The Ethics of Playing Provisional Players

General Go Discussion
41

I’ll be honest, this is a good point that I had not considered, because I had never noticed such a thing. I came here after reaching 3k in DGS which is purely correspondence and has no other mode, so it was my impression that OGS was not correspondence friendly. I guess it is all a matter of what someone expects from the game, so thank you for pointing this out :slight_smile:

I am genuinely not sure if playing correspondence is a good thing for most beginners.

On the contrary it is imho one of the best ways to test a LOT of moves and eventually distinguish good shape from bad shape. In a fast game, people do not have time for variations and, let’s face it, a lot of people do not review their games afterwards and just head for the next game.

On the other hand, a correspondence game is practically a game with an ongoing review happening, so in a live game you’d do mistakes and never review them or forget them by the time the game ends, in a correspondence game you can learn and review at the same time. Here is an example:

And that is just a teaching game where in order to pick the variations which you want to present, you have to play them out and see if they are useful and to the point. But in a real game, you actually go through variations for EACH move and then pick the best. :slight_smile:

I attribute my fast improvement mostly to this tactic. I am not particularly smart or anything. All I did was challenge stronger players, show how they played and tested all variations for each my moves while playing. Add a couple of books I read and a few videos and I got from 12k to 5k in six months.

there is usually no continuity between the moves

This is easily solved, at least in DGS where there is no analysis and you have to download the sgf. I usually just opened it with cgoban (KGS’s old client) and made all my moves and variations there, offline. This way there was continuity and a full review left in the sgf afterwards.

Sadly I no longer do that, but it is an efficient way of learning.

This especially makes reviewing the game challenging, since you probably won’t remember what you were thinking when making the moves.

True, but as I said, you do not have to review a correspondence game. You already did that while playing.

the temptation to consult an AI engine

I’ve never done that. AI moves are usually counter-intuitive and require AI strength to execute correctly. Even at 3k, I never look at the AI analysis because I do not think that I have what it takes to gain much from it. So:

especially for beginners, who are often confused and mystified by the game. After all, consulting an AI doesn’t feel that much different that just analyzing future variations on your own.

So, a beginner checking AI moves and thinking them as future variations of their own, is veeeery optimistic :slight_smile:

if you get attached to winning.

Now that you mentioned it, I should point out that my aforementioned system involved getting attached to losing, since I kept challenging stronger opponents on even games. I lost most of those games and I have much fun and learned a lot.

I think that everyone knows that having a machine play for you is cheating because botting is against the rules in every game on the internet.

There are several reasons why I usually don’t play corr, and the temptation to consult AIs is one of them.

Resist the urge and join us to the correspondence side … you are allowed coffee and cookies … heck, you can even cook while playing or take a nap and return to the game refresed :smiley:

42

That clause makes no sense to me. You can never predict the result of a game in advance.

1 Like
43

You know, this is not the first time I see you take a quote out of context and misrepresent it. It’s starting to feel like you are doing this on purpose.

It’s especially annoying since people coming here and reading your reply are probably going to be too lazy to go back and check that you didn’t misrepresent me, and they’re just going to assume that you’re arguing both in good faith and without a gross misunderstanding of what I said.

If you’re not doing it on purpose, you could at least read the very post you’re replying to, to check if maybe your objection has already been addressed:

Of course you can then disagree with this, but you literally just repeated a point that I had anticipated and discussed in my post immediately before the part you quoted.

Oh, and the complete sentence you only partially quoted is:

Which to me feels already much clearer than the part you quoted out of context.

1 Like
44

I don’t really understand what you mean, but I meant “continuity in time”. At any point you’re just seeing a board in a vacuum, and especially if you’re a beginner it’s difficult to remember what happened before to lead to the current configuration.
 So unless you do a backward review at every move, which is counter-intuitive and boring, when your mistakes are punished you have no way to realize and learn from it.

Apart from that, generally I agree that, if you do it “the right way”, correspondence games are possibly the best way to learn, especially for older people (because another factor is the enjoyment, and younger beginners usually enjoy playing fast more, and they’re also able to learn from live fast games more than an older person).

But I already thought this before writing the previous reply to you, and my point is that beginners are not necessarily likely to play correspondence “the right way” if you just let them be.

Well, keep in mind that many beginners are literal kids, so you can’t make too many assumptions about what they do or don’t know about the world, including the world of multiplayer gaming.

Similarly to how beginners often come here without knowing anything about kyu ranks, what a rating system even is, or how scoring a game works, they also often don’t know anything about what “sportsmanship” is, especially when applied to a game like Go.

I think without supervision, many beginners stumble upon extremely unsportsman strategies, and they just think of them as good strategies to win.

One way to justify this is that, after all, if it was against the rules, surely it would be impossible to do on the computer program, right? The computer program prevents you from playing self-capturing moves or recapturing a ko, so why would it allow to do other things that are against the rules?

Though this specific justification doesn’t apply to botting, since the program internally doesn’t know if you’re cheating outside of it (although I wonder if all kids understand that).

Go is a game that, especially in the West, tends to attract the most cerebral personalities. When kids with a cerebral personality discover an apparent loophole in the rules of a game, that allows them to win against every opponent who doesn’t know how to deal with that, many of them just keep doing it. “It’s all in the game” :laughing:

Much of this, I believe, comes from the pride of having figured out an out-of-the-box strategy to win. And this does apply to botting. Even if the kid isn’t winning by finding the single moves autonomously, they’ve found an out-of-the-box strategy to win. “I’m so smart, I can look at the moves the bot plays to figure out good moves that I wouldn’t figure out otherwise. I bet no one’s ever thought of that, I’ll be the best player in the world in no time”

While they might feel on some level that this is “wrong”, it’s easy for that feeling to be overcome by the pride in their lateral thinking solution.

And when they realize they’re not in fact the first ones to ever have thought of that, they still feel like it’s a good strategy that they figured out by themselves. Anyone who doesn’t do it is just a sucker. Nash Equilibrium, right? (ok, a kid wouldn’t know that term, but I genuinely think they would instinctively understand the concept, at least the kind of kids I’m talking about – I certainly did as a semi-autistic kid)

By the way, funnily enough I believe one can find their way to AI consulting regardless of whether they care about winning or not.

For example, my position on botting was mostly indifferent, before I realized the negative impact it can have on a system such as OGS, because I mostly care about learning in my Go experience.

If there weren’t rules against it, and if I didn’t care about the unwritten contract between me and my opponent, when playing (unranked) correspondence games I might consult the AI without thinking I’m doing anything bad. As you said, in a correspondence game you essentially review the game every move, and consulting the AI is part of my process of reviewing the game (although it doesn’t always work, but funnily enough, at my level it often does).

So if I’m playing to learn and not to win, and consulting the AI helps me learn more, it just kinda makes sense.

In fact, I’ve had the idea that people could agree to do that in an unranked game, but since I don’t play correspondence, I can’t do it myself :laughing:

Of course I would try to avoid doing this in every game for learning purposes, since it’s kinda like learning to swim with… uh, those inflatable things they put on children’s arms, whatever they’re called.

So my point is just that telling beginners to play ladders, aka correspondence, is not a “low-maintenance” strategy, because I think correspondence games are only conducive to good learning and good behaviour if done “the right way”, and “the right way” is also probably too boring for many young beginners.

1 Like
45

Yes, it is true that many beginners seem to have no innate sense of sportsmanship and very often do not read the OGS documentation. Some players score cheat in their first few games, probably due to frustration from losing. However, a much larger number play legally for awhile and then score cheat.

When I looked back at their histories, I often found that someone had successfully score cheated against them recently, they had not reported it, and the game was not annulled. This was a very common pattern, and I concluded from it that they somehow felt justified in their cheating because they had been cheated against. It is similar to the pattern of vandalism, which typically invites more vandalism if it is not dealt with.

The most extreme form of the “if it isn’t prohibited, it’s legal” philosophy was the player who argued that his score cheating was or should be legal because the system permitted it. Of course, OGS does prohibit cheating, but his argument was that because the system permitted his action, then he wasn’t cheating. I don’t know whether he was sincere or just being a “jail-house lawyer.” I made the obvious explanations and issued a warning. I don’t remember whether or not he continued to cheat.

3 Likes
46

I apologise if you were annoyed by my post. Annoying you was not my intention. I thought I was expressing constructive criticism of your proposal, by highlighting the clause that looked the most wrong to me and explaining why I thought it looked wrong.

In addition, my post was directed mostly to you, and intended as criticism. It was not intended as an argument against your proposal to be read by other people in order to influence them in their opinion about your proposal.

I quoted that sentence because that was the sentence that felt the most wrong to me. I can quote a larger chunk if you want, but that won’t make me agree with it any more.

Yes, I did read your post before commenting. Yes, I strongly believe that adding a clause in the TOS about allowing ranked teaching games only if the teacher wins is a very bad idea. It doesn’t matter whether that clause is out of context or not, it’s still a bad clause in my opinion.

Sorry if you felt that your post was misrepresented because I only quoted and commented one clause out of it. I commented that clause because that was the clause I disagreed the most with.

Now, I can assure you that I’m commenting in good faith, if you need me to explicitly state it. I can’t assure you that I hadn’t misunderstood your post, but if I had misunderstood it, then I must still be misunderstanding it now because my understanding has not changed. To be honest, I don’t really see what’s misunderstandable about that sentence, “Ranked teaching games are allowed only as long as the teacher wins and it’s clear that the teacher is the stronger player, so that the ranking result is “correct”.” It sounds very clear and very unambiguous to me, but feel free to re-explain it if you feel that I have misunderstood it.

I’ll finish this post by taking, again, a quote out of context, because I agree with that quote and it sums up most of the things I dislike about your proposal:

1 Like
47

Well, I kinda feel that you still haven’t really explained why you think the clause is bad.

In the vein of your reply, here’s my own piece of modern art:

(because if the two players have a large difference in rank, the most likely outcome is by far that the stronger player will win, so it won’t upset the rankings)

With the meditative reflection generated by putting together this masterpiece, I realize that the crux of the matter is that imho the TOS need to be vague to avoid bureaucratic loopholes and allow good common sense to prevail, but at the same time they need to constitute guidelines, which means there needs to be a good balance of vagueness and clarity.

There are two use cases for the TOS:

  1. Moderators need to be able to point to them and say “this is why we’re banning you”.

  2. Users need to be able to consult them to understand what’s allowed and what isn’t.

If it’s true that

  • the teacher is much stronger than the student

then it’s true that

  • the most likely outcome of a normal ranked game between them would by far be the teacher winning

which also means

  • it’s not a big deal if the strong teacher wins a ranked game against the weak student, because it’s the expected outcome

which means that reasonably

  • ranked teaching games are not a big deal as long as the teacher wins and they’re clearly stronger

and thus

  • ranked teaching games where the teacher is much stronger and wins (=RTGWTTIMSAW) are reasonably allowed

which means, that, IF

  • teaching games are considered a form of “outside assistance”

AND

  • outside assistance is prohibited as a general rule

AND

  • the TOS need to inform users of what’s allowed and what isn’t

THEN

  • RTGWTTIMSAW need to be pointed out as an exception to that rule

To which I see two (2) main counter-points:

  1. as GaJ and Feijoa discussed, there may be no need to consider a “teaching game” as “outside assistance”.

  2. People may disagree with my assessment that there does need to be a balance of clarity and vagueness in the TOS, or at least on the measure of that balance.

  3. I don’t really give a flying broccoli how the TOS are worded, as long as I don’t get banned for trying to help beginners in a reasonable way. I was only making the proposal since GaJ said (1) that they believe RTGWTTIMSAW are not too bad, (2) that the TOS needed to be amended to reflect that, (3) that maybe we should debate how to amend the TOS.

  4. The clause feels out of style because it’s much more bureaucratic-sounding than the rest of the TOS. If the TOS were a creative essay and included my clause, I would r8 them 1/3 2 legal 4 me :microphone:

We all thought the master piece of modern art ended half-way through this reply, turns out the rest of it was just a meta piece of modern art. And this sentence is meta-meta-modern-art.

1 Like
48

I agree that you shouldn’t get banned for helping a weaker player.

I don’t believe there is going to be much of a consensus about what “in a reasonable way” means, though. (But I don’t think that matters, if the game is unranked and both players are in agreement.)

I strongly disagree with your premise that the “expected” outcome of a teaching game is a win by the stronger player, especially since teaching games often come with handicap, either in the form of handicap stones, or in the form of the teacher restraining themselves, or both. If the game is played with handicap stones, and the teacher forbids themselves from outright killing a group, and the teacher tries to aim for jigo, and the teacher additionally gives advice to the weaker player, then a win by the weaker played doesn’t seem that unlikely. And I think such a game would be very reasonable and it would be sad if such a game was forbidden on OGS.

I’m about to quote one sentence from @espoojaram 's post. Disclaimer: I do not intend to misrepresent @espoojaram 's post by giving a quote out of context; please refer to @espoojaram’s post for full context of the quote.

Because it’s a clause that literally says that teaching games are prohibited by the rules if the teacher doesn’t end up winning. Since the teacher cannot guarantee that they will win, it effectively prohibits all teaching games. And I think that’s a very bad thing. Teaching games are good and I don’t want them to be forbidden. You said yourself that you didn’t care about the exact wording of the TOS and that what was important was that you could give teaching games without being banned. Or at least that’s what I understood from your post. Well, in that case, then this clause is not helping. If you want to be able to play teaching game, then don’t include a clause in the TOS that says that teaching games are forbidden.

1 Like
49

BTW, I completely support giving teaching games, and I thought the ethical discussion at the beginning of this thread was very interesting.

However, I did not like your proposed amendment to the TOS, which is why I expressed criticism of that. I did not realise it might sound like an attack and I’m sorry that it did.

Also, I don’t understand the point of all the quotes you juxtaposed in your last post, and I don’t understand the acronym “RTGWTTIMSAW” that you used.

1 Like
50

The handicap thing is a very good point. There is no clear expected outcome to a handicap game if the handicap is anywhere near correct. Including that in the TOS would make it even more convoluted, so I feel already convinced that it’s probably better to avoid it.

I mean, I could reply to your other points, since I don’t agree with some of them, but it’s irrelevant since I now agree with your conclusion.

The juxtaposed quotes were supposed to be kinda like juxtaposed cuts in an “artsy” movie, you either instinctively “get” what they’re supposed to convey, or you don’t. Art doesn’t need to be good to be art :laughing:

RTGWTTIMSAW (=Ranked Teaching Games Where The Teacher Is Much Stronger And Wins)

(also, Antimetabole)

2 Likes
51

So, to help out @GreenAsJade, is your opinion that the TOS should be written in a way that allows any ranked teaching game?

52

FWIW, I really liked both the art and the content/intent of the artsy post.

I think that a “ranked game where the teachers is much stronger and wins” is precisely the exception that we want to make to the rule “no assistance in ranked games”, even though it is a cumbersome term.

I think that “RGWTTIMSAW are allowed” is a case that might warrant a clause in the TOS, because it otherwise does cause the ethical dilemma that started this post.

Note that in practice, if the teacher accidentally loses one of these (which is possible when concentrating on teaching, and especially in a handicap situation), the expected action would be “request an annullment”, IMO.

I will also note that there is one other exception that we want to make, which is “casual” assistance from the opponent, such as “hey did you mean to play self atari” or “would you like an undo? (*)”.

I think that “casual opponent assistant” is something that just does’t come up as a problem in practice, and is not ammenable to a legalistic description in TOS anyhow.

(*) although this particular question is not especially recommended .

1 Like
53

Are you suggesting that the game can be ranked, but that the stronger player should request for the game to be annulled if they lose?

What is the logic behind this? If you’re not willing to accept a ranked loss, then play an unranked game.

Playing a ranked game with the intention to keep it ranked in case of a win and annulling it in case of a loss? That kinda sounds like the worst possible form of cheating, and, if done by a stronger player a beginner/newcomer, then it sounds like an abuse of trust/authority that might easily convince the newcomer not to stay on OGS, once they realise they’ve been abused.

4 Likes
54

This is an interesting discussion that I have been trying to follow.
I can see the points on both sides and I think there is a lack of clarity in some of the terms and concepts being used. Also I think there is a bit of a conflation of objectives with the ranked teaching game business.

It seems to me that there a desire to help beginners - great! This desire has two parts

  1. teach them
  2. get them an appropriate rank quickly

The problem I think we have is the desire to do both these things together - i.e. play ranked games with beginners and give them some kind of instruction and also beat them.

I’m not convinced though that the solution is to permit ranked teaching games. I think there are different views of what is meant by “teaching game”. I’m more leaning towards @ArsenLapin1 in that I think there are range of things that might be called teaching games and it’s far from the case that these will or should inevitably be ones where the teacher wins. Or in the jargon I don’t think that RGWTTIMSAW is a useful sub category of teaching games that merits special treatment in the TOS.

I think the problems of teaching beginners should be considered separately from the problem of getting an appropriate rank as quickly and painlessly as possible.

7 Likes
55

I’m firmly against the idea of ranked teaching games. In a teaching game, the primary objective is to collaborate on learning and the result is a secondary concern (and often the teacher might try to lose, in order to give amply opportunities for learning and to avoid discouraging the student). Hence, it does not make sense to include consideration of such games into the ranking system, since the systems aims to measure the relative strengths of players based on outcomes from fairly contested games.

I find the idea that the teacher should always win a ranked teaching game even worse. I understand the concern on the one hand that a teacher giving advice to their student would artificially boost the chances of the student winning, however, I think the best way to handle such contests is to altogether avoid using those for measuring the strengths of the players. If we instead say that the teacher must always win such teaching games, then that might be assuming too much that the teacher should always win. It’s conceivable that some “teaching game” might be set up between two players that are only a few stones apart, and it’s not impossible for the weaker player to actually win, even in a regular, fairly contested game. It’s also possible that the relative ranks of the players are not even so accurate, giving the false impression that one should always win. It especially does not make sense to assume a player’s rank or that they should lose a game, when they are still provisionally ranked.

I really don’t think it makes any sense to allow games with predetermined outcomes to be ranked and fed into the rating system. Why would we need teaching games to be ranked in the first place?

I guess the original premise of this thread was a bit different and seemed to be motivated by a perceived problem that many beginners started with too high of rating and thus need to lose a bunch of games to stabilize their rating. It seems that @espoojaram’s original proposal basically amounted to offering to take those easy wins in return for offering teaching during those games. However, I think that creates muddled situation involving, potentially incorrect assumptions about player’s rank and unfairly contest games that are being ranked.

Regarding making the TOS more explicit about all of these sorts of issues, particularly clarifying the expectations of conduct within ranked vs unranked games (e.g., it seems that we don’t even clarify that sandbagging or rank manipulation is not allowed), I do think that we could improve those. Actually, the TOS is a mix of both rules about general site/game conduct and a ton of the typical legal disclaimers, which might discourage people from even looking at the page (by assuming it’s just the typical legal stuff). Maybe it makes sense to put general rules in a separate page altogether.

9 Likes
56

I don’t have much to add to the last several posts. I don’t see any value in the proposal except to overcome a dubious legalistic interpretation that no moderator to my knowledge has ever imagined acting on. A teacher teaching is “outside assistance”? Really? The mods are more sensible than that, I hope. It could be argued, anyway, that “outside” refers to outside of the game players.

I am now and have before given correspondence teaching games in which I coach the opponent throughout the game, explaining my moves, answering questions, and sometimes even suggesting takebacks or pointing out specific lines of play. I consider this a real teaching game, rather than just giving a game review, because players are more engaged during the game than afterwards. This is “risky” from the standpoint of winning, but who cares if it is unranked. In addition an opponents might be sandbagging, or might improve significantly in the course of the game, exceeding the strength difference I expect before accepting a game. And then, there are occasional blunders. What teacher would want to play like this when, if they lost, they would suddenly become a violator (based on what I consider a false interpretation of the policy)?

2 Likes
57

I have been persuaded by yebellz’z post that it’s not a good idea.

But I want to explain the logic nonetheless.

The premise is that the teacher is much stronger. The goal is to do teaching and help the beginner get the correct rank.

=> If the teacher is much stronger, then clearly from a rank perspective, they are expected to win. If they don’t it’s not because the beginner is suprisingly strong, it’s a result of the teaching: not a valid input to ranking. Therefore it should be annulled in that situation.

yebellz points out that having a goal of “helping a beginner get a correct rank by playing a teaching game that they lose” is not the right way to approach the problem of getting a correct rank. I agree with that.

1 Like
58
As a general principle, do you believe ranked teaching games:
  • should be allowed
  • should not be allowed at all
  • should be allowed if they satisfy some conditions

0 voters

Clarificaton: I’m not asking “what should the TOS say”, I’m asking how do you believe players and moderators should act.

Of course, speak out if your position is not clearly covered by this poll, which seems to be usually the case :laughing:

59

I don’t understand why teaching games should be mixed with normal games at all.

1 Like
60

who is going to check which games are “Teaching” and which are “Normal with dialogue” ?

3 Likes