NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

[ffaacc03]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

And that is a testament of how valuable Da_Czar connection with our community is ...

In the 1st place it is a tremendous added value for us to have him and the devs work on updating teams playbooks and adding real plays after the season started as some of us have already stated before (thus maybe not as loud as we should), this should no be understated.

2nd, to fix the plays that need fixes in such a short time is even more reason to be grateful.

Let’s focus for a moment about what we are getting than about what it was and than about what it is supposed to be, as a way to fully let him and VC/2K know how much we appreciate what they are doing for us all.

We SHOULD KEEP highlighting/comunicating the things/plays that need to be looked at or addressed, as much and detail as we can, but more importantly, we need to STRESS the HUGE, UNFORSEEN GREATNESS that we are receiving, even more than we have had done before. The future is brighter with the direction that they are taking but we should never let the future or what needs to be improved on overshadow the present, let us never take for granted what we are receiving, even if we are paying for their product.

Here is a HUGE THANKS to Mr. President and to all the Devs and staff that has granted us not only hopes of an even better future but that have already granted us an improved experience ... we have been teased, so looking forward to seeing more.

[vannwolfhawk]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

Thanks for the clarification. I would imagine that your list of plays that CPU can run well "on its own," would still be meaningful to me (and others) because these would be plays for which the tuning of CPU off-ball action is optimized. I.e. especially effective off-ball screening for shooters.

But that raises the question: what differentiates between a human player's ability to execute a play, and the the CPU's ability?

Are you referring to plays in which, say, there is action by the ballhandler that is required to be "good," and if, say, a ballhandler does not use a screen properly, the play doesn't get executed well?

If a play gets initiated by a simple pass, for example, CPU should be able to "execute the play" as well as a human player, no?

I'll check out what Da Czar uploaded most recently.

I'm trying to embark on the headache of "charting" plays in verbal, written form. So that I have notes on the various branches. I try to not just describe the actions for myself, for reference, but also call out "features" of plays that are novel, such as the "final" pass to an intended target shooter coming from a big, instead of a PG/guard/wing. So that I can take advantage of stuff like, say, running Jokic or Sabonis and wanting to put him in position to make frequent user of HoF Dimer.

Or note which plays/branches create 3pt looks for players based simply on their position and who the playcaller/target is, so that I can have "3pt plays" for players who don't have the 3pt Play Type. (In MyTeam, we cannot customize a player's play selection/type.)

The headache is especially big with these revised playbooks, because I'll need to list out all the play names first. For what came default in the game, I have an Excel and PDF file from another obsessive maniac who posted his transription work on the Reddit sub (this year and also 2K18).

Awesome! You and me would get along great! Haha! Yes, that’s the idea have the CPU executing at a human level. Check my thread from last year out here at:

https://forums.operationsports.com/forums/nba-2k-basketball-rosters/917653-nba-2k18-real-synergy-scouted-team-specific-playbooks-complete.html

If you look at the spoilers it explains info about the project as well as what you are talking about doing with categorizing each action, play, etc. this was from 2k18 and listed on my lists were ONLY plays the cpu would execute well. I’ve done this since 2k12. It’s a 500-1000 hour project easy with the testing, documenting, & scouting but the end result is worth it.

I went in more detail in that thread, but I’ll give a quick synopsis with testing plays. Go into Run plays mode. I’ll go play type by play type and assign plays to a custom playbook. So I’ll put 50 cutter plays in the playbook and then assign it to a team. Then I’ll go in and go play by play. Select a play then hit demo. Watch the play over and over (10-20x). I’ll rate the play, action, spacing, is it fast or slow in developing in action, how often does the play break or how often does it succeed, and then categorize it. I will then test as a user as well. That’s how I decide what’s good and not. I use all the above. Some plays don’t initiate or start. Some pnr’s have bad roll angles by roller. Some plays have 3 guys in small areas with bad spacing, some plays the timing windows are off. Some plays a screener misses a pick and sets on air. Some plays there are delays from 1 action to the next. Some plays have players criss crossing in front of the main action and causing chaos.

Once you have tested and documented every play it’s easy to watch a game, see an action, and grab it and put it in their playbook. It’s just knowing what you want and what your looking for.

All these things are crucial in having the CPU and AI playing well. It’s night and day with how the CPU executes. 1 thing I forgot to add that is a huge importance is adding as many plays as possible with branches and multiple options in it. This way the CPU doesn’t break down. Even a simple play as a simple pin down with a pnr option behind it. I’ll use that as opposed to the simple pin down with nothing after the pass is made.

I did the same thing as you are talking to get guys like Jokic and A. Sabonis playing as they do through plays. Same thing with 3 pt. Plays without the 3pt play type. A good example of this is 08 punch 15 Fist. I think is name? It was in 18 and is in 19 as well. It’s a simple roll and replace play, but the stretch 4 will end up with a 3 quite often but the 3rd option after the pnr, throwback to stretch 4, is the high low look.

Hope all that made sense and helps with your project!

[Bexthelegend]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

Hey Vanwolf, I got a question. If you want a cpu controlled team to run primarily a 1-4 pick and roll ala Stockton-Malone. What playbook is best for that?

[youALREADYknow]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

YAK, hopefully you chime in here on the importance of testing plays and the impact plays have to CPU AI execution to create sim experience. It can’t just be me and I know we think alike on this. They need to hear it. You & I hang out on roster forums where most get it and understand it there but main forum there are a lot of people who don’t. Might be good to state your stance as you know this game inside and out and do great work and you are very thorough in testing. Great advocate and voice in this area. Hopefully you can chime in...

I’ve touched on it in the SIMNATION thread, but this game suffered through 3 months of almost unplayable sim basketball because ACE wasn’t customizable this year. We need the ability to configure Series, Freelance, and Set Plays at the Coach level in the Roster. Even with the great new plays added to Playbooks, it was actually some of the ACE player specific and team specific changes that had the larger impact over the past few weeks in my opinion. AI teams were just cycling through Series and Iso before these updates regardless of what was in the playbook. The ability to update these systems without game patches is a gift for the community. I just feel bad for Czar and the team because it’s too much work for them to take on while focusing on other areas of the game.

As for play testing, that’s as core to gameplay as any aspect of the game. There have been past 2K releases where up to 25% of the plays in the game have had broken/misconfigured actions. Czar and the team have done a great job with branching to limit the impact of one bad action, but I’ve been told they need to re-do play design every release which is a nightmare. That means the product team has decided to put more pressure on QA/QC to test these plays to ensure they run properly. That’s not an easy task since it takes some knowledge of real world basketball concepts and the underlying play system.

Vann and I have probably spent thousands of hours combined in Practice Plays and game modes just running through sets. That’s time wasted from enjoying the user experience and it can’t be the consumer’s job each year to discover those issues.

Hopefully the community recognizes the difference between gameplay now and before these playbook and ACE updates. I’ve played some of the best games ever in this series over the past week. It’s not even a similar game to the one that was shipped. Thank you to Czar and let’s keep it going for the rest of the league.

[alabamarob]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

As much as I hate to say this, since i thumbed my nose for years. Tendencies, rosters and playbooks affect the realism of the game more then anything. I played the 76ers about 10 times straight and watching Ben Simmons lead the team in shot attempts every game is a game killer. They run post ups and clear outs for him all game, and Embeid and his 63 o board rating eats only off of put backs.

It would nice to see Butler and Embeid get more shot attempts and actions ran for them.

[JoFri]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

The mav playbook is horribly unreal, just like the spurs playbook out of the box.
Luckily, there are two valid plays.

CUT LOOP 42 DIVE (Play Type = Cutter)
QUICK 43 GIVE LOOP (Play Type = 3 PT)

Both plays can be assigned to Mathews.

https://youtu.be/SDgmhLSAEyU

https://youtu.be/P8JUMg0f8Nk

There is a base set run by the Mav this year that looks like the Piston Snap action, but due to the placement of the centers and the cut actions, they aint the same at all.

Hope Czar can give some love to the Mavs, give them a series!

https://youtu.be/VfKOjfDkQVk

[JoFri]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

The lakers must have stolen this play from Rick Carlisle. Saw them run it in the Mavs game.

https://youtu.be/qcl60KvZcHw

You will need to add this play to the lakers playbook as they just run it recently.

PUNCH LOOP 25 (Post Up Low)
Assign to Ingram and have Lebron play as the power forward

[JBCAROLINAKID]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

Vann do u have a list of plays to keep for the teams czar has edited so far so I can keep only the ones that actually work?

Sent from my SM-T350 using Operation Sports mobile app

[Smirkin Dirk]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

The mav playbook is horribly unreal, just like the spurs playbook out of the box.
Luckily, there are two valid plays.

CUT LOOP 42 DIVE (Play Type = Cutter)
QUICK 43 GIVE LOOP (Play Type = 3 PT)

Both plays can be assigned to Mathews.

https://youtu.be/SDgmhLSAEyU

https://youtu.be/P8JUMg0f8Nk

There is a base set run by the Mav this year that looks like the Piston Snap action, but due to the placement of the centers and the cut actions, they aint the same at all.

Hope Czar can give some love to the Mavs, give them a series!

https://youtu.be/VfKOjfDkQVk

Ive been using the Piston snap with Dallas, but yeah it's not perfect.

If he gives them a series, it should be that action where a guard runs an Iverson cut in a wing pick and roll, has the choice of going either side, or kicking to the other wing where the other guard can run a pick and roll. So many options, seems perfect for a series. They've been running it for years.

1.40-2.05 of this video.

https://<iframe width="560" height="...reen></iframe>

[JoFri]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

I will suggest that they copy piston snap, modify it and call it Mav Snap. They already have Bos Snap isnt it? Lol


Sent from my iPhone using Tapatalk

[ffaacc03]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

Havent played nor been followig Mr. President as much, but does anybody knows if there have been any further updates introduced lately ?

I know it is not as easy or fast as we would hope, just to know if there are any more of them.

And again, cant stress enough how much I/we value these playbook updates !

[Smirkin Dirk]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball



Just played Houston, I was helpless.

They got 54 points from 3s.

72 in the paint.

7 FTs.

Which meant they only got 6 from the mid range.

I play the Spurs in my Myleague next, might post some of their highlights.

[Smirkin Dirk]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

I just started playing around the Bucks.

The 5 out plays look great until Lopez (or their 5) sets a pick. Instead of popping out to the arc, most of the time, he dives to the hoop, and junks up all the space. It defeats the purpose of having plays designed for Giannis to operate in space.

[harryl]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

I did the same thing as you are talking to get guys like Jokic and A. Sabonis playing as they do through plays.

Here's something you might find amusing: in MyTeam, in 2K17 and 2K18 (and seemingly 2K19), 2K gave Arvydas Sabonis play types as if he were, say, Magic Johnson: P&R Ballhandler, ISO, and Guard Post-Up.

Kind of interesting to run him as the initiator this way, but obviously very kludgy to try to have him use screens in FIST plays.

Last night I started digging into the Suns PB and while I knew this PB had some plays unique to it, I immediately encountered a handful of plays that are "weird," in interesting and fun ways. Didn't pay close attention but one action in a couple plays immediately jumped out at me, in which the initiator passes to a 4 around high post, and the 4 then performs a DHO with a wing cutting around from weakside wing or corner or something like that.

By the way, in case it might be helpful to you, over on 2KGamer.com, we have a member who made up a PDF with a breakdown of each PB, after the update:

https://2kgamer.com/t/all-of-the-current-teams-playbooks-and-most-common-plays/25860

Direct PDF Link: https://drive.google.com/file/d/0B80...HZmloM0dV/view

[Mikelopedia]

Re: NBA 2K19 Playbook Changes Are a Significant Test Case for Simulation Basketball

Have more teams been updated or did this project go the way of MLO stability?