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

[NoQ]

You are talking about making the borgs 12x harder to hit compared to a python.

This only applies to a single borg. A massive group of borgs will receive its 100% hits anyway, regardless of accuracy. Well, single borgs are usually not much of a challenge anyway.

[Searge-Major]

You are talking about making the borgs 12x harder to hit compared to a python.

This only applies to a single borg. A massive group of borgs will receive its 100% hits anyway, regardless of accuracy. Well, single borgs are usually not much of a challenge anyway.

I mostly agree with that, but it does seem assume that multiple borgs will be grouped closely. However, I also think that the accuracy stats will need to be upgraded to adjust for the new function, as machine-guns would become hopelessly inaccurate at maximum range, approaching 0%. With an increase to 60 or 65%, machine guns would become more like they are now, but with a better accuracy function.

This is rather unrelated, but I was thinking that perhaps the damage of mgs could be halved, and to compensate they could fire salvos of six shots at a time, maybe up to 10 for assault gun. This would have the effect of a lot more shots being fired, and compensating to some degree for lack of accuracy. A secondary effect would be a somewhat more realistic set of weapons.

Nevertheless, units like cyborgs will be harder hit. But, I think the overall effect would be not too noticeable. According to standard deviation, 68% of shots will land within one deviation, and at just over half a deviation, 50% of shots will land within that area. Naturally the value of a deviation would vary with individual weapon accuracies, but if mgs are given salvo fire, the negative effect should be countered.

If anyone doesn't like these ideas, that's completely fine, all I'm trying to do is to provide a solution for this problem.
Anyone else is more than welcome to come up with different ideas, or to improve on those already stated.

[Iluvalar]

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)
(that's reg's)

2) Make the miss "hit" misses in the x/y axis not around the z axis. You cannot miss behind a target !!

3) Make an hit an hit and a miss a miss. The accuracy roll should seal the faith of the projectile. Never draw a miss inside the target hitbox, and make all the projectiles autotargeting if the prediction is not efficient enough.

4) NoQ's gaussian distribution for the _MISSES_ seem nice.

5) A continous function for the accuracy seem better that the defined step. As long as the long range and short range at the defined range still follow mostly the actual balance.

6) DO NOT make the accuracy increase with target size !! DO NOT DO NOT!!! It will only result in balance changes to increase the HP or larger target and decrease the HP of smaller target and it will become tedious to assess the resistance of a unit. (assuming the magic modder will come on his flying carpet and do the "easy" balance change by solving the unit density vs gaussian distrubution problem [which is btw impossible unless size and prices are jammed together by a continuous function]).
6-1) Instead just pretend that the accuracy/size problem is solved into the HP mechanic itself. IE : bigger units have already lost HP because the are easier to hit.

[Searge-Major]

Would you be able to use this in your mod? Just that it would be nice to be able to test. I don't know what would be involved, but I think that the best way to choose a "winner" is to test them in practice. Certainly I agree with your comments about balance though.

[camellia0x++]

My suggestion is to do this. Use real location for it.

If the graphic hits, do damage. Accuracy should work as it does in reality... Only suggestion is to check for walls so you don't end up hitting yourself with motion prediction...

[Iluvalar]

Did you just read my point 6) ? Or do you just dont care about the game being playable ?

[NoQ]

I'd prefer to make smaller things harder to hit, but rebalance cyborg HP accordingly (but don't rebalance small body HP, since they aren't usable now, so this will hopefully make them usable). Well, then we will need some rebalancing, of course, but this might be worth it in the long run (?)

P.S. every function defined on a discrete topological space is continuous (:

[Iluvalar]

Yeah hopefully ! Games are balanced with hope and prayers .

Who cares Iluvalar say it's impossible ? He only made 2 techtree, a dozen of mods and the last balance changes.

The magic fairy will come, will make massive change that will totally alter the whole game keep the game as it is even if we completely change the core mechanic. And everybody will be happy of that. Of course, because she is magiiiic .

Right ? right ?

Edit : sorry that is uselessly aggressive...

[Reg312]

@Iluvalar: its your magic says "DO NOT make the accuracy increase with target size"
please don't adapt game to math equations, but adapt math equations to game...
current game 3.1 good enough, so i do not suggest dractic changes, just make accuracy working

[Searge-Major]

Well... looking over the original post, it seems that we have moved from discussing accuracy to discussing the missing.

If I may restate the problem (as I understood it):

The current accuracy system is bad, because it is independent of actual range apart from short/long range values.

Weapons like flamers can miss even when a target is directly in front of them, sometimes at ridiculous angles.

When artillery misses, it often causes damage anyway, as it generally hits quite close to the target.

So, the questions that need to be asked are:

Are all these still current problems in the latest betas?

And, if so, then what should be done to fix them?

Frankly, I cannot see anything wrong with the function I posted here, also, I don't know why graphs are unpopular , but the one I posted showed the accuracy stats for a weapon with 80% accuracy, and the function outputs an accuracy value, which was intended as input for the function which would do all the actual calculations for firing.

You can probably forget my posts about Gaussian distribution; as has been kindly pointed out by Iluvalar, this would make smaller units virtually impossible to hit, at least at maximum range. If my thinking is correct, larger structures and units like Dragon would be still easier to hit, if only because of splash damage from weapons such as howitzers, however, only a few "misses" would be able to do damage.

That's accuracy dealt with. (I think, naturally there'll be something I've missed)

Now, the misses. NoQ, IIRC, suggested Gaussian distribution for locating miss points, I don't see a problem with this, but there would have to be slightly different adaptations for artillery and df weapons, as artillery can and should be able to miss behind a target, whereas direct fire shouldn't.

So, conclusion. To determine accuracy, a function is used based on a given constant for each weapon. This defines a curve where the actual accuracy equals the given accuracy at optimum range, as expressed by a percentage or decimal in the graph mentioned earlier. The output is then fed to the relevant game mechanic. In the event of a miss, the distance from the target is based on range and a distribution, i.e. more ammo will land close to the target than away, but as the range increases so does the spread.

My suggestion is to do this. Use real location for it.

If the graphic hits, do damage. Accuracy should work as it does in reality... Only suggestion is to check for walls so you don't end up hitting yourself with motion prediction...

Did you read any of the previous posts? And more importantly, understand them?

Edit : sorry that is uselessly aggressive...

Maybe, but did it make you feel better?

[Iluvalar]

@Iluvalar: its your magic says "DO NOT make the accuracy increase with target size"
please don't adapt game to math equations, but adapt math equations to game...
current game 3.1 good enough, so i do not suggest dractic changes, just make accuracy working

It's not magic, I can explain why : The idea of making the shoot realist would create 2 new paradigm in the game.

1) Range vs Size = Some body size would be more efficient at different distances. The range of your weapon would need to fit the size of your body so the body could stay at his preferable distance longer. That effect would easily worth 50% of efficiency it would become more important than the good weapon modifier.

2) Price vs Accuracy = More expensive units will make less dense group and with the gaussian approach, the would receive less projectile when fighting bad accuracy weapons. Again the effect will be as much important thant the weapon modifiers itself. You will need to chose your unit price accordingly mostly to the opponent's accuracy.

That mean that a balanced game following those rules would require to offer a good array of choice in range and price to any player. Some lines will lack of such valid choice during some part of the game. Balancing the game would require EXTREME changes that could not fit at all with the setting. For exemple : The mg line WOULD require a longer range weapon to be playable, so the TMG would shoot at 11 tile. And they would need an expensive weapon, so the HMG would suddenly cost 250$.

That's the kind of balance change that would need to be done to fit the new mechanic. The game would also become extremely elitist because a "good" design will become neglectable compared to those new rules. The same TMG will become better on python hover when facing rockets, tracked light bodies when facing flamers, and scorpions htrack when facing other mg. And the good player that will make the good choice will suddenly become 2 time stronger than the opponent.

[Jorzi]

As a graphical artist I must point out that, while having different sizes for different units might be considered (although it will break the current balance), having the hit depend on the actual shape of the model certainly is not an option.
I make models and I do not want to be responsible for any ingame behaviour of units. I already had problems with my bunker hitbox being underground.

[Per]

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.

[zany]

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.

stop trying to dumb down the game into some other rts
accuracy calculations should stay

[Reg312]

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.

not sure if i understood your point of view.
but i hope you get fun when works on kinda clever game
i ask to implement some of suggestions from this thread