Points of Imbalance.
-
I build spreadsheets to analyse the fairness across units/factions. My opinions are only that, opinions, yet I shall when relevant or possible attempt to corroborate them with facts or figures.
Shimmer stun time is a point of potential imbalance. I shall now point out a myriad of different factors to consider when deciding if the Shimmer is indeed imbalanced using the Medusa as the ‘baseline’ for a balanced unit. Shimmer stuns for 2 seconds. Medusa stuns for 3.5 seconds. These values were measured in-game not taken from the unit database. With stun duration, we have to factor in stun frequency and stun area. Medusa has 2 radius while Shimmer has 4. Medusa hits target every 6 seconds while Shimmer without micro (meaning without hover bomb) hits every 10 seconds. The Shimmer in the unit data has a fire cycle of 5 seconds but a flight path of 10 seconds, resulting in a 10 second fire cycle. Compounding this data the Medusa stuns for 1.75x longer, fires 1.67x faster and hits 0.25x more area (<-simple geometry). Assuming simple multiplication can deduce effectiveness, the Medusa is 0.73x more effective than the Shimmer at stunning if units can't die. But units die, thus reducing their effectiveness. HP per mass the Medusa is 2x more efficient, has 0.683x more HP and T1 interceptors deal 1.8x the damage of T1 tanks for 0.935x the mass cost on average. Additionally the Medusa costs 0.34x more mass and 0.0735x more energy. To summarise, the Medusa is more “tanky” for a reduced cost. To calculate an adjusted effectiveness rating the standard effectiveness should be multiplied by survivability (HP ratio) and divided by cost (using mass ratio for simplicity's sake). This yields the formula: 1.75 x 1.67 x 0.25 x 0.683 / 0.34 = 1.467. At this point I expect the reader is drowning in values and simple calculations if they are still reading at all! To summarise, the Medusa is 1.467 times stronger at stunning stuff than the Shimmer. Perhaps the balance team could consider increasing the 2 second Shimmer stun too 2 x 1.467. Or 3 seconds! However, in patch 3704 Chrono Dampener stun duration was decreased from 3.5 seconds to 2.5 seconds because the longer stun duration caused units to lose their move orders. In order to not lose move orders, I suggest the Shimmer stun time be increased from 2 seconds to 2.5 seconds (not 3 seconds) to preserve move orders of affected units.
The Aeon T3 bomber is imbalanced (weak). Here are the facts explaining why. Currently the ratio between damage and payload radius is: Damage = 4500-(Radius*250). I shall assume this ratio has been tested to be “balanced”. Each T3 bomber follows this rule except Aeon, why? Subsequent question. Why are the AA ‘goodies’ on the UEF T3 bomber stronger than the Cybran T3 bomber? If the answer is because Cybran has stealth, why don’t Aeon or Seraphim have AA? What ‘buff’ counteracts this imbalance? Next topic. Each T3 bomber has a different set of stats which makes them more or less suited to fulfilling a particular role. First a table indicating the relationship between Mass Extractors and T3 Bomber bombs required to destroy said T3 Mass Extractor.
Faction UEF Cybran Aeon Seraphim UEF T3 Bomber 3 2 3 3 Cybran T3 Bomber 4 3 3 3 Aeon T3 Bomber 3 2 2 3 Seraphim T3 Bomber 3 2 2 3Next the radius for each bomber. Cybran=7, UEF=6, Seraphim=5, Aeon=4.
From the table we can see that the Aeon T3 Bomber is just as good at killing Mass Extractors as the Seraphim Bomber yet has 1 less radius. Imbalance. You may say “the Aeon bomber has more damage to compensate for this” and this is wrong. The Aeon bomber only has 200 not 250 extra damage over the Seraphim bomber and doesn’t follow the “balanced” ratio! Imbalance. Solution is to add 50 damage to the Aeon Bomber. Just do it.
Now onto the goodies (T3 bomber anti-air). Seraphim and Aeon don’t have goodies like UEF (their anti-air) or Cybran (anti-air + Stealth). Aeon is the closest to Cybran with only 100 more hp. Does having 100 more HP justify no anti-air or Stealth? No. People may say “but Aeon deals a whopping 700 more damage” but they also forgot that it has 3 less radius. Remember we are assuming the ratio for radius too damage is correct, so clearly the Aeon bomber is underpowered (provided the ratio is balanced). Now for Seraphim. They have 200 more HP than Cybran and 100 less than UEF and STILL don’t have anti-air! How can you justify giving the UEF T3 bomber AA and not the Seraphim bomber! UEF favouritism I see… It was suggested in this forum post (https://forum.faforever.com/topic/49/all-the-reasons-aeon-sucks-t1-worst-of-all) to offset the damage and lack of ‘goodies’ imbalance by giving the Aeon T3 Bomber’s bomb target tracking. However if the Aeon T3 Bomber was to get tracking on their bomb, I would expect the bomber price to increase drastically, or the tracking to be negligible/non-existent. That said, this would provide more factional uniqueness, compensate for reduced (imbalanced) bomb damage and offset the lack of goodies.Galactic Colossus. A reasonable unit with a suspected software error in its tractor claws. The Galactic Colossus tractor arm in-game only activates once per 12ish seconds when tested (GC idle against Percival’s moving towards it). In the unit data (https://github.com/FAForever/fa/blob/develop/units/UAL0401/UAL0401_unit.bp), each of the two claws is supposed to have a fire rate of 0.15shots/second which is about 6.6 seconds per activation per claw. Why the major discrepancy between what actually happens and what is supposed to happen? Conclusion, a software error. Can this please be fixed to make the Galactic Colossus more consistent?
If this is in fact, not a software error, please consider halving the activation time for the claws as the Galactic Colossus is more costly compared to the Ythotha (by 1000 more mass) and is less effective against large T3 unit groups because the Galactic Colossus has no AOE damage. If you are curious about this ‘fact’ I suggest you make a ‘test’ game and throw 30 T3 units (direct fire of same unit) at a Galactic Colossus, then repeat for the Ythotha. The Ythotha will have destroyed more T3 units than the Galactic Colossus.The Novax. Currently it costs exactly half as much as a Duke, has more than half the DPS of the Duke, is 100% accurate, has omni and radar while having unlimited range (via moving). The only reason not to build 2 Novax instead of 1 Duke is the area effect of the Duke. Please apply a very minor increase to the mass cost of the Novax to offset the additional benefits.
Talking about intel and balance Seraphim is lacking. Aeon has the Eye of Rhianne, Cybran has the Soothsayer and UEF have the Novax (despite this being way more expansive). What can the balance team do to balance this? Yes, the Iaselen T3 Spy Plane has Sonar unlike the other Spy Planes, but this is to offset the lack of a T3 Sonar Platform.Now onto the Percival. Its shots are almost impossible to dodge with T3 land units. Factor in the alpha damage (1450) + greatest range (34) every unit which attempts to tango with the Percival has their actual maximum HP reduced by 1450. This range, plus their high HP, plus high alpha makes them nearly impossible to counter with land units. This statement assumes equal mass investment into both armies and only into direct fire land units, not indirect fire. I can already sense people saying something like “just don’t fight them and attack elsewhere”. Such a statement is true yet also implies two things. One, that Percivals are unbeatable. Two, the non-Percival player is playing on a countdown to win before the UEF player reaches a critical mass of Percival’s. A suggestion to allow counter play by land units would be to reduce the Percival’s speed to ~1.8. This will mitigate the Percival’s range advantage, emphasise the “attack elsewhere” point and make ‘poke’ from Percival’s more easily punished. Movement across the battlefield for Percival’s will admittedly be affected, yet UEF possess the only T3 transport. Perhaps the speed reduction of the Percival will encourage the Continental to be used once again. A less favourable suggestion would be to further reduce the Percival’s muzzle velocity however this forces the opposing player to burn more APM in fights to dodge shots, resulting in a subtle advantage to UEF players.
Fun fact, the average range of the Harbinger, Titan, Loyalist, Percival, Brick and Othuum is 26.125 when including all their land-to-land weapons. The Percival has ~8 extra range on average at the T3 phase. Perhaps this is one reason why Snipers and T3 Mobile Artillery were the main T3 land army composition a few patches ago??I am exceptionally curious to hear feedback from the balance team and the community.
-
I have tried myself quite a bit on mathecatically describing balance. I am currently at the stage: Complex system even just for direct fire units you will only get to a precision of 10% of the balance mass cost at best. There are so many unit stats that change the strenght of a unit. Easiest to start is with effecitve DPS, HP and Masscost. But then you have a lot more stats like energy cost, movementspeed, range!!, firecycle (regarding alpha damage and overkill, which also quite depends on the army composotion you are facing.), turret turn speed, hitbox size, acceleration. And I am sure there are at least twice more stats than I have named here that effect balance. So a theoretical approach via those unit stats becomes very hard to do. It can be usefull when you have to first set a baseline for balance for a mod etc. I did help Dragun with his SCTA mod a bit with this and an efficiency calculation based on lanchesters laws. But that is only a base line after that you can only improve balance via playtesting.
I am going to talk about the medusa bomber stun issue as an example. I find it quite hard to compare the stun of a medusa and the bomber. Mostly because one is an air unit and one is a land unit. So the counters to each other are quite different. Also the bombers main advantage is that it is very fast and is not restricted by terrain. So it needs to be less efficient than a medusa because it can attack way behind the front line whereas the medusa is not capable of doing so. The stun is rather a gimmick for later game use of the bomber because it is cheap and can stun high tier units. Also medusa usually gets dodged, whereas t1 bombers attacking an army usually don't get dodged.
But the biggest issue in my opinion is that you are asuming all tiers of all domains of all factions should be equally strong. That is not the case and is not desired as far as I am concerned. Quite similar balance is required because of the map diversity present in FAF. Some domains have a huge advantage over other on some maps. This results in very similar factions if you want to have interesting games between all factions on all maps. But you still have a bit of wiggle room and in my opinion it should be used. So that is the reason there isn't even an intent to make all t1 bombers equally strong (even if it was possible). They need to be comparable but I am quite certain if you made a statistial analysis of all replays and just calculate the average vet of all t1 bombers in all games it would not converge towards the same value for all bombers. And quite often you don't even have an equivalent unit in the other faction.
Balance is messy and as far as I can tell it is currently done by a couple of the top players having a discussion about their intuition of what is op and what is not. Another problem on top of that is balance is distorted on what tactic (unit composition etc.) currently is in fashion. Also ease of use has a huge impact. Most memorable example that comes to my mind is the introduction of SCAU presets.
If you are interested this is the (old) combat efficiency of selected units calculated via a partial discretization of lanchesters laws wich only takes HP, Mass and avrg DPS into account:
-
"Balance is messy and as far as I can tell it is currently done by a couple of the top players having a discussion about their intuition of what is op and what is not."
Yeah, I think that's how we wound up in the current situation regarding T3 land. Assault Experimentals used to be in a place where they were niche units that complemented T3 land, but their utility dropped off in higher skilled games.
So they changed that, and now assault experimental units stomp on the role of T3 land formations, effectively replacing a large part of the game on some maps for some factions.
People that understood the complexities of balancing a game set up a system, and when the hand off to the FAF team happened, the intentions and understanding of how things were supposed to work were not conveyed. The FAF team then made a change that on the surface seemed very reasonable, but which had unintended consequences.
This is further exacerbated by the differences between the kinds of maps you find in ladder versus the type preferred in team games, where map control is often less of an issue. Maps like Isis which were once dominated by T3 formations are now dominated by T4 assault experimentals (unless the game goes late and the map closes completely), and replacing large diverse formations of many different units with a single unit is clearly not good for gameplay.
The point is that there is more to balance than just knowing what is OP and what isn't. There is vision, and If you lose vision, its easy to mistake the role of a unit and make choices that have far reaching unforeseen consequences.
How do you keep Assault Experimentals from stepping on the role of T3 land? You make them weak for their cost, but fast to build, so they can be used to take advantage of a large reclaim field quickly, or be rushed in a surprise attack. That's how you do it. It doesn't mean that they're underpowered.
If you don't understand that this was the vision behind them, and then make changes, you wind up undercutting a lot of other units and reducing unit diversity in the late game.
-
Most direct fire Experimentals can still be countered efficiently by a t3 army. Especially by Bricks and Percies. But usually people are just too incompetent to have a good formation against the experimental and suicide them one by one. Did the balance even get changed in that regard. I only remember there was a build power nerf on experimental and Heaven complaining that the dynamics of Experimental ping pong on most team game maps. got killed by it.
-
T3 nerf to bring it in line with the rest of the tech tree was an indirect buff to T4. Games between competent people shouldn't have experimental ping pong anyway. I mean, there is no ping pong between large groups of t3 either. Simply too much of a mass investment to throw away in a "ah yeah maybe it'll do damage lol" attack.
In fact the ping pong between
- player masses up t3 push and takes map control
- other player puts mass into a t4 concentration
- Other player retreats to build up a larger t3 mass to defeat t4 or retreats to get his own t4
Is more of the ping pong situation you describe.
One example would be a dude pushing with a mass of snipers/shields, meets a fatboy, has to retreat to some sort of firebase/own t4 to push and tank fatboy with/needs some major clump of t3 that can tank the fatty.
Also a lot of underselling of the balance team here lol. The GPG balance was absolutely terrible in this game. In fact, I'd go so far as to say that the balance in base supcom/FA was SO bad it's actually a fundamental reason why the game failed to garner any sort of real competitive scene. It had so many inconsistencies with tons of the unit rosters being either useless or bonkers OP with no real thinking needed to choose between options. Nothing that has happened due to the T3 rebalance has really been "unexpected" in the sense of broad meta strokes. Well, maybe the huge efficiency of t3 mobile arty wasn't accounted for. T4s may or may not need a slight adjustment in their stats, I personally don't think they do and enjoy their purpose in the current meta.
Would rather nerf garbage like RAS SACUs.
-
@harzer99 said in Points of Imbalance.:
Most direct fire Experimentals can still be countered efficiently by a t3 army. Especially by Bricks and Percies. But usually people are just too incompetent to have a good formation against the experimental and suicide them one by one. Did the balance even get changed in that regard. I only remember there was a build power nerf on experimental and Heaven complaining that the dynamics of Experimental ping pong on most team game maps. got killed by it.
God no, go test this. You have to really blow it to lose even a ML to bricks.
Literally spin up a game, and put a formation of bricks against a lone Monkeylord. With decent micro the Monkeylord has a really good shot at winning. You have to be careful and pick at the sides of the formation, but bricks just don't have the range or speed to punish a ML that nips at them.
Percies are a bit better. Only 2 range difference between them but between alpha - which means when they do connect they do decent damage up front before the ML deletes them and that 2 range putting more percies in firing range they do substantially better than bricks against a ML. Very small advantages add up to a very different result.
Then consider cost of production of large numbers of bricks, and unless you're on a map where you have to maintain control of a large area, the ML is just superior.
I mean just producing the cost of a Monkeylord in bricks takes forever unless you've invested a ton of mass into T3 production. Even after making those investments, the ML likely finishes faster.
Then the bricks are far more vulnerable to certain units - like T3 mobile arty or bombers. They get more health, sure but anyone with any sense will use splash against them.
-
Yeah, the first thing I do when I finish my T3 Land HQ is produce a T3 engineer so I can build an experimental. Sure I only have a single T2 pgen and a handful of T2 mexes at this stage of the game, but I can afford to give up the entire map and spend 15 minutes building an experimental with my one engineer and small eco. There is no way my enemy will attack me with his existing T2 army or build a few T3 units during this time and crush me. This strategy is foolproof. I can just invest 5220 mass into an HQ then another 20k in a Monkeylord and there isn't a damn thing my opponent can do about it. T3 units have no role in this game.
Will you just ban this guy already?
-
@FtXCommando said in Points of Imbalance.:
Would rather nerf garbage like RAS SACUs.
Agreed on that point.
So I agree that the balance team has mostly done a great job. I can only think of a few things I disagree with that they've ever done.
As for your other points. Ping pong?
I assume you mean a situation where you attack and lose your assault experimental so the other team uses the mass and builds one and sends it at you and they lose it?
I mean that does happen, and it makes sense for it to happen at any level, because oftentimes the damage done with the experimental is worth more than the reclaim left behind. You can do significant damage without getting a kill, and it can be worth it to leave the mass on the field in exchange for doing that damage.
If you wipe out someone's SMD as an example, it was worth it to use the Monkeylord - even if it dies and the enemy gets the reclaim.
I hope that the SACU changes will rectify the imbalance between T3 units and T4 such that nothing else is required, but I think its going to require more than is planned. Cybran in particularly looks like it won't benefit much versus T4. I checked the FAF Beta implementation and if the SACUs do grant a speed boost, its so small as to be unnoticeable.
Making SACUs giant deceivers won't do much to help Cybran T3 land avoid getting stomped by Chickens or Colossuses.
-
I don't play dudes dumb enough to put their SMD out in a way that it dies to an ML.
Don't even know why you bring up ML, it's like the easiest T4 in the world to absolutely throttle mass efficiently. Hell, get an ACU w/ gun+respective hp upgrade in a t2 transport and like 6 t3 assault bots or 6 snipers and you raped it.
-
@ThomasHiatt said in Points of Imbalance.:
Yeah, the first thing I do when I finish my T3 Land HQ is produce a T3 engineer so I can build an experimental. Sure I only have a single T2 pgen and a handful of T2 mexes at this stage of the game, but I can afford to give up the entire map and spend 15 minutes building an experimental with my one engineer and small eco. There is no way my enemy will attack me with his existing T2 army or build a few T3 units during this time and crush me. This strategy is foolproof. I can just invest 5220 mass into an HQ then another 20k in a Monkeylord and there isn't a damn thing my opponent can do about it. T3 units have no role in this game.
Will you just ban this guy already?
You have games where you played on Isis out there.
Did you spend money on 5+ T3 land factories so you could spend a significant portion of your mass on T3 units, or did you... ya know... push out some T2 and T3 and as soon as you could start spamming assault experimentals?
You do it too.
-
Also all this reminds me: this imaginary t3 phase of unit diversity never even existed. You had the generic early game cheese with harb/loya.
Then you had the absurd crushing of infinite loya stun spam combined with absurd crush of percy + shield + fatty.
Then you had cyb respond with mega spam because it's all they could do against percy spam (since no mobile shield). Then you had Aeon spam infinite gc because it's all they can do (since percy hard counters harb so quick it isn't funny). And Seraphim spammed infinite chicken (phim t3 is just utterly pathetic).
Cyb kinda did the same thing as UEF against aeon/phim with bricks. Or they could abuse loya but it isn't as hot against harbs. Aeon just ran around siege tanks and abused their shit range.
Now we have:
Titan/Loya used for raiding in massive maps with lots of loose mexes/map control to abusePercy/Brick far more even in combat with a slight edge still going to percy due to shield/alpha advantage
Arty mix able to capitalize on and respond to both snipers and immense buildup of assault units with no support.
Snipers able to act as a response to massive percy/brick spam rather than a tickle force that insta dies if you sneeze on it.
Bricks/Percies well used will still crush T4s other than particular ones built for kiting them like chicken. Fatty exists as a way to stop snipers from destroying percies so that they can keep pushing in. ML exists for the cheese bs it always existed for. GC exists to brutal tank land pushes and is a solid hp wall to hide your more dps heavy units behind (whether harbs or snipers) so that you maintain full dps while whittling down the enemy.
I'd say it's irrefutable that t3 land both early and late has the highest skill ceiling of all tech levels on land. You could make a totally reasonable argument it's the highest in the game if you wanted to. The only thing that really compares is naval combat.
-
I'm not complaining about how bricks do against percies...
I'm also not complaining about the balance between T2 and T3 units. The change wasn't all bad.
I'm specifically complaining about the T3/T4 balance.
You can't make bricks in large numbers on some maps, because if you form them up they just get annihilated by any assault experimental the other player(s) send your way. That didn't use to be the case. You needed a critical threshold, but after that threshold was reached, bricks were pretty damn good.
Now that threshold doesn't exist.
Its now a replacement for T3.
The only reason this isn't obvious to every single player is that T3 units still find utility on maps where map control is critical. On maps where you don't need units in 50 places at once, assault experimental are pretty clearly OP.
Don't you guys have statistics? How much mass is spent in high level games on Isis in T3 land versus assault experimentals? If you think I'm full of shit it should be easy to prove with data.
-
15 bricks crush an ML. 30 bricks crush a mega when it has nothing to retreat to. 22 bricks crush a GC. 21 bricks crush a chicken when it has micro as garbage as mine lol.
Not even sure why you say that closed games come down to T4. If I know where the T4 is coming, I can very easily stop it through mass efficient defenses against it and just keep making my eco sim. This can be point defense, arty, rambo SACUs, air. This is what happens on every canis/hilly/whatever I play that isn't decided by a brutal air crush.
This is also why T3 is so garbage on these maps. Because it's SO bad to make infinite unit spam since you are immediately made irrelevant as everyone and their mom ecos up and leaves you on your unringed t2 mexes totally marginal to the game state. To make t3 spam worthwhile you need a consistent transition between t1 -> t2 -> t3. Or at least some t1 -> t3 or t2 -> t3 aspect. If it's some closed garbage map then you skipped t1 and t2 and will more than likely skip t3 because it isn't worth the lump sum investment. Might as well as scout and abuse defender's advantage by rushing your own t4 faster than the enemy can make theirs + bring it to you. I'm already spamming hives for my quantum gateway for RAS SCUs so why make even more unneeded infrastructure?
This is not a balance problem but rather a map problem.
Actually no, it isn't even a map problem most of the time. Mostly it's just that players are pretty incompetent and can't be trusted to work as one. So you are better off just ecoing up and finding your own sieves to abuse and get ready to capitalize on anything some dude on your team manages to randomly do.
This is why 2v2s are much better to see how serious teamgames would go. You not only have easier coordination but a vast space to interact with which almost always correlates with more variability in strategy. 3v3 and 4v4 are also ok but it requires 20x20 maps pretty much.
-
Spin up a test FTXCommando, spin up a test and actually see how well a Monkeylord does against 15 bricks.
Takes 5 minutes.
You're guessing here and you're guessing wrong. I'm sure you can find a friend to help you test that. If you can't just message me when I'm online.
-
I did run a test. Unmicro'd, 10/15 bricks live.
With micro purely on the ML, I managed to kill all the bricks with a 500 hp ML left. This was because 6-8 of the bricks were walking out of range and stopped shooting.
-
@FtXCommando said in Points of Imbalance.:
I did run a test. Unmicro'd, 10/15 bricks live.
With micro purely on the ML, I managed to kill all the bricks with a 500 hp ML left. This was because 6-8 of the bricks were walking out of range and stopped shooting.
I think you're bad at tests Are you on now? Wanna join a test?
-
I think you're bad.
-
@FtXCommando said in Points of Imbalance.:
I think you're bad.
For anyone that wanders through here.
This is the real issue. If he ran a test he ran it badly. If you don't believe me, find a friend, and test this yourself.
How does a ML do against 15 bricks. See how you do. Really easy thing to test for yourself.
Talk is cheap.
-
@moses_the_red said in Points of Imbalance.:
Talk is cheap.
In that case, feel free to post the link to your testing at any time.
If you're going to just shit up another thread then I will ask for you to be banned from posting.@Arran mate, I think you've misunderstood something here.
You're working on the assumption that every equivalent unit must have the exact same level of power. This is not how it works in pretty much any game balance ever made. Doing so would instantly force a meta wherein the faction with more options will always be at an advantage, and I'm really sick of this permanent cybran meta already.All these minor stat changes do nothing to the game besides suit the criteria that you and you alone have created. You're again under the idea that the balance team use this similar forumla (because you noticed a couple of coincedental patterns maybe?) but this is also likely not the case. Stuff like adding 50 damage doesnt solve a problem with balance, or remove a point of frustration. I don't at all see why we should "just do it"
Stuff like the Novax and sniper meta reveals a lack of game knowledge. I don't really have the motivation to explain why that is though, moses will probbaly derail your thread with his hero complex anyway.
-
"With micro purely on the ML, I managed to kill all the bricks with a 500 hp ML left. This was because 6-8 of the bricks were walking out of range and stopped shooting."
funnily this is pretty much what happened in moses "test". somehow the bricks were never in any kind of real formation with 2-3 shoting at once and the monkey had infinite kiting space so it could just keep picking off flanks
