Editing Isotropic
Warning: You are not logged in.
Your IP address will be recorded in this page's edit history.The edit can be undone.
Please check the comparison below to verify that this is what you want to do, and then save the changes below to finish undoing the edit.
| Latest revision | Your text | ||
| Line 29: | Line 29: | ||
A key feature of Isotropic was its skill level estimates. Player skill was quantified using an adaptation of Microsoft's [http://research.microsoft.com/en-us/projects/trueskill/ TrueSkill] skill ranking, itself a variation of the [http://en.wikipedia.org/wiki/Elo_rating_system Elo] rating system. The system assigned players a level rating based off of their results against other players, factoring in the skill level of all players involved in each game. Among the attributes of the system: | A key feature of Isotropic was its skill level estimates. Player skill was quantified using an adaptation of Microsoft's [http://research.microsoft.com/en-us/projects/trueskill/ TrueSkill] skill ranking, itself a variation of the [http://en.wikipedia.org/wiki/Elo_rating_system Elo] rating system. The system assigned players a level rating based off of their results against other players, factoring in the skill level of all players involved in each game. Among the attributes of the system: | ||
* Beyond the win/loss status, end of game score has no bearing on rankings. A one point nail biter is worth just as much as a 100 point blowout. | * Beyond the win/loss status, end of game score has no bearing on rankings. A one point nail biter is worth just as much as a 100 point blowout. | ||
| − | * Multiplayer matches get broken down into individual head to head games for the rankings, so getting second place in a multiplayer game is still important. A 3 player game is broken into P1 vs P2, P2 vs P3, and P1 vs P3. | + | * Multiplayer matches get broken down into individual head to head games for the rankings, so getting second place in a multiplayer game is still important. A 3 player game is broken into P1 vs P2, P2 vs P3, and P1 vs P3. More discussion on this can be found [http://forum.dominionstrategy.com/index.php?topic=2289.msg39298#msg39298 here]. |
* A win against a highly ranked opponent increased your rating more than a win against a lower ranked opponent. Conversely, a loss to a highly ranked opponent damages your skill rating less than a loss to a very low ranked opponent. | * A win against a highly ranked opponent increased your rating more than a win against a lower ranked opponent. Conversely, a loss to a highly ranked opponent damages your skill rating less than a loss to a very low ranked opponent. | ||
* A player's displayed (in Isotropic lobby and on leaderboard) skill level was the result of subtracting a player's variance from their skill estimate. This provides a lower bound for a player's skill estimate. There is some debate (especially among elite players) on whether this is the best way to manage skill levels, and the Isotropic [http://dominion.isotropic.org/faq/#leaderboard FAQ] addresses some of these questions. For some players, their level calculations would actually fall below 0 because their variance was larger than their mean skill rating. dougz addressed this by creating a lower end cutoff of Level 0. | * A player's displayed (in Isotropic lobby and on leaderboard) skill level was the result of subtracting a player's variance from their skill estimate. This provides a lower bound for a player's skill estimate. There is some debate (especially among elite players) on whether this is the best way to manage skill levels, and the Isotropic [http://dominion.isotropic.org/faq/#leaderboard FAQ] addresses some of these questions. For some players, their level calculations would actually fall below 0 because their variance was larger than their mean skill rating. dougz addressed this by creating a lower end cutoff of Level 0. | ||
