[SimonFa]

I was playing a match point table with three basic robots.

---

If you follow the play you will see that at trick 3 I played Q♦ (all other tables played the 6) and East robots all switched to A♣. My opponent then switched again to 10♦ rather than playing a second club to defeat the contract, which happened at all the other tables with robots defending defending and 6♦ was played at trick 3.

When I made the Q♦ play it was more a reflex to muddy the waters than with any particular plan in mind. I'm not complaining that it worked but am wondering why it worked?

Was it one of the robot random plays we get every now and then or did it think it needed to shorten my trumps and if so why? Or maybe there was some other reason?

More importantly would it have worked against a human?

As always, thanks in advance for your thoughts and comments,

Simon

[BillHiggin]

SimonFa, on 2013-September-07, 03:36, said:

More importantly would it have worked against a human?

I am glad you asked about humans. We can use logic when we talk about what humans do. Not so much with the bots.
If West were a human, we would be thinking "clearly a novice". The rush to cash aces is very much a novice strategy.
When East won the ♦K, the human thinking would be something like "Clearly my partner has shifted to his doubleton diamond, now why his declarer dropping the Q". He would likely still return a diamond since this sequence of plays is absolutely inconsistent with a singleton club and four diamonds. But, your falsecard might shake him out of that line of thought since the Q play is so inconsistent with what he thinks is going on. Of course, it is tough for you to visualize the actual West hand on this defense for the same reason - this series of plays is just plain inconsistent with the cards he holds. If did know the actual holding, you would NOT drop the Queen and now East would have no reason to shift.

There is a name for a series of plays that are so bad that they will convince an opponent that something else is happening and cause him to go wrong. That is called a "Grosvenor". In a way that is what has happened here. West has committed a Grosvenor by shifting to diamonds instead of the obvious club shift. But East is so confused by the illogical shift that he thinks your play of the Q is the actual Grosvenor (never mind that you would have a hand inconsistent with your bids).

Back to actual bots. This sort of falsecarding does seem to short circuit their little bot brains fairly often. Sometimes they work against humans, but humans are more likely to catch the inconsistencies and figure the truth out.

[SimonFa]

Thanks, Bill, that made sense.

When I reflected on afterwards I thought my ♦Q if anything should signal shortage of clubs in West's hand.

I have noticed, through great frustration that bots don't get singleton leads when I make them so maybe they're programmed not to make singleton leads so East just didn't believe West had short clubs?

Against a human I suspect I would have received a club lead anyway so the false card position wouldn't have arisen.

[PhilKing]

You made a clear error at trick four - assuming they are playing standard count it is mandatory to drop a club honour. This would help nurture the illusion that West is 3623 with Hxx in clubs.

As pointed out, if West actually held a doubleton diamond, there is nothing you can do to talk them out of it.

[SimonFa]

PhilKing, on 2013-September-07, 05:51, said:

You made a clear error at trick four - assuming they are playing standard count it is mandatory to drop a club honour. This would help nurture the illusion that West is 3623 with Hxx in clubs.

As pointed out, if West actually held a doubleton diamond, there is nothing you can do to talk them out of it.

That's a good point, thanks, Phil.

[SimonFa]

Thanks, Bill, that made sense.

When I reflected on afterwards I thought my ♦Q if anything should signal shortage of clubs in West's hand.

I have noticed, through great frustration that bots don't get singleton leads when I make them so maybe they're programmed not to make singleton leads so East just didn't believe West had short clubs?

Against a human I suspect I would have received a club lead anyway so the false card position wouldn't have arisen.

[MrAce]

SimonFa, on 2013-September-07, 03:36, said:

I was playing a match point table with three basic robots.

---

More importantly would it have worked against a human?


Simon

Vs humans you could be defending vs 4♥.

Only Robots, and people who think they are programmed to make choices depending on bean counts would open 2♥ with W hand.

[PhilG007]

SimonFa, on 2013-September-07, 03:36, said:

I was playing a match point table with three basic robots.

---

If you follow the play you will see that at trick 3 I played Q♦ (all other tables played the 6) and East robots all switched to A♣. My opponent then switched again to 10♦ rather than playing a second club to defeat the contract, which happened at all the other tables with robots defending defending and 6♦ was played at trick 3.

When I made the Q♦ play it was more a reflex to muddy the waters than with any particular plan in mind. I'm not complaining that it worked but am wondering why it worked?

Was it one of the robot random plays we get every now and then or did it think it needed to shorten my trumps and if so why? Or maybe there was some other reason?

More importantly would it have worked against a human?

As always, thanks in advance for your thoughts and comments,

Simon

Your biggest error was trying to hoodwink an android(!) How do you make a robot blink(?) Answers on a postcard....(!) Artificial Intelligence will always triumph over natural stupidity