Chance to hit (accuracy) - does not sense in 3.1 (?)

The place to discuss balance changes for future versions of the game.
(Master releases & 3.X)
Post Reply
Reg312
Regular
Regular
Posts: 681
Joined: 25 Mar 2011, 18:36

Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Reg312 » 16 Feb 2012, 15:02

Chance to hit (accuracy)

In Warzone each weapon have 2 parameters:
1) Short hit chance %
2) Long hit chance %

Example: howitzer have 40% short accuracy and 50% long accuracy
this means what howitzer have 40% chance to hit target in close distance (from 0 to 15 tiles)
and 50% chance to hit target in long distance (from 16 to 39 tiles)

how works accuracy (simplified)
- if we are lucky, then weapon shot hits target, ex: howitzer lucky only 1 time from 2
- if we are unlucky, then weapon hits place very close to target, ex: howitzer have wide range of damage, so even if we was unlucky. howitzer hit target anyway
- missed shot in WZ is shot in random tile close to target

what wrong with accuracy:
1) Accuracy does not depend on range to target, we get some weird cases:
1.1 Flamer cannot hit 1 unit from 0 range (close up), because 1 tile of unlucky shot makes fire in another direction
accuracy angle fo flamer if from +45° to -45°
1.2 Artillery very precise.. Ripple Rockets can precise hit moving hover unit on another side of map
1.3 Ground shaker have 2 tiles of area of fire
so if ground shaker missed hit and will damage target anyway
1.4 if some modder will set 0% hit chance most of weapons still hit very precise!
1.5 its100% chance to hit target with large size, even if you set 0% chance to hit

What i suggest to do:
1) Change accuracy formula: Hit place of missed shot should depend on range to target
for example: if ripples rockets have missed shot then it hit to tile in wide area
..if range 10 tiles, then missed shot hits place in range of 1 tile from target (lancer, HPV cannon)
..if range 4 tiles, then missed shot hits place in range of 0.4 from target (flamer, mg)
..if range 70 tiles, then missed shot hits hit place in range of 7 from target (ripples)

2) <...??> remake accuracy system

Per
Warzone 2100 Team Member
Warzone 2100 Team Member
Posts: 3763
Joined: 03 Aug 2006, 19:39

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Per » 16 Feb 2012, 15:13

In my opinion, the long/short accuracy separation is insane and should be dropped. Nobody will remember them since there are too many weapons in the game, and the difference is too subtle, and they will therefore have no effect on actual tactics.

Reg312
Regular
Regular
Posts: 681
Joined: 25 Mar 2011, 18:36

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Reg312 » 16 Feb 2012, 15:56

Per wrote:In my opinion, the long/short accuracy separation is insane and should be dropped.
current accuracy formula insane a bit more.
why you setted 1 tile range of missed shot? why not 1.5 why not 2?? who decided why 1 and why he decided it?
Per wrote:Nobody will remember them since there are too many weapons in the game
Players dont need remember accuracy! if player uses artillery and see what his projectiles hit inaccurately its ok!
Accuracy is visible! its animated! it's clear to understand even for child.
for example damage (main and vital part of game) is invisible
unit can make 1000 or 100 damage with same animations
but if we have less accuracy - we see it
when we researched laser designator we can see what our units started shoot more precise
Per wrote:and the difference is too subtle
subtile? did you seen why Assault Gun 30-40% stronger in 3.1 version? possibly problem is increased accuracy
AG became OP weapon (or close to OP)
Per wrote:and they will therefore have no effect on actual tactics.
it has great effect on tactics, some examples:
1) unaccurate weapon (flamer) very good for large mases droids
2) fast units should have ability to avoid artillery file.
3) artillry very strong in high oil games because it have 100% accuracy and emeny armies have no chances
4) if enemy tank entered your base, then if you bombard this tank you could lose your base and it is OK
5) if you fight with enemy army and your artillery started fire you can lose also your army, because artillery is inaccurate

accuracy is nice way to make some thing is game more balanced, more real, less OP, and more variants of tactics and maneuvers

User avatar
aubergine
Professional
Professional
Posts: 3459
Joined: 10 Oct 2010, 00:58
Contact:

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by aubergine » 16 Feb 2012, 18:20

I have to agree with Reg312 in that accuracy plays a big part in the game, it's one of the few things where you see very visible and visual results from your research.

Think about mortars and MRB for example - their initial accuracy is very bad, they often completely miss the target. So you do research to increase accuracy and you actually see it taking effect in the game.

Whether accuracy has an actual impact on unit damage or not, it has a huge psychological impact in that you get direct gratification from your research efforts.
"Dedicated to discovering Warzone artefacts, and sharing them freely for the benefit of the community."
-- https://warzone.atlassian.net/wiki/display/GO

Per
Warzone 2100 Team Member
Warzone 2100 Team Member
Posts: 3763
Joined: 03 Aug 2006, 19:39

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Per » 16 Feb 2012, 18:43

Not sure why you spend so much time arguing that accuracy is important. Did anyone say otherwise? :?

One problem with accuracy currently is that when you see it missing, you don't know whether that is due to accuracy or simply because the enemy moved slightly. In games where projectiles always hit when your accuracy roll succeeds, you can tell how good your accuracy is by looking at your screen. In Warzone, you cannot. (Note that you could do that in 1.10, since its projectiles always hit their target on a successful accuracy roll.)

Another problem is, as I pointed out above, that the current short/long distinction is very arbitrary, complicated, hard to remember and also not visualized anywhere.

User avatar
aubergine
Professional
Professional
Posts: 3459
Joined: 10 Oct 2010, 00:58
Contact:

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by aubergine » 16 Feb 2012, 19:01

I think both myself and reg312 read your comment wrong, we missed the "short/long range" bit and read it like this:
In my opinion, the long/short accuracy separation is insane and should be dropped. Nobody will remember them since there are too many weapons in the game, and the difference is too subtle, and they will therefore have no effect on actual tactics.
Our bad.

That being said, and I'd have to go back a play test with a buch of base defences I've not used for a while, I think there are some weapons that do give strong visual feedback of them being useless at short range. IIRC mortars go mental when trying to hit a short range target, I think that's why I always try building them on hills or a fair bit behind walls. So maybe the distinction is still important, but people who've been playing the game for a while have learnt to deal with its effects by changing where they place defences, etc., and are probably no longer acutely aware of how those weapons behaved when they first started using them?
"Dedicated to discovering Warzone artefacts, and sharing them freely for the benefit of the community."
-- https://warzone.atlassian.net/wiki/display/GO

Reg312
Regular
Regular
Posts: 681
Joined: 25 Mar 2011, 18:36

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Reg312 » 16 Feb 2012, 19:09

ok, i've read wrong :)
long/short can be removed but how you set that mortar should miss in close up fight
my topic was about making missed shots area to be dependent on range or on specific parameter in stats

User avatar
Iluvalar
Regular
Regular
Posts: 1785
Joined: 02 Oct 2010, 18:44

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Iluvalar » 16 Feb 2012, 20:08

The game mechanic should be made so when the stat files say "70%" it IS 70%.

Not 70+(100-70)*atan(size of target/distance)*(100-70)*30%

or whatever it is right now...
Heretic 2.3 improver and proud of it.

RBL-4NiK8r
Trained
Trained
Posts: 132
Joined: 24 Oct 2007, 22:04

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by RBL-4NiK8r » 17 Feb 2012, 00:34

Well one of the cool things you can do is make weapons act differently, take the Cluster Bomb give it 2 reloads of 50 so bombs with a damage of 15 to 25 and 1 second burn time no burn damage and change the accuracy to 20% and watch all the bombs fly all over the place just like a real Cluster Bomb. As I recall you did have a few weapons that were better at long distance then the close it or was it the other way around.


4nE

Reg312
Regular
Regular
Posts: 681
Joined: 25 Mar 2011, 18:36

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Reg312 » 17 Feb 2012, 10:07

Iluvalar wrote:The game mechanic should be made so when the stat files say "70%" it IS 70%.

Not 70+(100-70)*atan(size of target/distance)*(100-70)*30%

or whatever it is right now...
Warzone is not RPG, it's unreal when units hits in all directions with equal chance but cannot hit in direction of target (case when accuracy 0%)
i think accuracy should be changed to be real, to use some kind of real model of accuracy

Searge-Major
Trained
Trained
Posts: 182
Joined: 10 Sep 2011, 03:36
Location: Great Southern region, Western Australia

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Searge-Major » 17 Feb 2012, 12:11

Reg312 wrote: Warzone is not RPG, it's unreal when units hits in all directions with equal chance but cannot hit in direction of target (case when accuracy 0%)
i think accuracy should be changed to be real, to use some kind of real model of accuracy
Perhaps a function similar to this would be realistic?

Given a base accuracy, ie. 80%.

Accuracy decreases from maximum to 80% at optimum range. (optimum range being best compromise between range and accuracy, probably around 70-80% of maximum range)

The decay function would ideally be a root function (either square or cube), to provide a slow non-constant decay inside optimum range and a faster (also non-constant) decay outside optimum range.
Accuracy function.png
Accuracy graph for a weapon with an accuracy of 80%, at 100% range, accuracy has dropped to 60%. Unless the weapon has no minimum range, the accuracy will never be 100%, unless base accuracy is also 100%.
Accuracy function1.png
The function in game would use f(x) = 1-(sqr(x)*(1-weapon_accuracy)*2). For a weapon with 50% accuracy, at absolute maximum range accuracy would be nil. Therefore a better function has to be found.
Accuracy function1.png (3.39 KiB) Viewed 6213 times
Find the target location, (for target computers, find the location where the target will be at impact, provided it's movement is constant), and set the firing angle based on that.

Note: the location found is the centre of the target, irrespective of size, meaning that base structures, esp. large ones, are practically impossible to miss.

Have a set maximum miss angle, maybe even one for each weapon, or perhaps define it by a constant multiplied by the inverse of the weapon's accuracy. If the constant is 60 degrees, for a unit with 80% accuracy, max miss angle would be 0.2*90, or 12 degrees, (better than 45 :) ).

Using standard deviation based on the decay function, and the max miss angle, generate a firing angle, and fire. :) Job done; hopefully someone understands all this... :wink:

This is probably best suited to direct fire weapons, but could probably be adapted for arty as well.
Now to find someone to do all the hard work of coding this... XD

:geek: Must learn a programming language... :geek:
I fully realise my input is just another drop in the proverbial bucket. It is my goal to make each drop count.

User avatar
Iluvalar
Regular
Regular
Posts: 1785
Joined: 02 Oct 2010, 18:44

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Iluvalar » 17 Feb 2012, 17:58

reg312 : I don't think it will be more enjoyable if we reduce the HP of the cyborgs by half, because they are smaller and make the weapons miss them twice as much. Yeah it would be more realistic, but at the end it would be just harder to compute for the players.

Searge ideas are good, as long as we fit the funciton with the stat data, I have no objection for an accuracy that decay over distance.

But I have objections against a game mechanic that try to miss behind the target :D .
Heretic 2.3 improver and proud of it.

Reg312
Regular
Regular
Posts: 681
Joined: 25 Mar 2011, 18:36

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Reg312 » 17 Feb 2012, 21:13

Iluvalar wrote:reg312 : I don't think it will be more enjoyable if we reduce the HP of the cyborgs by half, because they are smaller and make the weapons miss them twice as much. Yeah it would be more realistic, but at the end it would be just harder to compute for the players.
1. much weapons have splash damage
2. i suggested make larger area of missed shots
it still can have % of hit chance, but if we fail dice then it should miss more real
3. every projectile also can hit another target in cyborg masses
...
cyborgs is cyborgs, if they better by accuracy and not by modifiers its better

i believe is not so hard to implement kinda "real" accuracy model, i have not math background, so i dont know what model better
i can say it should depend on:
*target size
*range to target
*specific of weapon (rockets dont vibrate when launch, for example)
<...> any more?

Searge-Major
Trained
Trained
Posts: 182
Joined: 10 Sep 2011, 03:36
Location: Great Southern region, Western Australia

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Searge-Major » 18 Feb 2012, 13:36

Iluvalar wrote:Searge ideas are good, as long as we fit the funciton with the stat data, I have no objection for an accuracy that decay over distance.

But I have objections against a game mechanic that try to miss behind the target :D .
Sorry, either you misunderstood me, or you can see something in the function that I can't. :hmm:
Reg312 wrote: i believe is not so hard to implement kinda "real" accuracy model, i have not math background, so i dont know what model better
i can say it should depend on:
*target size
*range to target
*specific of weapon (rockets dont vibrate when launch, for example)
<...> any more?
Sorry Reg, I had trouble understanding your post, :( but I hope this answers your questions.

The accuracy my function was giving was based on the centre of the target location, or in the case of target prediction, where the centre of the target will be. This means larger units are automatically easier to hit, as the accuracy is based on a standard deviation, which basically means more shots will land close to the target than further away. (Basically, it's actually more complicated than that, but you should get the general idea :) )

Right.... any more questions? :roll:

:wink:
I fully realise my input is just another drop in the proverbial bucket. It is my goal to make each drop count.

User avatar
Iluvalar
Regular
Regular
Posts: 1785
Joined: 02 Oct 2010, 18:44

Re: Chance to hit (accuracy) - does not sense in 3.1 (?)

Post by Iluvalar » 18 Feb 2012, 17:19

who will multiply the HP with the size of the pie ? To fit the stats with your new reality ?

Searge, it's the actual behavior in 3.1 to try to miss trough the target. not your suggestion.
Heretic 2.3 improver and proud of it.

Post Reply