|
|
Post by shakes on Jan 11, 2019 16:07:25 GMT -5
How would you then convert that to a hit or miss?
|
|
|
|
Post by jcarter on Jan 11, 2019 16:37:12 GMT -5
How would you then convert that to a hit or miss? if attack < defense, miss if attack > defense, hit |
|
|
|
Post by shakes on Jan 11, 2019 16:45:33 GMT -5
Ah. So straight forward once you've extracted the probability number. I get it now. Satisfied.
My tendency to overthink code issues always rears its head.
|
|
|
|
Post by jcarter on Jan 11, 2019 17:00:55 GMT -5
Ah. So straight forward once you've extracted the probability number. I get it now. Satisfied. My tendency to overthink code issues always rears its head. the problem is how distributions are implemented. if you have a weapon that deals 2-12 damage, depending on a random number picked, you have a 10% chance to deal 2 damage and a 10% chance to do 12. the average damage is going to be 6, but that still also only has a 10% chance. if you roll 2d6, you get something approaching a normal distribution: you can still do between 2-12 damage, but the chance that you'll roll 2 or 12 is much less than getting a 6. if you just implement a normal distribution, you can set it up to fine-tune what chance you want to get those outliers. this image kind of demonstrates it, at least comparing 1-12 vs 2-12. for the comparison to hold, the 1-12 should be a 2-12 range and it should be 10% instead of 8%.  the point of it is to reduce the variability. |
|
|
|
Post by shakes on Jan 11, 2019 17:18:00 GMT -5
If you successfully reduce the variability, how then do you manage extra-combat functions such as bash, kick, flee? When I coded in a combat system, I included (overcomplicated perhaps) hit locations and damage percent reductions. If you were at 50% of your hp, you fought at 50% of your ability. Hit locations (arms, legs, torso, head) had damage modifiers but also individual hit locations had their own hp and could be both targeted and disabled. That was my attempt to give players something to do in combat that was more "arcade" than simply watching the rounds roll by. I also didn't obscure the numbers. It would read, "You swing at a bandit and miss (roll 7, need 10)." It was fascinating to code, but according to players and reviews, boring to play.  But part of that was maybe the PK atmosphere of the game. When combat's the goal and adrenaline the reward, you'd better keep it all flowing as fast as you can. I perhaps applied the wrong code to the wrong type of game. |
|
|
|
Post by jcarter on Jan 11, 2019 17:33:36 GMT -5
If you successfully reduce the variability, how then do you manage extra-combat functions such as bash, kick, flee? When I coded in a combat system, I included (overcomplicated perhaps) hit locations and damage percent reductions. If you were at 50% of your hp, you fought at 50% of your ability. Hit locations (arms, legs, torso, head) had damage modifiers but also individual hit locations had their own hp and could be both targeted and disabled. That was my attempt to give players something to do in combat that was more "arcade" than simply watching the rounds roll by. I also didn't obscure the numbers. It would read, "You swing at a bandit and miss (roll 7, need 10)." It was fascinating to code, but according to players and reviews, boring to play. But part of that was maybe the PK atmosphere of the game. When combat's the goal and adrenaline the reward, you'd better keep it all flowing as fast as you can. I perhaps applied the wrong code to the wrong type of game. i'm not sure what you're asking. changing the distribution and stats doesn't change extra combat functions. the complaint it's meant to address is that people are getting 'extremely lucky' rolls when stats are allegedly identical, and do not like that level of randomness. you can do that by using anything other than a uniform distribution. |
|
|
|
Post by shakes on Jan 11, 2019 17:53:28 GMT -5
I see it as a linear "increased complexity makes combat rounds take longer". Which I suppose doesn't have to be universally true.
But if it is, then you have extra reaction time for people to flee out or do other shenanigans.
|
|
|
|
Post by jcarter on Jan 11, 2019 18:06:03 GMT -5
I see it as a linear "increased complexity makes combat rounds take longer". Which I suppose doesn't have to be universally true. But if it is, then you have extra reaction time for people to flee out or do other shenanigans. it can make combat rounds take longer, or it can make combat rounds shorter. it's however you choose to set the means and distributions. |
|
|
|
Post by pinkerdlu on Jan 12, 2019 0:56:03 GMT -5
I mean Atonement and SoI did it pretty well and both were just as popular as Arm, back in the day...
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 12, 2019 22:23:15 GMT -5
You guys seem to know arm code better then 85% of all staff. Is changing these things in arm a major undertaking?
|
|
|
|
Post by shakes on Jan 12, 2019 22:26:33 GMT -5
A major undertaking? Not especially.
Balancing though? That'd take a lot of time and a lot of crying.
|
|
|
|
Post by jcarter on Jan 13, 2019 11:59:16 GMT -5
You guys seem to know arm code better then 85% of all staff. Is changing these things in arm a major undertaking? lmao ah yes, the esoteric knowledge of pointing out a random number generator. |
|
|
|
Post by jcarter on Jan 13, 2019 11:59:42 GMT -5
qwerty do 85% of arm staff really not understand what a dice roll is???
|
|
|
|
Post by shakes on Jan 13, 2019 12:51:34 GMT -5
Yeah, the combat code isn't really all that complicated in any Diku-base. And from what I saw of the code dump here it's all self-contained.
You could bypass it or just cut it all out with relative ease.
Compared to that dude's work on changing all the classes, changing the combat code would be a weekend project for him.
Where it would go wrong is Arm's culture of secrecy. They wouldn't want to expose the numbers and they'd invite only a handful of their trusted buddies in to do the testing, and they'd end up rolling out something that's probably far more unbalanced than the current model.
|
|
jkarr
GDB Superstar
Posts: 2,070
|
Post by jkarr on Jan 13, 2019 23:26:20 GMT -5
You guys seem to know arm code better then 85% of all staff. why do u think that |
|
