Range vs Accuracy: Let's improve it.
- Stratadrake
- Trained
- Posts: 197
- Joined: 07 Sep 2008, 09:43
- Location: Pacific NW
- Contact:
Re: Range vs Accuracy: Let's improve it.
I was thinking about this the other day and realized that we could probably eliminate the diceroll system on the collision detection end completely.
First, the game calculates a projectile trajectory that should land inside the target's hitbox. Then a diceroll is used to determine whether the projectile is a hit or miss. If it's a miss, then the trajectory is altered to land outside the target hitbox (taking target size into account), causing a miss. With a stationary target, this provides exactly the same end result as an exclusively diceroll system (e.g. original 1.x). Projectiles do still need to individually record their diceroll however, especially homing projectiles as the very nature of their flight path implies that they'll impact the target's hitbox, and we don't want that causing a hit if it's not meant to.
In other words, a diceroll determines whether the projectile is initially hit or miss, but it's physical collision detection that ultimately determines when and where a hit is scored.
I can draw up a visual aid, but it will take some time.
First, the game calculates a projectile trajectory that should land inside the target's hitbox. Then a diceroll is used to determine whether the projectile is a hit or miss. If it's a miss, then the trajectory is altered to land outside the target hitbox (taking target size into account), causing a miss. With a stationary target, this provides exactly the same end result as an exclusively diceroll system (e.g. original 1.x). Projectiles do still need to individually record their diceroll however, especially homing projectiles as the very nature of their flight path implies that they'll impact the target's hitbox, and we don't want that causing a hit if it's not meant to.
In other words, a diceroll determines whether the projectile is initially hit or miss, but it's physical collision detection that ultimately determines when and where a hit is scored.
I can draw up a visual aid, but it will take some time.
Last edited by Stratadrake on 31 Jul 2012, 19:34, edited 1 time in total.
Re: Range vs Accuracy: Let's improve it.
That's exactly how the patch work afaict. Unless you specify what is changed ?
except, the projectile trajectory is calculated just after the dice roll. But that doesn't affect anything except saving calculation time. And that the trajectory take into account the speed of the target. So it should hit if desired even fast movers.
Hmm, there is no diceroll at collision time on any real or proposed model so far.
except, the projectile trajectory is calculated just after the dice roll. But that doesn't affect anything except saving calculation time. And that the trajectory take into account the speed of the target. So it should hit if desired even fast movers.
Hmm, there is no diceroll at collision time on any real or proposed model so far.
Heretic 2.3 improver and proud of it.
- Stratadrake
- Trained
- Posts: 197
- Joined: 07 Sep 2008, 09:43
- Location: Pacific NW
- Contact:
Re: Range vs Accuracy: Let's improve it.
Visual aid. Did I get this right?
Re: Range vs Accuracy: Let's improve it.
And why do we want a dice roll in the first place?
(Since this is a strategy game, and not a simulation, answers concerning realism will be ignored.)
(Since this is a strategy game, and not a simulation, answers concerning realism will be ignored.)
Re: Range vs Accuracy: Let's improve it.
What do you want? an UNREALISTIC strategy game? Maybe we should remove gravity. Just because it is a strategy game does not mean weapons should be perfect every time. In the real world guns miss. In the game, it adds layer of strategy because getting closer lets you land more hits (but you take more hits). So your guns are more accurate and damagingPer wrote:And why do we want a dice roll in the first place?
(Since this is a strategy game, and not a simulation, answers concerning realism will be ignored.)
Proud former soldier of the 39th Infantry
BATTLE OF THE PLAIN OF REEDS - WE WILL NOT FORGET
BATTLE OF THE PLAIN OF REEDS - WE WILL NOT FORGET
Re: Range vs Accuracy: Let's improve it.
We don't want accuracy. We only want a third parameter to go with the DMG and ROF. It's crucial to have a decent array of control on the game.Per wrote:And why do we want a dice roll in the first place?
(Since this is a strategy game, and not a simulation, answers concerning realism will be ignored.)
Nice representation and that's why I fixed:
Image explanation (bug fix patch):
bug: Never "miss" in 90° behind again.
bug: Never "miss" inside the hitbox.
improve: Miss more often close
- Attachments
-
- Accuracy fix
- accuracyfix.png (8.92 KiB) Viewed 4513 times
Heretic 2.3 improver and proud of it.
Re: Range vs Accuracy: Let's improve it.
I'm all for that! Can it be something other than accuracy, though?Iluvalar wrote:We don't want accuracy. We only want a third parameter to go with the DMG and ROF.
Other games often have various armour piercing abilities. I think that might make more sense. Another attribute that is not changed through research is range. Not saying it is wise to let it be upgraded. Just pointing out the possibility. Yet another is splash radius... and I am sure I haven't thought of half the possibilities yet.
Re: Range vs Accuracy: Let's improve it.
In any case some kind of aiming parameter is needed. There is no substitude to SEEing your misile not hitting or hitting the target, when in range. After all this is a game, not a mathematical simulation...
Ok, Ilunavar, how about adding a trigonometry calculation to reduce BaseSpread according to distance? This is not realistic, but will give the desired results, I think ...
Ok, Ilunavar, how about adding a trigonometry calculation to reduce BaseSpread according to distance? This is not realistic, but will give the desired results, I think ...
Re: Range vs Accuracy: Let's improve it.
This game HAVE an armour piercing, Based on damage and armour. That's exactly why we can't leave the DMG and ROF alone.Per wrote:I'm all for that! Can it be something other than accuracy, though?Iluvalar wrote:We don't want accuracy. We only want a third parameter to go with the DMG and ROF.
Other games often have various armour piercing abilities. I think that might make more sense. Another attribute that is not changed through research is range. Not saying it is wise to let it be upgraded. Just pointing out the possibility. Yet another is splash radius... and I am sure I haven't thought of half the possibilities yet.
The range and the splash radius have no direct multiplicative effect with the DMG/ROF duo. It's not viable solution.
iap, not sure what you mean, in my patch I made the variable worstShot (distance of the landing miss from the border of the target) multiplied by the distance. Is that what you meant ? It make the misses act exactly like there was an angle of fire.
Heretic 2.3 improver and proud of it.
Re: Range vs Accuracy: Let's improve it.
1=
current accuracy system is totally bugged, so i cant imagine how Iluvalar's patch can be worse
if balance issues will appear, then it will be fixed easier
i think we need common methods and program tools for balance testing
for example script which creates 2 armies and forced them to fight over and over
many people concerning balance, even whose dont play multiplayer
this game have many features which concern realism
2=
i notice Artillery is ignored in this thread
Accuracy have higher strategic value for artillery
Warzone have many thing hidden for most players
1) Armor/Hitpoints - many players dont know diff
2) Splash damage, incendiary damage - most players dont know how it works
Accuracy is not hidden because players can see trajectories of projectiles
@Per
Accuracy should not be considered as Randomness
1) Some weapons can be more effective with lower accuracy
(flamers, heavy cannons, artillery)
2) Unaccurate weapons can work better against large masses on enemy units
hm, i see accuracy system conflicts with new prophetic targeting...
current accuracy system is totally bugged, so i cant imagine how Iluvalar's patch can be worse
if balance issues will appear, then it will be fixed easier
i think we need common methods and program tools for balance testing
for example script which creates 2 armies and forced them to fight over and over
many people concerning balance, even whose dont play multiplayer
lets community to decide what to ignorePer wrote:(Since this is a strategy game, and not a simulation, answers concerning realism will be ignored.)
this game have many features which concern realism
2=
i notice Artillery is ignored in this thread
Accuracy have higher strategic value for artillery
Warzone have many thing hidden for most players
1) Armor/Hitpoints - many players dont know diff
2) Splash damage, incendiary damage - most players dont know how it works
Accuracy is not hidden because players can see trajectories of projectiles
@Per
Accuracy should not be considered as Randomness
1) Some weapons can be more effective with lower accuracy
(flamers, heavy cannons, artillery)
2) Unaccurate weapons can work better against large masses on enemy units
hm, i see accuracy system conflicts with new prophetic targeting...
Re: Range vs Accuracy: Let's improve it.
Iluvalar, what I meant was to make the spread angle more accurate whhen the target is more distant. This means that correct calculations can make the spread be the same for near and far units. The difference from your patch is that it bypass the dice roll without making calculations more complicated, and without the need to be carefull not to shoot behind if the dice said miss, or to always hit big targets...
Re: Range vs Accuracy: Let's improve it.
Do you play Warzone 2100?Per wrote:I'm all for that! Can it be something other than accuracy, though?Iluvalar wrote:We don't want accuracy. We only want a third parameter to go with the DMG and ROF.
Other games often have various armour piercing abilities. I think that might make more sense. Another attribute that is not changed through research is range. Not saying it is wise to let it be upgraded. Just pointing out the possibility. Yet another is splash radius... and I am sure I haven't thought of half the possibilities yet.
Proud former soldier of the 39th Infantry
BATTLE OF THE PLAIN OF REEDS - WE WILL NOT FORGET
BATTLE OF THE PLAIN OF REEDS - WE WILL NOT FORGET
- Stratadrake
- Trained
- Posts: 197
- Joined: 07 Sep 2008, 09:43
- Location: Pacific NW
- Contact:
Re: Range vs Accuracy: Let's improve it.
Sounds like we have a debate over what the concept of "accuracy" actually is.
If it's simply a probability of a given shot being a hit or miss (which is precisely what we had in original 1.x) then the game engine needs to do its best to ensure that hits hit and misses miss. If the actual hit/miss is determined by hitbox collision then the game MUST ensure that misses are NOT fired at such trajectories that they will touch the expected hitbox. This means, like Illuvalar says, that a miss MUST land outside the hitbox, and cannot land BEHIND the hitbox, because both of those situations will cause a false hit, resulting in artificially boosted weapon accuracy.
If it's simply a "scatter" that determines how far away the shot lands from the intended target, then it's not a chance-to-hit at all because the actual chance-to-hit will vary based on variables like range and target size and cannot be predicted in advance.
I certainly wouldn't mind units having some kind of specified size modifier that is incorporated into accuracy formula (like a 0.75x accuracy modifier against cyborgs, since they really are smaller targets than tanks), but that is a separate matter.
If it's simply a probability of a given shot being a hit or miss (which is precisely what we had in original 1.x) then the game engine needs to do its best to ensure that hits hit and misses miss. If the actual hit/miss is determined by hitbox collision then the game MUST ensure that misses are NOT fired at such trajectories that they will touch the expected hitbox. This means, like Illuvalar says, that a miss MUST land outside the hitbox, and cannot land BEHIND the hitbox, because both of those situations will cause a false hit, resulting in artificially boosted weapon accuracy.
If it's simply a "scatter" that determines how far away the shot lands from the intended target, then it's not a chance-to-hit at all because the actual chance-to-hit will vary based on variables like range and target size and cannot be predicted in advance.
I certainly wouldn't mind units having some kind of specified size modifier that is incorporated into accuracy formula (like a 0.75x accuracy modifier against cyborgs, since they really are smaller targets than tanks), but that is a separate matter.
Last edited by Stratadrake on 01 Aug 2012, 18:37, edited 1 time in total.
Re: Range vs Accuracy: Let's improve it.
What is wrong with a scatter distribution? Simply specify a sigma for each weapon, and this determines how accurate it is.Stratadrake wrote:Sounds like we have a debate over what the concept of "accuracy" actually is.
...
If it's simply a "scatter" that determines how far away the shot lands from the intended target, then it's not a chance-to-hit at all because the actual chance-to-hit will vary based on variables like range and target size and cannot be predicted in advance.
High sigma = inaccurate.
Perhaps visualize the horizontal axes as angular offset from the "right" track, and G as the probability of a given trajectory.
Last edited by JJjopando on 01 Aug 2012, 18:29, edited 2 times in total.
Proud former soldier of the 39th Infantry
BATTLE OF THE PLAIN OF REEDS - WE WILL NOT FORGET
BATTLE OF THE PLAIN OF REEDS - WE WILL NOT FORGET
Re: Range vs Accuracy: Let's improve it.
Of course it's not a chance to hit, because chance to hit must have a dice roll!
However, maybe it can give more varied or nicer looking game dynamics, it might even solve the prophet droid bug.
Chance to hit is something of D&D games or TBS...
However, maybe it can give more varied or nicer looking game dynamics, it might even solve the prophet droid bug.
Chance to hit is something of D&D games or TBS...