionally high/low rating for a very small number of players. Examples: I think Ed Davis is a monster so he has what most might think is an overly high POT as I personally believe he could be a genuine star level player. Bargnani and Austin Rivers both have comically low POT...

Interested in your thoughts on this. Happy to provide updates and findings and answer questions on both my actual playing Assoc and further sim testing I plan to do...

[sword1986]

Re: A look at Association Player Development, problems and possible ideas..

Maybe I'm wrong, but isn't that because 2k changed the formulas for overall ratings a bit? I don't know if individual ratings are that much different.

They should have kept the formula the way it was.

[mango_prom]

Re: A look at Association Player Development, problems and possible ideas..

@charley farley
I agree with everything. The issue with the current player potential logic is that it's simply broken if you want both realistic young player development and regression for veterans.
Basically high potential is needed so prospects gain skills nicely, but older players don't go down in terms of OVR enough that way.

Associaton mode was designed about 10 years ago and left mostly unchanged, and it shows.
I have to mention Sega's Football Manager series again. Nobody knows that game over here in Germany, so here's a player ratings screen if you haven't heard of it.

http://footballmanagerstory.com/scout/wp-content/uploads/2012/11/olivier-ntcham-fm-2013-future-profile.png

The cool thing about how the game handles player progression is that it actually treats physical attributes differently from mental and fundamental abilities. So you have players in their late 30s without much stamina and athleticism left but with high mental ratings and great skills.
That's much more realistic than 2ks approach of a single potential value, where you'll see players in their mid-30s with prime Lebron athleticism.

It's clearly possible to create realistic player progression, but it has simply been not an issue for 2k. They won't care as long as most complaints are on the level of "omg plz fix the nets shoez" and "we need more dunkz plz"...
Association has been a topic for years, don't expect much to change for 2k14...