I went to bed and had some thoughts, and they eventually turned to how you would balance a game. In the end, in a game where you have a limited amount of inputs which results in a predefined output (press S to move in south facing direction at 1 tiles per second), with limits on attacks that can be performed, etc, there is a theoretical best game in every game of CM13, and possibly in any game ever created.
A tool assisted speedrun is a run which is designed to make the exact, frame perfect moves, every frame, to generate that theoretical absolute perfect run:
You could make such a tool to run a CM match, where you would input from the start the direction, ability uses, attacks, etc, of each and every marine and xenomorph that exists in the match. But why engage in the thought experiment? Well for one, this is a diagnostic tool for game balancers.
If you create a perfect game state, after some time, you would end up with a situation where every player has played âperfectlyâ with no room for any theoretical improvement. They have all reached the absolute pinnacle of what is actually, physically possible, in CM13.
And that results in a win condition.
And if that win condition is marine or xeno, it has implications for the current balance state of the game. So what can really be done to create a balanced team match? The solution might be to create situations which result in a stalemate 100% of the time.
Human beings are simply not capable of playing frame perfect matches forever, they will eventually make a mistake. And this will end the stalemate through cumulative imperfections, but if the game is designed in such a way that if every player was playing perfectly, and it always reached a stalemate. Thatâs probably the most balanced the game can actually be, isnât it? No side has the advantage, itâs all about the playerâs inputs.
CM heavily depends on RNG, unlike Mario or other games. In one instance all 4 buckshot pellets hit at two tiles, in other none do due to accuracy RNG.
But if one side requires much less effort to play close to perfection than the other, then that statement is completly false.
Individual xenos are much more powerfull than individual marines (with specs being exception).
So while perfect AIs playing xenos vs perfect AIs playing marines, the latter would basically always win, if you were to balance the game for those AIs by nerfing marines and then you put that gameplay for humans, xenos would always win.
As it is easier to coordinate with yourself, than with others using a chatbox without even TTS.
This is not âbalancedâ except by the term âbalanceâ. Itâs simply equal in outcome.
Balance is to have a state that is fun for everyone. Not a state that is 100% fair for everyone. If perfect fairness is predictable and boring, equality in outcome (in videogames) has no meaning and is just an arbritrarily defined âgood/desirable thingâ. Ultimately the desirable thing is the gameplay overall, the outcome is just a small, small part of the fun of a videogame.
CM, like chess, has a massive amount of outcomes. Calculating every single possible input into every single outcome is impossible. But if calculating by agreed upon patterns (tactics), there is always another solution, especially since both sides are independently controlled.
Thatâs why CM is ultimately a complex game with an incomprehensible number of inputs and outputs.
marines stop extremely hard, every single marine grabs extended barrel making it impossible for xenos to dodge bullets even if perfectly moving.
Marines just murderball with 1000+ DPS to any TAS-Xeno that walks onscreen, with the only counter being boilers who snipe from out of vision and delay marines for a few seconds.
Queen screech means very little when every TASMARINE can focus fire.
Mk2 does a total of 176 DPS assuming all shots hit, even if a TAS-Xeno can dodge 50% of bullets thatâs still 88 damage per second per marine with LoS. Even in caves, youâre still generally able to be shot at by 4 marines at a time (352 DPS) and die in 2 seconds, meaning that all melee xeno castes⌠donât really work at all, even TAS-Crusher would get melted as every TAS-Marine can instaswap to their AP mags the moment a crusher appears onscreen.
Not even mentioning CAS/Mortar/OB, TAS-Marines stomp 10/10 even if theyâre all limited to just the basic marine gear of MK2s + ammo + a few AP mags.
I think TAS-Marines would still win even if they were permadeath with no firesupport, thatâs how much the game is balanced around marines, yaknow, not having perfect accuracy/spacing/gameknowledge.
Surely this cannot be true. If you have a side which is advantaged to win in a hypothetical perfect play scenario, your game is fundamentally one-sided. Can you explain your views on this further? Fun has nothing to do with balance in my view.
This is true, and I did think about this, but then it would be a case of the average performance. Card games are similar, card deck draw is often highly randomized, but the average chance of a good deck to win is higher by reducing the RNG as much as possible and making the possible results more impactful.
All marine DPS can be averaged based on its actual chance to hit and spread, which should give you a number that would resemble âmarines will get this result in X% of casesâ, which is still very useful for balance purposes.
There is actually a term for this in chess and itâs a theoretical idea thatâs been around for a long time, itâs called the âsolved stateâ in chess.
Whilst it is hard to get there, there is a numerical limit to the total possible outcomes of moves in a chess game, even if that number is very very large. And AI have been mining that cavern for a very long time, before I was even born.
Hard to use does not actually amount to anything when we talk about good players. Think about many fighting games or other instances of a game having a supposedly âhard to useâ mechanic which, given you have a competitive scene, it becomes a standard that you will master it to the point that executing these mechanics becomes second nature.
Hypothetical perfect play scenario // Fundamentally one-sided:
Probabilistically, this will never happen. After one or two inputs you will have a completely different generated game from the last - and each game will have a different solution, with different problems every single marine and xeno face (and problems are arbitrary and depend wholly on what the marine/xeno interprets as a problem - which is part of awareness and pattern recognition). That is why this game has an effectively uncountable amount of outcomes - we cannot possibly perceive or learn of every outcome but just operate in a way to make patterns ideal, which might create any amount of predictable outcome that we can utilize.
If every single operator is acting in a way to make ideal patterns for both teams, that is an undeniably complex problem that I canât possibly begin to think of how to solve
The skill ceiling of this game is so high on both sides that noone even comes close to it - there is a currently uncountable number of fundamental patterns to solve for undeniable gains (tactics + strategy) in the face of undeniable counterplay (layers of counterplay multiplied by teamwork) based on and in response to enemy and personal resources (such as info, terrain, position, tempo, etc) which no player can come close to doing perfectly - to the point predicting or influencing the probability of other players doing something (player rng) is part of the skill ceiling.
All on top of an asymmetric game - one side is completely different from the other and has a completely different experience.
The reality of the situation too, is that if the player feels like they are enjoying the scenario, they are enjoying the scenario. It has nothing to do if they win or lose, except for maybe a feeling of helplessness or like theyâre unable to do enough for their team. Much of the fun of a game is getting better, learning more, and seeing the results you worked for in reality - it encourages you to find your own fun.
To keep the game âbalancedâ, ironically the best option is not one where it is fundamentally 50/50 (because this is arbitrarily applied and defined as good - itâs a fallacy), but a solution which maximizes the feeling of enjoyment for both sides. Itâs a fallacy to say balance is only of outcome when it is balance of all breadth of game aspects in the ecosystem of the game - in which considering this, there is many desirable solutions.
Defining something as âone-sidedâ is improper logic here - thereâs more to games then the outcome, and this is a fundamentally different thing then a 1v1 game. Even in chess, one side has an advantage, and it really doesnât matter, the fun is the back and forth and coming to solutions to the enemies tactics and strategy while perpetuating your own (and further in a competitive format they play as both sides equally which would not be unusual in a CM competitive format too).
Here it is even much different then a 1v1 game - everything you can do and use. You make the choice of how to play to do what you find most fun. In a niche-based ecosystem, you choose the niche you use for whatever reason you want to choose it - and making these niches fun, capable, but not too capable is the priority. You will never have the same capabilities as others, nor a guarantee of win, but thatâs fine because you have your own unique capabilities. We can only measure how niches go against each other to some degree of reliability based on the things that make those niches unique, not a perfect representation of every niche in every situation. Further, there is enjoyment in the depth of every niche which you can find by playing that niche.
And if a solution impacts desirable things such as the depth (and thus replayability) of the experience by making certain niches incapable of performing their desired playstyle (having fun), or too capable (reducing depth or breadth and enjoyment of others experience without systematic, strong, or a variety of interesting counterplay), that is when itâs improperly balanced. Because improper balance leans too much depth into certain aspects and leans the other side downwards in depth as a result - thus to build a stronger foundation we must make sure it is structurally consistent, even if uneven in outcome.
The main reason marines choke so hard is because the average player misses over 50% of their shots
Xenos being a melee faction means in a vacuum, they lose every time simply because 100% accuracy marines will always stomp before xenos even get into ability range half the time
That it will probably never happen doesnât undermine the logic beneath it, we can demonstrate this with much simpler games that have a reduced total set of outcomes, chess and CM13 are on a fundamental level, games that have a limited set of outcomes given each input from the first âturnâ, or in CM13âs case this would be represented as the roundâs very first game tick onwards.
You could show this with tic-tac-toe, which is a solved game that always ends in a draw given optimal play. (Of note, 3D TTT apparently always ends in a win for the first player).
Why is this relevant to your claim here? Well itâs because we can use TTT to demonstrate that even if a player starts their game differently and generates a different game state, the next optimal move by the opposing player has already been found, itâs a case of âif opponent does this, then pick thisâ, this works great for computers emulating a game because theyâll do it immediately and perfectly.
In a game of colonial marines, what this might look like could be represented by movements in a closed room. Say there is a xenomorph that is being hunted down by a marine, the marine has the advantage here in that they will win the fight if the xenomorph cannot regenerate their health enough at that time.
There is an optimal position for the marine to move which would prevent the xenomorph from finding cover, or the marine could outright destroy cover in that room. Any other option for the marine could be considered the suboptimal play in that closed and specific space. If we can find optimal solutions for specific puzzles in a smaller space of CM13, it goes that we have a hypothetical perfect game in a grander sense, it just hasnât been found yet.
It begins by waking up in cryo, marine!
Tool-assisted runs of games are sometimes made by entire communities who work together to try to figure out an optimal solution, and they can take many months. Like in the video, it describes a TAS as âa painting made over a long timeâ versus a short rehearsal of skills in the moment. Wouldnât you find it interesting what we could cook up given we had the tools to do so? Itâd probably be very interesting to watch that round play out.
This is true of speedruns as well, it can take lifetimes to get âthe perfect runâ and some people are sure no-oneâs going to optimize the time, but then someone does eventually come along and shaves off anything from milliseconds to minutes off the best score. Just because these strategies arenât the perfect ones in actual reality, it doesnât mean that they canât be used to get an ideal of what optimal play looks like in that moment.
You keep using quotations here when you talk about balanced. Balanced is very intuitive to anyone who plays competitive games. It means fair. In real life, what people will be heartbroken to find out is that there has never been a fair fight in all of human history, and it comes down to learning one of the fundamental issues facing humanity. No-one is born equal, no-one lives equal lives, and thus no contestant has ever arrived to an arena with the same advantages and disadvantages as another contestant. Weâve just never seen it.
Another sad thing people will realize is that you can take a hundred people in a room who all practiced âequal amountsâ and worked âequally hardâ for 10,000 hours and many will perceive them all as equally deserving of the reward of winning the competition. But there is only one world champion. And if everyone worked equally hard, then that didnât mean anything in the end, did it?
But even if we cannot make a âfairâ game, everyone who enjoys a competition wants to have a winning chance, and a 50/50 shot is as good as it gets. The only reason that people donât complain too much about whiteâs first move advantage in chess is that it isnât yet clear if it is that much of an advantage. But there are some people out there who do think itâs a farce to say that chess is fair or balanced due to it. If the advantage was causing a big enough upset (say competitions saw white winning 60%+ of the time) theyâd have to figure out a way to amend chess so that it was fair in reality.
If a game is in a solved state with the first move player always winning, it is one-sided. Thatâs not an illogical position to have.
This is your subjective opinion and I donât mind it, others do have issue with it, the more people invest their life into chess, the more itâs going to matter to them that they lose a grand prize with these kinds of imbalances in place. Then it becomes a financial issue for them.
If I want to queen in corpo dome thatâs a meme game but Iâm not going to pretend itâs a good idea or balanced. I think that showing people the ways in which some niche players can struggle is a good thing because then they can have some balance changes made to make their ânicheâ less punishing for both them and their team to use. These processes can lead to discovering who is falling behind as much as who is doing too well in a mechanical sense, so itâs worth doing for that.
But a tool assisted run doesnât even have to be made for the purpose of balance, sometimes itâs just fun.
I donât think this would happen at all, unless it was discovered that someoneâs ânicheâ was just a previously slept on absurdly overpowered mechanic which did actually deserve a nerf then even before it was discovered and then everyone starts to abuse it for being OP. As it turns out, if you find winning fun, being on an imbalanced game will be a fun experience for you because you will win more in the unfair matchup. I donât particularly have much sympathy for those kinds of people.
But do we? One good player on xeno side kills 30-40 marines alone. One good marine player (non-spec) working with semi-competent teammates kills T3 and couple T1s, then dies to alpha PVT PBing him with buckshot.
CM is not played excusively by pros. Your perfect TAS game doesnât take FF into account, which for me consist of about 50% of my deaths as a marine.
Decent xeno player is miles ahead decent marine player due to individual power. TAS marines could focus a xeno before it appears on the screen because one of them was one tile ahead and instantly relayed information to everyone around, while normal marine stays in place and starts typing, while others look at the chat and donât react to the screen action and others ignore chat focusing on action.
All TAS marines wear helmets as they have no downsides etc.
I mean for example, dancerâs balance problems have long been justified in the community as âif the dancer player is good, the caste is fine actuallyâ and weâve avoided giving dancer buffs because it would push those people who can pull it off over the cliff where it becomes a menace with fewer counterplays on the marine side.
Similarly we have the SADAR, who in the right hands can really shut the xenos down very hard. It is not the problem of the SADAR that he has shitty team mates, and the balance of the SADAR probably should not include the likelihood that he will be buckshot in the back of the head, even though marine armour already does mitigate this scenario to some extent and more ballistics specific resistance could reduce FF.
I donât think that balancing the game around a playerâs âlackâ of skill is healthy in the slightest. Friendly Fire from a xeno perspective is a blessing you cannot control and thus canât count on to win, when you pray for FF, you pray to your enemy that they enable their own loss, instead of giving the xeno player the appropriate tools they can use to ensure their own victory, you should never make a side dependent on the performance of the other side, those performance problems should more be seen as âopportunitiesâ that the player can capitalize, never their central strategy.
(I could segue here into how carrier is deeply affected by this due to marines running into traps being entirely outside the carrierâs actual control, but itâs not a topic about carrier so I wonât derail things that hard.)
It was a rethoric question, of course we do, but we should not treat everyone on the (lets face it, autistic) levels of gameplay.
High skill floor and low skill ceiling for xenos vs low skill floor and high skill ceiling for marines (potentialy). That is why TAS CM would give us nothing of use.
There is an old CM saying that it is better for something optional to be shit than something optional being overpowered. So it is for good that certain castes/strains/weapons/etc are weaker, otherwise we end up with vampire lurker, godstims and OT.
SADAR alone canât do shit. If you donât follow up on his AT rocket to T3/Queen, then he might as well not exist at all.
While balance of SADAR alone should indeed not consider the likelihood of him getting FFed hard, the overal marine balance should take that into consideration.
If not, then just add current B8 scope to every single marine gun, why not?
Wiki states that SADAR armor has the same bullet protection as regular marine armor, that is 20 + I think 5 from uniform. Wiki also states that at 25 bullet armor buckshot pellet does 44 damage, that times 4 gives us 172 dmg. 28HP to death from full HP. So unless his body falls on a tile next to a corpseman with advanced brute kit then he is dead due to oxy damage, otherwise almost dead and with a fracture. What if that PB happens on feet, or hand? Straight up death from full HP.
I wouldnât call that âmitigationâ, really.
And sure, you can add more ballistic ressistance to reduce FF, but in that case you have to take FF into balance, which TAS does not.
The only thing TAS CM could help are isolated 1vs1 maybe 1vs2 fights to determine certain rules.
Runner for example has exactly enough health to survive one buckshot PB from full health.
This is a fairly good example of what we could and would want to achieve.
But that can be achieved with raw numbers, no need for pixel-perfect scenarios.
Game has to be balanced around both marines and xenos fucking up, but because marines have bigger potential to fuck up, this should be considered more. Perfect TAS wonât take into consideration that. In TAS CM runner would be worthless, while in normal CM a big chunk of his damage is just running around marines who shoot eachother.
In TAS CM hedgehog rav would be almost worthless, while in normal CM PVTs shoot him with a shield on.
And that is why game developers DO NOT LISTEN TO COMPETETIVE PLAYERS.
Here is some vid on a topic if you want to watch:
TAS balance would go a step further. Just listening to pros is not enough, take feedback from perfect plays, surerly that translates to regular game, right?
That is just simply a wrong statement. Because majority of players lack skill. But they deserve to have fun, otherwise you will end up with 10 players server of pros who will get bored anyway.
Ever wondered how Quake Champions died?
You canât control it fully, but you can influence it. Runners and lessers excel at that. Hedgehog rav is the closest to actually controling FF.
Neither side is dependant fully on the perfomance of the other side, but partially they are and that is fine. Otherwise we end up with a game of âI win buttonsâ and the matter of winning is who clicks it faster.
As I said before, there are castes that rely on that more, but some donât at all. Lurker, or sentinel, they donât rely on FF at all, they can use it occasionaly if opportunity arises, but that is not their gameplay loop.
All traps work this way that somebody has to trip them. But you can influence this.
In short, I think that relying on perfect marine vs xeno plays is wrong and would be extremely unfun, TAS CM would give us little to no information that we couldnât gain from raw numbers anyway.
People make mistakes, marines FF and bodyblock, xenos bodyblock, SADAR misses a rocket, marines donât folllow AT rocket stunned Queen, xenos donât follow Queenâs screech.
The beauty of doing a tool assisted run is that it does not have to be for any utilitarian purposes, but it would have some, because even if it showed us which side was weaker, that is something you can address with balance PRs in the future.
⌠? Prefacing your post with references to autism like this is not sending me a good message about what the rest of any discussion is going to be like.
What are you yapping about? This is a hot take if I ever saw one.
Uhhh, they do though? It depends on whether the developerâs goals is to create a competitive environment. If youâre talking StarCraft, League of Legends or some other shooting game thatâs FOTM, of course the developers should (and typically do) listen to competitive players. Again, what are you yapping about?
I wouldnât say itâs a wrong statement but as someone who canât play dancer I donât necessarily feel like I would feel the same sense of fulfillment from finally getting good at it than I would if I had had the class dumbed down for me to perform better earlier on. This is more of a subjective take, though, from my view as a person on a team, I do want my team mates to be able to deliver what their kit is supposed to offer to use as easily as possible, and making it so players have to practice for some time to get their just gives me a shortage of competent dancers to use in that hypothetical.
As the person who is playing with a dancer, I benefit more from them being capable, yeah. As someone who plays dancer, who knows, maybe I want to feel like Iâm one of the few who can make it work.
You may not be aware of this, but you already exist in that world. As of right now, there is an unspecified top ten players on the server, and there always will be. They shift all the time for many reasons, new players come in, old players simply get old or leave, thatâs nothing new to gaming in general.
No you canât, I can only fling enemies into a trap as another caste, I can never make it so that an awake player will walk into a trap, and there are plenty of players who simply never will walk into one. Thatâs beyond my control, there is no âspecial sauceâ hiding technique you could apply here to change that, it is a mechanic dependent on the enemyâs skill issue.
Seeing a run happening in real time is not the same experience of creating a formula that can express the gameflow of CM13 because people do not intuitively see math ârunningâ in real time (at least I donât), what I mean is, when Cipher looks into the Matrix he sees red heads, I see green text flow, thatâs how a mathematical representation of a perfect CM game would be conveyed using a function like that. And it would give you the same answer, it is expressed much differently, though.
Imo TAS would absolutely be useless for getting any data because there are a truly unpredictable number of variables in a real round. You canât reliably optimise a game which revolves around 100 people beating the shit out of each other in different ways every round. No round is ever the same as another.
TAS works in single player games where everything is the same, e.g. Mario, Skyrim, etc. but is extremely limited by games such as Foxhole, Hell Let Loose, etc.
CM13 is on a completely other level in terms of player interaction, so goodluck.
if you realy want to simplify, xenos win, you all assume that in TAS you fallow the server rules, xenos could just camp LZ and even without weeds they could outheal the SUPER limited DPS that marines can put outside when they are in the DS and their DPS does not scale with pop at that position so you could have 40 xenos compleatly outhealing even 200 marines who are all cramed in one ship. If you do not agree with this, it at least limits the number of variables by a lot, but then you can go split drop jady jady jada who cares, optimal stratgy is nowhere near fun and does not tell you jackshit, balance is a lie, none is trying to make this game balanced, as you can not have balanced multiplayer without competetive play CSGO or LOL style so new players do not play with experienced ones, this is just pointless, the focus is to try and make the game fun as balance is a fucking lie and the only âbalanceâ there is, is who joins what side.
reply to below (slowmode is pain so I am just gona run conversation by updating this comment)
the sentryguns can not do much when each is imidietly surrounded by even three xenos (and this si not some outlandish clame, xenos are able to push LZ with sentryguns up) and as said the DPS that marines can output outside of the ship is super low and does not scale with pop, so xenos can choose that their perfect TAS expects 500 marienes and with that 80 xenos, 500 humans can not do any more harm there then 100 or 20, but the 80 xenos can just roatate the doors to weeds on other side of colony to heal. or fuck it 1000 umans to 200 xenos, what now, mariens are just fucked by the sheer amount of healing and them being limited to chokes where they can not scale. with that xenos can do the start on both LZ, and chutes scale slower then pop I think? did not check.list(âParachuteâ, floor(scale * 20), /obj/item/parachute, VENDOR_ITEM_REGULAR), okey this is much slower then pop gain 20 chutes per 50 marines, if you scale pop enough the amount of troops marines can put on ground goes to zero when compared to xeno numbers. if you scale the pop enough you can just choose that the ratio is not 3-4 marines per xeno, but 60 out of 150 parachute but are facing 50 xenos, so you are rather 1v1 ratio with you landing among the xenos and loosing all the advantage of range. also with such numbers you do not give a shit about lack of weeds, you do not have much use for all the t1s other then defenders and do not need that many of them, healer drone spam and gardener spam just chuging health from weeded area to lz
Also so far I did not assume any above avrage RNG and in PvP TAS the correct RNG is that evvery event happens in every 1/p instances I would assume, but tbh, I think that what I proposed works even with the worst rng possible for xenos, they can do all they do just with stuff that hits every time
Wouldnât the camping strat completely not work since not only canât you not weed LZs for 20 minutes after drop, but there are 4 turrets dropping with it, alongside up to 4 more turrets on the Alamo?
Not a reality based posted, for more reasons than one, but the easiest and quickest to identify being:
Is correct.
I think so to, as it stands right now only a few xenomorphs have the off-weed speed to prevent themselves being kited to death, and they canât keep themselves stuck on marines long enough off weeds that theyâll be able to keep marines down in a meaningful way. Focus fire here is assumed to be perfect, so xenos will melt very quickly indeed. Weâre talking in the timespan of 1-2 second TTK with optimized play.
