Every few months my KGS rank becomes something like [1d?], so I just press the automatch button and play 2-3 games. I didn’t feel that finding opponents was difficult.
It seems a long bow to draw to say that because you think that [?] on KGS has caused you difficulty getting games there, therefore on OGS it is a stigma!
I think much better evidence that it is a stigma is that there are games called “No ?” here on OGS.
And ongoing requests in the forum along the lines of “how do I avoid playing against [?] people”.
Right now, the OGS answer is “we discourage you from avoiding it, because those people need ranked games”.
I’m not comfortable with that personally, because I’ve seen the effect on SDKs of having to play [?]s who are basically just barely curious about Go, here for their first game.
It’s not pretty.
I’m not really sure what connection this has with sandbagging, kind of lost the thread there.
1 Like
Some players haven’t played ranked games for a long time, but play unranked games on a regular basis. Among those players, some of them are stronger than their rank. Let’s call them “pseudo-sandbaggers”: technically they are not sandbaggers since they don’t lose games intentionally, however the effect of playing against pseudo-sandbaggers may be unpleasant. Hence the proposal to display a [?] (like [5k?]) if they haven’t played a ranked game for more than 1 year.
The objections to this are:
- if a [?] is a stigma, then getting a [?] is unpleasant if your rank reflects your actual strength. A 5k who hasn’t played a ranked game for 1 year and becomes [5k?] may still have a 5k level.
- According to Conrad Melville, the problem of pseudo-sandbagging is tiny compared to real sandbagging.
Of course an easy way to avoid the [?] is to play a ranked game every year. I don’t find that difficult, that’s more or less what I do on KGS every 3 months, but people who dislike live games or who have rank anxiety may think otherwise.
2 Likes
Theres also some legit cases where i think ranks like “[5k?]” after x amount of time might be good solution
-
OGS not being the primary server where they play. People who mostly play on other servers or in real life might have improved considerably after their last ranked game here. Of course we can never know how much someone plays outside ogs, but the question mark after their rank would indicate that they might have gotten stronger (or weaker) since their last game here
-
Comeback accounts, people who havent been here at all since the server still used Elo-based ranking system, and then find their rank vastly inaccurate due the rating system updates when coming back to the server
-
So-called “heliumbagging”, these are mostly people who have once gained very high rank due some rating system updates, lucky timeout wins, one good winning streak etc., and are now unwilling to risk their rank dropping down so they either play unranked games or use an alt account to play their games. I’ve chatted with many ppl who have dan-level rank which they themself feel being too generous, but are still very happy to have the “d” after their name. Also many go teachers who mainly use ogs to play with their students will count as heliumbaggers, as they want to keep their rank high but do not play seriously in order to win as much as possible
-
Users who feel stressed about games being ranked while they just want to play casual games and have fun on an internet server. They might have played some ranked games in the past but lost because the added pressure they’ve felt, so they have switched into playing unranked games in order to enjoy their games more.
And yeah, of course there are those “real sandbaggers” who intentionally resign on games where they are winning because they want to have their rank too low for the purpose of getting paired with weaker players they can then crush. The [5k?] style of rank wouldnt affect those, but i feel like those people are only a tiny minority in the ogs’s player pool.
Also, there are some cases where people simply resign out of boredom. They might think something along the lines of “ughh im winning by 100 points but my opponent isnt resigning, i’ll rather just end the game here myself instead wasting half an hour on playing meaningless end-game because this isnt fun anymore…”. Quite often they might feel that their opponent is the one who is stalling, instead thinking themself as sandbagging.
1 Like
I think though, and I’m not sure if it’s being overlooked, but what was the addition of something like [5k?] supposed to be addressing in the first place if not “real” sandbaggers?
If you’re intentionally throwing games to lower your rank, the question mark makes no difference.
If you’re intentionally not playing ranked games so as to not have an accurate rank so you appear as 15kyu when you’re actually 5kyu, or whichever, and all you have to do to maintain after the update is play one ranked game (a year?) per period, I’m not sure the update helps the problem.
Instead really I think we see a lot of places that there is potential impact to players other than intentional sandbaggers where we’re not necessarily sure if the change will have positive or negative net effect to these players, because in some cases it’s a community perception that we don’t have control over. That is unless you widely publicise, banner the update and explain it to people, which generally doesn’t have save for really big updates.
It doesn’t change anything about how these players would get reported either, maybe some people believe that the new question mark means they shouldn’t trust the rank, but being annoyed after losing a game probably means if you’re going to report you’re going to report, question mark or no question mark.
2 Likes
If you’d like a concrete example to compare and contrast, supposing someone has the idea to reintroduce the flag on a players profile that they have recently timed out of a correspondence game
I think very similar arguments apply as to a lack of control over how the community uses or stigmatises such a flag.
Indeed in the past people have suggested that they found it harder to get games and wanted the flag to be manually removed either because of bugs or because potentially having to wait several months to end a correspondence game to naturally get rid of it was too long.
Reintroducing such a thing, it may sound like it’s helping players who don’t want to be matched with someone who times out, but it’s not clear it’s not just some cosmetic thing, not really addressing any actual issue and instead potentially creating a new stigmatic issue among the community, that we can’t control. (We the people voting or suggesting to implement such a change)
It’s worth noting that that “recently timed out” flag was buggy and stuck around even after consistent good behavior.
I think there’s a big practical difference between a “stigma” that you can get rid of (like a ?) and one you can’t get rid of (like a criminal record)
1 Like
If you mean sticks around like the game history, for concreteness,
do you also suggest or not to include the question marks in rating in the game history table - since the table there also includes historical ranks?
Because in both the ? and the timeouts, you can’t really change the game history. If you timeout of games people can see it or not, if you beat stronger players and lose to weaker ones people, or play only unranked, people can see it.
It’s a just a question as to whether you want to condense it into something easily stigmatisable or not (not even considering some buggy implementations)
2 Likes
no i don’t care about the game history
Ok.
So I’m not sure where we are then.
I’m saying it’s much easier to say things like (I won’t quote the post directly)
I do not want to play ? players when I use the “restrict rank” button…why is there no way to prevent this!?
or if you do not want to encounter such and such, don’t play with ? players, than it is to try to describe some more complicated pattern in rank or game history that potentially indicates something like sandbagging.
Hence attaching a single symbol to such a thing, makes it much more easily a stigmatisable symbol.
Yet
you call it a stigma you can get rid of (if you play the right numbers of games and get an accurate rank), however that doesn’t change your game history, so you can’t really get rid of it
and maybe the game history could even be stigmatised as such by some adamant to avoid potential sandbaggers.
1 Like
Here’s the thing- if you’re willing to scroll through someone’s game history, it doesn’t matter whether it’s a [?] or a [7k]. You can figure out much more about their skill by looking at their win record
So let’s just say I don’t think that column of the game history is very important in the context of this thread and I’ll let you decide what is most consistent for the rank column 
As for the highly visible [?] that follows a player around the site (including on the matchmaking page) that one is easy to change by playing ranked games. And that’s what I meant when I said it is easily removable.
I mean I can agree and won’t be pedantic about any more than maybe I am being. People will and can put in a lot of effort into selecting their opponents and no symbol will change that - maybe it’ll speed things up maybe it won’t.
But for the sake of clarity maybe we can answer the following regarding introducing such a thing as [5k?]:
- who is likely to be most affected by this change? (e.g active users, inactive users, users never to return, sandbaggers, unranked players - whatever groups of players you can think of etc)
- who is this change likely to help the most? (similar ideas of groupings I suppose)
- does this solve or go some way to solving an existing problem?
I’m surprised that go players, most much stronger than me, do not apply their “reading” skills to a problem like this. If OGS makes a move, such as instituting [rank + ?], the quasi- (not pseudo-) sandbagger will make a response. The quasi-sandbagger has several variations to choose from:
-
“Pass” (i.e., do nothing). He knows he will still get games (perhaps at a slower pace) because some players don’t care (they will play against a 9 handi in a 9x9 game), or because they want the challenge of playing someone stronger in an unranked game.
-
Play the required ranked games against players who are much weaker (vetted by their history).
-
Play the required ranked games against anyone and lose. This could be legitimate against stronger players, or faked. The latter takes some effort to do artfully, but when so done is practically impossible to prove. The result of this would be that the [rank + ?] rule has actually made the situation worse. The 5k? quasi-sandbagger has now become a 6k quasi-sandbagger, thereby increasing the range of the putative damage he can inflict.
Meanwhile, players who have topped out in their rank—the majority in the issue under discussion, IMHO—are unduly and rather meanly stigmatized. These players are NOT quasi-sandbaggers. They have chosen to play unranked because they believe they have no further room for improvement and would like to hold on to the rank they have, rather than risk losing rank by going up against the army of undetectable alt sandbaggers who infest OGS. Or, if they are stronger players, of losing rank by going up against the other army—of botters.
This is all silliness. The best solution is to check the game history.
1 Like
Aren’t you exaggerating? I’ve played about 2000 games on OGS. Except for one game, I didn’t see anything suspicious in my opponents’ moves or game history.
1 Like
exaggerating may be included in thoughts of those who don’t play ranked
Not about the alt sandbaggers; I dealt with them virtually every day when moderating (when they were still detectable).
As for botters, there were, and I have no reason to think there are not still, a steady stream of botting reports. I never dealt with the high-level reports because I was/am too weak to do so. I did handle a few botter reports from weak DDKs and found that in every case the loss resulted from very obvious strategic or tactical mistakes, not from botting. However, a mod who handled many high-level botter reports told me there were a lot of legitimate cases and that the phenomenon is under-reported, just as score cheating is hugely under-reported. I find this believable because botting is much less obvious and harder to prove than score cheating; in many cases the victim may not even realize it.
Could it simply be that this number if sufficiently large for the few mods to feel overwhelmed by reports, yet sufficiently small to be very rare for a given user in average?
For what it’s worth, my experience matches @jlt (despite the level difference). I think I faced this issue maybe once or twice at most, which is nothing.
I may be wrong but I find it hard to believe that one could legitimately be afraid to unfairly lose their rank due to “an army of sandbaggers”.
3 Likes
Many players who have complained in the Forums, most notably snakesss, have had terrible struggles achieving their rank in the face of sandbaggers.
It seems strange to me to argue that normal alt sandbagging is insignificant (“sufficiently small to be very rare for a given user”), while the quasi-sandbagging being discussed here is a terrible problem needing a radical solution. That doesn’t make sense. Moreover, alt sandbagging is mostly undetectable today.
As in the case of botting, many players are probably victimized by sandbaggers without even knowing it.
Those who do have big problems probably do something wrong, but they do have problems.
in perfect system even those who do something wrong should not have problems.