Illegal Exit 04

[sinewav]

I'll wait a few days before spamming all the other forums. See you there!

[Fate]

awww i'll be out of town :\ maby next time.

[Titanoboa]

Yes! I got a spot awesome.

(Hint to people who want to play but didn't sign up: sadly, there's a chance I won't make the tourney. But I hope I will.)

[sinewav]

...sadly, there's a chance I won't make the tourney.

And this is precisely why I have extra spots on the sign-up page.

[vov]

Shit, I missed it!
Can anyone sign me up for one of those extra spots? I'd really like to play

[sinewav]

You didn't miss anything. Just sign yourself up whenever you get the chance.

[sinewav]

I'm having thoughts about the flow of the tournament. Right now the rules are: matches end at 10 points, first to win 2 matches is the victor. But it's really hard to gauge how long that's going to take. It could be over in 15 minutes or it could take an hour. I think I have a better plan.

I would like to make this a timed event. No score limit, and a singe 30 minute match. Who ever has the most points at the end wins the game (and I can record these scores in the history if I remember to get a screenshot). I think 30 minutes is long enough to make it worth showing up, but no so long that people get bored and tired. Quick and fun tournament.

Also, if you want to get a feel for it, I've attached settings for local play. Because 12 players is a lot, size_factor is now -2 instead of -3 (which is what we use for half the players).

[ItzAcid]

I think that's going to be a great adjustment. Looking forward to it

[PokeMaster]

maybe add a round time limit too, because if somebody's ahead and they're one of the last two, they could technically wait it out until the end of the 30 minutes

[compguygene]

I am working on a multi-mode incam server. I was reading through the settings you just posted, and I wondered what CYCLE_WALL_NEAR does?

[vov]

CYCLE_WALL_NEAR (Maximum accelerating wall distance)

In these settings, it's set to 4, which means if you're closer than 4 grid units to a wall, you accelerate, otherwise not.

[Phytotron]

Are you sure it's grid units?

It's also, hmm, what's the term.... As we all have surely noticed, the closer you are to a wall (within that value), the greater the influence of the acceleration. There's a term for that.

But what also seems to happen as you raise or lower that value, is that it affects that influence. So, for example, Shrunkland has _NEAR 8 (as all the Breakfast servers had, I think), which means that you'll catch acceleration at a greater distance from the wall. But it also seems to increase how strong the acceleration is when you're right up on a wall. A low _NEAR seems to not only require you to get closer before acceleration kicks in, but also make acceleration weaker once it does. So, at the same distance from a wall, a higher _NEAR will provide greater rate of acceleration than a lower one. I think. Or, I could be delusional.

Z-Man?



EDIT: Wait a minute. Scratch that, reverse it.

[Jonathan]

Increasing CYCLE_WALL_NEAR will always lead to more acceleration. The slope is even identical. There is suddenly a good amount of acceleration coming out of nowhere with small CYCLE_WALL_NEAR, which may be what you're experiencing. You can compensate by decreasing CYCLE_ACCEL_OFFSET, which makes for steeper acceleration near the wall.

The basic equation, before any modifiers are applied:
acceleration = CYCLE_ACCEL * (1 / (distance + CYCLE_ACCEL_OFFSET) - 1 / (CYCLE_WALL_NEAR + CYCLE_ACCEL_OFFSET))

[Phytotron]

Ah, you posted while I was editing, heh. I was just in Shrunkland checking if my recollection was correct, and turns out it's the reverse. Doh! So, everyone, reverse what my edit describes and you'll get what Jonathan is replying to directly to me. The edited version now properly describes my experience. Everyone following? Heh.

[Jonathan]

I declare you not the king of page breaks, but king of edits!

By the way, CYCLE_WALL_NEAR and CYCLE_ACCEL_OFFSET are really in grid units. Well, actually, the distance to walls is measured diagonally backwards. But the sideways distance is still what matters.

CYCLE_WALL_NEAR is where acceleration begins.
-CYCLE_ACCEL_OFFSET is where acceleration is infinite.
The above is applied by translating the basic reciprocal function only; its scale is left intact.
CYCLE_ACCEL scales the result. It can't be defined as concretely as the other settings. This is about as close as you can get to capturing its essence:
acceleration at distance (1 - CYCLE_ACCEL_OFFSET) = CYCLE_ACCEL * (1 - 1 / (CYCLE_WALL_NEAR + CYCLE_ACCEL_OFFSET))
Or: d/ddistance (acceleration at distance (1 - CYCLE_ACCEL_OFFSET)) = -CYCLE_ACCEL