Suppose "Pat" is playing online. It might be Pat is playing unaided, or with help. Help might be in the form of other players in the same room watching the game, discussing it, and making suggested moves. It might be that Pat is using joseki database software. Or, Pat may be duplicating the game in software which suggests several moves, and Pat picks one.

Many who play online, including admins with whom I used to work, regard this as cheating. But, I have a different view: If Pat is consistent in getting help, I do not have a problem with this.

Those of us who chat online often see advice that people with whom we chat are not always whom they claim to be. Some people may even maintain more than one account on a server, and change persona when they change accounts. I see a parallel in chat and play.

Some users realize their quality of play might differ when playing blitz or playing slow games. So, they reasonably create two accounts: One to use for blitz, and another to use for slower games. Let us speculate further: Suppose "Sandy" is an alcoholic, plays go, and uses only one account. If Sandy sometimes plays while sober, sometimes after drinking, and this indeed results in different playing ability, opponents might perceive inconsistency in Sandy's play. Unaware of Sandy's health problem, they might reasonably conclude that Sandy's account is being shared by two people. Suppose, instead, Sandy makes two accounts: One to play while sober, and the other to play while under the influence. We would then give Sandy credit for being thoughtful. So, Sandy would have two playing personae, or be two "virtual players."

In my thinking, I extend that to those who play online under different circumstances, and use different accounts to match. Suppose Pat uses account "PatAlone" to play games unaided, and "PatJoseki" to play games with help from joseki database software. If Pat were to do this consistently, each opponent PatJoseki encounters would have a consistent experience when playing with PatJoseki. The same would apply with account PatAlone and opponents.

Let me emphasize that, in playing ranked games, using help or not using help must be done consistently with each account. If Pat uses account "PatPlus3" to play while discussing the game with three other players, that must be done each time PatPlus3 plays a ranked game, with the same advisers. If Pat were to use that account to sometimes play alone, and sometimes to play with advisers, the admins could and should treat the PatPlus3 account the same as they would any other account used to play ranked games by more than one person.

What are your thoughts?

It all depends on what your goals are. If you are playing on line you can do whatever you want. If winning is your long term goal you have to accept that the handicap system means your opponent's chance of winning will be about 40% to 60% of the time. Also, winning is determined more by your opponents' mistakes than by your superior play. When I play I want to play good moves, which I certainly often fail to find. I consider each move an opportunity to solve a puzzle. Using help such as a database or having a move suggested just deprives me of an opportunity to solve the puzzle.

So you're saying that as long as they're consistent, it should be okay. But is it really consistent? Different databases have/lack different variations. Are you drinking the same amount of alcohol within the same time interval in relation to starting the game? Are your friends giving you suggestions on every move and to what extent are you discussing? Does your computer go engine have weaknesses (ladders, ko, etc.)?

I think there's still room for skill fluctuation.

It depends on the "environment" you play in:
For example on DGS (turn-based) it's allowed to use joseki-database or books, but it's not allowed to let someone else help you find moves (like other humans or computer programs) ... unless your opponent agrees to some other arrangement.
In official ("real-life") tournaments though it would not be allowed and you would be disqualified.

So in general (if the "rules" allow it), in my opinion, it's not cheating, if your opponent does know about your "aids" and agree with that.

You must realize that online you are not playing a person: you are playing an account. The handicap is based on the current rating of that account and it does not really matter how that rating was achieved. Perhaps it is achieved by using a random number generator to select moves, perhaps it is achieved by consulting an uncle who is a pro. Or, more likely, it is achieved by consistently using external tools to help decide the move. In reality it is probably achieved by just playing the best move the player can think of.

Does it really matter which if these options is used to determine the moves? All that matters is that the option is consistently applied. The rating for the account will reflect the choice and you are playing the account.

_________________
Still officially AGA 5d but I play so irregularly these days that I am probably only 3d or 4d over the board (but hopefully still 5d in terms of knowledge, theory and the ability to contribute).

I agree. The consistency is the key here.

What is important is that when you set out to play a game against a 2d account, this account really plays like a 2d, within whatever the standard fluctuation for the 2d is. It matters not, nor should it matter, if the account is actually played by a irl 5k or an irl 5d.

Problem arise when you set out to play a 2d account and it suddenly starts playing like a 5d or like a 5k. This can ruin your evening, so its not good. But it it always cheating?

To give you an example: There were times when I used to play a lot of my games pretty drunk. It affected my game by about 2-4 stones downward people said (although I think I actually played better, heh.) Most of these games were played unrated, but still... Somebody was expecting a certain level of play from me and they were disapointed. Was that cheating?

What if I played all my games rated, my rank would be 2 stones lower because of the drunk games. And then, occasionally, I would play sober. Cheating?

Now what if insead I sometimes use books and sometimes not. Same rank fluctuations, from your perspective as an opponent might be not really distinguishable from drunk/sober or sick/healthy, or tired/rested, or whatever.

Anyways... this is what I think is interesting to discuss.
My personal opinion is that as long as an account plays more or less consistently, there should be no issue.

The exception to this would be in games that actually count towards some more specific things than ratings - like qualification games, and such.

_________________
- Bantari
______________________________________________
WARNING: This post might contain Opinions!!

There is also the question of how you portray the account. If you raise the rank and try to sell lessons on the back of using a jokes book and a computer AI - certainly that is cheating. Publicly stated that you are a man-machine is not exactly cheating .

_________________
North Lecale

I agree with you, and am pleased to see a KGS admin who can think for himself and does not assume the existing rules are already the epitome of perfection. Honesty is important, more so than consistency (I can beat KGS 5ds and lose to 2ks) but the existing KGS rules encourage dishonesty because those behaviours you and I consider valid are not allowed.

For example on the Go China trip Andrew Kay, Will Brooks and I made an account for collaborative play. We were stronger than our individual ranks. It was a valuable learning exercise and I see no harm in it.

I also once made an account in which I just built solid walls or ponnuki shapes as an experiment. I seem to recall I beat and lost to idiotbot and also beat a human before I was deranked for losing on purpose, which is inaccurate as I didn't lose on purpose, I played badly on purpose and this usually made me lose but not always.

P.S. I also have drunk and blitz accounts.

You are welcome to explore different styles of play with the weak bots, but making them Free games would save you and us a lot of hassle. It's not possible for us to read your mind to tell the difference between losing on purpose and playing badly on purpose, especially when the end result of both affects the ranking scheme.

Shortly before I retired as an admin, I was engaged in a chat with other gold star admins about deranking policy. One of them mentioned an incident in which he/she deranked a "fake-dan" account. Normally, a fake-dan account is created by cheating the rank system: Someone creates more than one account, which are then used to play each other, with the fake-dan-to-be account getting the wins. Or, the fake-dan-to-be gets some friends to lose on purpose. When KGS admins find these, they are deranked.

However, this case was different: A group of three or four, if I recall correctly, created an account. When playing games, they discussed moves. The result was an account that had a rank a few stones stronger than any of the individuals in that group.

At the time this occurred, had I both been an admin and aware of it, I would have argued against deranking. It seems to me, in addition to being a situation which I regard as not cheating, this is a useful way to study go. And, having the group earn a ranking likely provides interesting, and possibly helpful, information to them.

Perhaps this was the incident Uberdude referred to above?

Ranked bots typically create ranked game offers, and it is not obvious to some KGS users how to play a free game vs a ranked bot. So I thought I would add this tip, both to provide instructional help and as a reminder: To make a free-game challenge to a ranked-game offer from a ranked bot, KGS users may turn off their rank by using the "Edit Personal Information" tool under "User" on the main menu.

My assumption when I play on KGS, and I think it's a fair one unless otherwise stated, is that I am playing a person. I know that people's play is inconsistent, so I don't think that consistency is an issue for me. I think it boils down to whether or not I know that my opponent is using outside help. If I don't know, then I am being deceived, whether the rules expressly forbid it or not.

_________________
Patience, grasshopper.

In many areas of life, I am beginning to think that there are two sets of standards:
*Those standards which I apply to myself.
*Those standards which I feel others should adopt.

For me, the bar is more important for the former. If I think about my own standard, I am against using outside help while playing online. That's because there is no benefit I get from it. If I use help and win, what did I accomplish? It's less gratifying, and less fun.

Now regarding what I expect of others... Others can do whatever they feel like. They have their own ideas of what is fun, so whatever floats their boat is fine with me.

And personally, I don't care about consistency. If they make 2k moves, then 5d moves, that's fine with me. My job when I am playing is to play my best, no matter where they play. If the opponent plays poorly and I get an easy win, then fine. If they play a great move I didn't think of, also fine.

Of course, there is more to learn when the opponent plays great moves, so the only real concern is if they play 2k moves when they are supposed to be 5d. I guess that can be a bit disappointing, but players that aren't "cheating" do that, too, from time to time. If the ranking system is good, it should account for that overall (though, you get a boring match once in awhile).

_________________
be immersed

what's the saying?...

The whole point is that he created a special account for that.
If this account plays *consistently* in a certain way, it does *not* affect ratings.
So unless there is more to the story, from either side, it just doesn't make sense what you say.

_________________
- Bantari
______________________________________________
WARNING: This post might contain Opinions!!

My experiment was to find what rank such play would have (would it be stronger than some real humans who were trying their best?), so playing free games would not be successful. This was many years ago when you did actually have 30 kyu human players on KGS, and a decent spread through the 2x kyus too. My play on that account was at a more consistent level than many of my "real" accounts, so any argument that that account 'breaks' the ranking system doesn't hold water. Obviously were I to start playing to my full strength and crushing 28 kyus I should be deranked, but I didn't. In fact players improving 'breaks' the ranking system in that non-active accounts drift upwards in rank over time as the past opponents' ranks goes up. It would be most perverse for KGS to ban players from improving to prevent such effects on the ranking scheme.

I have a blitz account on KGS, ranked 3d. I play bad moves on it on purpose all the time because my purpose with that account is to overplay and have fun fights and win by huge kills. Should I be deranked?



My case was way back in 2007. I don't think we actually got deranked, we only played maybe a dozen games. I can't even remember if we hid the fact we were a team or put a note about that in the info. I know of another Go couple who have a ranked KGS account to play pair go, and they note this in the account's info. They use it for fun and also to practice before they play in pair go tournaments. I think such behaviour should not be banned.

Maybe his mistake was to use IdiotBot, which is something of a special case. Its rank hovers at 30k, so a couple of deliberate losses to it (or deliberate attempts to play badly that lead to losses, or whatever) will raise its rank to 29k. That clearly does affect the rating system.

OK, so the difference between 30k and 29k might not seem like much from way above, but to a beginner trying to get a rank by playing IdiotBot, it will affect hir initial rating. So that's why we monitor IdiotBot's losses and derank whenever we see someone losing to it when clearly stronger.

How so?
If a complete beginner starts playing against this IdiotBot a few times, thiw will raise the bot's rating as well, no?

I'd say that if the assumption is that IdiotBot cannot / should not lose, then just make it resign every game, or make the system count its wins as lose, or whatever. And if the assumption is that IdiotBot has to be 30k or else, then just make its rank constant, liek the pro ranks, or achors, or whatever. If the IdiotBot is part of the regular rating system, like every other playr, I don't see why small fluctuation like that would matter. Eventually it will settle back at its 30k, like every other player with same strength.

As I said, as long as an account plays at a consistent level, it should never affect the overall ratings. If it does, then the system is broken. Or there is lack of understanding by the people who monitor it.

Or again - there is something more which I do not know or understand yet.

For example, if Uberdude played such games against with an account that plays inconsistently at different levels, you have a good case for deranking. But I do not see anybody saying that.

_________________
- Bantari
______________________________________________
WARNING: This post might contain Opinions!!

This is giving me a strong urge to make a new account and close my eyes before I randomly click the mouse.

_________________
Still officially AGA 5d but I play so irregularly these days that I am probably only 3d or 4d over the board (but hopefully still 5d in terms of knowledge, theory and the ability to contribute).

First of all idiotbot is not the weakest possible player, although it mostly places stones randomly it understands atari and extends from it. So if I don't extend from atari I can be weaker and should lose to it and have a lower rank.

Also now I remember more to the story which could be why I got deranked, and that is I used this 30k account to kibitz at full strength. This was to see if people would take my comments for what they said or dismiss them and make ad hominem attacks based on my rank. Now there is some legitimate appeal to authority/rank with "my judgement is this result is good for black" type comments, but for something like "that peep was bad because you could have cut" the rank doesn't matter. Too often dan players poo-poo comments of kyu players because they are kyu players and not because they are wrong.

So perhaps what got me deranked was inconsistency between my playing and kibitz levels (though the admins gave the reason as "losing on purpose"). Maybe KGS would be a better place if this was applied more widely to stronger players who make dumb kibitz.