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

[Reg312]

my suggestion: simple change minOffset in accuracy formula

minOffset = MIN (target->radius, 256)
goal: missed shot can hit only another target or nothing (false for buildings larger than 2x2)

you know guys, this possible could improve cannons, because it seems better then cannons spread their splash damage in wider area

[Reg312]

hey, can someone suggest good accuracy formula?
i'm not so good with mathematical stuff
current accuracy formula is just piece of sh*t

requirements:
1) missed shot should not make damage to target
*exept big structures
*exept case when target stay close up (direct weapons)
*except splash damage and other effects like incendiary radius damage, penetrating damage
2) missed shot should not damage ground behind target (direct weapon only)
* when attacker shoots from maximum range it should be taken into account for direct weapons and missed shot should not go farther than max_range + 1 tile;
3) chance to hit should not depend on target size
*may be except very big bodies and case when target stay close up
4) missed shot should look real. but formula should look simple
5) radius of missed shot should depend on some parameters
* weapon type: direct/indirect, i think indirect weapons should have larger radius of missed shots
* range to target
* maximum deviation for missed shots: 2-3 tiles for direct weapons and 3-4 tiles for indirect weapons

aviable parameters:
chanceTohit 0 - 100%
range to target and target x,y
weapon type: direct/indirect

[Shadow Wolf TJC]

Personally, I'd like to adopt an accuracy system that's similar to Supreme Commander's, where weapons DON'T have any accuracy variables. Rather, they have a "firing randomness" variable that determines how far the weapon's angle can be offset, meaning that targets far away are less likely to get hit than targets that are closer. Of course, the weapons could have homing capabilities, or a splash damage radius that's large enough to compensate for said firing randomness.

Sadly, if such a system were to be adopted over our traditional system, then we'd have to rework a few things. For one, we'd have to give the shortHit and longHit variables in weapons.txt a new purpose, such as a chance to home in (for homing weapons), or a max angle in which turrets can be offset due to firing randomness. Also, we'd have to repurpose the accuracy upgrades in functions.txt so that they reduce said max angle for firing randomness.

[Emdek]

Or make such calculations a function that could be replaced using mutator and let mods define additional fields for different targeting systems.
But I would like to see experience being taken into account anyway.

[Reg312]

experience? it already working, as far as i know
http://guide.wz2100.net/experience

i'm saying about fixing buggy accuracy before final 3.1
just small bugfix, without changing weapons accuracy and meaning of hitChance
so please give me formula and i'll create ticket with patch/fix
noone have good idea how accuracy can be fixed without drstic changes?

[Emdek]

Reg312, sure, I mean in context of post by Shadow Wolf TJC, where it wouldn't matter.

[Reg312]

Reg312, sure, I mean in context of post by Shadow Wolf TJC, where it wouldn't matter.

in multiplayer experience does not work in most cases
in campaign is vital part of game

[Emdek]

in multiplayer experience does not work in most cases

Doesn't work or is just nearly not usable due to usually short unit life period?
In case of MP I would vote for returning like 30% (or maybe more, but not above 75%) of previous experience to build queue.
Without similar solution it doesn't make big sense there and probably it wouldn't change too much if it would be not used there at all (would simplify -> speedup calculations a tiny bit).

[Reg312]

exp bonus will add advantage to winner
winning player will be able to kill his enemy even faster

[Emdek]

Yes, it would encourage to attack others early instead of only building defenses.

[Iluvalar]

<out of topic>The exp is far more usable in NRS, some units reach level 5 . The key ? More variation in the units size and the possibility to spend more in defense than offense upgrade. There is no so much change that need to be done to make the exp fonctionnal in standard.</out>

Reg312 : here is my formula :
Roll the dice, if it hit it hit, if it miss, it miss. PERIOD.

[Cyp]

...

Reg312 : here is my formula :
Roll the dice, if it hit it hit, if it miss, it miss. PERIOD.

That's simple to implement — if it should hit, fire towards the tank, and if it should miss, fire straight up into the air.

[Iluvalar]

No, to be exact, i'd suggest to not even try to test hitbox for missed projectile. I believe that splash weapons still need to fall somewhere.

From gameplay perspective, it will be the simplest to understand and balance. And afaik, there is direct weapon in the game that rely on missed shot to hit.


Visualy, the big projectiles (like scourge) might visibly cut trough a unit once in a while. But I don't believe it would be a big problem. We could talk about how to draw missed projectiles better another time.

[Reg312]

Reg312 : here is my formula :
Roll the dice, if it hit it hit, if it miss, it miss. PERIOD.

i just dont uderstand how your formula can be used with current units and their stats
if target stay close up and attacker missed, projectile should fly away at 90 degrees??

current formula: it it hit it hit and if it miss it hit

[Iluvalar]

when the attack miss :

Graphical : do whatever you can to look cool.
Mechanic : Whatever graphics are, miss !!!!

In other words : don't care about hitbox for misses at all !