Likelihood of Quitting

We've looked at player fatigue and likelihood of quitting before. In this article, we'll drill down a bit more and look at which demographic and play factors are most likely to drive quitting. These differences would also hint at how a game population changes as a game matures, in terms of which kinds of player are most likely to stay in or leave the population.

One stable finding we've seen before is that female players are less likely to quit an MMO. In the question below, players were asked to rate the likelihood that they would quit in the next 3 months on a 5-point scale (1 = Very Unlikely, 5 = Very Likely). The graph also shows an age trend. Older players are less likely to quit compared with younger players. Or more accurately, male players are less likely to quit as they get older. Female players across all age groups are just as likely to quit. Once past the age of 35, male and female players seem to have converged on the same likelihood of quitting.

====

A multiple regression would let us tease apart the relative importance of these demographic factors (i.e., age, gender, hours played per week) as well as include motivations of play factors (i.e., achievement, social, and immersion) on the likelihood of quitting. The model that emerged showed that the relative importance of the factors goes like this:

1) Hours (b = -.17): Not surprisingly, the more a player currently plays, the less likely they will quit.
2) Social Motivation (b = -.10): What was surprising was how much the social motivations were related to likelihood of quitting. The more a player enjoyed socializing in an MMO, the less likely they will quit the game.
3) Age (b = -.08): Age comes next. As we've seen, older players are less likely to quit.
4) Achievement Motivation (b = .06): The result here is unintuitive at first. The more a player is achievement-oriented, the more likely they will quit the game. This may be due to grinding and "burning-out" factors, or that players who are goal-driven lose interest in the game once they reach their goals or if they perceive their goals to be not worth the effort.
5) Gender (b = .03): The somewhat striking gender difference in the graph above actually isn't a good predictor of whether a player will quit compared with the other factors listed here.
6) Immersion Motivation (b = .008): And finally, how much a player enjoys role-playing or exploration has no impact on their likelihood of quitting.

This set of factors implies that as a game matures, and assuming that there isn't a large constant influx of new players, the maturing player base will tend to be more socially-oriented, older, and less achievement-oriented. It'll be interesting to see this in an actual game with longitudinal data.