Disappointing Tendency Discoveries Roster Makers Please Read

[Baebae32]

Hey guys. The shot tendencies do not function as you would think. I was getting frustrated with the creation of my fictional roster due to the players not playing true to the tendencies that I had set for them so i decided to do some testing.

ALL TESTING DONE ON DEFAULT 2K ROSTER WITH ONLY TENDENCY EDITS

(1.)I have tested lowering every players "shot" tendency to zero.

-in cpu vs cpu
this didnt matter at all. Player still played as if there was no change. they would take shots that came at the end of plays (3pt point plays, pick and pop). they would take spot up and pull up transition 3's with no hesitation. and of course they would barrel their way into the paint and hit cutter after cutter for layups. again all shots were taken without hesitation.

-in user vs. cpu
there still was notable change aside from the ai hesitating a little to take perimeter shots but they would still take them after a small delay.

This test explains why i have read about you guys that play with the actual NBA players experiencing problems with guys like ben simmons taking way too many shots. The shot tendency does not do anything.

(2.) i have tested lower every player's "shot" tendency AND their "shot location" tendencies to zero

- in cpu vs cpu
the AI defaulted to barreling into the paint and spamming back door cutters but they would still take 3s if coming off a screen and when spotting up in transition.

-in cpu vs. user
this proved my greatest frustration.. the AI is hardcoded to paint spam. In this test, they would drive to basket relentlessly and spam backdoor cutters again. as for taking 3's they would take them if the pass was kicked out to them in half court or pick roll setting. i do have to note that they were very delayed in thinking what to do with the ball if they did not catch and drive immediately but they would still take 3's (curry, durant, thompson)


So in summation,

1. the shot tendencies do not function as we think or if at all. i think that whether a player shoots or not is actually determined by if the play was called for him

2. there is a default code for the ai to barrel its way into the paint and attempt layups with little thought for kicking the ball out.

Feel free to test this on your own in a cpu v cpu or a cpu v user setting. We might be wasting our time with gameplay edits this year.

[youALREADYknow]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

1. the shot tendencies do not function as we think or if at all. i think that whether a player shoots or not is actually determined by if the play was called for him

This is only partially true. Shot Tendencies were split into two types years ago by 2K/VC and they published gameplay blogs describing the changes.

There are Tendencies for the sim engine (Shot/Inside/Close/Mid/3PT) which are used for simulated stats.

The values that are used in gameplay are the action driven Tendencies (Spot Up, Off Screen, Contested, Stepback, Driving Pull-Up, Transition, Driving, Standing).

2. there is a default code for the ai to barrel its way into the paint and attempt layups with little thought for kicking the ball out.

This is also partially true.

I'm seeing the default code in the game which tells a player how to "win" in one-on-one dribble situations (drive forward until reaching a high tendency shot zone or resistance stops the active dribble).

The second part of that doesn't seem to be true since there are also Tendencies and Coach Settings which determine what the player does AFTER attempting to beat the defender.

Just like before, check the action-based Tendencies. Lower the Spot Up Drive, Off Screen Drive, and Driving Layup/Dunk Tendencies for any player and you will rarely see them attempt to dribble drive as long as the Coach Settings are set to a Normal/Balanced/Patient offense and Run Plays is set High (>80).

The Attack Strong on Drive and Dish To Open Man Tendencies also have some impact, but it is rare that there is an actual open man to pass to in this game since help defense (i.e. pre-rotate and help the helper) is so rarely used by the AI/CPU. I've seen a high Attack Strong on Drive make players dribble into a hard drive animation more often while guys dribble with their head up into the lane with a low Tendency set. Tendency Sliders also have an impact here.

As a roster maker, this year's game feels like a long battle against a few decisions and bugs from the devs. There are some behaviors that appear hard coded (ACE play calling, default dribble attack behavior, defensive help behavior, block/charge fouls) and we're trying to minimize the impact to the overall game as much as possible.

[STLRams]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

This is only partially true. Shot Tendencies were split into two types years ago by 2K/VC and they published gameplay blogs describing the changes.

There are Tendencies for the sim engine (Shot/Inside/Close/Mid/3PT) which are used for simulated stats.

The values that are used in gameplay are the action driven Tendencies (Spot Up, Off Screen, Contested, Stepback, Driving Pull-Up, Transition, Driving, Standing).



This is also partially true.

I'm seeing the default code in the game which tells a player how to "win" in one-on-one dribble situations (drive forward until reaching a high tendency shot zone or resistance stops the active dribble).

The second part of that doesn't seem to be true since there are also Tendencies and Coach Settings which determine what the player does AFTER attempting to beat the defender.

Just like before, check the action-based Tendencies. Lower the Spot Up Drive, Off Screen Drive, and Driving Layup/Dunk Tendencies for any player and you will rarely see them attempt to dribble drive as long as the Coach Settings are set to a Normal/Balanced/Patient offense and Run Plays is set High (>80).

The Attack Strong on Drive and Dish To Open Man Tendencies also have some impact, but it is rare that there is an actual open man to pass to in this game since help defense (i.e. pre-rotate and help the helper) is so rarely used by the AI/CPU. I've seen a high Attack Strong on Drive make players dribble into a hard drive animation more often while guys dribble with their head up into the lane with a low Tendency set. Tendency Sliders also have an impact here.

As a roster maker, this year's game feels like a long battle against a few decisions and bugs from the devs. There are some behaviors that appear hard coded (ACE play calling, default dribble attack behavior, defensive help behavior, block/charge fouls) and we're trying to minimize the impact to the overall game as much as possible.

Very informative. This years release just feels very frustrating and confusing with playing against the CPU this year. Between plays missing from the playbooks, ACE not working as expected, and player tendencies, this makes it difficult for roster makers. I'm tired of watching ever CPU player just try to force their way in the paint all game instead of running the offense and star players playing as they should. So as far as editing what is the solution to get the CPU to run more plays for the star players and have the star players take more shots than the non stars.

[Baebae32]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

1. Appreciate the response YAK. Will take what you said into consideration. It was just odd to me the effect that the shot/shot location tendencies had on gameplay.



2. i do feel as if they AI is coded to force passes into the paint rather than opportunities to swing the ball on the perimeter. The constant back door cuts are driving me crazy

3. What is your opinion/solution for the secondary break/early offense execution for the AI. i am getting the same thing every single time.
-if the ai cannot score on the primary break (spot up 3 or forced pass into the paint) the ballhandler will come off of a side pick and roll and try to force their way into the paint. interested to see what your stance is on this. maybe a change with the coach settings can help

[youALREADYknow]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

2. i do feel as if they AI is coded to force passes into the paint rather than opportunities to swing the ball on the perimeter. The constant back door cuts are driving me crazy

I'm seeing the same thing you are. Every action seems like the CPU is trying to find the easiest path to the rim, including dribbling and passing paths.

I've turned the Backdoor Cuts slider down to 15 and seen the constant backdoor passes slow down from defaults. I might even go down to 10. There isn't much help defense movement to create passing lanes on the perimeter, so removing backdoor cuts makes the AI offense run a bit slower than usual since more possessions end in dribble drive or contested shots. I wish there were more play designs with quick perimeter ball movement outside of a few Swing sets/plays.

-if the ai cannot score on the primary break (spot up 3 or forced pass into the paint) the ballhandler will come off of a side pick and roll and try to force their way into the paint. interested to see what your stance is on this. maybe a change with the coach settings can help

Maybe someone else has found a way to control the fast break transition into play calling, but I haven't seen anything yet. The secondary break seems like it's coded logic. Offensive Tempo and Run Plays for the Coaching Profile and POE will make guys shoot or back out to run a set in primary fast break mode. After that, I'm usually seeing them execute their default Freelance set in ACE.

[Baebae32]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

I'm seeing the same thing you are. Every action seems like the CPU is trying to find the easiest path to the rim, including dribbling and passing paths.

I've turned the Backdoor Cuts slider down to 15 and seen the constant backdoor passes slow down from defaults. I might even go down to 10. There isn't much help defense movement to create passing lanes on the perimeter, so removing backdoor cuts makes the AI offense run a bit slower than usual since more possessions end in dribble drive or contested shots. I wish there were more play designs with quick perimeter ball movement outside of a few Swing sets/plays.



Maybe someone else has found a way to control the fast break transition into play calling, but I haven't seen anything yet. The secondary break seems like it's coded logic. Offensive Tempo and Run Plays for the Coaching Profile and POE will make guys shoot or back out to run a set in primary fast break mode. After that, I'm usually seeing them execute their default Freelance set in ACE.

Hopefully a patch can correct the paint pass determination.

[jmarcguy]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

I haven't really played enough yet or looked at tendencies yet. That gets to be a draining process every year. However, thanks to this forum, I usually end up with a solid game. The tools are usually there to tweak the game as needed. Hearing how much is hard coded into the game this year is a downer.

[vtcrb]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

Quick question and may have overlooked it but did you edit Triple Threat, Drive, Attack Tendencies as well?

[youALREADYknow]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

Not sure it's 100% related and probably should go into the Sliders forum, but thought I'd share something that could help us all.

Last year, I found that the Sliders for Shot Tendencies (Inside, Close, Mid, 3PT) played a huge role in determining where AI/CPU players stopped their dribble drives when facing contact from a defender. The actual player's Shot Tendencies (Pull-Up, Driving, Stepback, Contested, etc) determined whether the player actually took a shot along the path.

For some reason I haven't been testing the Shot Tendency Sliders much in 2K19, but tried to use my Sliders from last year (25 Inside, 5 Close, 40 Mid, 60 3PT) and saw the best ball movement on the perimeter that I've seen in 2K19.

The logic is that the Shot Tendency curve tells the AI where to look for action. In theory, setting a high 3PT with a curve going lower through Mid/Close/Inside would stop everyone except players with high Drive/Attack Tendencies from trying to drive.

I'm going to keep testing with the extremes to see if there's any consistency, but it would be great if others gave this a try to see if the AI starts passing the ball around more than just driving to the basket.

[zhenhuay]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

Hi Just try to ask a question regarding tendency in 2k19. I found that as a newly registered account, I cannot create new post yet. Hope someone can help with my question.
Not sure if any of you notice that if you simulate the first season, most players' shot under center tendency will jump to 99 in the 2nd season. Even if you change them manually, it will jump back to 99 again in the 3rd season. My question is if there is any solution for this? And how shot under center tendency influence game play and simulation.
Thanks for your help in advance

[michdb8]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

I will like to know who much affect shots under the basket tendency on the gameplay and sim

I try in the coming days with different setting on gameplay and sim sats to check how this affects.

especially because on m league on every offseason there a tons players who get 99 on that tendencies and after 5 seasons thee like an 80 % of player with than tendencies at 99, so we should check how big deal is this bug on my league for longs terms

[STLRams]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

With 2k/Czar doing some recent updates to the game, has anyone noticed more balanced gameplay from the CPU A.I.? Havent played that much lately but was also wondering has the ACE system been fixed or should we still turn it off to force the CPU to run more plays?

[vtcrb]

Re: Disappointing Tendency Discoveries Roster Makers Please Read

With 2k/Czar doing some recent updates to the game, has anyone noticed more balanced gameplay from the CPU A.I.? Havent played that much lately but was also wondering has the ACE system been fixed or should we still turn it off to force the CPU to run more plays?

I see positives and negatives. I have actually started looking at it differently.... Instead of trying to mirror a particular player or team from Real Life NBA, I changed up approach and just started trying to make it look like a true basketball game instead. I even changed NBA Players names and Tendencies, just to be impartial. And I am enjoying it in the limited testing I have done since Da Czar put out the updates.