Joshua Donn
Joshua Donn

Source: BBO

On Sunday, November 18, at 1PM US Eastern Time (7PM Central European Time), we had a special 1 hour teaching session with star player  JDonn (Josh Donn). Josh used this hour to share his best tips for the players gunning for NABC Robot Individual glory. For the BBOers who were not able to attend the lecture, let’s dive into the review now.

Hi everyone.

Today I will talk about bidding and playing with misfits. But first I will do a plug for the online NABC event.

This big event is held in conjunction with the North American Bridge Championship that starts this week in Honolulu, Hawaii. The online event starts tomorrow, and it runs for three days. It is open to ACBL members no matter what country you live in. For $50 you get three full days of play, 24 boards a day. You play with GIB robot as your partner and opponents, and you try to outscore the other human players. You always have the best hand at the table, meaning there are no boring weak hands. There are already about 1600 players enter, and there will be even more by the time it starts. It should be a lot of fun, and it gives out big masterpoint awards.

To help everyone get ready a little, I will talk about misfits. First, the main key to bidding misfits can be summed up with one piece of advice: Pass as soon as possible! Let me show an example.

Partner opens 1 and you respond 1. Partner rebids 2. Let’s summarize what we know so far. You have 11 points, normally enough to invite on this auction. But the diamond void is a huge negative. You don’t have a fit in either major suit, and partner probably doesn’t have four clubs either. There is also a big misconception that diamonds won’t play well. On the contrary, playing in diamonds, partner’s long diamond suit will probably provide more tricks than it will in notrump. You have found a playable contract, and trying to improve it will probably just make things worse. Pass while you have the chance.

Let’s move to partner’s side of the table and try to play the hand.

The opponents lead a small spade. What I recommend when playing in a bad fit is to imagine how the play might go if you draw trumps, and then imagine how it might go if you don’t. So let’s suppose you win the spade with the ace, and play the ace and queen of diamonds. You look like you will end up losing two hearts, two clubs, and two diamonds. That means you will go down.

Is there another option? Well there are hopes of setting up a heart trick. It looks important to get that going right away, in time to discard a club loser. It probably won’t matter if you can’t draw trumps, since you can also score some tricks by ruffing in your hand. So let’s win the ace of spades and lead a heart toward the QJ (better than leading the ten, since the opponent on your left may go up with the king). Most likely the opponents will win this trick and continue spades. Now comes the key moment. Win the king and ruff a spade in your hand before leading a second heart. Sneaking ruffs in your hand will be key to scoring enough tricks for your contract.

Now lead the heart ten. Maybe the opponents will win and exit safely with another heart. Win that, discarding a club, and you will want to lead either the last spade or heart and try to ruff that in your hand as well. If you can score another small ruff in your hand you will be well on your way to scoring 8 tricks. At worst you will force the opponents to lead trumps for you, and get a free finesse.

Here is the hand:

Let me show another misfit hand that I played with and against strong players.

I held AQT432 AQ92 K87 -. My partner opened 1, and I responded 1. He rebid 2. Despite the misfit, I was clearly strong enough to go to game. Note a few things. First, we might still have a fit in either major. Second, I have extra strength beyond a minimum game force. I bid 2, artificial, to see what I could learn about partner’s shape. Partner bid 3, a natural bid showing four diamonds. I now knew partner had 10 cards in the minors, and he probably would avoid this bid with a doubleton spade. So I gave up and bid 3NT.

The opponents led the 2, and dummy came down – K83 Q654 KQT532. Two 6-0 fits were going to make this difficult. The diamond went to the J and K. Trying to set up either 6 card suit is a road to giving the opponents lots of winners. I decided to simply lead a diamond back to the queen. My left hand opponent went up with the ace, and he had a tough play as well. He couldn’t tell my spades were so good, and he played a spade. I threw a club from dummy and won the king with the ace. Now if the hearts were to run I would have enough tricks to make by setting up a club trick.

I didn’t want to play any hearts now since I wanted my entries in that suit to stay fluid. So I tried a diamond to the queen, as a club was thrown on my right. I led the king of clubs, which was won on my right. He led a spade through me. The spade jack must be wrong, as the opponent on my left would never have shifted to a spade earlier with nothing in the suit. I decided to simply play low, as maybe next the spade jack would fall. But the opponents defended well, overtaking and playing a club through me. This forced me to part with my spade ten. I played a heart to my hand, and took the spade queen. This happened to squeeze my opponent out of his fourth heart, solving my guess in the suit.

That hand is a good example of the sort of step by step approach you need to playing a misfitting hand. There often won’t be an obvious long suit to set up or source of tricks, so you have to take it one trick at a time and see what happens.

I hope this was informative. Remember to register for the online NABC event. You can find it in ACBL World, with all the ACBL tournaments on BBO. I hope to see you all competing. Thanks everyone.

Q1: I opened, raised partern’s spades in between opponents competed in hearts, then partner stopped competing while opponents then bid 4, which I doubled (+500), now pard pulled for -1100. How can I predict this behavior? Partner was a GIB.
A: It is hard to answer without seeing the whole auction. Make sure you always look at the definition of the double before you make it. Unfortunately, GIB perhaps doubles a bit too much and pulls doubles a bit too much. Just keep it in mind.

Q2: I want to learn how to read the robot’s signals when we are defending. Also, it would be useful to understand the robot’s signals when I am declaring.
A: The robot doesn’t signal at all. In theory it plays standard carding, but is not a reliable signaller. I actually don’t mind playing that way, because it is the same for the opponents as well. Just keep in mind not to take inference from GIB’s signals.

Q3: When you double for penalty at a lower level the robots always take it out. Is the answer just not to use penalty doubles?
A: They often do but not always. Make sure you read the definitions of the doubles. Doubling is a very difficult area for the programmers to work on, unfortunately. Lots of auctions, humans don’t even agree on the meaning of a double. Or you know it’s penalty because you can just tell, but it’s hard to give it a clear definition of why it’s penalty. But until more progress is made, just be wary.

Q4: What about cheaters in the online tournament?
A: OK so let’s talk about that. Firstly, BBO takes several major precautions to prevent cheating.

The main precaution is what is called deal pools. Pretty much, not everyone plays the same hands.

The hands you play are only played by people who don’t live near you, aren’t near you while you are playing, have never played with you before, and aren’t your friends on BBO.

You have no idea who else is playing the same hands you are, which makes colluding with another player essentially impossible.

On top of that, there is a group of expert players who review the hands played, especially by the leaders, for suspicious activity.

So I don’t think cheating is anything you need to worry about in this event. Cheaters are caught on BBO regularly. BBO is very good at investigating them. But this event is designed to make that kind of cheating impossible.

Q5: It seems like finesses fail much more often than the expected 50%. Is this true because the robots can select their hands as we play?
A: It’s not true. It seems that way because of selective memory. Humans remember the losses and disappointments much better than they remember the successes. So sorry I have trouble convincing some people of this, but you do not lose finesses more than half the time.

Q6: I have been reading about robot’s preset rules, can you elaborate on best hand? And lead preferences?
A: OK, first best hand.

The human at the table always has the best hand at the table, meaning the most high card points. It says nothing about distribution. Another player may have the same high card points, but never more. So if you have 11, no one else has more than 11. Most often you have like 15 or more. It makes playing more fun, you never have a 2 count or anything.

OK leads.

GIB is a very passive opening leader. On a double-dummy basis, this is the best way to play. But it’s not how most humans play. So remember two things.
– First, GIB is more often leading from weak suits and less often from honors.
– Second, humans often complain that GIB doesn’t return their leads. That’s because GIB also plays partner to be leading passively.

So in general, remember GIB is leading passively, and if you think a decision is close then lead passively yourself so your partner GIB knows what you are doing.

Q7: In a competitive auction, say I open 1, partner responds 1, and GIB jump overcalls 2 in 4th seat. How do I show that I want to compete in hearts in this auction without GIB playing me for a bunch of extras?
A: You can’t, just like you can’t with most humans. Best is pass, and if partner has anything extra let him come back in, then bid 3H. I would play this way with most humans as well. After all, 2 doesn’t mean they have a spade fit so you don’t want to overcompete here. Maybe partner is 4-4 in the majors.

Q8: Suppose I have a flat 11-count such as Axxx/KJx/Qxx/Jxx in first or second position. I wouldn’t open that hand in regular bridge. But in the robot game where I know my hand is the “best hand” in terms of HCP, should I do anything different?
A: You could try opening, and then passing partner’s response. Kind of like people do in real life if they open light in third seat. In third seat when you open light, you know you don’t have a game. Same idea, with 11 points your side has at most 22 so you probably don’t have a game. Let me tell you though, I have followed this strategy and done pretty well, but there was one huge disaster, so no guarantees. I opened with 11 or 12 balanced and passed the response, and missed a grand slam. Partner was 7060 distribution. So it’s a fine strategy, but like everything else do it at your own risk.

Q9: How should we adjust our bidding and play when playing with GIB?
A: Remember the things we have talked about today. You have the best hand, GIB leads passively, GIB doesn’t signal well. Always check the definition of a bid before making it, or a double.

OK I think that’s it for today. Good luck everyone, I hope to see you play in the event.

Don’t forget to follow us @