Now I know many people don't care about gameplay mechanics, but it would be nice if these Youtube guys could actually spend some time talking about substance instead of graphics, presentation, MyPlayer-acessoires, dunks and so on.
- Realistic generated draft classes to maintain balanced league strength over the years. So since 2k hasn't found the solution after 10 years of overpowered association leagues...if you limit rookies rated >70 OVR to 0-5 per year, the league turns out pretty nice. Right now, it's a mess.
- Fix distribution of tendencies/attributes/animations for generated rookies. This means more variety of different player types suited for specialized roles on certain teams instead of the generic crap we have right now. Also, not every SG/SF needs dunk >90 and NBA Jam packages assigned, just as not every white 6'0 PG should have layup at 90.
- Biggest issue for me, and won't buy 2k14 at all if it gets ignored the 4th year in a row...is AI playcalling fixed? First of all, the "run best play" script is useless. It limits the CPU to using only a handful of plays no matter how many you have assigned to a playbook.
Playcalling should be handled by a properly working touches scale
only!!!
Right now (maybe 2k devs don't know this), in fantasy seasons, as well as associations with trades and with changes to default rosters in general,
the AI doesn't recognize PGs properly and distributes touches according to some useless scripts not really related to the tendencies assigned to the players.
If 2k wanna figure that stuff out, simply start a season with a fantasy draft. Now watch a game with ball dominant PGs in coach mode and notice how the PG gets completely ignored by the CPU. You can spam the useless "run best play" option and you'll notice the AI suggested plays have no relation to how you scale touches.
It's a total mess and breaks gameplay against the CPU, especially if you work around the ridiculous PG iso drive spam in transition-script by setting the (1 year later still) revesed "run plays" tendency to 0 in order to try to make the AI play semi realistically. Why do we even need 50 play slots and a tendency that's supposed to determine usage% for players, if the AI can't use both?
I've tested it with lots of different scalings, and it turns out that with fantasy rosters and every player having the same touches tendency to eliminate any kind of team specific scripting, it simply doesn't work like 2k claims at all.
Is it that hard? Use some random generator to decide for which player the AI runs a play on each possession, with "touches" determining the odds of each player in the procedure. After that, why not pick the specific play randomly out of the playbook? I mean you could use the same probability system for finetuning distribution ( e.g by allowing us to assign percentages to play types for each player individually), but I'd be happy if at least the basics were working properly for once.
Also, it would make sense to remove "touches" from the bottom of the individual tendency menu and put it into a team management window, allowing us to distribute touches while looking at the whole team instead of scrolling though pages for each player individually.