Talk:Elo rating system/@comment-188.178.187.14-20120126081725/@comment-3014924-20120126145520

"Ha ha ha. That was a good one. I could put two team of monkeys to play LOL and that would have a win loss of 50%. That is build int the game. Claiming that playerinterference have had ANY significcant effect on the game, by quoting a win loss of 50% is quite strange.

What u have to do, is pick 1 moderately good player. 25% better that average. Match him with 4 random team mates, and 5 random enemies, and measyre the number of games that it takes to get an average rating of 25% better than average. Then you know the usability of the tool. My guess is that it will take more than 200 games. Just a guess though. The ten games initail ELO basis is a joke"

Ah, oh. It was not my point.

I was arguing that you could invoke the law of great numbers.

The external factors affects you for sure and randomize even more your win chances.

But eventually, you'll reach a point where the mean value of your results are close to your real value.

You even somewhat agree to that by claiming that it would take 200 matchs.

Let's say you play 200 games. Let's say your skill doesn't improve. The average ELO over the course of those 200 games should be very close to your real one.

Let's say it's 1300. But you could be, after those 200 games, be at 1200 (external factors applies). In truth, it's the average result that is close to your real level.

"Also u may have noticed that many players are very angry in ranked games. That is a typical behavioural sign of a feeling of injustice.

An injustice that is likely to have its ofsett in a genuine feeling, that the gamis is fun, but the ELO system sucks and is, by al standards wronly applied. "

Normal games have an hidden ELO. Technically, no great difference from ranked. Just people take ranked too seriously!