Posted on March 7, 2022 in Technology

Can Gameplay Balancing Ever Do It Right?

Gameplay Balancing has become a feature in many of the online competitive FPS games available today. You’ve probably encountered it if you’ve ever played OverWatch or BattleField online, for example.

The goal is clear – keep newbies from getting utterly destroyed, and don’t set up experienced players to get bored with players under their skill level. The most obvious balancing bases off of tiers, or ranks, where bronze players only go against other bronze players until they get promoted up to silver. Forms of this have always existed! Chess, for example, uses a couple of ranking systems with ‘points’ (usually the Elo system in America), and the higher that score is, the better you’re likely to play against others. A high score of 2000 is almost always going to wipe the floor with a score below 1200. Many sports do this as well. Gameplay balancing can be exclusionary, or it can let people into the game without crushing their spirit. It can be used as a weapon, or as a tool. Complaints of the systems in place for video games boil down largely to a lack of consideration for the smaller group of higher-level members and an inability to play the game as desired for any level due to wacky balancing AI.

But to do it Fairly

Let’s say we rank players with player points – the better you are, the more points you have. If you round up six average players, and they total 90 points together, that would be the same as five players who have 10 points each with a sixth player who has 40 points to himself, right? Right?? Well, actually, no. But some balancing systems behave as though that’s true, and the end result is a very frustrated 40-point player and several 10-point players who’ve been yelled at for poor performance in voice chat when they shouldn’t have even been paired with the 40-point (or put up against 20- point guys) in the first place. League of Legends, famously toxic on any chat option, often has this problem.

Or, let’s say that instead of points, we go on K/D ratio with a running average from the last five games. You, with a K/D ratio of 0.45, are put up against a new account that’s much better than you, for whatever reason. That’s okay – but it brings your K/D down to 0.25, and other players begin to accuse you of smurfing as you climb back to where you were before that truly disastrous match. For bonus points, some games weight that score on account age, based on the logic that new players shouldn’t be able to beat older players unless the older player is bad, so that one match against the new player now takes five matches with other, more experienced players to drag yourself back up.

Or, maybe, the game takes a sum total of your K/D across every match you’ve played in, combines it with rank and MVP votes in tournaments, and tries to sum all of this up into a usable, useful score with which to pit you against other players, and you’re still mad because you’re still getting stuck with two-pointers even with everything that’s supposed to be in consideration.

Measuring the Wrong Things

Automatic balancing can work – the problem is that the people in charge of designing for it are doing so with tools that aren’t properly calibrated, because they’re measuring for skill, not fun.

In the business world post Six-Sigma, there’s a growing wariness to using KPIs for everything, because what you measure, you encourage. Say you measure words typed in an email as a measure of client/vendor communication – somebody is going to cheat that system by writing five paragraphs of tough-to-read purple prose. Or, if they’re especially clever, they’ll write a bunch of text at the bottom of the email in white. The real goal, client happiness, is ignored in favor of the much easier to measure KPI, and everyone except for the top management (who thinks they’ve had a great success because the KPI is so high) is a little worse off for it.   

How you measure skill and how you measure fun are two different things. Skill, as mentioned above, can be summed up with K/D ratio, time in matches, points scored against an opponent, total damage dealt, etc. everything in a video game detects ‘skill’, because that’s all games really are. Fun, on the other hand, might require surveys, monitoring of emote usage, and other complications that aren’t already built into a system that has always measured skill.

High-Level Player Complaints

Of the complaints, the high-level players have more than new or lower ones do. Some streamers complain that they never get to have fun, because they always get put up against other high-level players and can’t win if they’re not playing at 100%, which makes the experience more like a tournament and less like the semi-casual playing that would allow them to talk to chat and play at the same time. Of course, by fighting to win, they’re not going to be going down in rank. You could go on a losing streak, but that doesn’t always solve the problem, it’s boring, and it’s an insult to a gamer’s pride – you play games to win. You shouldn’t have to lose to have fun. There’s the option of having a second account where you play more relaxed all the time, but that’s not always feasible or free to do.

The struggle, then, is finding a way to mix in more relaxed players with the sweaty ones to achieve that. You don’t want to grind down newbies by running them against high level players – if they lose badly enough, they’ll feel the game was unfair and leave. The same goes in reverse – high-level players will get bored of winning by watching newbies make beginner mistakes.

Balancing so everyone can play at 80% for most of the time is really, really difficult, because the game is thinking ‘all of these people want to climb the ranks at all times’, and it doesn’t know how to split the players who want to chill out and the ones who want to climb. You could ask the player ‘how do you want to play today’ when they sign on… executing that idea is tough, but not tougher than mind-reading or trying to guess how the player wants to experience the game, within reason.

Others complain that autobalancing expects them to keep a whole team of AFKs (away-from-keyboard players) or newbies afloat, which is also more stressful than fun – Overwatch players will regale you with stories about useless healers or defense classes, both of which are usually necessary against even a mediocre, but engaged, enemy team.

In Summary

The autobalancing works okay, but only for the mid and low players, which just by bell curve rules make up the majority of any player base. The top player’s complaints are valid, and trying to improve this system should come with some consideration of fun. Otherwise, what’s the point of the game? Speedrunners and no-hitters are trying to prove skill – everyone else is just trying to have a good time, sweaty or not, and sometimes the auto-balancing gets in the way.