Grenade/explosive damage fading
2013-04-14, 03:56 PM
#11
RE: Grenade/explosive damage fading
Is this what you are talking about
[Image: plot.html?g=p341317822.png&t=plot&db=r619579487]
or is this it
[Image: plot.html?g=p-710236706.png&t=plot&db=r-1070691486]
Post Reply Quote this message in a reply
2013-04-14, 05:46 PM
#12
RE: Grenade/explosive damage fading
Dang it! Then I meant (1-x)^2 or (x-1)^2

Best regards,
Victor
//victorz.ca
Code:
Your antithesis compares favorably with any high magnitude of pwnage. (-you > |p|, you < -|p|)
My antithesis compares favorably with _that of_ any high magnitude of pwnage. (|-me| > |-p|, |me| > |p|)
Post Reply Quote this message in a reply
2013-04-14, 06:06 PM (This post was last modified: 2013-04-14, 08:27 PM by ruler501.)
#13
RE: Grenade/explosive damage fading
those are equivalent.
I'll add that into the graph I posted

EDIT: Looking at the source it looks like the actual value it is cutting off at is 5(damage-rangeminus=230-225=5)
Should this be fixed? I have a patch for it currently I could commit.
Post Reply Quote this message in a reply
2013-04-15, 06:23 AM
#14
[split] 2.5.8 Changelog
Well I'd change the RPG's damage to the yellow/brown curve, and maybe change the grenades curve to the purple one ? In my opinion the grenade does too much damage at long range, so you should give the purple curve a try since the damage at close range isn't that different...
You should really test it, I mean both, if the grenade's not that cool anymore change it to the blue curve again ( I think it's the current curve? ).
And if the yellow curve isn't good for the RPG I've another suggestion.
The damage curve should cross P(3/100), it should kill people who are 3m or less to the rocket but the curve should fall much faster after that point. It should be kinda like ruler's purple curve which is still killing at 4.2 - 4.3 meters but not that overpowerd.
Post Reply Quote this message in a reply
2013-04-15, 03:18 PM (This post was last modified: 2013-04-15, 03:21 PM by Victor.)
#15
RE: Grenade/explosive damage fading
The cut-off is fine, and it used to be a bug that allowed 5 damage for the RPG at infinite range. There's no need to commit.

Fubizz, which graph are you referring to?

Best regards,
Victor
//victorz.ca
Code:
Your antithesis compares favorably with any high magnitude of pwnage. (-you > |p|, you < -|p|)
My antithesis compares favorably with _that of_ any high magnitude of pwnage. (|-me| > |-p|, |me| > |p|)
Post Reply Quote this message in a reply
2013-04-16, 07:45 AM
#16
RE: Grenade/explosive damage fading
I was talking about your second graph.
Post Reply Quote this message in a reply
2013-04-16, 08:56 AM
#17
RE: Grenade/explosive damage fading
I've actually been messing around in the source to try these out and I got the reciprocals working(I stop calculating damage after it drops below one). I think they are a nice curve. The reciprocal square root and reciprocal linear are too hard to work with. Reciprocal quartic looks nice though. I'll add these observations into my earlier graph. Here are my descriptions of the kinds of graph we have

Reciprocal Square max/((x/range)^2+1): It is the most realistic, but it drops off really quickly, but stays damaging for a long ways.

Reciprocal Cube max/((x/range)^3+1): It drops below 100 at roughly 2.25m for grenades so is slightly less lethal, but carries damage farther than current. Less range than reciprocal square.

Reciprocal Quartic max/((x/range)^4+1): It drops below 2.75m for grenades so is slightly more lethal. It also has a really quick drop after 5m, but still has more range than our current one.

Square Root max*(1-sqrt(x/range)): This is the current curve. It drops at a relatively constant rate(after it gets past the kill level)

Square max*(1-(x/range))^2: This is mostly the same as our current method, but it drops off slightly faster after the killlevel which is pretty much the same point

Cube max*(1-(x/range))^3: This would greatly reduce the effectiveness of grenades as the damage curve is very steep with very little damage being done at more than 1/2 range

Quadratic max*(1-(x/range)^2): This is the most deadly curve. It doesn't drop off till almost its whole range is over and then quickly goes to zero. This would make grenades much more terrifying in game since you'd need to be almost out of range to not be nearly insta-killed
Post Reply Quote this message in a reply
2013-05-09, 06:27 PM (This post was last modified: 2013-05-09, 06:56 PM by Victor.)
#18
RE: Grenade/explosive damage fading
Using reciprocals with an exponent in x will cause a different shape:

[Image: eWRzT5b.gif]
plot (x+1)^-1, (x^2+1)^-1, (x^3+1)^-1, (x^4+1)^-1 from x=0 to 10, y=0 to 1

Now, what if we put exponents into the entire denominator?

[Image: FrnpRLR.gif]
plot (x+1)^-1, (x+1)^-2, (x^2+1)^-2, (x+1)^-1/2, (x^2+1)^-1/2 from x=0 to 10, y=0 to 1

Also, for guns, the bullet has kinetic energy. If the energy loss is constant, the speed will be affected like a square root: (proposed fading for "bullet" weapons)

However, in the graph, 1 on the x-axis represents when the bullet stops.
[Image: DE3Umky.gif]
plot 100*sqrt(1-x) from x=0 to 1

Best regards,
Victor
//victorz.ca
Code:
Your antithesis compares favorably with any high magnitude of pwnage. (-you > |p|, you < -|p|)
My antithesis compares favorably with _that of_ any high magnitude of pwnage. (|-me| > |-p|, |me| > |p|)
Post Reply Quote this message in a reply
2013-05-09, 07:32 PM
#19
RE: Grenade/explosive damage fading
I do like the curve with that more. I modified them to make sure they drop below one at range.

rquar: max/(((max^(1/4)-1)*x/range+1)^4);
rcub: max/(((max^(1/3)-1)*x/range+1)^3);
rsqr: max/(((max^(1/2)-1)*x/range+1)^2);
rlin: max/((max-1)*x+1);
Post Reply Quote this message in a reply
2013-05-12, 07:57 AM (This post was last modified: 2013-05-12, 07:58 AM by Victor.)
#20
RE: Grenade/explosive damage fading
I realized one thing. Without moving the exponent to the entire denominator, it is missing 2x as a term.
[Image: QPBWRI8.gif]

By using the exponent on the entire denominator, we are basically "adding 2x" to the original.

Best regards,
Victor
//victorz.ca
Code:
Your antithesis compares favorably with any high magnitude of pwnage. (-you > |p|, you < -|p|)
My antithesis compares favorably with _that of_ any high magnitude of pwnage. (|-me| > |-p|, |me| > |p|)
Post Reply Quote this message in a reply


Forum Jump: