There are currently 2 point-based leaderboards in eterna: Most points, and most points earned in the last 30 days. There are problems with each of them, as was discussed in Eternacon:
Most points:
-
It is overly hard for new players to catch up to the older players, which can make them like they are out of the race to number 1 since the moment they signed up.
Last 30 Days: -
Certain players have an unfair (and quasi-random) advantage over others, since certain puzzles are worth much more points then others. For example, if I beat the last puzzle in the puzzle progression 30 days ago, I will never be able to use their 500 points in the “Last 30 days” section, which a new player (or one which signed up before puzzle progression was created) will be able to solve it to gain a big advantage over me. The same goes for easy Challenge Puzzles.
-
Because each day the definition of “the last 30 days” changes, it can be unclear if you are on your way to passing another player, since it could be that I solved a lot of puzzles 29 days ago I will lose a lot of points tomorrow, but if I solved non then I won’t lose any but the person that I’m trying to surpass will, which makes everything much weirder and feel random.
My alternative to these leaderboards, which solves the problems listed above, is seasonal leaderboards, which is mainly inspired by Diablo 3 and World of Warcraft.
A season is a preiod of time, for example one month or the time between the end of two labs, such that points received from all puzzles which were published during that period accumulates into a seasonal leaderboard.
This model isn’t without it’s fair share of problems, which is why I think it should run alongside the “Most Points” leaderboard (with “Last 30 days” being removed).
However, there are some modifications to the seasonal model that could improve it:
- We should add specific puzzles to be part of the season, such as the progression puzzles.
- We might want to have seasons last less than the amount of time the puzzles it covers span. An example of that would be a season which lasts 3 mounths, but whose leaderboard scores players based on puzzles released in the last 6 months. This could help to fill the start of a new season with seasonal content, and also make players not feel like they are starting from square one.
- Another way to include earlier puzzles would be to take sets of puzzles (not necessarily series such as Drake’s special knowledge, but rather a time span or a big random group) and add them to the current season. This allows solving old puzzles to be rewarded in the seasonal leaderboard, and can help obscure old puzzles to resurface.
A downside of this system could be that there would be less incentive to play older puzzles, but I don’t think that it is really that problematic.
First of all, this system should run in conjunction with the current “Most Points” leaderboard, which ensures that old puzzles always have value.
But beyond that, I don’t think that it is that important to give older puzzles a lot of value - the only scenario in which players don’t go back to older puzzles is if they have enough puzzles to toy with in the season.
That means that the only problem that could occur, at least that I could think of, is that someone will not play a certain interesting puzzle (or series) that he found because it doesn’t contribute to the seasonal leaderbord.
I think that this problem could be solved by either allowing changes to the previous seasons leaderboards (even if the winners are announced right as the season ends), which would give older puzzles similiar value to new ones, or by having other sought after things similar to the leaderboard for older content, like achievements, or other things that I can’t think about.