On a GCL Amoeba discussion, Eric Brosius started a discussion on his Staying Power metric he calculates for the games he has played. That interested me enough that I implemented the same metric in my own game stats.
The way it works is that for a start, it’s an average of number of times a game is played each year. To reduce the significance of years when a game is played a lot, the square root of the number of plays is used.
The years are also weighted with a formula (5/6)^(last year – this year) so that the current year gets a weight of 1, the year before that gets 0.833 and so on.
Finally, to balance things with games of different length, the resulting score is multiplied with the game length in hours.
Here’s my top 20 list based up to 2019:
- Love Letter (3.22)
- Magic: The Gathering (2.75)
- Agricola (2.27)
- Suburbia (2.01)
- Tarock (1.85)
- Fields of Arle (1.84)
- The Mysteries of Old Peking (1.77)
- Innovation (1.61)
- Dominion (1.41)
- London (1.38)
- Ora et Labora (1.38)
- Carcassonne (1.37)
- Battle Line (1.30)
- 1846 (1.23)
- Age of Steam (1.20)
- Afrikan tähti (1.20)
- Brass (1.19)
- Terraforming Mars (1.10)
- Dale of Merchants (1.09)
- Oregon (1.107)
I think this list is a solid list of games that have stuck with me. Some of these, like Dominion, are on their way off the list and new games are bubbling in, but this is a fairly conservative list that resists changes. New games won’t instantly pop on the list.