This is a classic response to a software feature request. “I don’t feel like I need it, therefore you shouldn’t need it either”.
It happens to every feature request known to man, discussed in a public forum.
Right now that is what you are doing: you are trying to convince me that I don’t need or want the thing that I have said that I would like.
I’m sorry, you are wrong: I would like what I have said I would like
I would like the graph to show me more emphasis on the vertical scale, and my reasons for wanting it are as valid as your arguments why I should not want it.
Note that the new graph also suffers from projecting improvement across time instead of games, so it penalises even further the performance of slower playing players.
Your current argument is that “nothing much in general happens to our ranks for ages - the climb from 25k to 7d is a long slow one - so there’s no point in making the changes as they go visible because they are insignificant, so let’s represent them as such”.
It’s an odd opinion from where I’m sitting, and I think the argument that “the journey is long so it is nice to celebrate achievements along the way” is equally valid.
Ironically, the old graph supported both use-cases - it had a zoomable vertical scale. Maybe that is what this graph needs…
I’m sorry if I have offended you in any way. I didn’t mean to come across as attacking your view. As I said, if you compare my two graphs… I OBVIOUSLY prefer the old one because it just makes me feel like the bees-knees! But I believe this issue is bigger than either of us and as I pointed out, for some players apparently crashing after a couple of losses can be devastating enough for them to quit altogether.
I do agree that we should be able to toggle the x-axis between months and games.
I just feel it necessary to push back on arguments I don’t agree with
I don’t feel that I want the old graph “because it makes me feel like the beez kneez”. In fact if you look at mine, there are significant periods where that would not be the case at all
I want an improved vertical scale because I want better visibility into my progress.
I think that “keep the graph flat to protect the mental health of the players” is not the right way to look at it.
You might as well say “hide the graph to protect the mental health of the players”, and the next step after that is “remove rankings, because a fall in ranking can be devastating”.
A better approach to managing emotions would be to point them at videos like this:
and encourage them to tackle the actual problem (overly caring about rankings) rather than hiding from it.
Then we could safely have graphs that acurately reveal what is going on with our ranking without having to tread softly for fear of feelings.
I’m also in similar situation (I play only corresponding/long games):
in old system:
• 1591 wins
• 3488 looses
• 2 draws
in new system:
• 600 wins vs. weaker opponents
• 526 wins vs. stronger opponents
• 140 losses vs. weaker opponents
• 3296 losses vs. stronger opponents
My correspondence rank in old system was 13k (897 - almost 12k) and in new system is 1063+/-97 (23k)
I think it would bear pondering long and hard whether this is in fact a “problem” that “should be tackled”.
We can see that the best way to rank up is not necessarily the best way to learn.
But who says that it should be?
Remember that we need people to play down as well as up, otherwise who do the lower people play against to rank up?
What this means is that if you have someone who is always playing up, and as a result learning a lot but losing a lot so not ranking up, then the indication is that it’s time for them to pass it forwards.
They need to play down some more, and as a result they will both pay back the favour they received, allowing lower players than them to play up and learn, and also reap the rewards of their learning by actually winning and ranking up.
So Julko … it looks like you owe the weaker players about 2500 games. By the time you’ve done that and won most of them, you will be ranked where you deserve.
While I agree that there is nothing morally superior about playing stronger opponents and not passing the knowledge on to weaker opponents, I do not think that it is the task of the rating system to influence these dynamics either way. The task of a rating sytem is to give the best possible guess of strength based on previous games. If it “punishes” people who choose certain opponents it does not do its task as well as it could.
However, we still are not sure if this is actually the case.
I play mostly in tournaments - many of them without handicap - so I play many games with stronger opponents. I do not choose them as defalult. Wulfenia says: “The task of a rating system is to give the best possible guess of strength based on previous games. If it “punishes” people who choose certain opponents it does not do its task as well as it could.” IMHO it’s the case.
Sorry if this was asked already. When the ELO ranks disappear, will our win/loss record bar also disappear? I like seeing how many games other opponents have played and checking my own record (sad as it may be).
You lose less points against a stronger opponent than one of your level of power and gain a lot more if you beat them… But there is no way your rank can go up after a loss. If you strategy is to play stronger opponents to fast track your growth, eventually you’ll be strong enough to beat them and your rank will go up.
not always, the person could just as possible have gotten a fairly good boost because of this new system, or opponents may “ease up” thinking a person is weaker just by looking at his/her ranking.
yes, not always… i have that tendency of throttling too… i played a “newbie”, but in fact that player is stronger in real life. i underestimated my opponent and paid the price. although, the price i paid wasn’t as as bad since i’m only a ddk.
if i win, i’ll be happy; if i lose, i’ll be disappointed.
if i learned something from my loss then good for me; if i didn’t learn anything from the loss then tough luck for me.
anyways, we’re not yet one month into using the new rating system. once the dust settles, i’m sure the system will reach an equilibrium.
It’s not the task of the rating system to influence in this way - agreed. One should not distort the design of the rating system to try to achieve that.
BUT - it’s actually just a happy fact that the well designed rating system, which “give the best possible guess of strength based on previous games”, actually has this effect!
And note that it isn’t the task of the rating system to “help people learn” either. So if people are only playing up, because they want to learn against better opponents, that’s not the rating system’s fault.
The best way to “give the best possible guess of strength based on previous games” is to have people play both stronger and weaker players than themselves.
If players only chose one direction, there is no real way to determine their actual strength. I’m sure that I can lose just as many games against a 1d as you can. Unless you play some people who are weaker than you but stronger than me, there is no way the ranking system can differentiate our strengths.
Bottom line: for the ranking system to work you have to play up and down. You can’t really complain about your rank unless you do that.
And as a side effect, that is good for other reasons.
Hi, I’m wondering why I’m being stuck at 25k and get negative points when winning against other 25k players. The old rating shows an overall rating of 19k. Should I try to play against 19k players? I can’t recognise which rating is the leading rating to find apporpriate oppontents.
