Leo LaSota

Not exactly the best customer service with BBO either... sent BBO support an e-mail about this ~ 10 hours ago and have received no response.

Interesting that this is for stars or players with 5+ platinum points; seems like I may qualify as I am a star and have >400 platinum points and a NABC+ victory. I did not receive an invitation :(

The robots do not cuebid 2nd round controls. This has been discussed previously in one of the lectures that I gave on playing games with the robots. You cannot have a constructive auction like you would in real life with a good partner. Instead, just take control and bid 4nt RKC at some point, then 5nt to ask for kings once they show 1 keycard (ace of spade). If they respond 6C or 6D, decide if you want to take a gamble at 7H or 7nt.

You were your own "third opponent" on this hand. You showed a stronger diamond suit with your double of 3D (KQ+ or better). Change your queen of hearts for the ace of diamonds and you are a big favorite to make 4D. The opponents have a good heart fit on this auction, and GIB is looking at a singleton heart. There is plenty of reason to think that not selling out to 3H is correct with the GIB hand. Your hand could have held 3+ clubs, so introducing the 5 card club suit is very reasonable.

While you may think that bidding 2C with a 5332 hand when partner opens 1H is "awful bidding", many would disagree with you. I would certainly bid 2C with the given hand in a partnership that plays Flannery, since there would not be a 9+ card spade fit unless the opening bidder holds a huge hand. Given that GIB does not play Flannery, there are some hands where it would be advantageous to respond 1S first. However, setting a GF at 2C leaves a world of room for slam exploration. Therefore, when you have a known 53+ heart fit it frequently works well to start with a 2C GF bid on 5332 shape.

I am not claiming that the individual in this situation did or did not look at the previous boards. However, I give this as an example situation that arises quite frequently. If you are having a very high score, and see that one other individual is having a very high score, you know that you can take a risk on the last board. If it works, you move to first. If it does not work, you will likely stay in 2nd place. By seeing the actual results from the other competitors, you can see where everyone stacks up through the boards that you have completed. When you see Rank 2 on the bottom, you may be surprised that you are not currently in first place. When you check the individual boards, you see that you and the individual currently in first place are well ahead of the field. Therefore, you know that you have a "free" opportunity to try something to reach first place.

The % of users that would ever look through other table results during an ACBL robot tournament is small. However, for those that do choose to look at other table results, this gives them a huge advantage in determining how to go about on the last few hands. In tourney #650, I finished second in my section with +49.34 IMPS; first was +50.03 IMPS. On board #12, the winner took an aggressive view and it paid off handsomely when they won 12.88 IMPS and earned them 1st place. I believe that the individual just took a blind chance and were lucky with the outcome. However, holding Axx KQ10x x Axxxx, the bidding went 1C (from partner)-2S overcall-Dbl-3S-P-P back to this hand. They have a wide cariety of choices at this point, including 4C, 5C, 6C, dbl, 3nt & 4S. The 4S cuebid or the jump to 6C both show more total points than this hand has. However, the person chose the 4S cuebid. This struck gold when their partner held a stiff spade, the Axxx of hearts, and the KQxxx of clubs.

You can check exactly who did what on any board while you are still playing the boards. It is only after you are done that you cannot see who did what on each board. I recently finished 2nd in my section in an ACBL robot IMP game with a +48ish IMPs. 1st in my section was +51 IMPs. 3rd in our section was way below 2nd (low 20s for IMPS I believe). This is a perfect example of where if someone is in a clear 2nd place with a couple of boards left, they can take an extreme action. If the extreme action does not work, they still finish 2nd. If the extreme action does work, they can finish 1st. My belief is that they should have the settings the same while you are still in the game. That is, if you can see other results, you do not know who did what on each board.

It is illegal in ACBL to view recap sheets for what everyone else has done on a board after you have played it, with the exception of the Barometer games in which everyone plays the same boards at the same time. Even in ACBL clubs that allow users to view other results on their bridgemate after they complete a board, everyone is forced to play the same number of boards at the same pace. Therefore, you do not have the advantage of always seeing what a specific player has done on each of the boards that you play as you complete them.

This capability needs to be removed from the ACBL robot games. There are reasons why reviewing results from other tables is never allowed in ACBL tournaments, and only a very small % of ACBL clubs still allow this capability. Why should a player be punished for playing quickly, having their results seen by everyone else as they finish their boards? As an example, suppose player A finishes all 12 boards very quickly. Player B is more deliberate in playing the 12 boards. Player B knows that Player A is a very strong opponent that frequently is on the leader board for an ACBL robot tournament. It is a huge advantage to Player B if they can see after 9 boards just how far ahead or behind they are from another player through the number of boards that they have played. Suppose you are Player B and you are +30 IMPS through 9 boards. If you see that Player A is only +28 IMPS through the first 9 boards, then you know that you are on pace to beat Player A as long as you get the final 3 boards right. However, suppose instead that you see Player A is +65 IMPS through the first 9 boards. In that case, you would know that your goal should be to prevent losing IMPS and dropping to 3rd place, as 1st place appears unattainable. While it is questionable already whether it makes sense to allow seeing your current rank, it is exponentially more unfair if you can check how far ahead or behind you are from someone else that has played the boards at least as fast as you have. You can use this information to adjust your tactics for the last few boards if you know the other players that you should keep an eye on for your competition.

As I stated in a previous thread, the only time that I stretch to have the robot declare a hand in the total points games is when I am doing something else while playing the game. Normally, I take the last 5 to 10 minutes off with an insurmountable lead and get other things done.

ACBL robot duplicate are usually 12 boards where you play with 3 robots. Your results are then compared against other humans. They normally grant the players up to about 59 minutes to complete the 12 boards. I only need about 7 minutes to complete the 12 board tournaments. I win about 85% of the $5 total points games. In these tournaments, the hands dealt are random and different for every participant. The only constant is that you are always dealt at least as many HCP as the 3 other hands. I am able to play well and play very fast in these, so I normally complete more hands than the other competitors and generally play the hands better than most. Therefore, I usually only lose when someone else is randomly dealt plenty of slam hands.

I am faster than the robots. I am known for being one of the fastest bridge players in the world.

When I play 2 ACBL games in an hour, it only takes me about 7 minutes to play each game. Therefore, I have over 45 minutes that I do something else anytime that I play a game on BBO. Since I win about 85% of the $5 total points games that I enter, I will not have to spend anything this year for my 2,000+ online points.

ACBL decided to institute the 2 game per hour rule rule at the beginning of 2011. I call it the "Leo rule" as BBO used to offer ACBL matchpoint games every 15 minutes and I would finish a game in 10 to 12 minutes, so I could play as many as 4 in an hour. After I won 3,340 online points in 2010, they decided to put the "Leo rule" into effect.

Perhaps should have clarified, but I believe that they normally bid 1D over 1 of a major with less than a game force hand and only a 4 card major. With GF values, I have seen them bid both 1D and bid 1 of the major.

Sure, defending seems likely to work fine with this hand. However, bidding 2S with a two suited hand wit ha side suit singleton is very reasonable, especially non-vul. Rather unlucky this hand that their partner only has 2 spades. Would have worked out much better next time when the opener has 4 spades and 2 hearts.

This auction by the robot is very reasonable. Indeed, 2S shows exactly 4 with a longer minor and you should bid 2nt, then pass 3C.

Passing after the t.o. double is reasonable. 3 small spades looks even worse with the t.o. double, plus the expectation is that the heart suit may not split with the t.o. double on their right. Of course the hands mesh well and the trumps split and 4H comes home easily this hand. Many other layouts where game will not make.

I believe that they are programmed to bid the major first if it is 5+ in length, but respond up the line if the major is only 4 cards in length.

There are more factors involved in matchpoint games that can play a key role in the outcome than at IMPS. For example, correct defense that holds a hand to -620 as opposed to -650 is huge at matchpoints but is only worth 1 IMP at IMPs. While GIB is well programmed on many of the important decisions, there are a fair number of key factors at IMPs that GIB is not well programmed for. Therefore, GIB is tougher at matchpoints than IMPs.

When GIB is looking at something like AKQJx of a suit in dummy and declarer leads the suit, GIB is programmed to play the 9 from 9 doubleton to give standard count. This is because the 9 is clearly NOT an important spot card in the suit.

This is nothing close to being worth a double. Zero aces; KJx of diamonds sure to be useless for the defense, three small clubs is bad for our side, ace fifth of hearts very likely on your left. I would not be surprised if they can make 5H looking at my hand. LHO could hold the ace spades instead of ace diamonds and make at least 4, possibly 5.

The robot does not give count by wasting a spot if the spot could be important. The 9 is a potentially valuable spot card, that is why the robot played low on the first heart. This is the same that one should practice in live bridge. Do not waste a potentially useful spot card.

Wouldn't it be simple to sort players in the ACBL robot tournaments by their averages in previous ACBL robot games? At least after a player has played a suitable minimum, this would be a good way to assign the players to different sections. The fact that Justin Lall uses an id in the ACBL robot tournaments that does not have a star designation does not mean that he should be placed below all of the stars. I think that he said he averaged close to 68% in 360 hands over 2 days recently.