Life In 19x19

Cynosure wrote:o.o lots of responses! I'm not gonna go undo willy-nilly, but I do want to have a bot interface/prog/whatever that accepts undos for experimental reasons, also. It's not just because I lose games for misclicks, etc. I've since put on the anti-slip, I didn't know that existed. Should help in regular people games too

More or less Ed wrote what I was going to write.

Let's keep it in mind: we are talking about playing a bot. Your finger slipped and you have to resign? So what? Play another against the bot if you feel like showing you are better than the AI. Or play against a human, it's slightly more fun.

The utter worst thing that can happen is that your rank gets a minor shake. So what...?

Well, they're free games, and as I mentioned I'm not necessarily playing to outright win. Bots can teach. I asked for bots/interfaces that allow undos, and the only answer I got so far was drago.

Cynosure wrote:Well, they're free games, and as I mentioned I'm not necessarily playing to outright win. Bots can teach. I asked for bots/interfaces that allow undos, and the only answer I got so far was drago.

All the applications allow undo - many faces, drago, leela, etc. You specifically mentioned KGS, however, and that's where there is a problem, because its important to the integrity of the KGS rating system that the bots don't get abused. Bot abuse is impossible to avoid if you permit undo's (for rated games).

Cynosure wrote:... Bots can teach. ....

I fully disagree with this. Even in 9x9 learning from bots is close to impossible. In 19x19 bots play like, well, bots.

RBerenguel wrote:

Cynosure wrote:... Bots can teach. ....

I fully disagree with this. Even in 9x9 learning from bots is close to impossible. In 19x19 bots play like, well, bots.

If the bot is strong enough relative to you to punish your misplays, you can learn something about fighting no? Even if it's just, that move doesn't work here.

RBerenguel wrote:

Cynosure wrote:... Bots can teach. ....

I fully disagree with this. Even in 9x9 learning from bots is close to impossible. In 19x19 bots play like, well, bots.

Eh? I played 100s of bot games when I was a beginner, and I credit MFOGO and Gnugo with getting me from 20k to 10k in a few months. Not the only way, sure, but it was a way that worked for me.

You will learn plenty from bots, albeit mostly tactical stuff, but that's OK. Tactics are important, perhaps more important than anything else (a debate for a different topic tho).

RBerenguel wrote:

Cynosure wrote:... Bots can teach. ....

I fully disagree with this. Even in 9x9 learning from bots is close to impossible. In 19x19 bots play like, well, bots.

Bots are a tool for improvement, rather than a teacher. They cannot explain ideas to you, but if you played 100 games against a strong bot, I suspect that most people would end up stronger than they were before.

They cannot explain ideas to you

If only we can wait long enough...

I think some chess programs, even a decade ago,
already provided "simple" instructions, tips, or ideas, as an option.

As Stephen Hawking recently said, there is no physical law
that forbids any particular arrangements of atoms such that
the resulting structure is vastly superior to the human brain.
Some day, a bot may be able to give the top pros 2, 3, or even 6
or more stones. They may also be able to teach Go better than
any humans.

Obviously, we're not quite there yet.

All engines will "suggest" a move, but all they are doing is what they normally do: calculate a strong move. There is no explanation of the position, or why it is a good move.

Some of the best commercial chess applications had some form of tips and instruction, but this seemed to be designed into the interface by the programmers, rather than come from the engine directly. Although, you might say that this is not an important distinction.

EdLee wrote:Uber, I suspect we were both typing our last posts at the same time,
and I ninja'd you. I believe my last post already made my position clear.
If there are parts you and I disagree, that's OK.

Mike Novack wrote:It might be better to step back just one bit.

Mike, and daal, started with a nice, civil tone.

And then things suddenly changed... Or did they ?

Uber, HH: about your posts 7, and 10, I'm sorry I cannot read your tone of voice over the internet. To avoid (more) misunderstanding, what was your tone in those questions ? Would you say your tone was leaning toward:

• positive, supportive, inquisitive (nice) ;
• completely neutral (just the facts, ma'am) ;
• negative, attacking (nasty)

?

Mostly very surprised. You gave an answer that was, IMO, very extreme. No shades of grey, the world is black and white, there is no notion of undo in go. I did not think you were the type to hold a world view as black and white as you had written it, so I posited a simple example of a situation in which, in my opinion, it is very obvious that the notion of undo is very real, and very reasonable. An opportunity for you to clarify whether you feel the world is really as black and white as you had posited, or not.

Now as to the original question, I think it would be good if bot writers wrote their bots in such a way that it would always accept a single undo for any and every move.

HermanHiddema wrote:Now as to the original question, I think it would be good if bot writers wrote their bots in such a way that it would always accept a single undo for any and every move.

Assuming that you're referring to KGS bots here, should I conclude from that that you also believe every (rated) KGS game should also allow a single undo per move? i.e. for human-human games too?

quantumf wrote:

HermanHiddema wrote:Now as to the original question, I think it would be good if bot writers wrote their bots in such a way that it would always accept a single undo for any and every move.

Assuming that you're referring to KGS bots here, should I conclude from that that you also believe every (rated) KGS game should also allow a single undo per move? i.e. for human-human games too?

if you're playing for rating points, then that's a completely different situation. That's like playing a tournament game, and there are no undos in a tournament.

quantumf wrote:

HermanHiddema wrote:Now as to the original question, I think it would be good if bot writers wrote their bots in such a way that it would always accept a single undo for any and every move.

Assuming that you're referring to KGS bots here, should I conclude from that that you also believe every (rated) KGS game should also allow a single undo per move? i.e. for human-human games too?

It is everyone's own choice, and I do not think it is wrong for people to deny undos, but personally I generally see little reason to ever deny an undo. I think it is a friendly approach to the issue, and in the case of bots I think it is a good idea to make bots as friendly as possible. I do think it is reasonable to have a sort of time limit, e.g. you must request the undo within 10-20 seconds or some such limit, so you're not wasting the bot's time (other people may want to play it too).

xed_over wrote:

quantumf wrote:

HermanHiddema wrote:Now as to the original question, I think it would be good if bot writers wrote their bots in such a way that it would always accept a single undo for any and every move.

Assuming that you're referring to KGS bots here, should I conclude from that that you also believe every (rated) KGS game should also allow a single undo per move? i.e. for human-human games too?

if you're playing for rating points, then that's a completely different situation. That's like playing a tournament game, and there are no undos in a tournament.

And yet, KGS (and most other online Go servers) allow undos for rated games (albeit at the opponent's discretion).

HermanHiddema wrote:It is everyone's own choice, and I do not think it is wrong for people to deny undos, but personally I generally see little reason to ever deny an undo. I think it is a friendly approach to the issue, and in the case of bots I think it is a good idea to make bots as friendly as possible. I do think it is reasonable to have a sort of time limit, e.g. you must request the undo within 10-20 seconds or some such limit, so you're not wasting the bot's time (other people may want to play it too).

I'm not opposed to your friendly attitude, but I still don't see how you can apply one rule for bots and another for humans, since they are (at least currently) involved in the same rating system. If bots have a KGS rating, and allow undo for every move, then I would expect some players to abuse it.