Page 4 of 13
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 05 Mar 2012, 19:28
by iap
Hi, sorry to burst in like this, but why not calculating the exact ballistic targeting, and then just add some randomness to the initial angle and power?
Having my units "run away" from a long distance artillery shot directly to their location is best! Seeing an enemy unit maneuver itself out of the targeted hitting place of my artillery of my weapons looks good too. That's the reason I always build artillery in masses.
As for direct weapons - angle variation should do the trick. Hi, slight angle change give small distance near by and large distance from far away - so there is no need for short and long distance difference.
I personally don't mind small units like cyborgs being harder to direct hit, after all, all the long distance weapons have a splash radius.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 08 Mar 2012, 03:30
by effigy
Per wrote:My personal opinion is that accuracy and misses are unnecessary. Just make everything hit and watch a whole category of quite hard issues go up in smoke. Then we can have fun playing the game instead.
It's hard to say if this would negatively impact the game long term, but would need to come with a major rebalance, imo.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 17 Mar 2012, 09:30
by bendib
Accuracy should depend on distance, as an absolute.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 17 Mar 2012, 10:48
by Emdek
bendib, for me it should depend on experience too (only a bit, but more experienced units in real world usually target better
).
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 17 Mar 2012, 16:05
by effigy
I think increased rank giving better accuracy is suppose to be reward for keeping that unit alive.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 17 Mar 2012, 17:29
by aubergine
Per wrote:My personal opinion is that accuracy and misses are unnecessary. Just make everything hit and watch a whole category of quite hard issues go up in smoke. Then we can have fun playing the game instead.
If weapons hit all the time, the game becomes too mechanical. It's a warzone, not a fairground "shooting ducks" game (and even there you miss some of the time).
I think misses should be dealt with in as simple a manner as possible for performance reasons (if it missies, I don't really care by how much) but should not be removed from the game.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 17 Mar 2012, 17:57
by iap
Emdek wrote:bendib, for me it should depend on experience too (only a bit, but more experienced units in real world usually target better
).
Fully agree (except that I think it should be more then a bit)
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 17 Mar 2012, 18:00
by Emdek
This part is relative, we could get a poll for how big it should be alone.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 20 Mar 2012, 21:47
by Jorzi
aubergine wrote:I think misses should be dealt with in as simple a manner as possible for performance reasons (if it missies, I don't really care by how much) but should not be removed from the game.
I partly agree, but this is not the case for artillery though, you want to have a certain accuracy-dependent spread.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 23 Apr 2012, 23:32
by Reg312
found following code for missed shots in combat.cpp:
Code: Select all
int missDist = 2 * (100 - resultHitChance) + minOffset;
Vector3i miss = Vector3i(iSinCosR(gameRand(DEG(360)), missDist), 0);
predict += miss
this code means: radius of missed shots depends on chance to hit?
just saw code in debug, seems it works as i expected, but i did not check it deeper
for weapon with 75% hit chance all missed shots will go to area with 55 radius near target (its lesser than 1/2 of tile)
may be i'm very wrong? i dont know
i'm not good with mathematical stuff
e.g. machinegun have 50% chance to hit => we will get missed shot in radious of 105 points to center of target? its lesser than 1 tile (tile = 128 points)
so for most weapons we have only 1/2 tile radius of missed shots and this means 100% accuracy for most weapons?? hm
[upd]
seems thing with missed shot radius depending on hit chance was in all old versions,
this still does not explain why units became more accurate in 3.1/master
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 23 Apr 2012, 23:48
by Giani
That would explain why researching acuracy for artillery doesn't change anything.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 24 Apr 2012, 00:30
by Reg312
i hope Cyr or Per can say what i'm wrong or explain why chance to hit target also defines radius of missed shots
i still dont understand what was done with accuracy in 3.1, i want get it
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 24 Apr 2012, 00:53
by Giani
Reg312 wrote:i hope Cyr or Per can say what i'm wrong or explain why chance to hit target also defines radius of missed shots
i still dont understand what was done with accuracy in 3.1, i want get it
Bug maybe?
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 24 Apr 2012, 00:56
by Giani
bendib wrote:Accuracy should depend on distance, as an absolute.
Emdek wrote:bendib, for me it should depend on experience too (only a bit, but more experienced units in real world usually target better
).
I agree whit both. And about distance if it is very close(1-6 tiles) and VERY far it should miss 3/7 of shots IMO.
Re: Chance to hit (accuracy) - does not sense in 3.1 (?)
Posted: 24 Apr 2012, 01:08
by iap
Giani wrote:That would explain why researching acuracy for artillery doesn't change anything.
What? This is not the only thing that doesn't change, or is it?
Damn, researching upgrade is like pseudo pill …