terrenblade wrote:QUOTE (terrenblade @ Nov 25 2006, 02:01 PM) Elo could care less about 'sunshine' it's mesureing how well the earth can support grass, by looking where grass grows, and how well.
And your right, wether a single game wins or loses is no indication of if a player is good or bad. But if you watch the trends, see if he wins more games then he loses, you'll see if he's more valuable or not.
Basicly it asks the question "Is what this player doing causing the team he is on to win more often then not." and it does it without the dev team or whom ever else spending months or longer figureing out which actions are worth points, which are not, and how much points each is worth.
Would some one like to say the same thing again in a difrent way for the fith time? or do I get another go. /tongue.gif" style="vertical-align:middle" emoid=":P" border="0" alt="tongue.gif" />
You fail to see the flaw in that logic so once again I'll explain. You are making an assumption of a players skill by game outcome. If you and I flew 1 vs 1, we could not deduce who the better player is with only one game. Since the map doesn't always favor each team equally, and since there are so many factions to choose from with so many possible combination of tech's researched and game settings.......we must play many games of EACH to finally determine which of us is the better overall. With a 1v1 game, we have a different role to play in the game than we would have if other players were involved but for the sake of arguement lets say whoever wins those games is "Better". OK, so now we have a relative rank, one is better than the other. Now lets add another player. He must now play each of us all those games all over again to determine where he ranks compared to us. You cannot assume that if you were better than me overall, and he beat you, that he would be able to beat me so he must play each of us before we can determine his rank relative to ours. Now lets suppose with just these three players, we add a new variable, 40 other players! Where is the rank of the original 3 compared to the rest? Well, we'd each have to play 1v1 against each other in every possible combination of games possible in order to eliminate all those variables the game imposes on us. The amount of games necessary to play to finally determine each other's rank relative to the other's would be a factor of the total number of players. Using a 1v1 game structure, this alone would take forever to do. Now lets add a new variable, we instead of playing 1v1, play 20 vs 20. We must now play every combination of every game playable with these players NEVER changing in order to find out how each one ranks to the other. Now we add even another variable, the commanders! They are a huge influence on game outcome. So to eliminate their influence we'd have to play every combination of every game possible with every commander using always the same players to determine player rank. Since each player has varying talents, they too must be considered to determine relative rank...how is this going to be factored in? Since the games being played will ALWAYS have changing players, the constant needed (us) in order to eliminate all the variables (game settings, factions, commanders, technologies....) is gone.
Again, the only logic offered is that somehow over time, after a bizillion samples, somehow ELO will sort out all the details and eliminate these variables to find constants that aren't constant. No matter how many samples you take, if you continue to use all these changing variables without elimination as well as ignore the commander's enormous influence, then you will never find anyone's rank.
When you repeat over and over a logic that makes no mathematical sense, then yes, I'm going to ask you to explain it until it DOES make sense or you finally see that it DOESN'T make sense.
being a top player doesn't necessarily mean good judgement of other people, just that they are good at a game.