Forums » Suggestions

Variables to Consider for Weapon Balance (guns only)

May 22, 2003 Acierocolotl link
Perhaps our visions of weapon balance are significantly different. At worst, this post can be glossed over. Presented below is a bit of screed on the mechanism I would use to balance weapons. I hope it will be of use.

There are several variables that need to be considered to properly balance weapons. Many of these variables are already obvious, and others may be less so.

Here are the variables; I'll comment on them afterwards.

Shot speed
ROF
Energy cost per shot
Damage per shot
Ammo considerations
Shot visibility
Effective range
Maximum range
Autoaiming radius
Precision
Specialty effects
Weapon cost

Shot speed is probably the singlemost important variable for any gun that relies on autoaiming. The faster a shot travels towards its target, the less time its victim has to correct travel. It extends the effective range of any gun as well, making guns that shoot fast-travelling shots excellent for standoff purposes (ideal for fast fighters against slow targets).

ROF is the singlemost important variable for any gun that relies on unaimed spray to hit a target--hosing an area down increases the odds of hitting the target. This has only become important now that autoaiming can be turned off, which makes it possible to (theoretically) use the light gatling to spray down barrel-rolling targets, which from what I've gathered, mystically messes up the the autoaiming reticule.

Energy cost per shot factors into two separate considerations:
1: (ECPS*ROF-50)/250 is the number of seconds the gun may be fired continuously, on a full battery (negative values mean the gun may be fired indefinitely). Continuously pumping shots at an enemy improves the odds of hitting, as constantly forcing them to dodge shots reduces their overall velocity, making them easier targets as autoaiming needs to compensate less.
2: (Damage per shot/ECPS) rates as the general efficiency of the gun--DPS*ROF represents the theoretical damage per second, although getting a second of sustained, 100% accurate fire on a target is rare and exceptional, representing ambush situations primarily. It is still a consideration. (The modified graviton gun has an efficiency of 44.4, versus the tachyon's 50, and all other factors are equal--the tachyon remains the weapon of choice.)

Ammo considerations: If a gun needs to carry ammunition (like the railgun) then it will eventually need to go back for reloads. This can put serious limitations on the operational efficiency of a gun. I would like to think that a gun can carry enough ammunition to kill at least one, if not two ships, or possibly even more if the gun is likely to miss a fair bit. (As it stands, a single railgun requires 26 hits to kill a Prometheus, but carries only 30. This either requires two railguns, seriously affecting the flexibility of the fighter, or unnerving precision.)

Shot visibility: If you can't see it, you can't intentionally dodge it so readily. Big, fat, huge glowing globs of energy are much harder to avoid than tiny little nonluminscent little pellets. This is a qualitative assessment, and for the most part, fairly moot--I don't recall any energy shots that aren't lit, although red or yellow shots might be confused, briefly, with the red and yellow targeting information around a ship.

Effective range: The range at which a target might be hit in combat circumstances. This is a bit of a subjective criterion, as the effective range of a gun will vary based on the target (acceleration and pilot skill) it's being fired at, and a desired rate of hitting. For me, my rule of thumb has been the shot's speed * 1.5 as extreme effective range--about 270 meters for the tachyons. The vast, vast majority of shots will miss at that range, but a "stalled", ponderous target will take a hit or two. The effective ranges of the gatling cannon will be greater, as their spray factor and extreme ROF can ensure lucky hits at longer ranges.

Maximum range: This isn't much of a consideration, so long as the maximum range (which has been about 500 meters for some shots, 750 for others) exceeds the effective range of the gun. A highly accurate gun with a very large maximum range could be used to pick off targets with a constant velocity, however.

Autoaiming radius: This is pretty self-explanatory, but comparing this to other variables is, again, subjective. I have my suspicions about the lead reticule, but that will go in another post. In short, however--*if* the lead reticule is perfectly accurate, then the wider this radius is (ie: the advanced gatling has a large radius, the gauss cannon has a medium radius, most guns have a small radius), then this helps a gun fire accurately. Inaccurate shots are, generally, a complete waste.

Precision: Does the shot hit the target the gun's aimed at? A gatling cannon will miss a stationary target very frequently, making it awkward for minesweeping (but good for other reasons, already discussed). If it sprayed even more, it might be utterly useless. A tachyon will go exactly where it's aimed; targets will be forced to move, and mines tremble.

Specialty effects: This is something that hasn't appeared in guns, being more of a domain of mines and rockets. It's a catchall, however, that might yet appear. Weapons that, say, slow up a ship, or render the pilot blind, or fire warheads that detonate within a certain range of the target, or render the target incapable of turboing, or in some other means, make the gun more desireable, would be considered here.

Ultimately, what all of this boils down to is a "real damage/second-of-dogfight". As there are an impossibly great many variables to consider in dogfights, this isn't easily assessed. However, all of the above variables come into play, as well the obvious ship manoueverability and size, and pilot skill factors.

Cost: The cost of a weapon should (barring RP considerations) reflect its desireability. A gun that costs twice as much as another should be, in theory, twice as good--in short, doing twice as much "real damage per second". The cost should also take the game's economy into consideration. The difference in most guns is only a few hundred to a few thousand (excluding the advanced gatling), which is too little to promote the use of lesser guns. A few thousand credits is pocket change to even a fairly new player.

I do look forward to future weapon balancing, and I hope that a consideration of the above variables will take some of the guesswork out of a fairly (but not completely) subjective system. At present, the weapon balances have made the gauss cannon more desireable, though they've quite some distance to go yet. I expect that people will still use advanced gatlings in heavy ships, and tachyons in light ships. (This naturally excludes rockets, et al. That's another kettle of fish altogether.)
May 22, 2003 Vlad link
Yep, those are exactly the variables we looked at. (Except for shot visibility, I had mostly neglected that one. I think all of the current weapon effects are sufficiently visible.)



May 22, 2003 Celebrim link
I think Acierocolotl made a very excellent post, and if posts were to be archived in a Hall of Fame of posts, that one would be a serious candidate.

Some people have a gift for tact. I don't.

I am beginning to think that I don't quite understand the vision of the developers. I sometimes even wonder if the devs know what they want. The changes in the weapons have been surprising to say the least and seem to almost indicate that there isn't a general desciption of each weapon somewhere which says (for example): 'slow velocity, hard hitting, average efficiency, moderate cycle rate - use detailed graphic effect.' or 'high velocity, low damage, low efficiency, high cycle rate - use simple graphic effect' or some such thing. That would be ... discouraging to me to say the least.

Take for example the current situation with the graviton and the tachyon. Prior to the latest build, the tachyon was clearly the better weapon, but the graviton at least had a few things going for it that made it theoretically superior in some situations. Certainly it was a reasonable choice of weapon on a centurian, or any ship mounting only a single energy weapon. However, it was still most certainly weaker than the Tachyon. One would suppose that in a 'balance' release one would try to modify the Graviton to address that problem and improve it compared to the Tachyon without destroying the uniqueness of either weapon. One would suppose that, but that isn't what happened.

I'm doing this from memory, but here are the new values:

Tachyon: Velocity: 180 Damage: 600 Cycle: .18 Energy: 12
Graviton: Velocity: 180 Damage: 800 Cycle: .25 Energy: 18

It should be readily apparant that those two weapons are basically identical now. You might as well do away with one because one of them is redundant and its graphic could be better used somewhere else. If its not reaidly apparant consider this:

Tachyon: Damage/Second: 3333 Energy/Second: 67 Efficiency: 50
Graviton: Damage/Second: 3200 Energy/Second: 72 Efficiency: 44

That's basically a wash. The two weapons are close enough in effect that most people won't be able to distinguish between them and they will certainly 'feel' and be employed the same. The Tachy is still the superior weapon numerically, and if anything its advantage is clearer than before. At least prior to 'rebalancing' the Graviton had a theoretical edge in damage per second meaning that it was in some situations superior at close range with short bursts. Now, the tiny edge it has in 'first strike' is basically insignificant and unable to overcome the Tachy's advantage in both damage and energy consumption over the long term.

I want to focus in some detail on two areas of Aceirocotl post. The first I agree with very strongly and want to elaborate upon, the second I disagree with and wish to suggest a better balancing strategy.

"Ultimately, what all of this boils down to is a "real damage/second-of-dogfight". As there are an impossibly great many variables to consider in dogfights, this isn't easily assessed. However, all of the above variables come into play, as well the obvious ship manoueverability and size, and pilot skill factors."

If it were possible to access a weapons 'real damage/second-of-dogfight', we wouldn't need to play test the game. But, and I cannot stress this too much, play testing cannot lead you to good design nor can it in and of itself turn bad balance into good balance. Play testing is only going to help tweak the numbers. If the numbers are wildly off to begin with, play testing will not be able to tell you anything but that the numbers are wildly off. You can't just erratically change designs over the course of play testing and expect play testing to provide you with any sort of useful feedback. There has to be design before these weapons are released. Play testing might tell you whether your concept is feasible, but it can't tell you what a good concept is. You have to sit down and do some numbers and think of some concepts.

Please.

I don't know what you are doing but I have an Excel spreadsheet with alot of little formula on it of varying subjectivity but which at least give me some idea of whether the change I'm suggesting is actually the improvement or penalty that I intend it to be. I don't claim to have the one true formula, but some things ought to be obvious:

1) The real damage is proportional to the damage per second.
2) The real damage is proportional to the damage per shot (because you can always make the first shot, even if you can't make the second).
3) The real damage is proportional to accuracy of the weapon.
4) The accuracy of the weapon is proportional to its velocity.
5) The accuracy of the weapon is proportional to square of the proximity radius of the weapon divided by the cross sectional area of the average target. Something close to the proximity radius of the weapon's exact effect on the accuracy can be deduced by proposing a number of 'typical dogfight situation'. For example, propose that my target is 100m away from me and moving away from me at a 45 degree angle and that we are both moving 100m/s. The accuracy of the weapon is proportional exactly to the size of the angle between which the attacking vessel must point order to hit the target. Naturally, the size of the angle depends on distance to the target and the width of the target. The distance to the target is going to depend on the velocity of the weapon, and the width of the target is effectively the sum of the proximity diameter and the width of the target. If you wanted to make this calculation for acceleration and 3D then you can more or less exactly calculate the effect of proximity fuses for a given size target, but personally I think it would be enough to acquiant yourself with these effects and come up with some reasonable subjective numbers - because in game situations (target size, relative velocities, relative accelerations) are going to be too diverse to have one perfect number anyway.
6) The accuracy of the weapon is proportional to the width of the automatic correction ASSUMING the correction is always accurate.
7) The real damage of the weapon is proportional to its ability to be employed ('wieldiness').
8) The 'wieldiness' of the weapon is proportional to its efficiency.
9) The 'wieldiness' of the weapon in a ammo consuming weapon is proportional to total expected damage that the weapon will yield in a average combat. This value depends on damage/shot * number of shots stored, but can be lower than this value for strange conditions like low cycle rate and high energy consumption because the number of shots likely to be made before the conclusion of an average combat may be then lower than the ammount of ammo stored. Acierocolotl correctly pointed out that a single rail gun probably lacks the theoretical capacity to 'win' against another fighter. I don't think I'm amiss in pointing out that with a 1.6 second cycling rate and a energy consumption rate per second higher than the best energy supply, that its unlikely that a ship armed with rail guns will be able to make all of its shots before the decisive moment in combat. The rail gun therefore is now not only operating with a low damage per second, but with a significant 'weildiness' problem.

"Cost: The cost of a weapon should (barring RP considerations) reflect its desireability. A gun that costs twice as much as another should be, in theory, twice as good--in short, doing twice as much "real damage per second". The cost should also take the game's economy into consideration. The difference in most guns is only a few hundred to a few thousand (excluding the advanced gatling), which is too little to promote the use of lesser guns. A few thousand credits is pocket change to even a fairly new player."

I want to quibble slightly with Acierocololt's wording with regard to cost though. I agree that a weapon's cost should reflect its desirability. I just don't agree that that cost and desirability are linearally proportional. In fact, I believe that cost should grow exponentially with desirability. A weapon which is literally twice as good as another weapon should literally be an order of magnitude more expensive. A weapon three times as good as another weapon should be two orders of magnitude more expensive. If an average weapon costs 5000 cr, a weapon which is three times as desirable should cost probably not less than 20,000 cr and depending on the economic situation might need to cost 500,000 cr before cost alone could drive balance. This is because desirability is directly proprotional to utility and in the end not only does utility pay for itself in a combat situation but in terms of winning cost is sometimes no object. Imagine a situation where a weapon was literally 100 times better than another? How much would it be worth to be able to take on 100:1 odds? How willingly as a designer would you allow a community of 100 players to dominate the 'game' (the 'life' if you will since this is an RPG) of a community of 10,000? If your answer is only '100 times the cost of an average weapon', then something is clearly wrong. On a smaller scale the same is true for weapons only two and three times the desirability, and it might be considered smart design simply not to have weapons that widely separated from each other in value or desirability.

RPG's have understood this from the very beginning. A +5 sword really only does about twice the damage of a unmagical sword, and really only hits about 50% more often in common situations. Yet, despite being only about 3 times as good as a unmagical sword (in most situations) a +5 weapon is generally priced about 2000 times the cost of a normal weapon and about 25 times the cost of an ordinary +1 magical weapon. Weapons in Diablo (and clones) also increase exponentially in value as their desirability goes up, and if anything it was this adherence to just this one principal of RPG design (and the corresponding diversity) that made an otherwise lame RPG into a classic.

But given Vendetta's evolutionarily superior game play to existing RPG's, I think that wide separation in the quality of weapons is not something we necesarily need.

The gauss and the plasma are finally starting to make some sense, but I really think you need to sit down and rethink your overall approach to balance.
May 23, 2003 Acierocolotl link
Let me just summarize up a few weapons here, define whatever acronyms I'm going to use, and then head into the discussion.

The primary point of contention lies between the tachyon and the graviton gun. So:

(Apologies if the chart doesn't come out right.)
---------DPS-EPS-Eff--ROF--TDPS-EPSF--SFD---Vis
Tachyon 600 12 50.0 6.66 4000 80.0 08.33 Bright Green
Graviton 800 18 44.4 4.00 3200 72.0 11.36 Dimmer Blue

DPS: Damage Per Shot
EPS: Energy Per Shot
Eff: Efficiency (Damage per point of energy expended)
ROF: Rate of Fire, in shots/sec
TDPS: Theoretical Damage Per Second (assumes all shots hit)
ERng: Effective Range (shot speed * 1.5 seconds; rule of thumb guideline)
EPSF: Energy Per Second of Fire (constant fire, of course).
SFD: Sustained Fire Duration. How many seconds of fire a full fast-charge battery can manage.
Vis: Visibility.


EDIT: I was on crack at first. Legibility improved a little, one calculation corrected.

All variables that were the same are removed. The Tachyon, irony of ironies, has become the Graviton of 3.2.4; it shoots faster, does more theoretical damage per second (assuming a lot of hits occur), and runs out of energy faster. The Graviton is better for long, stand-off situations, involving lots of plinking and then running away.

The weapons aren't really the same, although they sort of fill the same general role. Whether the graviton is superior to the tachyon now is debateable--the price implies a modest superiority, but I have my doubts. The guns are "about equal". The graviton's lesser energy footprint allows snappier movement, but does less damage at the same rate of accuracy.

So, uhh, I don't know what's up. But I say, "have faith"; I know the speed the developers work at is pretty slow, but it's traditionally been this way.