Inconsistent bowling physics

[ilipitagaon]

So after the guide has been posted on this forum, I spent two hours trying to test how consistent bowling is now that we can control all of the variables and the result is:

SUPER INCONSISTENT

Without changing any angles or applying spins, I shot the bowling bot in a straight line multiple times and got different results.

The bowling ball would be placed right in the dead center, shot in a straight line (mouse disabled, controlled by keypad) with zero spins.
It results 1 ~ 4 pins (random) still standing, and sometimes even a strike.
The test was done on two empty official servers (30 minutes each), on a wired connection with zero packet drops.

Could we make the bowling consistent?

[macdguy]

First off, I am assuming you are talking about the guide where you line up the roll to the side of the screen. I’m working on a fix for that.

We use our own math to calculate the angle and speed the ball will travel and that’s consistent everytime, however, we use PhysX to simulate the pin collisions. PhysX is not fully deterministic.

[ilipitagaon]

@macdguy

First off, I am assuming you are talking about the guide where you line up the roll to the side of the screen. I’m working on a fix for that.

That’s great to know. It will draw the players back to the Casino when that gets fixed, but hopefully not by much.

@macdguy

We use our own math to calculate the angle and speed the ball will travel and that’s consistent everytime, however, we use PhysX to simulate the pin collisions. PhysX is not fully deterministic.

That’s a shame. Ugh, PhysX.
Are there any plans to fix the inconsistency?

[macdguy]

@ilipitagaon

Are there any plans to fix the inconsistency?

Not at this moment. Making deterministic pins means writing our own physics simulation just for Bowling and it’s quite a large task as it needs multi-body physics collision detection.