Need help with possessions/pace factor (CPUvCPU games)

[Tandur]

So I've been running a massive tournament using the Ultimate Base Roster, simming games as CPUvCPU. I spent over a month using various sliders, tweaking them to fix things I thought didn't work, but I've never been satisfied.

Finally, it occured to me one day, the first thing I should look at correcting when trying to create a realistic and watchable 2k12 is possessions. If the number of possessions by both teams is not similar to those of their real life counterparts, then no amount of tweaking FG%s, turnovers, or defensive ability, etc, is going to make the stats come out correctly.

So I've started by just simming games using straight default sliders, and the pace factor the teams tend to average is about 83.7 possessions per game (aka pace). The forumula I use to calculate this is from Basketball Reference, which I also use to set player rotations for each team, current and historic. This also allows me to compare the pace factor of my games versus the historical pace factor of the teams I'm simulating with.

Now, in 2012, the league average pace was about 90.4 possessions per game. However, 20 years before that, the pace was well above 100 possessions per game. Seeing as how none of my games can even break 90 per team, I doubt I will ever reach 100 for the more historical teams.

So my question is, does anyone have any ideas on how to increase the number of possessions each team gets per game? Some ideas I have:

1. Game speed does not affect pace. I've tried setting it to 100, and to 0, and neither changes it. Game speed affects the clock speed along with the player speed. For example, with GS set to 100, 24 seconds in real life only takes 18 seconds in the game (I'm just paraphrasing, that's not exact, but it is what happens).

2. Framerate does not affect pace. I've tried lowering all video settings and screen resolution down to 800x600, getting a FPS close to 60. However, in the few games I tried like this, the pace of the game was amongst the lowest of all my tests.

3. Offensive rebounds. In the BR formula for calculating pace, or most other metric formulas, ORs are subtracted from the total, since they are merely extensions of a possession the team already had. Therefor, logic would dictate that the more offensive rebounds teams have, the less total possessions a team will have, since each OR eats up time from the total game clock and does not create additional possessions.

4. Turnovers. I have noticed consistently that turnovers are usually very low in almost all "authentic" sliders I have tested. While the number of steals might be accurate at times, what many people forget is the number of loose ball turnovers that happen in a game. Bad passes, fumbled dribbles, and rebounds that go out of bounds, these all add additional possessions. My idea here is to lower pass accuracy slightly, leave hands alone (ability to deflect passes), decrease steals (ability to control those deflected passes as well as steal the ball from the ball handler), and lower ball handling slightly. On ball defense needs to be dropped a bit, IMO, to minimize the massive amount of turnovers that are created unrealistically when a player drives into the paint and the ball hits another player's body.

5. 3 point shooting. One thing I noticed while studying the league averages for seasons in the 80s versus seasons in the 2000s is that while pace is quite different, the biggest difference in the stats is 3 point shooting. Now a days, teams in the NBA shoot 3s like they will become illegal tomorrow. Back in the 80s, nearly nobody shot the 3. How this quite creates more possessions I am unsure of, or how it may even be related.

I would love to hear any ideas from the pros here, and I think this discussion could be applied to NBA2k13 once we all get our hands on it.