What needs to be done.

[ginn]

Skoskav is doing a pretty good job for somebody who doesn't play the game... However, as he doesn't know how the game is played, the weapon balance is off (noticed especially in the 1.5.2 -> 1.5.3 wm)
As the weapons get stronger and stronger, the game play gets slower and slower, this is a really bad trend.

How do you fix weapons balancing?
Well, Bullets should have "Bullet mass", which would decide how much bullet push a weapon does. So for example, Minimi could be less damage, but have high bullet push. AK having slightly more dmg but next to no bullet push.
Same would go for mp5 and steyr.

There's also the damage modifiers, as it is now it's almost completely up to luck if it regs as a hs/miss/torso/legs. Fryer suggested an interesting idea, however, I don't think there should be damage modifiers (not for nades either, instead they should always do something like 60% dmg).
Neither should any weapon have movement inaccuracy, as this too is just slowing down game play.

Now, if all this is made, the game will be very clean.
An issue you might start to see now is that ruger would be impossible to balance. However, I believe the balancing should rather rely on movement speed instead of "head shots".
In order to 2 hit kill with ruger, you'd have to meet a certain speed. Now, you might have to make the inherited speed add extra damage, making the formula something like this: bulletspeed*(playerspeed*2)*bulletdamage/1000, or just simply make a Playerspeedmodifier in the weapons balance.

Anyhow... that's the way I think soldat should head.

[Adam]

and how does this make gameplay any faster

[ginn]

camping won't give you the aim advantage, and rushers won't die in half a second, especially with no damage multiplier for nades.

[Bonecrusher]

spas is op

[Fryer]

bulletspeed*(playerspeed*2)*bulletdamage/1000

I suppose you mean (bulletspeed+playerspeed*2)*bulletdamage/1000? Even if this is considered, I hope we can find a better solution than multiplying the player speed by some modifier, since that is IMO a bit too far from realism even for a game like Soldat.

[Clawbug]

Problems I see:
1) Gameplay isn't deterministic enough, too big "luck" factor in form of hit registration and weapon inaccuracy.

2) Weapon balance is highly related to map in question. It is very hard, if not impossible, make  "one balance fits all" weapon balance because player distance and position determine the outcome along with the chosen weapons, as it should be. However, as these change from map to map, and as the nature of new maps are small corridors and quarters, this skews the weapon balance to favor weapons which work well at short ranges. In CTF on modern maps there's very little(compared to older maps) use for weapons which work well from longer distances for example.

3) Currently defensive gameplay has huge advantage especially with certain weapons. This is because of bullet push and the nature of the weapons and the fact that a stationary floating/standing player is far more accurate than a rushing one.

4) Grenades play a huge role in combat situations, especially in modern, small maps. Combat is more about hitting your grenade AND being lucky that it registers, than anything else. Their potential is far too significant.

Solutions I have:
1) (luck, non-determinism):
  Reduce "random factors" to a practical minimum. Get rid of bink, make weapons shoot as straight as practically possible, improve network code.

2) (map size):
  Reduce the impact of bullet speed on weapon damage or reduce the relative bullet speed differences among weapons. Preferrably both. In other words, reduce the difference of "short range" weapons and "long range" weapons. Also, decrease Spas pellet spread in order to make it more balanced over longer distances.

3) (defensive advantage):
  Reduce bullet push, FIX knife not having player velocity added to it.(This was originally probably a bug anyway). Get rid of any form of movement inaccuracy.

4) (grenades):
  Get rid of hit/damage modifiers for grenades, so every grenade hit is as lethal as others. Reduce grenade damage so that they aren't insta-kill even if they hit feet. To compensate the potential slower gameplaty, completely remove the delay of both knife and LAW.

Thankyou.

[Adam]

Problems I see:


Solutions I have:
1) (luck, non-determinism):
  Reduce "random factors" to a practical minimum. Get rid of bink, make weapons shoot as straight as practically possible, improve network code.

+1 but shoot straight as possible ehh would be kinda too easy

[Clawbug]

Problems I see:


Solutions I have:
1) (luck, non-determinism):
  Reduce "random factors" to a practical minimum. Get rid of bink, make weapons shoot as straight as practically possible, improve network code.

+1 but shoot straight as possible ehh would be kinda too easy

No, it wouldn't. Shooting straight in this context means "as little bullet spread as possible". That wouldn't make things any easier at least. The contrary, more about skill and less about luck.

[machina]

spas is op

Yea, horrendously (in realistic)... And it wasn't fixed since two versions or more...

[Bonecrusher]

I agree with overall reducing weapon power[bullet speed/dmg], this would refresh the gameplay. Hopefully also create better gaming enviroment.

3) (defensive advantage):
  Reduce bullet push, FIX knife not having player velocity added to it.(This was originally probably a bug anyway). Get rid of any form of movement inaccuracy.

Two proper cannonballs and knife could be thrown with a half of the map lenght. I wouldn't like that

[Clawbug]

I agree with overall reducing weapon power[bullet speed/dmg], this would refresh the gameplay. Hopefully also create better gaming enviroment.

3) (defensive advantage):
  Reduce bullet push, FIX knife not having player velocity added to it.(This was originally probably a bug anyway). Get rid of any form of movement inaccuracy.

Two proper cannonballs and knife could be thrown with a half of the map lenght. I wouldn't like that

What are these cannonballs? You call already shoot LAW half of the map length. Or use M79. Or Barrett. Or Grenades. How would knife be something you'd dislike? Or do you dislike the other weapons too?

[Bistoufly]

3) (defensive advantage):
  Reduce bullet push, FIX knife not having player velocity added to it.(This was originally probably a bug anyway). Get rid of any form of movement inaccuracy.

Two proper cannonballs and knife could be thrown with a half of the map lenght. I wouldn't like that

Of course not Bonecrusher. Don't worry.
Knife throwing would simply be made similar to nade throwing.

[Adam]

why not just make a server with your custom made weapon balance and let others test it

[Clawbug]

why not just make a server with your custom made weapon balance and let others test it

This has been done many times, the beta team plays(or at least used to play, when I was in) actively different kinds of weapon balances and gathers player experiences to make further decisions on the weapon balance. This alone isn't sufficient though, the weapons aren't balanced when we start talking about "damage over time" and other statistics we can gather and their influence on the weapon balance.

In short, balancing the weapons is an art and science on it's own right, it's impossible to get "right", and needs constant adjustments. It is never perfect, although perfection should be desired at all times.

[Bonecrusher]

@combat knife

I think it should stay as it is, what needs to be fixed is nades and wep balance. Knife is short range 1hit kill wep and I dont see why it should be changed. Plus the 'adding players speed to knife throw' thing was discussed before and from what I remember we decided its not a good idea.

[Clawbug]

@combat knife

I think it should stay as it is, what needs to be fixed is nades and wep balance. Knife is short range 1hit kill wep and I dont see why it should be changed. Plus the 'adding players speed to knife throw' thing was discussed before and from what I remember we decided its not a good idea.

The way knife works promotes defensive gameplay over offensive and maintaining height-advanage over enemy, I feel that this unnecessarily complicates gameplay and makes balancing weapons harder, is illogical and is probably based on a bug rather than being an intended feature in the game.

Making knife work like other weapons, by adding the player velocity to the knife, would fasten the pace of the game, make knife more usable, allow nerfing grenades damage and hit modifiers, removing or at least greatly reducing LAW startup time and perhaps even delay in knife. All of which would contribute to more fluid, fast paced, predictable gameplay.

[13th_account]

1) Gameplay isn't deterministic enough, too big "luck" factor in form of hit registration and weapon inaccuracy.

Weapon inaccuracies will be handled better in the next version. No more 33% hit chance when your Barrett is super-binked. Stacking inaccuracies (e.g. bink + moveacc) works better. So in that regard the game will be more "deterministically random" rather than "quirky random," because based on what gun you're using, what speed you're moving at, what you're doing and whether you're being binked or selfbinked, you can estimate how likely it is that your shots will hit. As for hit registration issues, that will be dealt with at some point in the future.

2) Weapon balance is highly related to map in question. It is very hard, if not impossible, make  "one balance fits all" weapon balance because player distance and position determine the outcome along with the chosen weapons, as it should be. However, as these change from map to map, and as the nature of new maps are small corridors and quarters, this skews the weapon balance to favor weapons which work well at short ranges. In CTF on modern maps there's very little(compared to older maps) use for weapons which work well from longer distances for example.

Challenge accepted!

3) Currently defensive gameplay has huge advantage especially with certain weapons. This is because of bullet push and the nature of the weapons and the fact that a stationary floating/standing player is far more accurate than a rushing one.

I've recently been playing with builds of Soldat that instead of having all bullets give constant amount of bullet push, has their push based on what the damage would be, so basically the bullet's momentum. It would have the effect of making autos and Spas boost less, and semis more. Example of push over time when normalizing on AK (so its push remains unchanged):
Before: http://skoskav.org/etc/soldat/etc/bulletpush/1.png
After: http://skoskav.org/etc/soldat/etc/bulletpush/2.png

Though having it be moddable in the WM independently from the damage and speed is another, possibly better, option. It's certainly a more flexible one.

As for moveacc, it will be replaced on some weapons with a new WM attribute called BulletSpread that works as a constant inaccuracy. So e.g. autos' bullets are always going a few pixels up or down, not only when the players are moving. For weapons that shoot multiple bullets at once (Deagles, Spas) it works as the spread between the bullets. So they can be given a tighter or looser grouping through the WM. Some weapons (Ruger, Barrett, Minimi) will still keep their moveacc in order to stay balanced.

As for knife I would first need to hear thoughts from other clan players on tweaking its velocity, as they could easily rage.

4) Grenades play a huge role in combat situations, especially in modern, small maps. Combat is more about hitting your grenade AND being lucky that it registers, than anything else. Their potential is far too significant.

I would also need to hear thoughts from other clan players on tweaking nades and startup, as they could easily rage.

[ginn]

well, clawbug seem to be suggesting practically the same things as I.

Anyway, movement acc shouldn't be in the game, as imo, it just ruins it. Bink however, should stay, as it balances certain weapons.
High potential guns like ruger, spas and DE shouldn't just rely on aim, it should rely on speed as well.

The problem with bulletpush being decided by damage won't change anything really, seems like it's only gonna reduce it slightly.

[STM1993]

As for moveacc, it will be replaced on some weapons with a new WM attribute called BulletSpread that works as a constant inaccuracy. So e.g. autos' bullets are always going a few pixels up or down, not only when the players are moving. For weapons that shoot multiple bullets at once (Deagles, Spas) it works as the spread between the bullets. So they can be given a tighter or looser grouping through the WM. Some weapons (Ruger, Barrett, Minimi) will still keep their moveacc in order to stay balanced.

Then I would like to point something very important out: Certain fireintervals will cause your bullets to naturally spread (I should note that it starts happening at fireinterval 8 and below, with 8 being extremely bugged when you crouch). I consider this, along with the odd inaccuracy when you initially go crouch/prone, an old bug related to animations.

The most obvious in-game example of what I've described above can be seen in the Steyr Aug's fireinterval 7. Your first shot always goes straight, but the 2nd shot onwards always angles slightly upwards. A not so obvious example would be MP5's fireinterval 6: your shots always alternate up and down.

If you wish to add this BulletSpread function, then the above should be addressed as well.

[ginn]

I don't see why weapons should have any movement inaccuracy at all, all it does is to promote "camping".
Rather making weapons to a set amount of bulletpush instead.

bulletspeed*(playerspeed*2)*bulletdamage/1000

I suppose you mean (bulletspeed+playerspeed*2)*bulletdamage/1000? Even if this is considered, I hope we can find a better solution than multiplying the player speed by some modifier, since that is IMO a bit too far from realism even for a game like Soldat.

Yes, that is what I meant. I believe making rugers fireinterval higher and balance it's damage very carefully could fix it - forcing players to not go straight up vs autos.
Otherwise changing the formula to (Constant+playerspeed)*bulletdamage/1000, where constant is a value similar to the bullets initial speed.