Originally Posted by GnilRettemHC

Read the thread?

As above: "There's this fun thing called "placement" matches, where if you show at the beginning that you're a great player, you'll transcend a rank that for some it takes thousands of games to reach."

Every single time ELO is used they have some kind of calibration method, if you have played any of the hundreds of games that use ELO you will know this. Placement matches are very very common I can probably name a dozen games that use it (Valve is particularly fond of it, as are Blizzard).


The downside is that we have to change the ranking system. This is a considerably downside so in lieu of actual advantages there is no point in changing............

Originally Posted by SmallBowl

I have read the thread, I think you are confused about what the Elo system is and what calibrating the system itself means. The placement matches are not part of the Elo system per say they just decide what your base Elo is. Calibration would be changing how much Elo you win or lose per game based on more/different factors.

Originally Posted by SmallBowl

Ok, now that not only are placement matches not a method of calibration,

Originally Posted by SmallBowl

but they have been shown to not always work as if you play badly in 3 matches you have to play hundreds of hours to get the rank you would otherwise, and deserved to, have been at. In addition placement matches are far more common in team ranked games than 1v1 ranked games.

Originally Posted by SmallBowl

Please do cite a calibration method if you do have one.

Originally Posted by SmallBowl

Plenty of advantages are brought in, particularly the validation of the competitive side of the game.

Originally Posted by SmallBowl

In addition changing shouldnt be that much work since the source code is freely available for Glicko-2 and afaik only 1 variable has to be determined and changed in the code.

Originally Posted by SmallBowl

Why dont you go play a game that uses Glicko-2 as its ranking system and judge for yourself whether its better?

Originally Posted by SmallBowl

I mean alarm bells should be ringing when the game Elo was created for switches over to Glicko-2... Surely even you would admit that?

Originally Posted by GnilRettemHC

There are already methods for calibration with ELO, so thinking that a grind from 1200 (or whatever your base ELO is) is integral in ELO is simply not correct. Giving an undue bias for new players and allowing them to ELO camp at very high ranks with few games played is not very good.

If all you are concerned about is calibration, then there's no need for an entire overhaul:

...



On a side note: is posting in color no longer banned?

But besides that, what game even uses glicko? Almost every single major competitive game in the world uses ELO.

Originally Posted by Boredpayne

Most chess federations and CS:GO and GW2. Microsoft uses TrueSkill, which works on the same principles.

Originally Posted by Boredpayne

Glicko-2 is a derivative of ELO with changes to account for statistical uncertainty. It's in many ways a straight upgrade.

Originally Posted by GnilRettemHC

CS:GO used to use ELO but used some "ideas" from glicko-2. If you want to construe this as "they use glick-2" then that's a huge leap...

The CS:GO competitive ranking system started with ideas based on Glicko-2 rating model and improved over time to better fit the CS:GO player base.

Almost every single major competitive game in the world uses ELO.

We also track our uncertainty about your MMR. New accounts and those playing in Ranked Matchmaking for the first time have high uncertainty. Higher uncertainty allows larger adjustments after each match, and lower uncertainty leads to smaller adjustments. Together, the MMR and uncertainty can be interpreted as a probability distribution of performance in your next game; the MMR itself serves as the mean of this distribution and the uncertainty is its standard deviation.

So we are left with 'most chess' and gw2 (gw2 is a major competitive game? Hmmm, ok then...)...