Exponential Distribution: (Team Fortress 2, Skyrim, FTL: Faster Than Light, etc.)
Games that fall into this archetype are ones that can be played without a specific endpoint. Gamers can choose, based on their personal enjoyment of the game, when to stop playing. Making a value prediction for these games is difficult because of the wide range of play times. Many gamers wait to buy these types of games until the price drops to a point where they would be comfortable losing their money if they dislike the game.
Normal Distribution: (Metro: Last Light, Dishonored, Bioshock Infinite, etc.)
Higher priced, linear story games fall into this archetype. When people pay more for a game, it appears to be a motivator for completing a single play through. A linear story helps to group the distribution around a smaller range resulting in a distinctive hump near the average play time. The level of confidence in how long it will take to play through these games is much higher than the other archetype.
There are a few interesting observations regarding these archetypes:
- It appears that as a game ages, its distribution decays into an exponential distribution. This appears to be a factor of more gamers buying a game as the price is discounted and relying on their personal enjoyment to dictate when they should stop playing.
- You can have a greater level of confidence in the estimate for your expected play time when a game has a strong normal distribution, unfortunately this doesn't mean that you will enjoy the game.
- You can tell if there is a fanatical group of gamers that really love a game by looking at the range of the distribution. The longer the "long tail" is, the greater the fan base that really loves a game.