Lawlzz!!! Sorry Jovai I just had to laugh at this. But my finance example is true. Everyone knows it isn't the right way of doing it but it turns to be the right way when no one has a better idea.
That's probably true, Calmdown. That statement might even be a bit nice.
Note that I've updated Part 3 (post #3) on the first page with additional detail about how pure, blind luck affects the system in both negative and positive ways.
This is excellent, very comprehensive if a bit long-winded. I'm glad somebody took the time and effort to properly explain the rating system. Hopefully if we can get this stickied we'll see a couple less threads whining about how broken the rating system is and less people thinking of it as a "score".
Honestly, rating has a lot more to do with good matchmaking than anything else. It's put in place so that people can be playing against others in their skill "range", which as explained here is quite a wide category. The only thing I think is somewhat off is this idea of the top 20 highest rated players getting recignition as the "best" in a hall of fame, when in fact there may be many players in the same skill range who just didn't have the same luck with the rating system.
I see the marketing appeal of having a "Top 20" or "Top 50" list or whatever, but ya if you read my part 3 closely you'll see why such a list is essentially worthless for comparing our epeen to each other. Especially if you snapshot the list at the end of the season just before a ratings reset. It's totally luck whether you'll end up in that final snapshot, and therefore the "hall of fame" is meaningless and should be called the "hall of luck".
What would make a LOT more sense for epeen comparison is to track each player's PEAK RATING throughout the season, and use that peak rating to populate a leader board.
+1 to get this stickied.
+1
enlightening post shannong.
Thank you for the info! Brings a whole new perspective to the rating system and how much value some players might wish to see in it. I imagine it must be very difficult(stress-wise) for some of the players at the top.
Doesn't seem to have gotten stickied. Meh, I'll bump.
Great read! As an Xbox live player i knew and read about their True-skill Rating (think Halo Reach) and this seems remarkebly identical.
It's a great system, once you understand how it works, and it makes you really proud of having a high rating (especially since it's temporary most of the time) There's just one drawback, i always wished i could see my sigma value..... Maybe if we pool data from matches with the ratings and the changes of ratings, we can make a rough guide on how to calculate your sigma value?
Or maybe you can figure out a better way?