Idea for the statistics system

[SilverShark]

What if instead of ranking players by score, it would rank them by how high their Av. MVPs is, or at least add such an option to the Sort by: menu?

 

I suggest this because score does not necessarily reflect a player's skill, it just means they've played the game alot. Average MVPs tells much more about one's skill than score.

 

However, this setting would be problematic, because there are players with 100% MVP rate, because they have played the game only a little, yet they have managed to become MVP. Perhaps the Average MVPs could be made relative to the player's score/game count, so this wouldn't be a problem? It could also cause some problems with the rank titles.

[Pushwall]

An "average MVPs" scaled by game count would basically be the same thing as total MVPs. I like the idea but not sure of a good way for scaling to weed out "1 game 1 MVP" types.

[Einstein]

Count total games on the server, excluding totally empty ones and rank people by percentage against that number?

[Jeod]

Count total games on the server, excluding totally empty ones and rank people by percentage against that number?

 

Close. Say I've played 40 rounds on the server, out of 120 rounds played on the server total. That's 33%. Now of my 40 games, I've been MVP 12 times. 30%. Now we factor in the exclusions. Exclude the following:

 

• Games where I achieved a score of X while the player count was below Y

Let's say that excludes three of my twelve games. That leaves 9/40, or 23%. However, this doesn't protect against the 1 game 100% MVP exploit. Thus, introduce a restriction where a player must have played at least ten rounds total that meet the conditions for being a complete round.

[moonsense715]

Less than 2 or 4 player games don't record stats for any players.

[OWA]

We can certainly look into adding a separate tab that ranks players by the number of MVPs that they have. This could sit alongside the score rank quite nicely.