NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

[LorenzoDC]

<iframe src="https://www.youtube.com/embed/oRhIVXsBVZE" allowfullscreen="" frameborder="0" height="315" width="560"></iframe>

This is the October 31, 2018 roster for the CPU controlled Cavaliers in MyLeague. There are other teams overloaded like this with centers, and not enough teams with depth at SG especially. So through a combination of drafting, trading and signing players, this is the result.

The CPU is probably using logic that combines ratings with roster balance, but for some reason this league year is completely packed with high rated centers. As a result, without intervention, there are teams without enough PG's, SG's and SF's to fill out a season's game rotations.

I used CPU generated draft classes so if the league is overloaded at one position then this is also a matter of CPU logic for the mode.

UPDATE: This problem seems to be greatly minimized by using my MyLeague slider set found here. It's not completely solved, but it gets down to where managing it with some user rosters edits here and there every year is not a big job.

[AIisBroken]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

Not at all surprising but still pretty LOL bad

[trekfan]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

I wish I could say I'm surprised, but that would be a lie.

#30teamcontrolFTW

[apollooff320]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

Game just feels like it's in beta mode still. I'm seeing a ton of issues that should not been over looked in testing. The CPU is not affected by a lot of things while the user is, and even testing out the trade logic before I seen this thread left me shaking my head. Like really how can you let stuff get back into the game u fixed the following year? I'm glad I rented the game from Gamefly this year.

[LorenzoDC]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

Game just feels like it's in beta mode still. I'm seeing a ton of issues that should not been over looked in testing. The CPU is not affected by a lot of things while the user is, and even testing out the trade logic before I seen this thread left me shaking my head. Like really how can you let stuff get back into the game u fixed the following year? I'm glad I rented the game from Gamefly this year.

I see no improvement in trade logic either.

I've gone back to 30 team control with trade approval on, and I auto-deny all of them.

Some of the ones that allow me to negotiate - and I have no idea why some do and some don't with 30 team control - I check into. Some of them, maybe 1 of 3, make some sense. The rest make no sense at all. I mean, you have teams with five back court players trading guards for more front court in four player swaps. You have contending teams offering up starters.

I just don't think the franchise modes team gets many resources because the mode does not generate additional revenue the way the VC pay to play and upgrade modes do. I think Leftos and the whole crew want to fix these things. They just don't have the resources to do it because they are told to prioritize adding marketable features, like expansion.

In fairness, the team devoted massive resources to it a couple of cycles ago. But the things that got neglected when MyLeague was created have not been fixed since.

[BA2929]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

I wish I could say I'm surprised, but that would be a lie.

#30teamcontrolFTW

It's sad that we have to resort to controlling all 30 teams just to make the rosters somewhat normal.

Takes forever to do something for all 30 teams as well.

There should be some sort of position number limit hidden in the game to prevent a CPU controlled team from having more than 4 centers at once.

[trekfan]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

It's sad that we have to resort to controlling all 30 teams just to make the rosters somewhat normal.

Takes forever to do something for all 30 teams as well.

There should be some sort of position number limit hidden in the game to prevent a CPU controlled team from having more than 4 centers at once.

It takes awhile, but not as long as it did last year. The ability to toggle between teams in most screens with the right stick is incredibly time-saving, I feel that's a concession that the devs made to allow us to do 30-team control without taking 24 hours to go through a quarter of the offseason. I literally had to take a weekend last year for every offseason I hit in MyLeague. I started in the offseason this year and got through it in a weekend too, taking my time to learn the menus and new features, but I'm almost sure I can do an offseason in a day now thanks to that toggle ability.

Bottom Line: MyLeague AI is basically Billy King and that AI will ruin rosters and organizations with its dealings.

Friends don't let friends hire Billy King. Say no to King.

[33JAF]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

Oh man, not again. Us offline guys are in a world of trouble.

[flo182]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

FIFA has the same problem and it ruins the career mode.
Luckily 2k gives us 30-Team control, this year it's even nicer because you can just do the free-agency and then let the AI play the season.

Anyway, I'm always wondering if it's really that hard to give the AI some kind of blueprint that controls how they build teams.
e.g. 3-4 bigs, 3-4 wings, 3-4 guards.

If they'd stick to such simple rules we wouldn't get teams that spent 90% of their salary on 5-6 centers and 1-2 guards rated in the low 60s...

[Grandview2015]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

It is 20 times quicker than years past.

It's a minor price for your sanity

[trekfan]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

If there was any doubt the CPU was bad at its job, this should put it to bed. I'm just looking to see what I might get for Gordon for laughs and here's Boston ...




I'm an expansion team, 8 games under .500, and Gordon isn't happy that I benched him. He's on a 1-year deal and Boston has no injuries, they're sitting pretty at 30-22. There's no reason for this trade, especially considering the glut of guards Boston already has.

[Gramps91]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

Yea, this is kind of a big deal for me. I'm almost exclusively an offline MyLeague player and I usually just devote all of my time to one team. It's disappointing to see unrealistic rosters form like this over the years. Really hope they can patch this. I'm loving the gameplay and a lot of the new features and presentation but things like this can really effect my experience with the title in a negative way.

[kaeferirre]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

Yea, this is kind of a big deal for me. I'm almost exclusively an offline MyLeague player and I usually just devote all of my time to one team. It's disappointing to see unrealistic rosters form like this over the years. Really hope they can patch this. I'm loving the gameplay and a lot of the new features and presentation but things like this can really effect my experience with the title in a negative way.

The worst thing of all is in my eyes that the say every year the same thing. "We'll make a great improvement for AI-things in MyLeague". Literall every year. And the following edition kept the same issues over and over.
Hope they can fix a few things with a patch.

[BluFu]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

CPU trades are always subjective to me in terms of what's realistic or not for the most part. The big (and should be fixable) issue is CPU positional stacking. When the CPU is stacking up 5 PGs in free agency when they don't have a big man.. it's a clear logic issue. That absolutely needs to be fixed.

[joosegoose]

Re: NBA 2K17 MyLeague CPU Roster Logic Still Really Bad

Close to the theme of this thread, contract amounts have been discussed at length, with the general consensus being that amounts are almost universally low across the board and there are way too many $1-2m deals. That being said, I think/thought I found an additional consequence of these low valuations.

I was looking through free agency for youngish players I could throw a little money at, when I saw:

Spoiler

They...declined his team option? Wait a minute, there are a ton of rookies with declined options on their rookie contracts...do teams really value players that little, where $2-4m isn't worth it? Well not quite...

Spoiler

Uh...yeah. They declined his option, making him a UFA (giving up YEARS of control in the process) only to pursue him in free agency? Honestly there's nothing more to say here.

There are plenty other players like this. Another sort of category of weirdness:

Spoiler

Seems way too high (never thought I'd say that) for an unproven player, right? Well consider that there's no reason he should even be a free agent right now anyways, that's only because his stupid team declined his option only to throw all their money at him.

A different example:

Spoiler

LOL. So many things wrong with this. First of all, those numbers are absurdly low (minimum I'm pretty sure). Second, they're all 1 year contracts for a 26 year old quality role player. Third, they're all THE SAME. Nobody is trying to outbid the others, and it's not like these teams are all capped out. Just a couple thousand dollars at this point could 'outbid' the other teams. Fourth, he's an RFA. Really, that's how you're going to prevent his team from matching? UGH.

Speaking of RFAs, I still don't think they work properly. Sure, as a user we have the option to match when another team signs our player to an offer sheet, but I'm pretty sure CPU teams don't have this basic functionality. I'm have a good feeling an RFA is no different from a UFA as far as they're concerned. Not to mention the fact that the offer sheet/match is instant, no 3 day period where money is tied up until the other team decides to match (kind of like cap holds which don't exist, but I'm much forgiving on these features not being in the game yet until the basic stuff actually works).

Another quick hitter, I hate decisions on team options coming all at once in the offseason, rather than a deadline we have to meet the prior season. It results in almost none of the players that would typically be trade actually available to trade on draft day, because they are 'about to become free agents'. This is a big contributor to draft day trades making no sense, I think.

I really could have gone on and pointed out everything that didn't add up, dissect every signing that didn't make sense. But honestly, so few things went at all like one would expect in the real NBA, it got overwhelming. It's just so bad in the current state; I guess I would start with increasing contracts to their proper amounts and getting rid of the abundance of $1m deals. Even that will have problems though unless teams actually value their cheap rookie contracts though, as teams will eat up their cap space signing these guys in early free agency leaving little to no room to sign the actual free agents.