I’m a brand new player, and after a couple games for some reason the site gave me rank 9k… I’m slowly getting slaughtered and having my rank get to where it belongs (25 probably), but it’s annoying being paired with people who think I’m a somewhat even match for them…
No. But after losing a few more games your rank will become more realistic.
No, unfortunately OGS does not let new players say they are beginners, so you need to get crushed several times whilst your rank drops. Hopefully this bad initial experience does not discourage you from continuing to play Go which is a wonderful game. I have long campaigned for a change to this system to give beginners a better initial experience.
Thank you for bringing this issue to our attention. It’s true that moderators can’t directly adjust player ranks. But we can annul games, and I have done so for two incorrectly scored games in your history, which brought your rank down to 16 kyu. You should have a much better experience around there until your rank stabilizes.
Thanks I’ll get back at it and hopefully earn a higher ranking soon~
Generally this is not the case. If you start playing, your first opponents will be (other) double-digit kyu’s, and if you lose those games, you’ll find yourself ranked similarly to any beginner within 2 or 3 games.
The problem arises when you win one or two early games by error, for example when the first 11k and the second 4k opponents resign because they don’t want to play against beginners. The system can’t see whether those games were legitimate or not, and as a result the new player gets ranked as if they are a 1d or something. It’s the best estimate the system can get. Garbage in, garbage out.
On the other hand, if we allow people to say they are beginners, this will inevitably make sandbagging easier, thus you’d still have beginners getting crushed several times.
Don’t new accounts come in as 6k humble ranked to 12k? So if you automatch you will pair with 12ks, but if you use the line on the game challenge page that’s at 6k. So 12k is a double-digit kyu but 6k is not, and they are both a lot stronger than a 30k so those loses would still be crushing, and 3 still counts as several in my book. So what you’ve said happens could still be described as “so you need to get crushed several times whilst your rank drops”.
This is such a reasonable and common thing to happen that it shouldn’t’ be considered an aberration, but a natural consequence of the current system for new accounts. So the moderator workload of annulling all those games should be considered when comparing to the moderator workload of other systems proposed. One objection to the previous choosing initial rank system was a high moderator workload of rank reset requests. I would be curious to know how that workload compared to the current system’s annulling-people-resigning-who-don’t-want-to-play-beginners (most of which go unreported and result in upset crushed beginners most of whom won’t post in the forum), but seeing as I propose a system in which you can choose initial rank (with very high uncertainty so if adjusts quickly, and has little effect on opponent’s rank but big on yours) but can’t reset rank the rank reset workload in my proposal is nil.
Hence why I think new accounts should be able to choose a rank (with the instructions and caveats I’ve detailed before). Allow more (admittedly not perfect) information in, better information out. And most importantly, a better initial experience for beginners on OGS which is more important than the hallowed rank system (and I’m not even convinced it’s worse for the rank system).
It will make it easier for a 12k who wants to sandbag as a 30k to get a 30k account. However, what is the true rank distribution of sandbaggers? I suspect quite a lot of them are stronger than 12k and mid SDK or even dans. So all they need to do is make a new account and they are sandbagging already in the current system. People stronger than 12k are currently forced to sandbag (using the broad meaning of playing at a rank lower than their true one, rather than with malicious intent losing to lower your rank and do that) because they cannot provide the less garbage information in that they are stronger. I would be curious to know what fraction of active accounts in matchmaking around 12k are real 12ks vs new account 12ks. Are we talking half-half?
I dunno how many testimony from crushed beginners it will take so that OGS change.
Maybe a chess streamer with masses of followers tries to learn Go on OGS, get pissed off at getting crushed and bad mouths the server and the game would do it?
There is currently no compelling reason to change the initial rank system. While there may be problem cases, it remains to be established that the alternatives yield an overall improvement.
All proposals need to at the very least address the following questions:
- Is the new initial rank method more accurate than the current one? The intuitive answer may seem to be “yes”, because who wouldn’t choose their own accurate rank? However, OGS used to let you choose your rank and ended up moving away from that after some experience.
- Will the system remain stable overall? If new players come in at one end of the rank distribution rather than the mean, they could pull everyone to that side, causing a shift in the actual strength marked by kyu and dan labels.
- Is it really an issue for new accounts to have a handful of skewed games? I would say that someone who absolutely cannot take 3 losses in a row (or even 10) is not cut out for the game of Go. Even less if it is a completely new game to them.
- Does the proposal address the issue? It might seem obvious that a 30-kyu is evenly matched against another 30-kyu. But what are the actual odds that the first, second and third game will be even? People are just different beyond the grasp of the ranking system, so there will always be cases like the OP no matter where you start them at.
If you heard of Go and went to OGS to try it, I see 2 mayor possibilities:
- You get paired with a player who plays good enough to not self atari and knows how to score → you lose because you have no clue of the game.
- You get paired with a player who lacks either skill → you play anything but a game of Go.
Yes, I’d rather we get more players than put people off with the current hazing ritual. I guess I’m not cut out for this game (despite being the strongest player on OGS for some years and British Champion) becase I liked playing other beginners at 30k when I started on KGS and wouldn’t have liked getting crushed 10 games in a row.
I do sympathise.
Both with the uncomfortable feeling of not standing a chance at first, and with the desire to play relatable opponents who are also new.
It is simply a fact of life that beginner Go players start at the bottom of the skill ladder, unless they are a genius outlier. Your proposal hinges on the idea that the newcomer expects to have a shot at winning in competition from game one, and maybe some really do. But the expectation is unrealistic. There is no need to cater to it.
For those who (understandably) want an easier time until they are comfortable with competition, there are unranked games where anything goes.
You underestimate the problem with beginners coming to play on OGS. You could go read back the many topics here, mostly emerging from beginners asking why here.
That is much more as 3, was established over 10 in another topic. That’s not losses. That’s crush. Many just want to play from the start fair games with players of their level. Which can even not happen by the system overrating them.
That was debated in length, establishing that the system won’t be affected. The main resistance comes from fear of sandbaggers, but that could be easely avoided.
I won my 2nd game as a 30k beginner on KGS: https://www.gokgs.com/gameArchives.jsp?user=uberdude&year=2005&month=8. Which is not surprising if beginners play other beginners someone has to win.
Wins are important for motivation and positive feedback.
In a large majority of online games, you go play rated and don’t have to be humiliated like on OGS.
Which topic was it? I found this quote from not too long ago:
Also I vaguely remember it being 7 games on average to stabilize. If you can find the source for that or for your number, I will be happy to know.
Which games? I don’t remember exactly what happened when I first played Starcraft or Arimaa or the rather obscure KDice–all strategy games in one way or another. I believe it was humiliation every time, followed by improvement with a few rounds of experience.
Thanks by the link i could easely retrieve the summary i made a few years. I hope you may go through a bit more.
I quote you again because that’s infamous elitist words which just prove how far away you are from the beginners world.
Losing streaks are not a beginner thing. They happen to all of us at all levels. That is not elitist, it is just life.
Are you closer to the beginners world? Take the OP in this thread. He doesn’t seem too bothered by the way of things.
Not “this is outrageous; I quit!” Stereotyping beginners into victims for your cause is not the way to a better OGS.
The OP got his rank modified manually by a sweet admin, thanks to him. That’s not what i call a system working. Besides there are more testimonies, some of them from the forum and the main chat too, reported in the links.
You step away by saying all this stuff doesn’t matter. People can suffer.
Can’t you imagine players of lower ranks coming to discover go through a pleasant experience, playing dozens of fair games thanks to a larger pool of beginners?
When i started go, i made a friend and we played many happy games together. What should i think of your preponderous words, i was not made to be a go player?
Another perspective: you open a beginner class at your school, what is your plan for the first weeks? Bring some sdk to play simultaneous crushing games? Or let them have fun playing together? Which will bring you success?