My wish/suggestions are all related to MyGM / MyLEAGUE modes...
1. MISC (i'll get the quick ones out the way first)- Create a staff member (coaches, GM's, scouts, etc)
- EVERYTHING to do with the NBA landscape should be editable before starting a MyGM or MyLEAGUE mode. It'd be awesome to be able to edit staff members' skill grades (A,B,C,etc) and their style preferences myself. I'd like to be able to edit the owner attributes aswell (i.e. the 5 star things). And i'd like to be able to edit what freelance offense a team uses outside of a game, in the roster edit menus. It should merge into the roster file / be saveable. I wanna be able to customize the NBA landscape before i go into MyGM mode, then i can fully trust it.
- Coach preferred styles should be viewable in MyGM mode. They currently aren't. They should also be editable in roster edit mode.
- Keep working on dat generated rookie A.I. Maybe come up with a few hundred "player types" and generated rookies are given ratings + tendencies based on the template that got randomly assigned to them.
- Please allow more than 10 save slots for draft classes and rosters. I like to download many draft classes and edit them myself and *then* use them in my MyGM or MyLEAGUE mode. Surely we have the technology to allow for more than 10 save slots. How about 50? Also, can we have more than 150 create-a-player slots?
- Please allow us to edit body type and tattoos in create-a-player.
- Create a coaching philosophy mode? Where you create your own playbooks, coaching tendencies, freelance offenses, etc. And you can save it to your system, share it, and attach it to any team (and the player's "play types" will sort out who does what for the team) in either exhibition modes or MyLEAGUE modes?
2) IMPROVE PLAYER TENDENCY PROGRESSION A.I.Currently, players can improve their ratings but it wont necessarily improve their PPG stats if their tendencies are too low. We need to change the way tendencies interact with attributes progression MyLeague mode. Or something.
Perhaps: Can we make the
Shot and Touches tendencies "glue" to a player's collective scoring attributes a bit more? So that when a player progresses, he naturally gets more offensive responsibility, because he deserves it, rather than Touches tendencies becoming outdated a few seasons simmed into the future.
No more players who can't average more than 10ppg their whole career (despite having the ability to) because they were given a low Touches tendency by whoever edited the roster, please. (or how they were generated)
Or maybe make the CPU coach gameplans ignore the tendencies a bit more, and make smarter decisions based on cpu calculations to do with player abilities/attributes instead.
Either one.
3) MORE REALISTIC DEVELOPMENTAL CURVESAlso, some tuning of player progression and regression is needed:
Progression/regression is too gradual/incremental, so the overall landscape of the NBA is too static from season to season. Not enough change like in the NBA. Change happens, but it is too gradual. In particular, vets don't regress sharply enough; stay relevant too long.
There should be a sharper jump up or down within 1 or 2 offseasons at some point before/after their peak. Young players have breakout seasons. They make striding leaps. Old players have sudden drop offs. I'd love to see that reflected in 2k, so that the NBA landscape breathes in and out like the real NBA does.
E.g. in 2k at the moment a 20 year old who's rated 70 overall with an 82 potential will go something like 70 - 73 - 76 - 79 - 82 up until their peak age. I.e. small steps that take too long break out. It should be something like 70 - 75 - 79 - 81 - 82 instead. See what i mean? The growth happens early and fast, then steady from there until they reach their peak/cap.
Same with old regressing players. At the moment a player who is rated say 82 overall passes their peak and then each year just gradually regress 2 pts, then 2 pts, then 2 pts. He stays relevant for too long, and then the young players never get the opportunities to have the baton passed down to them timely enough.
Old players get to a certain point and then it's like BANG - they're completely ineffective all of a sudden. After their peak year, a better example would be something like 82 ovrl at 32 yrs old, 80 at 33, 75 at 34, 69 at 35. See what i mean?
AN IDEA JUST CAME TO ME: Add the ability to set the player's developmental curve.
We already have peak start & peak end. How about we attach a new value to either side of the peak?
For each side of the peak (start & end) you can have a different gradient:
1. Sharp early
2. Gradual
3. Sharp late
(maybe
sharp is the wrong word, but you get the idea)
It should be editable in a player edit screen.
^ If you can imagine what i'm about to say in terms of plot graphs then this will be easy to explain. (note: graph is just an example)
So for e.g. DeAndre Jordan might have a "gradual" or a "sharp late" curve before his peak start, because the center position takes a while to learn, and his skills are raw initially. And then he might have a "sharp early" after his peak end, because once he loses his athleticism he wont be anywhere near as effective.
Whereas maybe a skilled scoring star like Curry will have a "sharp early" curve before his peak, because he already had an established skillset, and just needed to get his feet wet in the pros. And then he might have a "gradual" or a "sharp late" curve after his peak, because his game isn't based on athleticism so much. Skill-based players age better.
So in summary...
BEFORE THE PEAK:
1. Sharp early = basketball prodigies who boom (or plateau) quickly (Durant, LeBron, Tyreke, etc)
2. Gradual = small incremental steps towards their peak (most players?)
3. Sharp late = Late bloomers who breakout closer to their peak start (Korver, Dragic, Lowry, most centers, etc)
AFTER THE PEAK:
1. Sharp early = Players who drop off quickly and suddenly after their peak (usually players who rely on athleticism)
2. Gradual = incremental steps towards fathertime
3. Sharp late = They extract every last little bit of basketball from themselves, have longer lasting career, then drop off suddenly at the end (PG's, all-timers, guys who don't rely on athleticism)
There doesn't have to be massive differences between the 3 (or 6?) options, but just enough for there to be a distinct affect.
With the ability to set the "overall cap" (i.e. potential), when the cap starts, ends, and the shape of the curve towards the start and end, that would allow the ability to COMPLETELY shape a player's progression/regression curve!
4. SMARTER CPU TRAINING DEVELOPMENTSmart CPU training schedule A.I. for the aspect of player progression that's based on training. (right now player progression of cpu players is too rounded. players get +2pts in all of the attributes, sort of thing). Training regimes should be based on player position +
STAFF PREFERRED STYLES! (i.e. a SG playing for a coach who prefers "skill" over "defense" is obviously gonna get bumps in the shooting and handling and passing categories)
5. GENERATED ROOKIE A.I.One of its flaws currently is that the tendencies lack connection to the attributes which lack connection to the badges etc. It's better than in years past, for sure, but it's still sorta random. Players lack a specific identity.
The whole reason why people spend so much time creating fictional draft classes is because it means we can create players with a specific identities. Their values connect with their own values.
I'm not sure how exactly you should tweak the generated rookie A.I, BUT i think it'd be good to think in terms of
SYNERGY. Generate players whose attributes and tendencies etc are synergistic with each other. Guards shouldn't be performing dropsteps on people. Athletic defensive players shouldn't be attempting step back dribbles.
The generated heights are fine, the generated positions are fine, the generated overall ratings are fine, the amount of badges generated is fine, the balance of good and bad players is fine, the way they look isn't perfect but it's fine (compared to years past), lots of things are fine. It's mostly good, but i'm so hungry for generated players to have specific identities that make sense. I don't wanna see that the 7 foot center i just drafted has a 78 rating for speed but a low rating for dunking but a nice 3pt but shot blocking badge shot etc etc. Some things just don't make sense together. Think about which kind of player abilities are synergistic with each other and tune the generated stuff a bit?
E.g.
I'm not sure how rookies are currently generated, but what you could do is have a "player DNA" set before he is "born" into the draft, so to speak.
So, ok, this draft is gonna have an "aggressive scorer", and "faceup four", a "craft offensive guard", and an "interior defender", etc. And then after that, certain ranges should be in place for what attributes or tendencies make sense for that player. (the ranges can be loose, but their must be ranges) I'm not sure which should come first - the attributes or the tendencies - but the next step would be to generate one of the two within specific ranges, and then based the NEXT thing (whichever remains left after choosing one of attributes or tendencies to come first). Then, we have either tendencies that are based on attributes, or attributes that are based on tendencies (whichever creates better results). THEN, only certain signature animations should be avail for the type of body type, athletic ratings, and play style. Not tooooo limiting, but at least ignore big man animations for guards in the generation process.
^ I don't know if that helps at all. But at the end of the day, some sort of genius automation process must be come up with so that our generated players feel like actual unique and believable NBA prospects.
But like i said, it is better in 2k16 than it ever has been. But i still feel it is necessary to edit generated players to death, which is really time consuming. And unless i create the class beforehand, i can't even edit players in MyGM mode.
There will have to be a lot of "if" codes for generated prospects.
Examples...
- If a player has a 60 for vertical, then his range for speed is 40-75.
(then lets imagine the computer generated a number of 70)
- If a player has a 70 for speed, then his range for acceleration is 50-85-ish.
- If a player has a weight of 211 lbs, then his range for strength is 48 - 65.
- If a player has a height of 6'10", then his range for vertical is 25 - 90. And the cap goes up +2 with every inch of height he loses.
Or maybe that's too strict.
But you get the point.
There needs to be more synergistic identity within the individuals of a generated class