Friday Game Design Open Forum: Solve Population Imbalances

How do you prevent, mitigate, balance, and/or resolve population imbalances in games with persistent team-based PvP?

Examples of what this is not: FPS or MOBA games where one team is way better than the other; temporary imbalances caused by uneven numbers of players in small proportions; PvE class imbalance. Let’s take some examples of population imbalance. Please fix one or all of them in the comments.

Assume that factions are balanced or symmetrical (or else we have a different problem) and that most players will not intentionally self-select to a losing team instead of the winning team (empirical). The problem we are trying to solve here is the spread of the population, which is mostly a matter of quantity but may involve quality, although there is no a priori reason why the pros would join one faction and the scrubs another.

Assume also that population imbalance is a problem, particularly for the survival of the game. Sheep leave if they are just being fed to wolves, wolves leave if there are no sheep.

In Dark Age of Camelot, there are three factions on each server and multiple servers. You cannot play for multiple factions on the same server, but you can across servers. The population imbalance problem comes when one faction on a server pulls ahead. Players are more likely to leave a losing faction/server and join a winning faction/server, so victory tends to snowball over time. It is not necessarily the same faction winning on each server, so Server A might be Albion-dominated while Server B is Hibernia-dominated. Mitigation measures already in place: “Darkness Falls” is valuable PvE content open only to the winning team, which both then draws PvE players into PvP for the losing factions while draining them from the winning faction; “relics” are useful buffs that factions can capture from each other, but your faction must own its own buff to capture others’, which reduces the incentive to pick on the weakest faction because #2 cannot afford to fight both #1 and #3 (although #1 and #2 may trade #3’s relic back and forth, they are at least doing so with war in their territories, not stomping back and forth across Poland in WW2).

In Guild Wars 2, each entire server is on the same team, but servers have vastly varying populations. Mitigation measure already in place: servers are grouped to fight against similarly sized servers. As discussed elsewhere, this rarely leads to even three-way matchups because the servers would need to be in neat triples of similar population sizes (AAA, BBB, CCC) or else one server will have an easy victory or certain loss or both (AAB, BCC, CDE). Players can switch servers, the effects of which are significantly mitigated by the changing matchups based on server wins and losses; rising in the ranks means fighting similarly full servers, and changing the order or size distribution of servers only has a big impact on the overall picture if it evens out the tiers. Mitigation measure already in place: there is a cap on how many players can be on a particular WvW map, so there is a limit on how much force you can bring to bear even with a huge server.

In Shadowbane, PvP was a free for all with one server and no predefined factions. Players defined their own factions. As one group of players became the bully on the block, players either joined or left. As the sheep population falls, repeat with the smaller wolves and so on. Until one faction wins and servers shut down. Whatever mitigation measures that were tried failed, although EVE Online does very well with a similar structure. One EVE mitigation measure is the huge size of 0.0 space: no matter how big a bully you are, you cannot take and hold half the galaxy unless a similar proportion of the playerbase is in your faction.

In Ingress, there are two factions and one server. That’s about all I have for problem specification there, beyond the initial assumptions. I know of no mitigation measures, and population flows about as you would expect when one faction has more players. Exacerbation measure: players are frequently encouraged to invite their friends, and the game was invitation-only for a while, so the larger faction both had more players and had more ability to attract new players.

: Zubon

10 thoughts on “Friday Game Design Open Forum: Solve Population Imbalances”

  1. The obvious answer is to reward the “weaker” faction. Players on the losing side maybe can gain a buff (that stacks per loss) on the following rounds (provided they do not transfer about). An example would be +5% total damage output.

    Eventually even a sheep who is constantly curb stomped on the losing team might have enough power to one shot the elites of the other side with their weakest ability for a rare victory. The winning team loses the buff but gains access to whatever – maybe a dungeon with a chance of getting good gear until the next round is complete.

    This means that eventually each side -should- win as there will be more and more powerhouses wandering about with each loss giving the sheep a reason to stick around since no one side can dominate forever, while the elite dudes who win all the time will be challenged to keep doing so against foes who continuously grow stronger.

    Obviously this will only work in games with predefined “rounds” and set rules like GW2. No hope of getting any sort of balance in Open PvP, yet another reason it is simply poor design.

    Edit: While re-reading this it occurs to me some might question “rewarding” the losing side. You need to provide incentive for people to stay. If the stronger guys are rewarded by getting stronger while the guys they defeat gain nothing, you will ultimately lose the interest of both parties.

    You could go the more sadistic way too where you punish the winners but I don’t think that’s a better alternative. An example of this might be: Players on the previously winning faction (regardless of which side they fight on now) can get all their gear looted and/or be perma killed. Players previously on the losing side don’t have that disadvantage.

  2. I think the winning team gets rewarded by winning, mostly. You definitely need mitigating feedback to counteract the reinforcing feedback that comes from people shifting faction, but I recall that this doesn’t help if the teams have a distinct population imbalance. Wintergrasp in WoW had a hidden variable where the more times a faction won, the easier it was for the other side, but often this meant that the dominant side would occasionally lose and then take back the fortress the next time it became available.

  3. I’d remove stress on an obsessive focus on overall faction victory and/or slaughtering the opposing team with superior numbers to be the only victory conditions worth mentioning or having fun with.

    Create lots of personal achievements for demonstrating increasing mastery of various faction or gameplay roles at a variety of skill levels from newbie beginner to absolute professional expert. This gives players mini-wins and feelings of overall progression even if their faction is facing an uphill battle. This should be account-based.

    Make it attractive to create and play different characters for different roles and perhaps even different factions, so that altholism can run amok and players can feel a sense of variety if they choose, a la playing different heroes in MOBA games, enhancing game longevity. People might scream about spying, but it’s already happening even in heavily restricted faction games, and anyhow, if we make the overall win less important, and more about having fun with comrades, it may end up more tedious to go through the trouble of spying.

    Perhaps take a page from Path of Exile’s races (something I’m currently obsessed with, even at a completely non-competitive level) with some points given for participation at a reasonable level (eg. points are awarded if you can stay alive and get to various levels, like 5 and up), some points given for taking routes/roles that preclude others (if you sidetrek to clear a side area or a quest mob, you’re likely not going to win the most experience/levels race, but you get points for doing that too) and points for playing competitively/winning too.

    Random offhand example: Points for getting a number range of kills in various regions of a map (eg. rewarding offense or defense), points for holding control points for an increasing amount of time, points for gaining a range of pvp ranks/levels, points for achieving optional objectives within the match time, whatever.

    Have a range of terrain/siege (or whatever equivalent) that allows for open field fighting in large numbers, small group fights in more confined/chokepointed or maze-y areas, and ways for a smaller outnumbered group to score ‘wins’ in a local region (sniping/ambush/trap spots) and/or reasons for a big group to split up on occasion (simultaneous objective capture and holds, for example.)

    It’s a bit of a sidetrek, but I stumbled on this page while googling for info on how people can play and enjoy sports (basketball, hockey, whatever), even while they are at a non-super-competitive top-athlete level and will probably never get to such a refined level in their lifetime.

    I found the exhortation that kids really just want ‘ice time’ and to be able to play rather than sit on the bench illuminating. I think adults would really just want to be able to play their pvp game too, rather than log in to see an outnumbered buff and face unbeatable odds with zero allies (which then encourages them to log off rather than play – or rather, switch sides so they can play, and win!)

    So the idea would be to try and always create opportunities for play and having fun, even if/when your overall faction is losing. Of course, building in chances to come back from behind would also be good for morale, rather than force people to endure a slow inevitable defeat.

  4. The really ‘big’ pvp games all do one thing in common – have very little gear disparity – ignore the MMO for a bit and look at real pvp games that cater to large crowds for years – Call of Duty – Battlefield – Starcraft – heck even chess. The issue with MMO pvp is that the entire system inherently is based on the ‘winners’ getting even stronger through gear – in some games (like WoW) that gear difference can be greater than 100% (that is a new guy that just hit max level with no special help can be 100% less powerful than someone who has been gearing up for PvP for a while) – there is no incentive to keep loosing when the entire game punishes you for not only loosing – but also getting weaker (and thus harder to win) the entire time. The looser doesn’t feel that there was even a chance in the ‘100%’ scenario – and because they lost any gains they make are small and marginal vs the guy who one keeping the gap throughout the life of the ‘season’ – and none of that involves skill – it’s just numbers.

    Looking at the games that can spawn massive pvp for years only using a few maps (again say call of duty) you find that skill is 90% of the formula – and because everyone is even in terms of gear the guy that lost still gains actual experience in how to respond and get better – to the point that after a week or two he’s not being curbstomped and even if he looses he feels like he put up a fight and had a chance.

    MMO’s never give the looser that feeling of ‘had a chance’ except at the very start when everyone is at (or near) gear par – or at the very end when new gear hasn’t been released for a year (or more) and everyone is (again) at or near gear par. The great dessert in between is just horrible for the loosers – as they face the online version of Sisyphus.

  5. Probably the simplest answer is to give the winners (generally the side with the higher population) an incentive to go do something else. And, PvPers being what they are, that “something else” had better also be some form of PvP or mad whining will ensue. So … success in a battleground gives bonuses for minigames, success in minigames gives bonuses for some hybrid PvP/PvE zone, and success in the PvP/PvE zone gives bonuses in batlegrounds. As each bonus is achieved, some of the winners will move on to the other activity, providing the weaker team(s) the opportunity to experience some success before the perpetual winners come back to stomp them again.

    Incidentally, a game should only rarely give PvP rewards that improve PvE, because that annoys everyone involved: the PvPers hate the rewards because they’re not PvP-related; the PvEers hate the rewards because they feel compelled to engage in an activity they quite often despise in order to acquire BiS gear for their preferred gameplay. Conversely, you also can’t reward success in PvP with gear that makes players significantly better at PvP (or, more precisely, gear that makes players significantly more lethal at PvP), because that leads to increasingly higher barriers to entry for new PvPers. Best compromise is probably to give rewards that give successful PvPers more flexibility (think CC breakers, resistance to CC, hybrid abilities with significant situational utility) rather than just raw numbers and killing potential.

  6. DAoC had two other mitigation factors, methinks.

    Firstly – there was no long-term score or rewards to worry about. One faction might have more keeps, but that didn’t translate into any sort of “points per tick” or end of season rewards. To a great many players, it was irrelevant who had keeps and relics – even if your faction had none of them, there were still fights going on in Emain that you could get stuck into.

    Secondly – your choice of faction and server was “sticky”. There were no server transfers, paid or otherwise, and levelling up a character was a significant effort, especially in the early days. It was by no means easy to jump on a winning bandwagon, which meant people were more inclined to stay and fight for the team they were on. While I personally left Hibernia on one server to play as an Alb on another server after a year, that was a significant wrench and if it hadn’t been driven by guild drama I doubt it would have happened.

    Unfortunately, in the current MMO field, I don’t think that a game that says “come pick a server and class – and then you’re stuck with it” will fare very well. Likewise, very few devs seem to have the courage to claim that their PvP is so much fun that people will play it without any rewards for doing so.

    1. PvP is almost never enough fun for the sheep to get involved without rewards. The wolves might do it without rewards, but they generally give up when they run out of sheep. So, even if you want to keep the wolves happy, you still need to provide rewards, Even EVE operates on a similar model; the difference is that many (most?) of the “sheep” are actually alt accounts of wolves, optimized for economic activities (which are necessary in order to support the wolves’ preferred activities). I’m pretty sure true sheep don’t stick with EVE for very long.

      As for Dark Age of Camp-a-lot, the Holy Grail of PvP MMOs, even there the vast majority of PvP activity consisted of 1. zergs, 2. ganking, and 3. what I call “avoidance PvP,” i.e. trying to hit undefended keeps and then run away when actual players showed up to fight.

      And no, I don’t think DAoC’s model would do very well today. Modern audiences have a lot less tolerance for the sort of leveling tedium that game entailed.

      1. The question is – what’s wrong with zergs, anyway? :) While it’s always been fashionable for those who just want to hang with their tight group of leet buddies to sneer at anything on a larger scale, I’ve always thought that idea of player vs player in a massively multiplayer game really ought to mean mass battles.

        1. Two reasons, really. A. zergs are a massive performance hit on both the server and client levels, and B. zergs reduce the skill required for PvP to almost zero.

          Of course, let’s be honest here, most PvPers really have no interest in a “fair fight” or a “challenge;” most PvPers just want to pwn someone by any means possible and then mock them afterwards. But even for that purpose, zergs aren’t that great; you generally have no idea who you just pwned and “ha ha, we brought more people” is a pretty lame putdown.

  7. I’m not english speaker, so maybe I did not fully understand the deeps of the question (and forgive me any grammar errors); anyhow I want to give my 2c.
    Problem: player base partitioned by servers => imbalances between servers’ population
    Possible solution: add one or more partitioning dimensions coarser than the servers (eg: server divide player base by 24, new dimension should partition by less than 24) and give them meaning and purpose that fits in current game system

    quick hypothesis: leverage the colours system and let player guest other matchups on the side of theirs colour, give colour some point/goal/reward system.

Comments are closed.