How you can use metrics to predict when players will quit

In a new Gamasutra feature, Innova's Dmitry Nozhnin writes how he applied analytics to MMO Aion to predict with 90-plus percent accuracy when players would churn out -- but there's one surprising caveat to that. The team's first goal was to predict when players would churn out of the game. Rather than waiting to try and recapture inactive accounts, the team wanted to incentivize players to stick around the very day they would quit. They got there. "With such high precision/recall we can be confident in our motivation and loyalty actions," Nozhnin writes. But there's another goal they didn't reach. "Have we achieved our second goal, determining why players churn out? Nope. And that's the most amusing outcome for me -- knowing with very high accuracy when a player will leave, I still don't have a clue why she will leave." It turns out that since MMOs are complicated games, it's hard to understand why some players don't stick around for more than a week -- the length of the free trial offered by Innova. However, predicting that they won't churn out is much easier -- if they hit certain gameplay metrics, he writes. "The key metric in this research appears to be the number of levels gained during the first day of the trial. Fewer than seven levels -- which represents about three hours of play -- means a very high chance to churn out." The next most important metrics are mobs killed per levels, quests completed per levels, playtime in minutes per levels, and playtime per day, writes Nozhnin. The upside is that once the reports run every day, "these are just-in-time results; at 5:30 am, models get processed, new churners are detected, and they're ready to be incentivized the moment we come into the office in the morning." The full feature explains how Innova devised and refined its metrics methodology from scratch, and is of interest to anyone working on a game with a strong need for player retention. It's live now on Gamasutra.