Separate ELO for every variant General forum

17 replies. Last post: 2020-09-13

Reply to this topic Return to forum

Separate ELO for every variant
  • ypaul21 at 2020-09-12

    As players of pure strategy games, you are all probably already aware that even a slight difference in board size can have a huge impact on the game strategically. Right now, not only do we have various board sizes lumped together in the same figure, but also completely different games (especially in draughts and shogi).

    I am well aware that this is a very old site with a lot of history, and implementing such a huge change retroactively is probably not possible at all, but I thought that it’s worth a discussion anyway. How do you guys feel about this? Am I the only one who’s always felt that this is the one thing that I’m constantly annoyed about in this otherwise perfect place for playing the games that I love?

    Thanks!

  • metzgerism at 2020-09-12

    I’m pretty cold on ELO – if someone is protecting their ELO rating, there’s no reason to play.

    Some sort of weighted wins system with decay is preferable to me... I’d rather Richard consider that than splitting ratings right now.

  • ypaul21 at 2020-09-12

    I’ve found it to be a useful gauge of my ability, and it’s also nice to be able to see a quantitative improvement in my progress.

    I’ve been playing a lot of Mini Shogi recently, and this game is in no way representative of any other Shogi variant. I don’t like that my Shogi rating is changing even though I haven’t been playing actual Shogi. 

    About people protecting their ELO: I don’t see why I should care about what other people decide to do. I play because I like the games and I want to get better at them, and I believe that a lot of other people here feel the same way. Seeing your ratings go up is just an added benefit/motivation. If there are people who want to make that the main thing, then by all means.

  • ypaul21 at 2020-09-12

    And say 15x15 Hex and size-7 Catchup has just come out. This changes a lot of things. By your own argument, people wouldn’t want to play these variants because they don’t want to damage their ratings. I personally think it’s a big issue. 

  • ypaul21 at 2020-09-12

    I’m genuinely curious: why are Hex 19x19 and Reversi 10x10 considered their own games? Was there a reason for this, if anyone here was around for long enough to know?

  • Ray Garrison ★ at 2020-09-12

    Originally, when I joined in 2005,  there were no variants except there were two reversi, two hex, and three go categories each with their own rating.  But then the webmaster combined all the go categories into one rating but did not do the same thing for hex and reversi, which I though was inconsistent.  Then he started adding variants and new games.


  • ypaul21 at 2020-09-12

    Oh. So Hex 13x13 was actually added after 19x19? Interesting. 

  • ypercube ★ at 2020-09-12

    No, 19 was added after 13.

  • ypercube ★ at 2020-09-12

    For the question asked here, I agree.

    I would like to see different ratings per size and variant (Torus Go for example deserves a different rating) but I would strongly prefer if we also kept an overall rating per game – as the one we have now for Go – and add an overall rating for the games that don’t have now (Hex, Reversi, etc)

  • ypercube ★ at 2020-09-12

    Having said that, I have no idea how hard would the implementation be.

  • Ray Garrison ★ at 2020-09-12

    I like ypercube’s idea of individual variant ratings and overall rating for a game type.  But some games should be moved, for example, dameo should have it’s own category and should not be in the draughts category.   I also think numeric wyps would be best moved to  it’s own category away from the  many word variations of wyps.  (or, possibly combine all qyps and wyps into one category for consistency)  Are there any other changes that should be made to game categories?

    I remember Richard the webmaster did figure out how to combine all three different go variants into one rating many years ago (I used to have three different go ratings at this site).  Was this complex for Richard to combine?  If he were to adapt ypercubes rating plan, he would have to reverse the process and split the go ratings into three different ratings again (and keep the combined rating system)



  • metzgerism at 2020-09-12

    My comment above was a little too short to tell anyone anything...sorry, been busy most of the day.

    I agree that splitting ratings up by variant would be ideal, but I’d personally prioritize finding an alternative to ELO...something that doesn’t harshly penalize losses and encourages people to play. If you just play to get an ELO score and have it stay there, you begin to get rusty and you’re no longer as good as your rating says. I don’t have something firm in mind yet.

    Richard switched up the Infinity leagues to be simpler and untethered from a numerical rating – it’s a lot more fluid. So far I think I like that, and most people’s largest apprehensions about the change were addressed well.

  • ypaul21 at 2020-09-13

    I feel that punishing players for inactivity is even harsher. Perhaps you may be right that it’s fairer, but I would be really put off by a rank that constantly ticks down over time. It’s going to remove a large part of one’s sense of achievement. It would be sad any account would eventually see their account with every rating at 100. Being forced to remain active would make playing on LG would feel more like work than a hobby. I honestly prefer it to stay the way it is. I really see no issues with people trying to immortalise their achievements if that’s what they want to do.

  • ypaul21 at 2020-09-13

    Okay, it seems like people agree that different games that are put under Shogi and Draughts should be split up, but think that close variants of the same game (LOA setups, board sizes, maybe different languages of word games?) should be grouped.

    That’s fair, but don’t you think that size-4 Havannah and Size-8 Havannah, for example, are really quite different games? I think it would give people motivation to play more than one variant too if they all have separate ranks. Right now, people have no reason to really try playing another board size if they’re already really good at one. But of course, if you are only interested in playing a single variant, that’s fine as well because these changes still wouldn’t affect you at all. 

  • vstjrt at 2020-09-13

    Maybe we could implement rating floors. When you cross some rating you never go down below for example more than 200 points. This will inflate rating system. To encourage people to play (and prevent inflation), We can decrease small number of points every month (like 1 or 2 points down every month).
    This will motivate people to play, because their rating will be reducing every month, but there will be sort of achievements, that will be rewarded.

  • ypaul21 at 2020-09-13

    I don’t know. That feels completely arbitrary, and if it’s not done correctly, it might be too punishing to keep players on this site. I’m still not convinced that it’s a good idea. 
    How about we implement a site-wide ranking for day, top X players for every game, and only players who’ve completed a game within the last three months will be listed in that table? I think that’s what BGA does. Sometimes it’s motivation when you see that you’re the top 20 player in the site for one particular game. After some amount of time of not playing, you won’t lose your ELO, but you’ll no longer be listed as a top player. 

  • ypaul21 at 2020-09-13

    Actually, I may have misread ypercube’s comment. So there should still be one average rating for each category of games. Yeah, I can see how that can be beneficial. 

Return to forum

Reply to this topic




Include game board: [game;id:123456] or [game;id:123456;move:20] or [game;id:123456;move:20;title:some text]