Page 6 of 6

Re: feedback on the latest exp modification

Posted: 23 Dec 2008, 16:10
by Black Don
Rotonen wrote:IMO levelling should indeed go somewhat smooth at first, but then pick out the enthusiast grinders from the rest of the bunch.

After you have done all the quests all there is left is to grind or chat.

Re: feedback on the latest exp modification

Posted: 23 Dec 2008, 16:36
by Jaxad0127
Quiche_on_a_leash wrote:INT(SQRT((HP)*(LV*0.5)+((ATK1*(0.25+(STR/50)))*(ATK2*(0.25+(LUK/50))))*STR*DEF))
I think it's a good start, but we should also factor in AGI and the others.

Re: feedback on the latest exp modification

Posted: 23 Dec 2008, 18:29
by Crush
There are also a lot of other factors which have to be taken into account. Speed, aggressive or not, helper or not, usefulness of drops, area and density where the monster occurs and so on.

I think trying to calculate monster exp using a formula won't get us far.

Re: feedback on the latest exp modification

Posted: 23 Dec 2008, 18:39
by Jaxad0127
Since monsters can be anywhere, and we should ignore drops as they may change, using the rest of the mob DB should be enough for exp.

Re: feedback on the latest exp modification

Posted: 24 Dec 2008, 04:47
by radiant
I've tried to make some improvements to the equation, from the perspective of figuring out a proper "XP per HP" ratio for each monster. After putting in terms for several facets that contribute to monster difficulty--how much damage it deals, how hard it is to hit, and how easily it can pursue someone who tries to get away, I found that a cubic (or at least a 3/2 power) curve makes for a pretty good fit, and offer the equation on this sheet.

Right now enemy VIT isn't represented in the equation at all (it seems like such a minor term compared to DEF anyway), aggression is simply a flat 1.1x boost with no 1.05x or anything for the evil mushroom (how do monsters that have both helper and aggressive flags set differ from those that just have the aggressive flag?) or for the santa slime behavior that's not in the mob DB and appears to be implemented as a spell. Also the flower is still enough of an oddball that the regression line might need to account for it a bit more.

Re: feedback on the latest exp modification

Posted: 24 Dec 2008, 16:41
by fate
Hi,

(apologies for my extended absence.) The principal factor I used to decide the XP for medium-level monsters (rudolph slimes to grass snakes) was the number of attacks an optimised very-high-level character would need to kill that monster-- at the end of the day, our monsters are so weak (overall) that this is the only thing that really counts.

How about a more radical rebalancing that greatly reduces the number of monsters and makes them more dangerous on average (particularly the medium-level ones)? They could then give an order of magnitude more XP, too.

Another idea: mutations right now (for a mutation value of n) have their average at n/2, with a triangular probability distribution (mirrored at the zero point). How about a Gaussian distribution with the mean at zero and the variance at n, or a triangular distribution with the tail ending at n and -n, respectively?

-- fate

Re: feedback on the latest exp modification

Posted: 24 Dec 2008, 16:49
by Elrrohir
I agree with doorsman and MK and others when it comes to the "not able to level up alone any more". Since the change of the def points for armors developers have been coming out with krapy ideas. In this case, with the older schema i knew how many arrows it'll take to kill a snake and the exp reward, now i can't deal with more than 4 snakes alone in the desert cause i'll be dead meat for sure and, in the worst case, i'll use 6 arrows to get 120XP points :( SUCKS GUYS! even worst than the armor def points. I would like to understand what are u trying to do exactly, but i can't... as far as i can see u are forcing us to play in groups, there's no upgrade for the archers weapon, i spend lots of money in arrows and i still miss lots of shots even having dex to 99, i have been hit by snakes out of my range and killed too many times. so far, theres no difference between an archer like me in lvl 60 and a swordsman in level 40... he even have more chances than i. It's clear that the only thing developers want from us players is feedback regarding bugs, but they will not change anything else no matter what. Ok, is your game after all.
I've played this kind of games for over 10 years and i have never seen something like this.

do u want this game to be played by many players and make it popular or what?

Elrrohir

Re: feedback on the latest exp modification

Posted: 25 Dec 2008, 09:33
by Arilious
fate wrote: Another idea: mutations right now (for a mutation value of n) have their average at n/2, with a triangular probability distribution (mirrored at the zero point). How about a Gaussian distribution with the mean at zero and the variance at n, or a triangular distribution with the tail ending at n and -n, respectively?
I'm able to understand your meaning, but perhaps you could explain it in less mathematical terms for the not-so-math inclined?

As for Gaussian, this is a normal bell curve shape, with a standard deviation of sqrt(n), or the square root of the mutation value.

Image

Where normal monsters are at n=0, which would be the center of the graph.

I think people would prefer it if perhaps you set the mean at something slightly higher than 0, even at .5, therefore you'd have only a slightly higher probability of getting a monster with greater exp, than with lower exp. Otherwise I'm not sure what a Gaussian distribution would do except make things perfectly balanced, if not actually lower the chances of getting a good mutation.

P.S. Fate could you give the probability density function currently used?

Edit: After further thinking, perhaps the probability distribution function for this new curve could be relative to the player level versus monster level? That way it wouldn't be the same, and higher level monsters could be adjusted according for higher level players? Just a thought.