Greg's Design Blog

A collection of posts by game designer Gregory Carslaw, including mirrors of all of his blogs maintained for particular projects. A complete index of posts can be found here:
 Thumb up

Feeling Balanced vs Being Balanced

United Kingdom
flag msg tools
Original Post

In my articles on how people experience probability and cumulative probability I've highlighted different ways in which a gamer's subjective experience of the probability of various events differs from their actual probability. This can be used by designers in various ways, but it creates an odd problem. If the probability of an event is one thing and the perceived probability is another then the game is likely to either be balanced or feel balanced.

If the players beliefs about various probabilities are that they lead to a balanced game and these beliefs are false then they game is probably unbalanced. Conversely, if the game is balanced then the players will believe that it is not. An ideal game would be able to overcome this effect, for instance each player would have equal access to and reliance on an option that intuitively seemed better than it was so both perceived balance and actual balance are achieved. However a great many players enjoy asymmetric games in which players have access to different options over the course of the game, so the options with unintuitive options will be available to different players to different degrees. As soon as this happens, this sort of seamless balance gets much harder to achieve.

Once a game exceeds a certain level of complexity designers move away from balancing them with mathematical models and move towards balancing them with playtesting. Here things get interesting, as the differences in human experience will cause the different types of data obtained from the playtest to conflict. Observing how often players who choose a particular character will win the game will produce different results to measuring how many playtesters express a feeling that this character is unbalanced. In an ideal world these problems would always be resolved by bringing perceived balance and actual balance into line, but game design is very often about compromise and sometimes it will be necessary to choose between having actual balance or perceived balance.

To understand this decision, it's important to go back to why balance is desirable in the first place. As with most things it comes down to how people feel about playing the game, it's good for players to feel that they have a lot of interesting choices (so those choices should be balanced or they are not choices at all) and it's nice for players to feel that the outcome of the game is a consequence of the things they did (so it shouldn't be down to them or their opponent having some unbalanced power.) Ultimately the players' opinions of the game will reverse with repeat plays, perceived probabilities become closer to actual probabilities with repeated exposure. If the game started off seeming balanced it will be revealed that it is unbalanced, if the game started off seeming unbalanced it will be revealed that it had been more carefully balanced all along.

Every player who plays your game will play it for the first time, but not every player who plays your game will play it for the second time. So there's merit to developing a game to show off its best qualities during the first game. On the other hand, once the switch has happened it will be that way forever, a player can only play so many games in the initial state but once they become experienced they can play a theoretically limitless number of games with their new knowledge. I don't know if there's a right answer here, is it better to make a game that everyone plays and loves a few times or one that most people don't play many times but those who do discover amazing things?

That's all I had to say on the topic of feeling balanced versus being balanced, but I stopped pretending that I'd keep my blog posts down to 500 words weeks ago and have something else I'd like to talk about while I've brought some of these ideas up. I mentioned a moment ago situations in which balancing by playtest could be contradictory, I'd like to talk about a way in which it can be outright misleading and in which it's better to be mislead.

Different sorts of players are attracted to different sorts of options. Some players will seek out complicated options with interesting interactions and clever little tricks, where others will stick to something simple. Sentinels of the Multiverse has characters that cater to both of the groups in this false but useful dichotomy. The thing is that players who are attracted to tricky options tend to be better at playing the game than those who aren't, the attraction to that sort of thing is driven by a desire to play clever moves that in turn inspires such players to get good at doing those things. The very act of playing such characters makes a player better at the game as it forces them to think of ways to adapt their options to achieve the best possible effect. As a consequence, if playtesters are allowed to choose which characters they play, it's likely that these characters will initially seem overpowered as more skilled players use them and they will need to be toned down. The result is a game in which the characters are not strictly balanced and that if characters were assigned randomly the characters will less direct force and more tricks would be revealed to be inferior.

Having witnessed a whole bunch of games of Sentinels in various contexts I strongly suspect that this is what has happened to that game. It seems that some characters can spend a long time manipulating their hand, board, deck and discard pile to unleash effects that ultimately fail to contribute anything significant next to straight up powerhouses just playing one of their cards. The thing is, I don't think it matters when this happens. Whether in Sentinels or other games, characters typically aren't dealt randomly. Players choose them based on their preferences, which will lead to the characters performing roughly as well as each other for exactly the same reasons they did in the testing.

The take home message from both of these topics is that while most of design is playtesting, playtesting has plenty of ways to produce misleading data. It's worth thinking as deeply as possible about the information that your tests produce.
Twitter Facebook
Subscribe sub options Wed May 22, 2013 3:13 pm
Post Rolls
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}