Ce diaporama a bien été signalé.
Nous utilisons votre profil LinkedIn et vos données d’activité pour vous proposer des publicités personnalisées et pertinentes. Vous pouvez changer vos préférences de publicités à tout moment.

Modeling of players activity by Michel pierfitte, Director of Game Analytics Research at Ubisoft

4 233 vues

Publié le

Publié dans : Technologie
  • Soyez le premier à commenter

Modeling of players activity by Michel pierfitte, Director of Game Analytics Research at Ubisoft

  1. 1. 1Modeling of players activityJune 20th, 2013Michel PierfitteDirector of Game Analytics Research
  2. 2. 2Lifetime RetentionDay 0 1 2 3 nGame Busa cohort getsin the busMetaphorLifetime = time spent in the bus, Retention = % of remaining users at each stop• Lifetime is a random variable, X = last active time - first active time• Retention(t) = Pr(X > t), probability of lifetime greater than t
  3. 3. 3Lifetime Retentiontypical lifetime retention curves of non-paying and payersnegligibledrop-offsignificantdrop-off50% on averageKPI : first day drop-off (50% on average)
  4. 4. 4Lifetime Retention model?horizonLife to date operation of the game modeling retention curvesR(t) = 1 – d * t1/αtparameters d and α are found with estimation techniques• The area under the retention curve is the average lifetime• KPI : quality of retention Q = log(area)
  5. 5. 5Lifetime Retention benchmarkWebMobileFacebookHD Online Multiplayer6 13 months5 5 months4 55 days3 20 days2 7.4 days1 2.7 days0 1 dayQ average lifetimeCriteria for launch : Q ≥ 3 (black line)
  6. 6. 6First day quitters in a mobile gameZOOM in the first day of the lifetime retentionDecomposition of the 21% drop• 3% leave within the first 15 seconds• 4% leave during the next 4 minutes• 14% leave during the remaining 24 hours• A lot of variation between games• Can help designers to understand whyusers leave
  7. 7. 7Playtime Retention• Users with same playtime canhave a very different lifetime,depending on the intensityand the frequency of play• Example : hardcore user10 h / day on average !Lifetime viewPlaytime viewactivity event• Playtime is a random variable, X = total active time of a user• Retention(t) = Pr(X > t ∣ lifetime > 1), probability of playtime greater than tfor users with lifetime > 1
  8. 8. 8Playtime Retention of a F2P gamenon-paying payers• We only consider users with a lifetime > 1day, complementary to first day drop-off• Impossible to read on a linear time scale• Playtime follows approximately a log-normal distributionKPI : median playtime
  9. 9. 9Population #1 : 39%, mode 0.8 hPopulation #2 : 21%, mode 11.7 hPopulation #3 : 40%, mode 21.9 hPlaytime Retention of aHD single player game of 20h• Modeling of the playtime retention by amixture of 3 population with log-normalplaytime distributions• Automated resolution using excel solver• Gives information to perform classification ofusers (supervised learning)mode #1 mode #2 mode #3
  10. 10. 10Revenuesfrom June 4th, 2012 to June 3th, 2013quicklystabilizedgrowthRpU = CR * AP * PFRevenueper UserConversionRateAveragePaymentPurchasingFrequency= * *= * *
  11. 11. 11quickstartslowstartachieve potentialPurchasing Frequency (PF)• Trend is known in 5 daysof observation• Potential PF is predictedby a model based on thecurrent known value• Can’t predict wether thepotential will be achieved• When the curve turnssharply, most of the timeit’s because of poorretention of payers= current value
  12. 12. 12Probability of Purchaseprobability of 1st purchasing day = CRKPI : probability of 2nd purchasing day• Spiral of probability of (re)purchase : 30 days dialrepresentation• Each probability point is the % of payers relative tothe previous point• The interval between two points is the median time• The probability to purchase increaseswith each purchase• 1st & 2nd purchases are critical to success
  13. 13. 13Purchasing DaysKPI : percentage of one-shotsone-shots (single purchasing day)
  14. 14. 14Progression• Ideal case: flat histogram (constant acquisitionof users who keep leveling up)• Outsanding bars signal levels where users quitthe most• Main reasons to quit (based on experience) : unpredictable time interval between levels peak of difficulty in the gameplay boredom• Very often the CR reaches 100% for high levels :this is a symptom of efficient monetizationhooksKPI : no outstanding bars in thehistogram of levels
  15. 15. 15Summary of KPIs• first day drop-off• Q : quality of lifetime retention• median playtime• RpU : revenue per user• CR : conversion rate• AP : average payment• PF : purchasing frequency• probability of 2nd purchasing day• percentage of one-shots• outstanding bars in the histogram of levels
  16. 16. 16Thank youfor your attention