CIU weapon balance program - Part 4: Photon Swarm

@BoggY Yes, it is unique mechanic indeed(i mean the color change, lighntning fryer has >50% dmg increase), but is it worth it gameplay-wise? I think that increasing the number of orange photons is a fair trade for that.

@anon27929001 It’s not that I dislike the damage spike from level 10-20. In CI2 it worked well, of course it did! When all weapons are designed this way, there is no problem in that. My problem with the 50% power increase is very simple. As I stated in part 2, I am trying to balance out current weapons based on the “Utensil Poker & Neutron Gun model”, because those two guns are most succesful right now.

Having a 50% powerspike weapon in the roster together with one that hasn’t the powerspike simply cannot be balanced. Either lvl 10 of the powerspike weapon will be to weak or lvl 20 will be too strong. I decided to stick with the current Neutron/Utensil model for various reasons (less changes, based on sth that works right now, easier calculations), but TBH I think that the CL2 model was indeed much better. Long story short: All weapons have huge powerspike or none has, no middle ground.

I also want to answer the RNG aspect that @Stardrone brought up, but now I don’t have enough time to answer this in-depth enough. I will come back to this topic tomorrow.

5 Likes

the neutron gun has changed a lot. the thing that i would love to see on the new neutron gun on max lv is the spread that old neutron gun had.

I assure you the old neutron gun was better than this one we have now.

I am aware of that @isp , but changing all weapons so drasticly to fit CI2 standards is above my capabilities. It would require solid dev team to do that, because it would probably mean redesigning all weapons from a scratch.

Huge powerspike between 10th and max seems good in series because you start from 0, different in Universe.

For flare, reduce its luminous should be helping for visibility.

For me, it’s not a problem with “chance of spawning charged photons” if for example 75% means it always spawns 75 charged photon out of 100.

And I suggest for balancing Hypergun and Boron Railgun to make them have linear growth of Damage per Volley and have constant time to overheat. That helps alot.

1 Like

A complete change in color is what makes Photon Swarm a unique weapon of choice. Imo, I don’t wanna change that.

RNG-based particle proc bears similarly to The Riddler’s spread fire (you’d argued about this as I remember).

Edit: I don’t see any problems using the weapon (in CI4/CI5) as I could pause for a short while to check if there were falling eggs (and to cool down a weapon) besides it’s fairly decent damage.

Nice joke :rofl:

Well, personally I don’t have that much issue with this “feature” of Photon Swarm, but based on your and several other user’s replies I decided to include this aspect in my post. It seems to be a common thing that draws players away from this gun, so it has to be fixed. You can read the new section of my post, in which i talk about this, and decide if you like it or not.

Finally I have enough time to get into this RNG topic. I designed the statistics for the proposed changes with the RNG aspect in mind. Originally i was going to include the “RNG talk” in the original post, but I decided not to do so, because it would take a lot of time to explain and because the RNG is a difficult topic in general. But apparently, it became the most controversial part of this post, so I have to explain it anyways.

Let’s start with the RNG itself, because most people think that RNG in games is simply a bad thing. This may be caused by the fact that propability in general is a very counter-intuitive concept that may be hard or tricky to fully understand, and by the fact, that from time to time in games there are bad RNG implementations that are very visible and ruin the player’s experience. This is often the case in competetive PVP games for obvious reasons.

Main example of such bad RNG implementation is situation in which your success depends only, or mostly on your luck. This includes random miss chance, random one-shot chance and random crit-chance in case of slow firing weapons/weapons with small amount of projectiles, where you depend on a small number of hits to accomplish your goal and RNG can mess you up.

I haven’t ever played Pokemon, but i have played other games in more or less competetive environment such as “World of Tanks” where missing a crucial shot due to terrible luck is infuriating. Or getting wrecked in “League of Legends” because of getting “1% crit” by an enemy early on. Both those cases (and many other) follow the slow fire rate requirement. As I said, I have never played Pokemon in any way or form, but I am pretty sure that the Pokemon situation you meant didn’t have anything to do with rapid-fire weapons or damage-over-time mechanics. (Correct me if I am wrong, I can’t be 100% sure.)

The fact that RNG can be badly inplemented, doesn’t mean that RNG is bad as a concept. In many cases it is almost neccesary, in other it can simply make things better. Right now CIU has a lot of random things such as velocity of an egg, moment when egg spawns etc. Yolk-Star™ has randomized attack pattern, which arguably makes it much more difficult and interesting than other bosses. Without RNG, those features couldn’t be as interesting as they are.

So with all those things said, it’s time to talk about my Photon Swarm RNG implementation, and why I am pretty sure that it wouldn’t cause any negative impact on the gameplay and it shouldn’t be called “RNG based weapon”. Of course this weapon doesn’t have any miss/hit chance mechanic, but it indeed features something that could be called “crit chance” mechanic. But there is a very important aspect, which is the fact that Photon Swarm is a rapid fire gun with huge rate of fire and projectile count.(especially on the higher power levels, where gun depends mostly on orange photons)

Because of how the probability works, the more instances there are in which something is calculated. (The more projectiles with crit chance there are), the final outcome is closer to the average value. (dps is more constant). Proposed lvl 10 of the weapon would rely on orange photons (crit hits), so someone could think, that RNG could mess the gamaplay up. But it also has 19 shots/s fire rate with 4 projectiles fired each shot, which gives us 76 projectiles/s. Let’s do some calculations: Let’s say that getting 2 only-blue volleys in a row is a serious problem.

Now if we look at the probabilites, it doesn’t seem problematic whatsoever. Probability for a volley to be blue-only is 0,25^4=0,39% Probabilty for 2 volleys in a row to be blue-only is 0,25^8=0,0038% .Two blue volleys have only 600 dmg difference from the average value, but their chance for appearing is already incredibly low. The chance to get a “low dmg streak” that would actually negatively affect the gameplay is so low, that it is negligible. To test things out I messed around with some RNG generators in internet and made 100 instances of 25% experiment many times. Average and expected result is 25 of course (25%*100=25). Vast majority of my results was in range of 23-27, with only few 22’s and 28-29’s. So I don’t think that there is anything to worry about. If Photon Swarm fired one projectile per second, there would be a problem, but with rapid-fire nature of the weapon, chance to get your run ruined by the RNG is impossibly low.

So to answer this question: Not always, but effectively close enough to make sustained and reliable DPS.

But someone could ask: What about low power levels? The fire rate is much lower there.

Well, as I said, I designed the changes with RNG in mind. Lower power levels rely on the blue photons more than on the probability of having orange ones. The main source of DPS increase on lower levels is from the rate of fire increase, not from the increase of crit chance. The more important the orange photons are, the fire rate of the weapon is higher. This way low power levels that have low fire rate, do NOT depend on the orange photons, but higher power levels, that do depend on orange photons, have fire rate high enough to make sure, no sudden drops in DPS happen. This is kind of complicated to explain. Basically : the more weapon depends on orange projectiles, the more rapid-fire it is, so the dps is more reliable.

Of course there is one argument left, which is “why not a pre-made pattern?”. From my experience I think that things with repeating pattern tend to look a lot worse and to be less enjoyable in situations like this. Pre-made pattern wouldn’t add anything to the high power levels, because the DPS would be stable anyways even with RNG, but would take away the element of straight fun and surprise from encountering unusually strong shots on low power levels.

Of course, all this is based on my calculations and experiences. I cannot be 100% sure that during playtesting some problems emerge, but I think that’s unlikely. In that scenario, pre-made pattern would be a go-to solution. As much as I don’t like this concept, I think it would still be much, much better than what we have right now. So yea, if this post doesn’t convince @Stardrone and others, that RNG in this case wouldn’t be that bad, pre-made pattern is good option. If most ppl wanted it instead of RNG…why not.

4 Likes

I read the new part and it may work

No but in those games critical hits gives a way huge advantage (damage multiplied x1.5 in the most recent x2 until oras) and is possible to miss 90% attack but hit
alwais 60% attacks

Agree and disagree.
As you stated a randomized attack pattern (and even powerup droprates) are a good thing to keep the game challenging but not frustrating (if is made well)
But in my opinion the player should have control over its weapons (Bloom mechanic in fortnite…) to avoid unfair situation.
By the way an rng aspect CAN be added on weapons.
On a mobile game called “War robots” (don’t install it… is unplayable for Free2players if you hadn’t started 3 years ago) there are sets of weapons that has a side effect with 25% or 20% of probability (that is low) to be given.
And returning to pokemon… a lot of moves have side affects.
What I am saying is not “RNG is bad overall” but on chicken invaders doesn’t have to stay on main weapons as a crit mechanic but as a secondary effect (imagine a weapon that gives poison on chickens)

Ok… I don’t want to do the maths at 21:23 but I trust you.
It may be accettable.

Well I didn’t changed my position even if I can agree with the math.
But this statement:

This…
Is totally not fair for the player.
In my opinion a weapon that states that rely on RNG is not good (I said it a lot, I know…) and can make a lot of unpleasant situations.
It may look fun,good and everything but NO.

3 Likes

I’m way too tired rn to try and make a coherent point of any kind, but since you mentioned WoT, I’d like to further prove your point by saying that the 3cm autocannon is, infact, pretty damn good at setting tanks on fire.

Well, I for one like the idea a lot. It makes the weapon more interesting and, as you have pointed out, the chances of getting screwed by the RNG are next to none. If I see a decent number of people opposing it I’ll make a poll and we can put it to a democratic vote, I guess.

I propose something, how about photons change in hue depending on power level instead of just having two colors?

I’d also like to drop this article.

That could possibly work, unless you actually want to keep the current colors while having the photons gradually turn from blue to orange.

morebloodyphotonssmh

I hue shifted the blue one and as you can see, the colors aren’t in the same order.
Also we’d probably need another rebalance without the RNG, since I don’t think it’d be a good idea to have two different projectiles that look the same.

That is true, however note that we also don’t have any weapons that can fire projectiles of two different colors at once. So this change would technically still keep it fairly unique, just not in the same way as it is right now.

5 Likes

The whole point of the part that you quoted was to show, why weapon performance wouldn’t rely on RNG enough to disrupt the gameplay, despite having RNG as a core mechanic for generating volleys. You said, that it is unfair for the player, tell me what element exactly you find unfair. The same with the “unpleasant situations”. What situations can be unpleasant? Elaborate on this a bit, because at this point I can’t really answer properly, because I don’t know what exactly you have in mind.

Unfortunately I cannot agree with this statement. Why is poison chance as a secondary effect any better than the flat damage increase chance? Both use RNG to determine if additional damage will appear or not. At its core, it is the same mechanic but with different name. Where is the difference?

In my opinion, %chance for special effect has a potential to be even more dangerous than flat damage increase, depending on the stength and importance of the effect itself. Because you either get it, or you don’t. In case of damage increase, you always have the damage, just sometimes a bit more (Of course I am still talking about high fire-rate weapons, that tend to not have problems with such solutions. In case of slow firing weapons it is just as bad as a %effect chance.)

Well, I am a bit skeptical of making the poll. I don’t think that should be decided by the majority based on guessing or calculations alone. Poll would be useful for understanding how people feel about it, but regardless of the outcome, the decision should be made by playtesters.

If majority wanted pattern, but playtests showed that nothing is wrong with RNG, throwing it away because of the poll would be a waste of the idea. On the other hand, if majority wanted RNG, but playtesting found out that it is not working properly, keeping the RNG would be a disaster.

Right now it is only an idea. Who knows how exactly it would work in game (Based on maths, RNG should be fine, but still, we will never be sure unless we play it.) Even though I am really against the pre-made pattern, after playtesting I could change my mind. Who knows. That’s kinda why I think right now making a poll would be kinda pointless.

Well, that could work (aside from the hue problems listed by @GgWw1175), but It would have to come together with big changes (that are entirely different from proposed here). Btw, I have this mechanic in mind, but for a different weapon. Maybe not exactly the same mechanic, but very similar.

Ah and I will answer part of one of the older posts in the thread that I accidentally ignored.

I have all that in my plans. However it turned out that giving Hypergun a linear damage growth would be mathermatically impossible without making radical changes to its projectiles and volleys (that was actually the thing that took me the most time)

2 Likes

I’ll be waiting for that I guess.

Regarding my previous reply:
To decrease the randomized effect, the number of supercharged photons fired is distributed evenly alongside the normal ones, until they’re all supercharged at power level 20. Take for example, at the lowest power level, 15 supercharged photons are made within 200 photon shots (fractional numbers of shots aren’t possible anyway). The average damage output will still stay the same this way.

It should be an another hue on every power-up and not distribued on a sample shot. Based on the article I shown earlier, the shot’s color gradually change in hue from red (representing low frequency EMs) to violet (representing high frequency EMs). It will still have the same amount of damage output starting from 400, etc.

1 Like

I doubt that the result would be any different from the solution I proposed. If among 200 photons, 15 are guaranteed to be orange, chance for a photon to be strong is still 7,5% and it doesn’t prevent weird stuff from happening. Theoretically all those 15 oranges could be amongst first 30 photons, and then next 170 would be blue-only. Of course the probabilty of somehing like this happening would be next to 0, just like with standard RNG. Even if you had 200 shot sections, each having 15 oranges, nothing would prevent 200 shot section with like 7 oranges from happening. (For example second half of first section has next to none oranges, and first half of the second section the same. Then between them there is section with next to none orange photons) So yea, the solution you proposed would probably work fine, but wouldn’t be any different from mine in case of preventing “bad situations”. It may sound different, but it essentially is the same thing. Each photon still has 7,5% chance to be orange, and the only thing that would change is the " ideally constant DPS on the long run", which ain’t that important anyways.

Besides, why are we so stressed about the orange photons anyways? On levels 1-3 they are basically the bonus feature that makes some bullets do 100 damage more. Core DPS comes from the blue photons and the fire rate. Even if oranges didn’t spawn on those power levels whatsoever, nothing bad would hapen. It is mainly an aesthetic effect. And on later power levels, where oranges actually become important, fire rate and the %chance are high enough that encountering sudden drops in DPS, or “low damage streaks” is impossibly low, thus negligible. RNG cannot ruin rapid-firing weapons. At some point it starts to work more like a stable DPS. That’s just how math works. It’s not the random “win or lose” mechanic, like crit chance, so I don’t think it should be even called this way, because it is misleading.

3 Likes

How can you read this? There’s no pictures!

I’m sorry I’m putting your balance program away for now, but since I’m not invested in the discussion, this really feels like a book assignment from school :S

3 Likes

Yeah, fair. And to be honeſt, I’ve been treating it in the ſame manner.

3 Likes

I am sorry that this part turned out this way. I really wanted to avoid getting into all this RNG stuff, but since it was brought up I felt obligated to explain everything step by step. About lack of pictures… well, it’s basically math and math is kinda hard to depict. Especially in case of the probabilities.

I think it’s high time to end this discussion, because you guys are starting to get annoyed by this. I’ve already said more than I initially wanted to. At least I have some good news. Next parts aren’t going to include any similar mechanics, so this “assignment from school” shouldn’t happen again.

4 Likes

Nah, I’m not annoyed; there aren’t penalties for not finiſhing the project (or ſtarting it, as the caſe may be).

2 Likes

If we must go with visual change in Photon Swarm I would rather prefer pre-made pattern. It seems more stable for some reason. But yeah, would be good to test it in the game.

2 Likes