Game Mechanics: Resist results on even level and lower players here

[pickled_heretic]

Originally Posted by pickled_heretic [You must be logged in to view images. Log in or Register.]

i posted a more articulate response in the other thread this retarded graph was posted in. this is called a hyperbolic function. you can generate a hyperbole with the formula y = tanh(x) and you can make it conform to any pattern by adding coefficients in the right spots.

for instance you could make a fairly nice EQ pvp resist formula like this:

y=tanh(x/100+(10*n)

where y is the resist rate, x is your MR, and N is the difference in levels.

assuming there's no difference in level, your resist rate would be about 29% at 30mr, 76% at 100mr, and 96% at 200mr. EQ always rounds down so there would never be a 100% resist rate, but it would start rounding to 99% at about 270mr.

with a 5 level deficit, your resist rate would drop to 19% at 30mr, 58% at 100mr, and 87% at 200mr.

with a 5 level advantage, your resist rate would be 53% at 30mr, 96% at 100mr and you would be over the 99% cap at 200mr.

this is just a sample, you could tweak the formula however you wanted. but i'm sure nilbog, rogean etc already know plenty about maths and have a good idea for a formula in mind.

[You must be logged in to view images. Log in or Register.]
since people lurv their graphs, i went and generated mine in excel. this graph is reproducible by the formula y=tanh(x/100+(10*n).

y axis is resist by percentage, x is MR. the blue line is resist values with a 5 level advantage, red is with no advantage and green is with a 5 level deficit. you could easily change the values by modifying the coefficients but it would be really stupid to use any other function than tanh to do this.

[Pescador]

You can use (1 - some exponential function) as well for essentially the same result, which is how I generated my graph.

But I agree, and I'm sure the devs know, that resists should fall off a cliff when the caster outlevels the target by more than a few levels, and you shouldn't see frequent resists unless the target outlevels you or has 100+ resists.

[pickled_heretic]

Originally Posted by Pescador [You must be logged in to view images. Log in or Register.]

You can use (1 - some exponential function) as well for essentially the same result, which is how I generated my graph.

i'm sorry what? this doesn't make anything that resembles a hyperbole, exponential functions make graphs of exponential functions. resist rates would not approach 100%, they would blow right past them.

[Pescador]

1-e^-kx, where k is a function of level and resistance

[Lazortag]

Here's what I came up with (using an exponential function): http://www.wolframalpha.com/input/?i=100-100%284%29^%28-0.01x%29+from+0+to+400
(this has an asymptote at y = 100)

You'd need to add constants that factor resist modifiers (like with lifetaps) and the ratio of the two players' levels (ratio instead of difference since I'd think a level 10 should intuitively stomp a level 2 much easier than a level 50 stomps a level 42). This is a little bottom-heavy also (having 50 MR gives you roughly 50% resists with this which is kind of ridiculous) but I just wanted to have fun like everyone else.

[tmoneynegro]

Originally Posted by gloinz [You must be logged in to view images. Log in or Register.]

should curve at 100 resist imo then never hit 100%

PvP resists capped at 98% during classic through velious, it's in the patch notes.

[Pudge]

Originally Posted by tmoneynegro [You must be logged in to view images. Log in or Register.]

PvP resists capped at 95% during classic through velious, it's in the patch notes.




Listen pal, I've already gone through this. I know for a fact that crowd control spells resisted over 90% of the time with 120MR.

When crowd control resists max out that fast, using a non-linear graph is wildly inaccurate to represent resists actually seen on PvP servers. They either give you too much resist early on, or not enough at the upper end. My linear graph is MUCH more accurate to what was actually seen on Sullon Zek for crowd control resists.

Please do not post graphs if you've never actually played high end PvP before in EQ.

nice i was about to mention the 95% thing.

i think resists just need to be tweaked with each era, to account for better gear. knuckle came up with a resist system he posted on vztz that didn't look bad. didn't agree with the exact numbers, but thought the general idea looked good. included partial resists too.

knuckle you have that post somewhere?

[tmoneynegro]

Originally Posted by Pudge [You must be logged in to view images. Log in or Register.]

i think resists just need to be tweaked with each era, to account for better gear.

I don't care what they do with damage spells as long as crowd control spells function like normal EQ PvP and not "some random guy's wildly unbalanced implementation".

This is why it's absolutely stupid to have those spells land:

If you're playing a solo caster outdoors and there's a zerg chasing you, soon as the root lands, you're gonna spam gate or shadowstep and gate and probably get away easy.

If you're playing a melee and that happens, you just die as they chain cast it over and over. Making it so casters have more survivability than tank classes is utterly stupid and whoever thinks that's a good idea doesn't need to be let anywhere near the server code.

[Harrison]

I think we've found wehrmacht

[tmoneynegro]

[You must be logged in to view images. Log in or Register.]

Here is the problem with this guy's graph he posted. Crowd control resist rates were parsed at 87% with 150MR during Luclin. This was after the patch to let spells land easier so resist rates were obviously higher before that. It was definitely over 90% with 120MR during Velious and previous expansions. Then *AFTER* the Luclin patch, they once again made magic resist even more effective because 87% resist rate against CC still made it too easy to kill pure melees that have 0 channeling skill to dispel it.

http://www.thesafehouse.org/forums/s...ead.php?t=7879

The second problem with his graph is that it shows a 30% resist rate while a character is naked with 25MR. When someone was naked, the resist rate was somewhere between 0-10%.

If you can create a chart/equation to show 0-10% crowd control resists while naked (25MR), around 90% resist rate with 100MR, and 95%+ with 120MR, then you probably have an accurate chart.