The Fortress Guide

Anything about how you get those awesome core-dumps, or why you don't get them...
Post Reply
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: The Fortress Guide

Post by Z-Man »

Oh, all right. Yes, with basic settings, grinds are indeed cumulative, and because rubber regenerates between the first and second grind in a DG, a DG to 4.9 will be deeper than a single grind. There are, of course, countless settings that can counteract that and they are somewhat active in the fortress sample configuration (because they're active in the total default configuration). CYCLE_RUBBER_MINDIST_UNPREPARED/PREPARATION increases the minimal distance to walls for the second grind, CYCLE_RUBBER_MINDIST_RESERVOIR is a linear, non-cummulative component of the grind depth, and CYCLE_RUBBER_DELAY(_BONUS) can make more rubber get used during the second grind because you hit the wall too soon after your last turn (not used by default).

The default settings are:

Code: Select all

CYCLE_RUBBER_MINDISTANCE_RESERVOIR .005
CYCLE_RUBBER_MINDISTANCE_UNPREPARED .005
Those values need to be compared to the depth of grind. The speed of approach on top of the mindistance is governed by
distance to wall = e^{-rubber_speed*t}
and sets in when the speed given by that (rubber_speed times the distance) gets lower than the cycle speed (let's say 30), and rubber_speed defaults to 40/s.
so:
distance to wall when rubber sets in = cycle_speed/rubber_speed = 4/3 m
We arbitrarily label this point t=0 in the distance forumula.
Rubber stays active for about cycle_rubber/cycle_speed, about 5/30=1/6s
that makes the distance to the wall just as rubber runs out:
max_single_grind = 4/3 m * e^{-40/6} = 4/3 * e^{-20/3}
which is approximately 0.0017 m. Smaller than CYCLE_RUBBER_MINDISTANCE_UNPREPARED.

So, math sez: if the code is working as indented and default settings are used, a nice and precise pure grind should be deeper than a double grind. But it needs to start with untapped rubber. Even a rubber meter starting out at 1 can ruin the grind.
User avatar
ppotter
Match Winner
Posts: 451
Joined: Sun Sep 28, 2008 12:45 am

Re: The Fortress Guide

Post by ppotter »

From the 10% I understood of that post, you're confirming that a single grind on the initial grind in fort will be deeper, right? :P
owned
Shutout Match Winner
Posts: 876
Joined: Sat Jun 09, 2007 11:01 pm

Re: The Fortress Guide

Post by owned »

Yeah, he is potter

@Z-man What does rubber_speed mean?
syllabear
Shutout Match Winner
Posts: 1030
Joined: Fri Oct 13, 2006 1:37 pm
Location: UK/HK

Re: The Fortress Guide

Post by syllabear »

Wouldn't this value change with the distance traveled forward before double grinding? for instance, turning back quite soon after would therefore produce a shallower grind than turning back a bit later than normal, since more rubber would be recovered in the meantime?

Also, didn't understand alot of your post, so it may be that distance travelled is irrelevant
The Halley's comet of Armagetron.
ps I'm not tokoyami
User avatar
Clutch
Shutout Match Winner
Posts: 1008
Joined: Fri Aug 28, 2009 5:53 pm
Location: A frozen wasteland

Re: The Fortress Guide

Post by Clutch »

Z-Man wrote: distance to wall = e^{-rubber_speed*t}
and sets in when the speed given by that (rubber_speed times the distance) gets lower than the cycle speed (let's say 30), and rubber_speed defaults to 40/s.
so:
distance to wall when rubber sets in = cycle_speed/rubber_speed = 4/3 m
We arbitrarily label this point t=0 in the distance forumula.
Rubber stays active for about cycle_rubber/cycle_speed, about 5/30=1/6s
that makes the distance to the wall just as rubber runs out:
max_single_grind = 4/3 m * e^{-40/6} = 4/3 * e^{-20/3}
which is approximately 0.0017 m. Smaller than CYCLE_RUBBER_MINDISTANCE_UNPREPARED.
:o
Boxed
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: The Fortress Guide

Post by Z-Man »

syllabear: no, the distance traveled parallel to the wall when DGing does not matter much, as the DG depth is mostly limited by the mindistance, which doesn't care about how much rubber you regenerate.
owned wrote:@Z-man What does rubber_speed mean?
It's the decay rate of your distance to the wall in front of you while rubber is active. In ln(2)/rubber_speed seconds (about 1/60 with default settings), your distance to the wall halves.

Also, I forgot: rubber_mindist_preparation is .2. That means that even your standard grind does count as a little less than half unprepared, moving the DG disadvantage from 0.005 to 0.0025, dangerously close to the rubber mechanic value. So it's really tough, the way it should be. Only a very good initial grind (timed with one or two frame accuracy) will be better than a regular DG. If a center attacker wants to beat the start grind, however, he can do a fully prepared grind with the 0.005 depth advantage. An expertly timed center attack grind (starting out with unused rubber, well prepared, turned at the last possible frame) should beat both the perfect single and double grind.

TL;DR: perfect attack grind > perfect single start grind > perfect double start grind > sloppy double start grind > sloppy single start grind.

And we've totally ignored ping_rubber. There's more rubber than 5, usually, and every little bit helps and makes timing easier.
PokeMaster
Match Winner
Posts: 638
Joined: Mon Jun 14, 2010 5:36 am

Re: The Fortress Guide

Post by PokeMaster »

A few relevant, but different questions.

1. Ping rubber doesn't change the total amount of rubber, right? For example, a 3 rubber single grind at speed x for person A = a 3 rubber single grind at speed x for person B, regardless of pings and cycle_ping_rubber, right?

2. How does the speed affect the depth of a dig? Is the dig totally time based, and not rubber based, or is there some sort of mixing of the two?

3. Is a grind from let's say, 1-3 rubber (starting at 1, and ending at 3) the same as one from 2.5-4.5? Because I feel that I've seen people (including myself) escape tough digs because their rubber meter started at a higher value. (Note: I've only noticed this in CTF I think).
Image Image Image Image Image Image Image Image
Image Image Image Image Image Image Image Image Image
Word
Reverse Adjust Outside Corner Grinder
Posts: 4258
Joined: Wed Jan 07, 2009 6:13 pm

Re: The Fortress Guide

Post by Word »

Z-Man wrote:An expertly timed center attack grind (starting out with unused rubber, well prepared, turned at the last possible frame) should beat both the perfect single and double grind.
yay, that's me!
User avatar
ppotter
Match Winner
Posts: 451
Joined: Sun Sep 28, 2008 12:45 am

Re: The Fortress Guide

Post by ppotter »

PokeMaster wrote:A few relevant, but different questions.

1. Ping rubber doesn't change the total amount of rubber, right? For example, a 3 rubber single grind at speed x for person A = a 3 rubber single grind at speed x for person B, regardless of pings and cycle_ping_rubber, right?

2. How does the speed affect the depth of a dig? Is the dig totally time based, and not rubber based, or is there some sort of mixing of the two?

3. Is a grind from let's say, 1-3 rubber (starting at 1, and ending at 3) the same as one from 2.5-4.5? Because I feel that I've seen people (including myself) escape tough digs because their rubber meter started at a higher value. (Note: I've only noticed this in CTF I think).
In Cheers it's set up so just touching the wall when your rubber is at 9, will be a harder grind than grinding from 0 to 8.9, but obviously fort isn't the same
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: The Fortress Guide

Post by Z-Man »

1. Not quite. If there is rubber to start with at all, the effect of ping_rubber is equivalent to adding more total rubber. The implementation works differently, though; the efficiency of the rubber you have is increased, so we don't have to think about what to do with the HUD. So a 3 rubber grind of a HPB is deeper than a 3 rubber grind of a LPB.

2. With default settings (distance based rubber), the higher your speed, the shallower your grind, given the same rubber usage.

3. Yes, bar subtle effects from the mindistance settings.
User avatar
Lacrymosa
Round Winner
Posts: 286
Joined: Wed Feb 15, 2006 6:44 pm
Location: Heaven or Hell...?
Contact:

Re: The Fortress Guide

Post by Lacrymosa »

Very interesting insights we get here, thanks z-man. I think this discussion probably deserves an on thread so it is easier to find for others.
User avatar
Phytotron
Formerly Oscilloscope
Posts: 5041
Joined: Thu Jun 09, 2005 10:06 pm
Location: A site or situation, especially considered in regard to its surroundings.
Contact:

Re: The Fortress Guide

Post by Phytotron »

You all should be aware that when Z-Man refers to "distance," that's usually (barring some unusual or extreme settings, e.g. a large _MINDISTANCE) a visually imperceptible distance.

With respect to _RUBBER_SPEED, if you really want to see its effects, make it a low value. Even at the default of 40, you may notice that your rubber meter will begin to move even if you haven't made actual contact with another wall, seeing a little gap left if you immediately turn. At low values it almost behaves similarly to a raised _RUBBER_MINDISTANCE, gameplay-wise.

An example in an active server would be the Yellow Submarine servers. I noticed they set _RUBBER_SPEED to 29 (and wonder if they know what it actually means), with the result of sort of initially repelling one from the wall when making a grind—and most annoyingly, when preparing to beat a grind.


2. To elaborate on Z-Man's response, there's a setting (new to 0.2.8.x) called CYCLE_RUBBER_TIMEBASED. Typical values to use would be 0-1 (binary off/on, with mixtures in between).

A value of 0 would be the traditional behaviour of rubber usage being based on speed (or, the distance you would have traveled had that wall not been there to stop you), not time. Like he said, faster you go, the more quickly rubber gets used and the shallower a grind.

A value of 1 would make rubber usage solely based on the time spent sitting with your nose on a wall, regardless of speed. (What I don't know is if and how the rubber value will equate to, say, milliseconds.) Shrunkland uses this value. I tend to prefer it because it minimises rubber as a gameplay mechanic/feature.

Values can also be set above 1, resulting in rubber being used up more slowly, and grinds deeper, as speed increases. Mud Puddle uses a value of 1.4. The result is that, firstly, the basic _RUBBER of 6 feels much greater (I find it's comparable to Yellow Submarine's 12 _RUBBER with 0 _TIMEBASED; Yellow Sub may even feel less forgiving). Secondly, and totally back-asswardly, one can barely graze a wall at high speed, using hardly any rubber, but then someone can come and try to beat that grind at a slow-to-moderate speed, use up all their rubber, and still not beat it. Grr.

Bebop in Harlem uses a value of 0.8 at the moment because myself and others swear that when it was at 1 it seemed like there was more rubber at higher speed. Probably just perception and conditioning from speed/distance-based rubber, but I don't trust that 1 is actually 1, heh.
User avatar
compguygene
Adjust Outside Corner Grinder
Posts: 2342
Joined: Thu Aug 21, 2008 12:09 pm
Location: Cleveland, Ohio
Contact:

Re: Average speed

Post by compguygene »

Thank you Z-man and Phytotron for elaborating further on the interplay of these settings. I have been tinkering with them with a rather incomplete knowledge of them. I copy/pasting your text to serve as a guide when I am tinkering with some servers.

[ I believe you wanted to put this here --epsy ]
Armagetron: It's a video game that people should just play and enjoy :)
https://bit.ly/2KBGYjvCheck out the simple site about TheServerPharm
Word
Reverse Adjust Outside Corner Grinder
Posts: 4258
Joined: Wed Jan 07, 2009 6:13 pm

Re: The Fortress Guide

Post by Word »

Phytotron wrote:Secondly, and totally back-asswardly, one can barely graze a wall at high speed, using hardly any rubber, but then someone can come and try to beat that grind at a slow-to-moderate speed, use up all their rubber, and still not beat it. Grr.
That's how i learned to play (and the reason why I die mostly in situations comparable to this. i don't die for doing the same thing in yellow sub). :<
User avatar
Phytotron
Formerly Oscilloscope
Posts: 5041
Joined: Thu Jun 09, 2005 10:06 pm
Location: A site or situation, especially considered in regard to its surroundings.
Contact:

Re: The Fortress Guide

Post by Phytotron »

Did I phrase that well?

What normally (i.e., _TIMEBASED < 1) happens is that a pure grind (i.e., using all of one's rubber) made at slow speed should easily beat (i.e., get behind) a previous grind made at high speed where only a small, grazing amount of rubber was used. Indeed, even with with equal amounts of rubber used, a grind made at slow speed should beat one made at high speed.

However, with TIMEBASED > 1, it's the opposite. Someone makes a glancing grind at high speed, uses hardly any rubber. Then someone else comes along at slow speed, does a hardcore pure grind, and still can't beat the first one. That's whack, yo.

So, yeah, I'm continuing to lobby for _TIMEBASED 1 (or less) in Mud Puddle. He had it at 1 for a time, and while it's still high rubber overall, it was so much better. :)
Post Reply