This is why you have a low rank

So i que like any other game.... and i catch a rare..offer fq.. he reject , fine with me ill play

So after i win a pretty easy game ( which he made so many mistakes i can't count ) he starts flaming me on PM


Id copy paste the chat but too much work blacking his name out all the times....

Summary... i suck , my winrate is only 50% atm... while he is at like 240ish wins , 160ish losses ( Superior by far) .... Rank doesn't matter thats why im limited and he is rare ( his words) , and win % is what matters not rank , this is all his words

If you want to improve rather then flaming your opponent , try and look at your mistakes... most times you didnt lose simply because the enemies deck is "OP"
You can beat OP decks with skill...from practice

 

Then your username is H0spy hehe

 

You shouldn't entertain things like that. People have been making so many threads of people raging at them because they beat their opponent. You guys are just giving them attention.

 

maybe true , but i tried to use this experience as a chance to educate others.

 

Im curious now. Please tell me his name when I ask you in game. This whole thing sounds funny.

 

This thread almost deluded me into thinking rank is an accurate representation of skill.

 

I really don't get this community's obsession with win rate. I understand that the rating system isn't too reliable with its constant resets and all, but it's still a lot more meaningful than win rate. Aside from the fact that given a large enough player base (which isn't the case of course), the win rate should always gravitate towards 50% the more games you play, it is even downright possible to "game" the system if you would be so inclined. And I put "game" in quotation marks because it really isn't gaming anything unless people are silly enough to pay any attention to win rate.

For that same reason I believe that rating resets are a bad idea in general. I would replace it with a rating accuracy variable. What this means is that the more games you play, the more confident the rating system becomes about your rating. If you don't play for a while, the accuracy is reduced. The higher your rating's accuracy, the less it changes and the more effect it has on other ratings. If you have this setup in place, then after major changes you can simply reduce the accuracy of all ratings so that changes happen more quickly until all ratings are stable again.

 

Welcome to my world

 

- Lack of population dilutes the value of the "best" rank rating systems currently...and then unfortunately get dismissed as bad but it's really just the bad sample.

 

But... but..

 

Rank these days have much more to do with how many games a player can spam rather than if he´s acctually good or not.

 

you want to know how to get to limited? You log onto to pox and you play one ranked game & win. Then you do not play any more ranked games that day. If you can do this till you in the top 20 you really do not have play that much at all and you'll always be in the top 30 because when you get that high in the ranks it is just the same people trading places and not hoards of people trading places like in the lower ranks. i have done this trick the last few drums of wars.

 

Which would by the way be far less true if the rating system would be allowed to mature more. The current top player has a rating of 1315, which really isn't that different from a beginner rating of 1000. This means that you still get a lot of rating points for beating weaker opponents.

Compare this to chess, where the best players have ratings around 2800. The difference to an average player is so huge, that "spamming a lot of games" against random lower rated players would not make a noticeable difference to their rating whatsoever (starting at around 400 points of difference, you get practically no points anymore for a win because the win expectancy gets close to 100%).

Right now it's really quite ridiculous, with the actual skill gap between a newbie and a top player rivaling the diameter of the sun, yet the skill gap as represented by the rating system is more like that between a complete noob and a slightly advanced beginner. If you keep resetting the rank this is never going to change, so please please please stop doing that.

 

last rank reset was vital due to decay.

I think it would be reasonable to give an edge (expwise) to already high ranked players so there are still some "starting tiers".

ladder was always favoring quanity of games a lot. all games. all type of ladders. it's a ladder afterall...

 

Please explain? Whatever it is, I am certain there is a more elegant solution than a rank reset. If there is an artificial decay to penalise those who don't play a lot, then that's another really bad idea.

Rating accuracy as described above would ensure that those who don't play often don't have their rating frozen. And if somebody doesn't play at all for a while, simply remove them from the ranking lists until they become active again.

No, that's not true. In this case "ladder" simply refers to a sorting by rating, of course that does not automatically favour quantity of games. And ELO rating is designed to create a representative and stable rating over the long term. It just simply can't work well with frequent rank resets, it's impossible. It's almost a farce to even use ELO ratings like this to begin with.

Basically the more you mess with the ELO system, the worse it gets. Sometimes you really just need to let it sit and ripe.

 

personally , id make the decay at top 100 be stronger then other ranks and less time needed to start to make limited difficult to hold without playing...... i play 5-15 games a day (depending on how busy/free i am) most days to keep climbing .... people shouldn't be able to do what G1 said.. reaching limited and just never playing

 

The problem is lack of transparency in how much exp is gained or lost after a ranked game.

 

i highly agree with this..... i don't even bother estimating i just play till i feel its enough wins for the day

 

Any kind of decay with Elo systems is obscene. The whole idea of Elo is to create an accurate representation of your skill, not to reward you for playing. The more you mess with the formula, the more the system will fall apart.

A system that includes ratings reliability like the Glicko Rating will ensure that your rating changes quickly if you only play one match a day. You can still sit on your high rating by not playing at all, but that isn't really a problem, especially if players are simply hidden from the ladder if they are inactive for too long. The higher the rating difference between a newbie and a top player, the harder it is to reach that high rating to begin with. If you are not a top player, you can't do it.

If you want a different type of ladder system that rewards participation, then make it separate from the Elo rating. Elo rating should remain as a true indicator of skill. Mixing the two concepts is absurd, and if people would rather take win loss ratio as a true indicator of skill than the rating, that's really quite telling.

Seriously, what do we have to lose by giving "true" Elo/Glicko ratings a try? Or would you say that the existing system has worked wonderfully well?

 

Spot on.
I gotta say, bud, your posts are always a nice read - are you new around these parts?
Either way, keep it up, it´s always good to welcome another voice of reason (that is, until you nerfcall ST )