[armant2k]

I/N players learning basic defensive partnership rules/maxims - 4th best leads, signaling, returning partner's suit, discarding, etc., etc. - and using those inferences as Declarer, require playing lots of bridge. There is no way around it - you learn and improve by playing LOTS of hands.

But it is impossible to gain that necessary and valuable experience by playing/practicing with the GIB basic robots because they do not do these basics or follow these maxims. Instead, newbie pairs practicing against robots will randomly get 'rewarded' with good score results for NOT applying them. As a result, they learn nothing, or worse, convince themselves to expect good results from bad decisions. Basic defensive partnership skills/habits are NOT reinforced.

Even the way the GIB robots follow suit (e.g. tossing the highest card of two equivalent remaining cards) is NOT what IN players will encounter from human opponents, or partners.

I do not recommend advising newbies to rent and play against BBO robots, nor do I think most advancing players enjoy playing against or partnering with GIB robots for the same reason - you cannot use the 'muscle memory' skills you are trying to develop. Perhaps the Advanced robots have these skills (can anyone speak to this?), but I refuse to pay the $$$ BBO wants to find out.

Does anyone agree/disagree the GIB robots are doing a disservice to the game we love?

[mycroft]

Everything you say is true. I remember Helene here saying that playing with the robots too much affected her RL play, and she is not I/N.

But I would say that the "training" they get from the robots is a) about as dangerous as double-dummy analysis of the boards on the hand records, and b) frankly isn't better or worse than what they get in I/N games. Which is one of the reasons I hunt for "good players in the 299er" and try to push them out - I'm not trying to save the newbies and no-hopers from them, I'm trying to save them from the retraining time required when all the skills they honed beating up the fish turn out to be bad habits punished by flight A (or even flight B).

And everything you say about the robots and everything I say about the limited games the "true flight As" and "wannabe-experts" over on That Other Site say about playing in club games - "as long as you don't go in expecting bridge to be played, you can have an enjoyable time." And they're not wrong either.

But all of them - the robots, par analysis on hand records, the limited games, the 12-table Tuesday random in Chapala - have a real purpose that transcends any flaws they may have for players' improvement. And I/Ns can be warned about the bad things they might pick up from playing with robots - there's no comparison to the degredation in skill that comes from not being able to (or not wanting to) play bridge at all.

It's great that I can take out my phone and push a few cards around waiting for my doctor's appointment or on the bus to Guadalajara. It certainly would be a disservice to the other players if I joined a table, played 3-7 hands, then disappeared halfway through one - so if the alternative is no robots, then the alternative is no play. And that's the true disservice to the game we love - stopping people from playing it.

[smerriman]

Agree with mycroft. You could replace 'GIB robots' in your post with 'any weak players' and exactly the same is true. Read literally any of AL78's posts about the quality of bidding and play at his club, and the horrible bad habits that people there have picked up that keep seeming to work against him. Or kibitz any human table in BBO's main bridge club..

Obviously nobody should be told to learn how to bid and play from GIB, just as they shouldn't learn how to bid and play from these humans. But saying this means they are a disservice to the game completely ignores the huge positives that they bring, allowing people quick access to the game at any time they like, without the bullying (and much worse partners) that comes with many other forms of it. And with the right mindset, it is possible to learn a lot playing with GIB, even if it's of the form why not to do what GIB always does.

And I would go further and say mycroft's point a) is far more damaging than what you "learn" from robots. Seeing people say "oops" solely because they go down in a contract that double dummy says is makeable (and vice versa) drives me insane, and means they never notice or learn from the true mistakes. But again, it's a good tool when used correctly, but easily misused.

[pescetom]

smerriman, on 2023-March-30, 13:03, said:

Agree with mycroft. You could replace 'GIB robots' in your post with 'any weak players' and exactly the same is true. Read literally any of AL78's posts about the quality of bidding and play at his club, and the horrible bad habits that people there have picked up that keep seeming to work against him. Or kibitz any human table in BBO's main bridge club..

Obviously nobody should be told to learn how to bid and play from GIB, just as they shouldn't learn how to bid and play from these humans. But saying this means they are a disservice to the game completely ignores the huge positives that they bring, allowing people quick access to the game at any time they like, without the bullying (and much worse partners) that comes with many other forms of it. And with the right mindset, it is possible to learn a lot playing with GIB, even if it's of the form why not to do what GIB always does.

And I would go further and say mycroft's point a) is far more damaging than what you "learn" from robots. Seeing people say "oops" solely because they go down in a contract that double dummy says is makeable (and vice versa) drives me insane, and means they never notice or learn from the true mistakes.

But it also implies we all deserve better robots for our money, which I think was OP's main point.
Although as someone wise once suggested, be careful what you wish for because it might come true.
There is a huge gap between "current" bridge "AI" and current Chess or Go AI, and I am not convinced it is inevitable rather than just disinterest in bridge.
I was intrigued by the recent article about cheating in Go, where it was clear that a new profession of teaching humans to exploit AI and to understand AI play is emerging, a role you are familiar with already
I think it could be fun and perhaps even the rebirth of bridge to have better AI, but we need to decide what we want and fix the Laws now, because AI when it wants to beat us with the current Laws will be able to formulate, exploit and probably even "disclose" agreements that are well beyond most human ability to comprehend, let alone formulate a defence (be it with mikeh or someone from al78's club).

[smerriman]

pescetom, on 2023-March-30, 14:52, said:

But it also implies we all deserve better AI for our money, which I think was OP's main point.

Can't disagree with that Though I would point out that my (and NookAI's) experience is that robots who signal accurately are considerably *worse* bridge players, because they are far too easy to exploit. So it depends whether you're wanting competition, or a learning tool.

[pescetom]

smerriman, on 2023-March-30, 15:01, said:

Can't disagree with that Though I would point out that my (and NookAI's) experience is that robots who signal accurately are considerably *worse* bridge players, because they are far too easy to exploit. So it depends whether you're wanting competition, or a learning tool.

I don't see why a decent AI should be any easier to exploit than a decent human, in terms of signalling. None of us would accept a partner who signals "accurately" to the advantage of opponents.
My hope is that AI will provide an appropriate level of competition without either boring or humiliating me, a learning tool with a capacity to explain my errors and direct my growth, a willingness to play with me following a system I choose, a capacity to discuss the current situation of play during practice games and to perform a diplomatic post-mortem, just for starters
My main concern is that it might be the final straw for the ideal of understanding the opponents' independently formulated agreements and that whatever replaces that might make for a less interesting game.

[smerriman]

I think AI is a long way off "figure out if playing spot card x vs spot card y is likely to give interpretable and helpful information to my human partner and more than what it costs by giving the same information to my human opponents" yet. To be honest I'm not sure if it'll ever get there, but if it does, it will be fun to see..

[thepossum]

I just regard it as a completely different game

I do feel they do very little to enhance much of real Bridge skill - skills (if any) I feel I have mostly lost

Maybe a bit of bidding and declarer practice but not much

[steve2005]

pescetom, on 2023-March-30, 15:23, said:

I don't see why a decent AI should be any easier to exploit than a decent human, in terms of signalling. None of us would accept a partner who signals "accurately" to the advantage of opponents.

Signalling is so crucial to good defence I accept that occasionally you will gift declarer
As you learn, you can learn when to not signal or even lie.

As it is, the ridiculousness of GIB leading a suit and switching or not leading it again when lead is regained because it can't tell if partner like it even though declarer is wide open is appalling!

[thepossum]

I think progress everywhere would be achieved by scrapping the term AI for almost all technology in existence at this time

I feel it has become somewhat diluted as a term

[thepossum]

I do think one place where GiB competitions could help learners (for want of a better word) is seeing how better players bid sometimes - even with GiB

It is rather cruel when some play their own preferred system with the robots all playing 2/1 (GiB style)

[pilowsky]

It depends how you approach it.
If you just want something that will improve your card memory it's hard to beat.

OP complains that GIB is awful and should be shunned because it doesn't do a bunch of things that he considers important.
One of which is "returning partner's suit".
Personally, returning partner's suit is very low on my list of things to think about when seeing the opening lead.
One of the great things about playing with GIB is learning that defensive maxims aren't everything they're cut out to be.

On the plus side GIB does allow players to do all kinds of helpful stuff: count cards, understand a basic bidding system etc etc.

[pescetom]

smerriman, on 2023-March-30, 16:19, said:

I think AI is a long way off "figure out if playing spot card x vs spot card y is likely to give interpretable and helpful information to my human partner and more than what it costs by giving the same information to my human opponents" yet. To be honest I'm not sure if it'll ever get there, but if it does, it will be fun to see..

I suspect we are talking about a number of years you could count on one hand before AI can do this and everything else necessary to completely outplay a pair of humans at anything but Bermuda Bowl level (and probably even then).
If it hasn't happened already it is because bridge is irrelevant, not because the problem is any order of difficulty superior to Go.
What happens then will be fun to see only if we have already thought about it or don't really care about the result.

[johnu]

smerriman, on 2023-March-30, 15:01, said:

Though I would point out that my (and NookAI's) experience is that robots who signal accurately are considerably *worse* bridge players, because they are far too easy to exploit. So it depends whether you're wanting competition, or a learning tool.

The obvious question to ask is whether the declarer advantage in knowing distributions is worth more than defensive blunders caused by lack of signaling. I have no idea which is better (or worse). And of course, a true AI bridge robot would know when signaling would only help declarer, not that I consider GIB an AI robot capable of such decisions.

[thorvald]

I would be nice if GIB at least har some understandable rules about leads later in the play. I think I have seen almost any combination, as it seems pretty random, when the simulation gives identical results.

This would not be an advantage for declarer but for the partner (player)

I had this example today

---

The bot lead the ♠T and this dummy hit the table

---

Now partner used ♠ K, followed by the Ace and you got the ruff (Declarer following).

For advanced players the sequence KA8 is showing interest in lowest suit, and the sequence ♠AK8 is showing interest in the highest suit. This is way beyond robot-logic, so it will just make a simulation ending up with a ♥-lead.

In https://www.bridgeba...ystem_notes.php it is stated:

When it has a choice of equivalent cards, it will choose based on leading and signaling conventions.

I haven't found the leading and signaling conventions, but I would assumed that when leading a ♥ here it would play 4th highest.

What do you think?

Spoiler

Lowest from a 6-card suit
Perhaps it is showing an even number?