For many (most?) people, I believe rank is just a metric of progress. Any relation to handicaps or whatever is coincidental.
It’s nice if it follows other organisations/servers’ ranks for easier comparison and I definitely appreciate if the ranks don’t drift too much. I’m totally happy to have the site calculate the appropriate handicap for my games and I have no strong expectations of how that should relate to the players ranks or ratings, whether integer or decimal.
None of the proposals change handicap or your actual rank, only the visual display of the decimal portion.
I feel like this has gotten lost in proposal limbo while there was good public support for either system A or B and against the current system.
Based on the polls, I’d say System A hasn’t gotten much more support than the current system.
Based on those polls, System C (current system) is disliked more and liked less than either A or B. The OP (proposer of System A) is in favor of B over the status quo.
Meanwhile there is a suggestion of next ranking update when there is no ranking update in the works.
Seems like a clear case of inertia to me.
In the meantime, more votes came in. This is the current situation:
I’d summarise it like this:
System A makes more sense than System C, but both of them are pretty confusing.
System B makes only a bit more sense than System A, but it is clearly the least confusing.
So, system C is still clearly inferior and only propped up by incumbency advantage.
I would never have guessed that it would be so a debate in itself.
Yes, it seems there is not much support for System C despite its incumbency advantage.
As for System A vs System B, I feel kind of bad that System B ended up favoured more than System A. I wasn’t out to beat System A of OP with my initially offhand remark about potential contestor System B. Am I now responsible to carry System B until implementation on OGS?
I think this factor is not to be written off though. The motivation for such an update is to reduce confusion, and changing the system is itself a source of confusion.
For this reason, I support the proposal to introduce any change in the next ratings adjustment.
Right, those real ranks (displayed next to usernames) represented elegantly by the unmoving color bands won’t move. But the decimal portion should still consistently communicate rank at the same level as display rank. The fundamental anchor of a rank is the handicap stone.
What I don’t like about the decimals and ranks comes down to 1k being one rank step away from 1d. 0k or 0d have ambiguous meanings since they don’t exist. Contrasting when we learn math we expect 0 is one unit away from 1 and one unit away from -1. Decimals can imply rank some sort of relationship with math numbers which work everywhere on the ranks scale except for the dan/kyu border.
With system A, a 1.1d gives zero stones to a 1.1k and one stone to a 2.1k. It distorts the intuitive scale of distance of a one stone difference between a 1d and a 1k. For every rank a player progresses through they can count on giving one more stone to the previous rank. For this threshold the difference oddly becomes compressed.
With system D, a 1.1d gives two stones to a 1.1k. So for this threshold the progression difference is suddenly stretched.
Note that the 1 rank gap between 1k and 1d is a bit similar to the 1 year gap between the year 1 BCE and the year 1 CE.
The year 0 is not used in historical notation, although in astronomical notation 1 BCE is denoted as year 0 and 2 BCE as year -1 etc. That’s a bit like writing 1k as 0d, 2k as -1d etc.
And that’s a problem with the year system, too. Obviously we shouldn’t change the meaning of 1k and 1d, but just because another commonly used system has the same problem doesn’t mean it’s not a problem
Maybe we’ll get a nice video game called 0 Kyu one day
But systems A and B are the same in this regard—both have a gap around 0, it’s just set up differently. As I’ve pointed out above it will be the case unless you allow ranks like 0.0d, which we can all agree would be horribly impractical and we must forget about. Whatever the system we choose, there will be a gap at 0.
Your intuition seem to be to think of 1.0k as the average 1k, which is where the oddity lies. As I’ve said above, finding system A meaninful all boils down to accepting that 1.1k is a strong 1k (almost at the end of the scale, 1.0k), so one that’s very close to dan level and who shouldn’t necessarily be given handcap by a 1d player.
Eventually I’ll be fine with whatever but as someone who prefers system A (more visually pleasant, consistent with AGA, keeps the current dan rank display as it’s already fine, and IMO altogether truer to the inherent properties of the kyu/dan system), I’m a bit frustrated to see people come up with this fallacious argument as for why it would be less intuitive than system B! I feel like I have yet to see valid arguments as for why system B would be better…
Is there a next ratings adjustment, though?
