Idea: automatic weapon balancing

It doesn’t matter if the changes are very small.

1 Like

Generally if we all play the game without any external factor we should reach a stable state where all weapon using rate in a range from 45% to 55% ideally (or 90% to 110% depended on how we calculate it, the number is not mandatory). But that’s clearly impossible. There are many ways to keep the changes stable, here i propose some:

  1. Give small changes in long period of time: A small constant number of damage (say 5 or 10) is given to 1/4 of weapons that were used less and reduced to 1/4 of weapons that were used more each 2 weeks (one week is too short and one month is too long). An anchor should be put on a general weapon like Fork, if Fork is in 1/4 used more group only the 1/4 used less group is buffed and so on.

  2. Use AI to calculate: I really don’t know how to do it but AI can clearly do this task, since data for it is clear, and many specific factors like how strong a weapon is when facing a boss/wave or an enemy can be learned if iA collect data about deathrate and flight time/normal time.

  3. Filter and standardize the data: players that don’t play game often should be filtered, they can’t feel the changes so they will only make auto balancing harder. Also what type of mission and what skill level and difficulty should be chosen and that is a hard thing that Im not really sure but iA. If iA want to have Plasma strong in challenges, Vulcan strong in comet chase, Absolver for boss rushes,… etc each weapon has its own game, i think we can standardize the missions and skill levels. For example in data we have in game: 1000 invasion missions flighted and vulcan was played in 100 missions and fork in 200 missions; 2000 boss rush flighted and vulcan was played 100 missions and fork in 600 missions, 500 comet chase flighted and vulcan was played 100 missions and fork was played in 50 missions. We take the fractions of Vulcan’s using rate in three kinds of mission which are: 10%, 5% and 20%. Fork’s using rate are 20%, 30% and 10%.So Vulcan’s using percent is mean(10,5,20) ~ 12% and it’s field is comet chase with 20% using rate. Fork’s using percent is mean(20,30,10) ~ 20% and it’s field is boss rush. Now we multiply the percent for the number of weapon need to be balanced say it’s 6 weapon, then we have Vulcan (mean, max)=(72%, 120%), Fork (mean, max)=(120%, 180%). The balanced zone should be around 100% but i think 80% to 120% is already acceptable. So if iA want equal’s using of weapons we say Vulcan should be buffed and Fork should be nerfed, and if iA want equal’s of weapon’s specificity Fork should be nerfed. For my ambient videos I upload to youtube I run a script to calculate mean and max volume of the audio then I adjust the total volume by delta=20dB-(mean+max)/2 and the audio sounded good for me, so this is also a way you can consider.

My mistake here the percent of max using fraction can go over 100% so we need to calculate max using fraction of all weapons to know the mean.

Yes I won’t use corn in rookie because lightning is a lots easier and faster to kill enemies.

1 Like

I still support the idea and the previous posts I’ve made in this thread. Popularity is the measurable result of balance and having all the weapons used for roughly equal amounts of time gives more flexibility in playstyles.


Whenever the topic of auto-balancing comes up, I always cite this post:

I don’t like to just blatantly disagree, but for the reasons I’ve stated above, I simply have to.
Not everyone picks a weapon because it’s “objectively” good.


Everyone use Fork before do you agree with it? Now Fork is not the meta anymore not much people use it anymore. Edit: Yes sure not eveyone picks a weapon because it’s the best and I do use corn only even before the buffs, but the percent of people like me is very small, can you agree with it?

1 Like

Based from Galbatorix’s weapon virtuousity data we can pick Photon Swarm and Laser Cannon for examples. Photon Swarm becomes the #2 most popular due to homing mechanic, but still not the best performing weapon. While Laser Cannon has a bit broken damage output (almost not overheating when manual fired) but it’s unpopular due to design and the fact it’s better to use manually. (People tend to save their fingers)

Photon Swarm has good performing in all kinds of missions. Meanwhile Neutron has bad performing but we don’t buff it because of the damage. Edit: I mean the high damage doesn’t prove that a weapon is efficient.

Laser still didn’t finish yet we can’t use it because of the lack of vision.

You forgot Fork too fast.

For people trying to argue about the damage I’d say we can totally compare the weapons for SSH >100% only, then the damage outputs should be balanced. Weapons like corn, neutron, ion don’t work well in lower difficulty because they require time and overkill enemies, while weapons like lightning and photon can ace everything very fast and strong enough to kill bosses, so once all weapon are equal corn, neutron and ion are the very less used weapons in total playcount but still the highest ones in damage. Again it’s iA choice.

Really? I recall that a while back there was an argument over whether Lightning ſhould be fired manually.

Why would you try to balance all weapons with a max difficulty miſſion as a ſtandard model, given that if the difficulties were ſet properly only a tiny fraction of people would fly the hardeſt miſſions?

No I just say we can balance the damage output if we compare them in max difficulty.

I think it would be good to have ſome weapons that perform beſt in eaſy miſſions as well as ſome that perform beſt in hard miſſions, which would in turn make it harder to compare them directly.

1 Like

It can’t be a specific, narrowed-down data set for technical reasons, I imagine. I don’t think iA will want to make a separate statistic for storing these values - it’ll be the “weapon virtuosity” that gets used, if this gets added. It will multiply whatever the base DPS value of the weapon is.

And if we are excluding stuff from these calculations, then you could make a gargantuan list of what “doesn’t count” - the immediate thing that comes to my mind is comet chases, which have virtually no reliance on DPS and the “best” weapon will remain stable through pretty much everything because firing pattern is all that matters there.

So Vulcan’s going to get wrecked under this system, surely. Anything that’s good for single-target damage (like Plasma) will also suffer far beyond the point of reasonable viability because people will still be desperately trying to use whatever scores best in their three dailies, and remaining competitive with pecking orders and meritorious marksmanship is probably going to pay off way more than changing to a “viable” weapon, so they’ll grow painfully weak to compensate.

And then there’s Absolver, which is likely to elicit some sort of backlash regardless of if it gets stronger or weaker. I suspect it’s going to get buffed by any sort of auto-system, but that’s only going to be a disaster because damage is literally the only thing about the weapon it doesn’t need more of. It will just grow in boss-destroying power while remaining unusable for many classic waves.

tl;dr - An auto-balancing system is not going to be smart enough to even figure out why a weapon is good or bad, let alone why it is popular. I can see no way to automate balancing that wouldn’t be more complex than just doing it by hand anyway. It’s unfortunate, and I do wish it could work, but I feel like it’s just the nature of the problem.

I think the WBP has done a very good job in a very short time, and similar manual tweaks are going to be much more reliable and productive than letting a spreadsheet micromanage the weapons.


Especially now that the firing system has been reworked and fire rates aren’t dependant on framerate anymore. It opens up a path to microbuffs and micronerfs that rely on subtle fire rate changes that wouldn’t be possible otherwise.

1 Like

Short term yes, but I think long term an automatic ſyſtem would be good; ſooner or later new enemies or whatever will ſhift the balance in favour of different weapons, and it would be nice if it didn’t all require rebalancing every time.

Well this is proposing that the game rebalances all the time, 24/7. So while I agree that balancing is tedious, an automatic system is only going to make balancing continue indefinitely, like playing a game of ping-pong with the meta rather than trying to flatten it out.


Yeah now think about it I don’t think comet chases get affected by weapon damage so we can just filter them.

As I said before the data need to be standardized. Also if iA want Plasma to be only strong in challenge they’ll add it in the table, but i don’t think it’s a good thing because competitive is different from normal play. If you play LOL you should know Ryze champion, he’s very strong in pro play but weak in normal play, and company of LOL want to balance him in both pro and casual and he got over 5 reworks in one year and his winrate in casual matches are always the worst of all. From Ryze case I think balancing for both competitive and casual playing is near impossible so it’s totally okay if we don’t count challenges in balance.

1 Like

Which are we trying to balance? Competitive or normal?

Because if you’re correct, and we can’t have both, we either need to balance one and skew the other, or compromise in the middle (which, at least out of those options, I’d prefer).