Casual vs Pro Game Balance


#1

So in an effort to keep the other thread clean, I was curious about asymmetric game balance in general. @niaccurshi brought up a series of good points in that thread which I linked above, with that being said, I think this would be a good discussion to be had.

Long story short, it’s a discussion about the game balance in the context of Casual vs Pro. How would you do it? You can use any game you’ve played as an example, but let’s try to keep this in the context of Asymmetric gameplay (as Evolve is generally the only game we’ve commonly played amongst all of us).


#2

How I would do it, would depend a lot on the “goal” of the game.

In the other thread I brought up Dead by day light as an example. Its asymmetrical as well- but it has some key differences.

Yes its survivors vs a killer.

But the survivors dont win or lose as a group. The killer isnt worried about the survivors, killing him.

This changes a lot of things. As a survivor, your buddies can suck- and you can still escape and win. As the killer, short of an extreme match, its less about IF youll kill the survivors, more about “how many” youll kill.

Point is- Theres always something to “be happy about” at the end of the match.

Evolve was far more 1 dimensional in its end-goal.

Either the hunters killed the monster,

Or the monster killed the hunters (Or blew up the relay, but the result is the same- you “beat all the hunters” at once).

For evolve I wouldve personally gone with a dynamic handicap system.

Evolve had “permanent progression” for both sides, albeit asymmetrical (as is evolves way) in nature. “Downs” against the hunters- Or permanent health loss for the monster.

Id monitor the “permanent progression”, whatver it may be for your game- And use that to dictate a sliding handicap system, depending on which side is “further ahead”.

The trick would be scaling it back as the game got close, with a window of tolerance- So at the end of the match the match APPEARED close, but you still allowed the side that SHOULD have won, to win.

The goal is simply to mask how much they wouldve beaten the other side by.

Evolve, due to its complexities, often had one sided rofl stomps- And i think these were above and beyond the most frustrating thing at the end of the day for people. Its hard NOT to feel this or that “IS SO OP”, even if its not, when a better player stomped you into the dirt.

So force the games to APPEAR close.

I would however limit this to the “Casual” mode of the game- Turning it off in competitive, and making it an option for custom games.

Id also “Dress” the mechanic up with in game lore, to help with suspension of belief and what not.


#3

I don’t think I’ve seen a game try something like this before. There are certain “soft” mechanics that lend themselves to dynamic balancing, like TF2/Overwatch payload gameplay bringing the fight closer to the defenders’ spawn point the more ground they lose.

But to do this in an overt way is extremely risky, in a quiet way. Apart from strange and contrived ripples in gameplay caused by players trying to obtain/deny the bonuses, my main concern is the feeling of being punished for doing well. (And, conversely, having your losses aggravated/your comebacks cheapened by the given handicap.) Even if the adjustment is minor or negligible, the negative feedback will feel real and have a psychological effect on how people play.

Evolve has inadvertently hit many of these “feels-bad buttons” as part of its core design: it feels bad losing to one guy as a team, it feels bad losing to a team as one guy, it feels bad being the focus target pinata, it feels bad getting CC-ed 24/7. I don’t think it can afford hitting more buttons.

Yet, this is exactly what team-based asymmetrical games need. So I would applaud any attempt to try this, so long as it makes sure the psychological reversal is taken care of.

So there are two approaches to this: hide the effects and tell no one, which is effective but morally unsound, or openly show the handicap, which means the dynamic balancing must be very, very well designed.

To that end, I would suggest explicit, in-game rebalancing, like #4 in niaccurshi’s post to be ruled out entirely. (Implicit is always good, so long as you can fit it naturally in the game mechanics.) So long as you tell the players, X did too well so Y gets a penalty shot, you can’t avoid the punished-for-winning feeling.

On the other hand, I’d support a mix of #1 and #2, as something shown openly to all players at the start and then never changed through the match. I think it is feasible and useful to account for the following factors:

  • Party or solo status: each individual gets a bonus depending on the size of party he is in, so a 2+1+1 scenario will see small bonuses for each in the pair and larger ones for each solo.
  • Each person’s MMR compared to the match average: lower players get a mistake-mitigating bonus like DR or JPR, and higher players get a performance-capping penalty like lower ability damage.
  • Each chosen character, depending on whether match MMR is high or low: used to correct pubstomper or pros-only characters, for example Gorgon in low leagues can have lower Spider health while high-level Wraiths can have slightly increased burst damage.

All these effects should be summed together (no component contributions shown) and presented to the players at the loading screen. This is so that people don’t nitpick but are still given full awareness of all changes.

Miscellaneous remarks:

  • This system (maybe minus the character-specific part) can be extended to ranked mode, as many games do personal rank adjustment based on which team is favored, this can be a replacement for that system by affecting power instead of rank points.
  • I didn’t include character counters in the system: for one, it’s pretty subjective in that counterpicks don’t work to the same extent for everyone, depending on playstyle or proficiency, and secondly I believe counterpicking is better balanced by means of a drafting system.

#4

Oh, after reading Sideways’ post on the other thread, I remember For Honor has a “revenge mode” mechanic, where if you are outnumbered or outmatched, blocking hits and taking damage fills your bar that activates a super buff that lets you even the odds.

I don’t know how it actually pans out, but apparently the intended gameplay is to smooth out outnumbered situations and make them less one-sided: the in-game tips advise the disadvantaged player to just focus on blocking knowing “his turn” is coming.

May be a useful reference.


#5

In the context of Evolve and the game I’m making, this can be rolled into objectives only that side can complete. E.X. The monster takes a nasty engagement, they can recover some health by traveling to a certain location and killing a creature that lets them recuperate some of their health in the process, or the Hunters have to activate a med-bay of some sort to remove a single strike from one of their own; purely an example. Another example is that the monster’s getting endlessly harried by the Hunters, they complete a quick objective that gives them a speed boost or stamina regeneration boost, etc.

This comes to the handicap that is integrated into the gameplay rather than subtly working in the background. The good about this one is that it can be used in both League & Casual.

On the other hand the dynamic balancing can be used in Casual play, but not pro. It’s not necessarily a bad idea, but it’ll take a hell of a lot of code to make it work. You’re talking about monitoring a player’s damage output with a certain character, with a number of abilities, and perks, etc. That’s a lot of data to track.


#6

I was more or less talking about monitoring total strikes, versus monsters health, then giving an increasing buff to the side that’s “behind” (with an eventual cap on it).


#7

So this is a good point to elaborate on. TF2 had a very strict balance system in this regard. Yes when the attackers got further in to the final objective, they were also close to the defender spawn, but the reality here was that the respawn timers were also altered each change of section to re-balance these fights.

The problem was that these respawn timers were based entirely on the sides being equal. If the attacking side was actually just slightly poorer than the defending side then the respawn/travel time for the attacking team was too long (mad complicated by engineers, no doubt). If the defending team were slightly poorer then the defender respawn was too long. The game could have benefited to some degree by recognising the overall power levels of the each team, the momentum of the cart, and giving some slightly different values to respawn times.

Not so much that it actually changed the game result, but enough that the game felt like it was more of a fair fight, but also enough that an entrenched team that is riding some particularly strong meta can’t boost themselves over an otherwise competent opponent either.

TF2 could have benefited from factors in game changing dynamically to the realities between the two teams, and payload isn’t even a truly asymmetrical mode!

As I said in my first post, there should be some fairly easy comparisons to make at regular intervals between teams, and between expected performance (however that is best defined) compared against time to achieve. It doesn’t have to get too specific, just recognise red flags that can be raised for the hunters falling a bit too far behind the monster and vice versa.


#8

I don’t think having things change in game without the players understanding that something is different is a good idea, mainly because I think it leads to confusion as to why one game felt different to the last. If you’re going to change the balance during a game, which I genuinely think is the best route to go down, then it has to be explicit that it’s happening to keep the trust of those playing, to show the people who are doing bad that they have something that they can use to get back into having a fighting chance, and to show the people that are doing good that they can’t expect the next fight to be the same as the last and to prepare accordingly.

Edit: To be clear though, I don’t mean this to mean literal message saying “You’re bad, here are the buffs you’re getting”, but to be something that is somehow conveyed in game to alert players to a change of normal conditions.


#9

Perhaps you don’t mean any harm, but I think we must give players enough credit that they can perceive this, that they are getting buffs for having their asses handed to them, or nerfs for doing well.

Also, people tend to take things the worst way possible and get triggered by it, and losing due to these buffs will only double the salt. On top of that, a community can fan each other’s flames of hating the feature, which means a possible hate campaign on your forums.

All in all this is why I’d rather stay entirely clear of mid-match balancing, though if anyone wants to try it I will watch the results with great interest.


#10

A bunch of loud mouths can create a storm in a teacup, it’s true. But nothing destroys a player base more than actually negative play experiences. You will never make a game that is free from people that think you’re doing it wrong, the important thing is to make a game where enough people have fun to stick with it and keep coming back for more.


#11

I’d think a negative player experience involves being punished for doing well or rewarded for not doing well. It would feel to me like a nanny-controlled play, where no one is allowed to pull ahead without being slapped back in line. More importantly, it would feel like actions taken matter less, because if you gain an advantage it is partial canceled, if you make a mistake and fall behind it is partially canceled. “Nothing I do really matters” is a leading criticism in another game I’m playing, that leads to it feeling dull, unresponsive and same-y. If you dynamic balance during a match, it will be in the same vein.


#12

We shall have to agree to disagree. No-one is saying that people that want a “pure” experience shouldn’t be able to do so, but I fundamentally disagree that casual players are put off of a game more by the idea of dynamic balancing than by being steamrolled game after game.

I also think you’re being a little too strict in your interpretation of what I’m saying, because at no point am I suggesting that one team that is clearly better than another team should ever be in a position where they don’t win, only that the team that is losing should feel like they’re having fun while they do lose, and part of that is feeling like they are still in with a chance (even if there isn’t an actual chance).

No-one is ever going to advocate for a system that ensures that everyone is always 50/50 balanced at every point in a match, because that would be ridiculous :wink:


#13

Perhaps I am being too much of a purist in this, and someone might do it “right” and get great results. But this is still something I’m against on principle, though I won’t stop anyone from trying save for having a vested interest, since I’d love to see how it turns out.

Ultimately, what I am giving is input from one kind of player, and it is the developer’s job to consider other kinds of players - a game should certainly not be designed around me alone unless it wishes to be super-niche. But I do hope my opinions are considered:

  • Doing well or poorly should make a significant difference, and it’s effects should not be diminished except for the most egregious cases.
  • A player must never feel like his accomplishments or failures have been reduced by a non-player. (Perception is important too, any visible counterbalance will immediately ruin the experience.)

That’s about all I have to say, anything else would probably be me repeating something from somewhere.


#14

Ok, just to try and give some context here to what I believe the problem is and exactly why the original comment that spawned this was about having a casual mode that uses this kind of technique, and a competitive mode that does not.

The reason for this is the reason why playing football in the park with your mates a game, but playing for a formal team, in a formal league, is a sport.

Some people want to play a game, and this means enjoying the act of playing, usually by making sure it remains competitive. Players will swap sides of one side is clearly too good, players will change positions to handicap themselves. They do this because the primary reason to play games with your friends is the collective enjoyment of everyone involved. If someone isn’t enjoying it, the mood is lowered, they may not come and play again. If many people don’t enjoy it then you can guarantee that your friendly kick about isn’t going to happen much longer. It’s mutually beneficial, as a social unit, to “dynamically balance” even if that means good individuals are playing down.

This also happened in evolve, of course, monsters that were part of the community didn’t do things that people hated, unless the hunters did things that monster players hated!

A sport is different, a sport is about competing and getting enjoyment from being part of that competitive atmosphere. People who play sports dedicate their time and their lives to it, because that’s the level they are engaging with the game that they play for sport.

Sport is ranked mode, sport is tournaments.

Similarly, if we look at the nearest proxy to video games, board games, we’ll see that “catch up mechanics” aren’t uncommon these days, precisely because there is an audience of game players that are not interested in spending their time playing a game that they will sit around for significant portions knowing that they’ve lost and they are just making up the numbers.

Ultimately everyone knows games are generally with their winners and losers, but it is the journey that matters to some… Are they just making up numbers, or is the game there for them to enjoy even if they aren’t the ones that have come out victorious?

I don’t discount players that have your view, but I also think players that have your view are the ones currently catered for by just about every online multiplayer game, and it’s time we explored a true casual mouse that respects the act of game play rather than the result of it. :slight_smile:

Thanks for your views though, they aren’t unappreciated, I just wanted to try to explain better where I’ll coming from and why I don’t think you have anything to fear from this kind of experiment should someone take it up!


#15

Balance for Casual, fine tune for Competitive. If the foundation isn’t there, you can’t build a house.


#16

Hah, you don’t need to worry about my feelings. I can tell you’re actually listening even though you don’t agree, which is a pretty rare sight.

I like this analogy, but I have to point out its flaws. The most prominent one is that the “prioritise the collective enjoyment” social contract only exists between people who know each other (across both teams). In a pub game, this isn’t happening at all: most players are in it to enjoy the game for themselves. To put it in your analogy, pub players are signing up for “sport” with faceless enemies, not social games with friends.

Additionally, since this type of game is very intense and competitive, and most players’ goal is winning, most people would unbalance it further in their favour, if they can. Just look at various games which allow you to switch teams in casual games, most pub players that switch are switching to the winning team.

As the developer, of course your goal is the opposite, to make things even. But to do this in a visible way in response to events where one side takes the lead, would antagonise your players who, as above, want the opposite. They would feel it is unfair to decimate their achievements with rebalancing. (And certain more puritan players would be unhappy about their mistakes being compensated as well.)

That is why I suggested rebalancing only at the beginning of the match. Firstly, players generally expect a good matchmaking system which grants a 50-50 win chance each match. This is different from a 50% average winrate which can be achieved by giving equal rates of 100%-win games and 100%-loss games, obviously not good quality matches. Therefore, players are likely to accept MMR-based minor corrections, partially because it stokes the ego of the “higher-skill” players taking handicaps. It won’t diminish their enjoyment of the match because the events of the match still matter with 100% unreduced weight to the outcome of the match.


#17

imo you should balance for the pros… but you need to make sure that one sided stomps dont happen frequently.

a good rank mode, smurf protection, punishments for ppl who derank… dont let ppl play vs premades all the time on lower elo.


#18

So one of the reasons that some form of dynamic or multi-tier balancing is a good idea is that the theory of what you’re saying, @teaching_stones is spot on, the reality on the other hand is quite different.

It’s worth remembering that a lot of the problems that people coming in to this game had, especially at Stage 2, was because of a discrepancy of experience vs hunters/monsters that were seasoned players from Legacy and (and this is the important bit) a lack of diversity and size in the overall player pool.

It’s 100% fine, in theory, to say that ranking algorithms with some protections or tweaks for premades would be enough. In practice the maths meant that for the largest part of either version of Evolve people did not have the choice about who they would end up being partnered with, or play against.

In this scenario (especially a problem when looking at Asia/Australia, or South America) you would have a tiny preferable pool of players, and that was before sorting for rank! So… having a good way to balance your game in the event of the two sides not being similar in skill (or for whatever reason being better in a game than the other side), it’s generally just good sense to stop a death spiral from happening in your localised communities within the game’s wider community.


#19

I think in any sane world you rely on your ranking algorithm to do that hard work here, with perhaps some modifications to smooth the curve when it comes to the difference facing premades or all solo queue randoms. If your ranking algorithm is working correctly then the win chance should be 50/50 at the start of the match.

However the chance of winning is a different situation from the enjoyment level of a match. I think everyone agreed that when two sides of similar skill faced each other in Legacy that Wraith was probably underpowered. However many people also agreed that facing Wraith was an experience they didn’t enjoy even if they did win. Now, that’s a little off topic because that’s a game design conundrum rather than a balance one, and one that TRS continually worked to fix from their initial concepts.

It is worth noting though that Wraith underwent these changes because of the enjoyment people had (or lack of enjoyment), not because Wraith was actually an overpowered monster, and it’s worth noting how many people (even though they said it was because Wraith was OP) badmouthed the game and must have also quit the game because of their experience with Wraith.

I feel that the common, casual, player is somewhere in between my analogy of friends at a kick about, and your feeling of someone that is joining up to play a competitive game and feel that they are in it for themselves. Being selfish is a common trait of your average gamer, when it comes to their own enjoyment, that can’t be denied. The question is whether by doing something that improves the enjoyment of all involved, you really then lose that feeling of the game being a competitive one.

If you can’t do dynamic balancing (which maybe I should stop using as a term, because at no point are we talking about making sure that there is a 50/50 chance of winning at any given point in the match, just making sure that the win chance for the dominant side never gets above, say, 90%) and keep the game feeling like it is one where you’re competitive… one where, like I’m sure many of us felt with evolve, you have the adrenaline surging and you feel like you’re deeply involved in this experience, then you shouldn’t be adding that kind of balancing/handicapping feature in, I agree.

But at the end of the day if the games are meant to be balanced (due to ranking) going in, then the games should feel at no point like they are lost causes either, because in doing so you make people question the balance of the game, and the effectiveness of the ranking system (which we saw both of during Evolve’s life). And as I’ve said above, when you can’t guarantee a diverse enough, localised, pool of players to make ranking and balance work from the point of entering a game, then actually responding to what the experience is for the players during a game feels like a very neat solution to a couple of other problems games face too.

… and I’m not saying you can’t take one of the options I described that does significant algorithmic calculations before a match to change balance for that match, I just think it’s an opaque technique (not to mention complicated and hard to test) and would cause confusion to a player base that do tend to like to know “my gun does X damage, I can rely on that”


#20

This is one of the reasons why I suggest individual corrections based on individual MMR, even though an ideal matchmaker should theoretically make things even… But let me digress a little, to case-study Heroes of the Storm.

Heroes has a dearth of matchmaking issues, despite being a 5v5 fully symmetrical game and thus supposedly having an easier time of it. This is in part due to low population, especially outside the US and EU, which is a problem very pertinent to us as well. Ignoring the team composition-related ones, the ones I want to discuss are:

  • Rainbow lobbies
  • Party versus solos

Rainbow lobbies refers to games where one team (usually just one) in a match will have wildly varying skill (bronze+silver+gold+plat+diamond, hence “rainbow”), as a result of trying to balance the average MMR of each team, so the middle band goes to one team and both poles go to the other, creating a team with high deviation and one with low deviation. Of course, this is only balanced on paper and the normal team will curbstomp the rainbow team, as the rainbow team has wildly different expectations of meta and strategy. The weaker players will be a total liability against a team of betters, and the best players will be unable to dent a cohesive team alone.

For Evolve and similar games, this problem is even more unavoidable, not only because the monster “team” has zero deviation by default (thus making equal deviation impossible), but especially with the “prefers X role” system where you can’t predict who is on which team until the match starts. This means we even have a amplification to the problem: there is no reliable way to ensure the monster player is the median-MMR player in the match. In fact, this is very unlikely to happen, so most of the time team strength will be both average-unbalanced and deviation-unbalanced. Therefore, somehow correcting each player’s MMR to the match average might be a good way to help make sure teams are both balanced and have low deviation, and by extension, matches are more often close.

As for party versus solo (quickmatch), Heroes also does this pretty often in low-population regions and sometimes even high-population ones, because there just aren’t enough parties. Obviously, this is a massive advantage for the party in a teamwork-based game. (Heroes has an additional aggravating factor where solo players can’t see their team before picking, whereas parties can, so you get nonsensical team comps against a well-planned one. Thankfully this is none of our business.)

To offset this, a party’s MMR is rated with a bonus during matchmaking so they face higher-ranked opponents. However, this system that seems good on the surface merely creates more problems. Firstly, it cannot account for how coordinated the party is, from a bunch of randos or friends casually playing together to a hyper-coordinated scrum group on voice chat with a coach looking over their shoulders. As a result, the uncoordinated parties are rated higher than they should be and the coordinated ones, due to the uncoordinated ones pulling down party-vs-solo winrates and thus the MMR bonus, have the same easy wins as before.

I’ve got no solutions to this one, it’s a problem we are likely going to have as well.

The second issue is, it’s not always feasible to find appropriate people to match with after correcting the party’s MMR. Since it is a teamwork game and the MMR value of teamwork is very high, the entire upper half of all parties will have no solos to match with after the MMR bonus: the required MMR will be outside achievable range. So, after an absurb amount of time, they are just thrown into “whatever goes” matches against 5 low-MMR sacrifices, and essentially get a walkover. On the other side, low-MMR solos matched against parties will also be stomped because there are no parties with a MMR much-lower-than-low.

For this reason, stat corrections rather than MMR corrections might be better to avoid this “unmatchable after bonus” problem. It also comes with the added benefit of being able to pick and choose which stats, in accordance with which areas weaker players need help in most. (In Evolve, probably survivability.)

For the monster specific one, I don’t mean use it to balance OP/UP monsters, I mean to balance monsters (actually, hunters too) that have disproportionate performance in different levels of play. For example, we can probably agree that Wraith had a high performance in “noob” games and didn’t do as well in “pro” games, which means that we can give minor nerfs if the match has a low average MMR and minor buffs if the MMR is high.

Heroes actually does a fairly good job of this (on maps that aren’t BHB, Alterac or Temple), but it is built into the game mechanics and not something the game does of its own accord. More specifically, you’re never truly out of the running due to the hidden confluence of many factors like increased kill XP against higher-level enemies, longer late-game death timers, and heroes outscaling buildings and creeps with time. It used to be that way (active correction), where the losing team got a large stream of passive EXP, and that had a rather negative community impression, as they felt that leads meant nothing as they self-destructed with time, and the losing time didn’t need to do anything at all to close the lead.

So if you do want to keep losing teams in the running (which is great), you absolutely need to make sure the game is not producing active correction mid-match (which is terrible), but instead has mechanics what allow a losing team to rally and comeback if they do win an event or two.

A little off-topic but does anyone actually go “OK I can kill the monster in 132812 shots of my AR and 43.9 seconds of Lightning Gun”?