BBO Discussion Forums: Bots Don't Cheat But - BBO Discussion Forums

I'm sure there are other threads on this topic. I'm not sure I've ever actually posted on the BBO forum before.

I am not a consipracy theorist. I don't believe bots are secretly programmed to cheat. But I can't help but wonder.

What do you lead from:

J32
J98
732
T742

Against 2NT - 3NT?

By bot opponent found a spade lead, which is of course the only lead to break the contract. On another hand a bot holding something like:

Q
xxx
xxxxxx
xxx

Lead the spade Q on a blind NT auction, finding Jx in dummy and AKTxxx with partner.

I can possibly understand the second. It could pickle the spade suit but it could be the only thing to aid partner, who is more likely to be the one with enough tricks to defeat the contract.

But it seems like bots find a LOT of blind leads from short suits that hit partner's long suit. It's almost like you want to play the partner of the opening leader to be the danger hand in that suit.

Now, I could be wrong in what I'm about to say. On the first hand I gave, the bot could be doing a simulation and if you simulated a zillion hands then a heart lead would almost have to be better than a spade, but if you only simulate a hundred or a thousand there could be plenty of random noise to made a spade more successful. But it seems like bots hit partner's long suit a lot of the time.

I've been online for a zillion years and heard whining about the randomness of online dice and cards forever. Hell, we all remember when people used to complain that the computer deals in ACBL tournaments were non-random. Most of the time, when people whine I challenge them to state specifically what they think is crooked and to collect data in a scientifically valid experiment to prove it. Nobody ever does. Bad players just like to blame losing on something other than their lack of skill. I haven't done a scientific study.

But - I still don't get it - can someone explain? Am I halucinating? Not about these two specific hands, but in general, it really seems that when I'm declarer in NT, RHO is more likely to be long in the suit led than LHO.

redtop, on 2015-August-22, 13:29, said:

bots like to like to lead majors on auctions like 1NT-3NT and 2NT-3NT if Stayman not used and they have no decent suit

Bots tend to avoid leading from broken honor combinations. They don't want to give up a trick by finessing themselves. This is especially true in 1N-3N or 2N-3N auctions, where declarer is stronger than dummy and finessing partner on the opening lead is just giving declarer a finesse they would've taken anyways.

This means that bots tend to lead away from their short suits, not their long ones.

The only reason that humans lead from their long suits is because they have stupid partners who can't look at dummy and think about the bidding and figure out what their hands must be.

like they said, bots just don't make the same lead any sane human would lead. even if partner has bid a suit it might not be lead!

steve2005, on 2015-August-22, 16:10, said:

What he said. Sometimes if you want to make a lead directing bid, don't make the bid because GIB might lead your suit by random chance.

As others have mentioned, GIB like to make passive leads and frequently doesn't lead their own decent suit. In general, I usually like GIB leads when I'm declarer because they are frequently in one of your own longer suits or in a suit which you have well stopped.

OP - You were definitely unlucky on those hands but even a blind squirrel finds a nut once in a while.

Read the book "Winning Notrump Leads" by Bird and Anthias. GIB uses pretty much the same strategy to decide its opening leads as they used to produce their recommendations, except that they used many more hands in their simulations (they didn't have the time constraints that a bridge robot does).

barmar, on 2015-August-22, 16:59, said:

Once I bid a suit, Gib raised it and opps are in 3N. Leading our suit beats 3N any other lead (like the one Gib chose ) and they make. I'd like to know where that is in Bird and Anthias' book?

Play Money Bridge on BBO and you will see lot of funny things GIB Does.

glen, on 2015-August-22, 14:05, said:

4) a passive attack is more likely to succeed if it is carried out in a major suit instead of a minor one because it is clear that it prefers to conclude in 3NT -instead to get at five level - hands characterized by greater strength in minor suits [lacking other indications between spade and heart is preferred spade lead] (from "Il libro completo del bridge" by Guido Barbone pag.381).(Lovera)

steve2005, on 2015-August-22, 17:50, said:

I can't find my copy, but I don't recall if they had much about competitive auctions that lead to the opponents being in 3NT.