We need a proper map popularity thread

[mibi]

I would love to see some rankings on which maps are popular and which have largely failed to attract the publics attention.

I used to compile such a list but based on total pages of active games. Cairnswick also had a thread about map movement stats but it sorted by total games played.

Given that the number of territories on a map has an effect on the length of play involved, and the number of active games going on, i think someone should come up with a formula that is based on the number of active games, but weighted to factor in larger maps.

i.e There are 259 active Super max games and 148 active Doodle earth games, however I would say that doodle is more popular, its just that its way smaller so games end quicker. so doodle would have to be handicaped in the scoreing.

anyways, i think that with so many maps out there, we really need to know what is popular and what people are attracted to.

[The Neon Peon]

An easy method to determine popularity is: (number of games on the map)/(days the map has been available to play).

It would be interesting to see the results.

[owenshooter]

I would love to see some rankings on which maps are popular and which have largely failed to attract the publics attention.

yeah, that might be something the foundry should look into doing and housing with a sticky at the top of that forum...-DDB

[mibi]

An easy method to determine popularity is: (number of games on the map)/(days the map has been available to play).

It would be interesting to see the results.

that is still skewed towards smaller maps as a player can start and finsh 10 doodle games in the same period of time they start and finish 3 supermax games.

[Incandenza]

that is still skewed towards smaller maps as a player can start and finsh 10 doodle games in the same period of time they start and finish 3 supermax games.

Seems like there could be some sort of variable that takes into account map size: .5 for tiny maps, 1.5 for huge maps.

Better yet, with some hard-core data mining, the constant could simply be the average number of rounds per game per map (tho one of the problems with supermax, at least in team games, is that the last several rounds tend to be mopping-up of cell blocks when the outcome is no longer in doubt).

[e_i_pi]

that is still skewed towards smaller maps as a player can start and finsh 10 doodle games in the same period of time they start and finish 3 supermax games.

Seems like there could be some sort of variable that takes into account map size: .5 for tiny maps, 1.5 for huge maps.

Better yet, with some hard-core data mining, the constant could simply be the average number of rounds per game per map (tho one of the problems with supermax, at least in team games, is that the last several rounds tend to be mopping-up of cell blocks when the outcome is no longer in doubt).

Hard core lol, you've obviously never done 3-year recursive analysis on 10000 stock tickers But yeah something along the lines of:
P = R / (G * D)
...where...
P = popularity
R = Total number of rounds played in all games
G = Number of games played
D = Days since the map went live
...that would be pretty easy to implement. Each map might need one more variable in the SQL tables, but since everything can be done by PHP/SQL, it wouldn't even need to be updated manually. It could be updated on refresh, or to reduce server load, updated via a cron job once a week / month. Now where's a coder when you need one?

As an example, here's some easy to understand stats:
Small Doodle-Like map: avg 5 rounds per game, 100 games played, since it went live 5 days ago
Medium Europe-Like map: avg 10 round per game, 100 games played, inception 10 days ago (same # of games, over twice the time, as it takes twice as long to play)

P(S) = 500 / (100*5) = 1
P(M) = 1000 / (100*10) = 1

That takes out the time per game factor, bringing it down to strict popularity...

[e_i_pi]

Hmm, you might need to take into account number of players, and teamsize as well though lol...

[Incandenza]

Yeah, that's the sort of thing I was talking about.

[AAFitz]

Well, after looking at this briefly, I think perhaps any formula like this will more define what popular means, rather than define what map is actually popular.

I think Popular, can only really mean, How many unique individuals play a particular map to be the exact definition of popular. If a small map is handicapped, then the formula is essentially penalizing the map for perhaps the very reason it is Popular, and then the number doesnt really mean Popular...though I suppose it could mean Relatively Popular.

Also, The fact that Ive played world well over 1000 times does not make it more popular...so I should only count as 1.

True Popularity = the number of people that play a given map. (thouogh even this does not account for people that played it once, and hate it.

I suppose a complex formula that accounts for true popularity, and relative popularity could produce the closest thing to what might be considered the most popular...It will be subjective...but if both numbers are shown, it will be close enough to be useful

True Pop= # Unique Players Playing 1 game on map
Relative Pop= # Games played on map / #Days available to play
Daily Popularity= # Games played on map / Day since map was established
Relative Popularity = # games Played on map/ # Days available * Size factor of map(or speed factor ie, AOR is a big map, but you get dropped 2 places, and its over quick usually.

Total Estimated Popularity = Combining all of these factors together in some weighted formula

[Gypsys Kiss]

Would you discount tournament games, seeing as you dont always get a choice of what map you play(i.e.Map Blaster)? Being forced to play a map shouldn't make it more popular.

[Supermarioluigi]

Would you discount tournament games, seeing as you dont always get a choice of what map you play(i.e.Map Blaster)? Being forced to play a map shouldn't make it more popular.

And not to mention Random Map as well, which we have no idea which games were Random...which throws a big problem into the mix.

[Incandenza]

I doubt that tourney or random games would significantly change the outcome, as they don't represent more than a relative handful of games compared to the 5.6 million or so total games.

[MrBenn]

A popular map will be played by lots of people, lots of times...

I'm too tired to think how to calculate it, but surely the map with the highest number of different people playing games on it each day = Most popular ?

[the.killing.44]

I like pi's equation, but it rewards build games too much.

[Crazyirishman]

Don't forget about how spoils will effect the length of games. flat rate/no spoils games can easily last 100+ round while escalating rarely go over 30.