smerriman

nige1's first hand is close to the actual layout - but note that in his hand, winning the first spade and playing a club has you going down straight away. If I'm correct the main line suggested so far relies on West having the heart queen (less than 50% based on vacant spaces). Can someone come up with the best line assuming East has the heart queen? (I could well be wrong, but suspect it might be better.)

Left2Right - I'm afraid you have me completely confused :/ Firstly, the opponents have 5 trumps between them, so I'm not sure what you mean by a 4-3 trump break. If you're referring to my earlier post where I said a 4-3 spade break isn't possible, that's because West bid 1♠, and GIB doesn't overcall with 4 spades. Regarding the rest of your post, I'm not asking if the hand can be made 100% of the time. I'm asking what you believe the best line is, single dummy. I analysed the hand double dummy, figured out what line I would have needed to take, and thought that it looked quite reasonable and that experts may have come up with it. Thus asking what people thought the best line was, to see if anyone would.

A 4-3 spade split is (virtually) impossible given the bidding. Try again, Right.

[hv=http://www.bridgebase.com/tools/handviewer.html?sn=smerriman&s=SA762HAK654DAT8C2&nn=Robot&n=S95HJT3D743CAK765&d=w&v=e&b=16&a=PPP1H(Major%20suit%20opening%20--%205+%20!H%3B%2011-21%20HCP%3B%2012-22%20total%20points)1S(One-level%20overcall%20--%205+%20!S%3B%208-11%20HCP%3B%209-12%20total%20points)2H(Free%20major%20raise%20--%203+%20!H%3B%206-10%20total%20points)P4H(5+%20!H%3B%2021-%20HCP%3B%2017-22%20total%20points)PPP]400|300[/hv] GIB leads the spade King. I went down in this contract, while double-dummy analysis shows it's makeable. What line would you take? If you duck the first trick, GIB switches to the 9 of hearts, East playing low.

Yep, posted about this last August, can't have come up enough times to be looked into yet.

Agreed. I'm just not sure how GIB is designed to handle it. Past threads have shown it's clearly not capable of assuming slight variations are acceptable, and here it's massive variations (aces, kings, queens, and points all wrong). I agree with your surmise in another thread that there are some fundamental bugs in how GIB works. I doubt we'll see a change there though.

GIB's play relies heavily on simulating what hands you might hold. You've shown a balanced hand, 24 HCP, 2 or 5 keycards, the king of hearts, and the queen of spades. The only part of that which was truthful was the balanced hand, and most of the rest was completely impossible. I'm not surprised it led to some rather unusual outcomes. Crazy bidding (like opening 2♣ with nothing) will certainly get you some good results with GIB opponents, but usually is going to lead your partner astray more than anyone.

My guess is that people finesse in dummy regularly on the opening lead, and find it loses nearly every single time. That of course is correct, but nothing to do with non-random deals.

Exactly - that was my point :) Having different hands is necessary, but it's a unique form of bridge - thus may well be better suited to a unique form of scoring.

Are you actually looking for how to bid in a goulash context, or as if it were a real hand? Because when playing goulash, you should be using completely different systems to normal bidding; but if it somehow popped in real life, you're probably going to have to admit that no matter what you do, it's probably not going to work out well.

You play 8 hands. Each hand is against a different group of people. Your average matchpoint score is your total score. If you look at the top scorers in every tournament, you'll see their win is always due to all 8 boards being 'swingy'. The last couple of times I've gotten in the top 200, one or two hands were in the 50-60 percentages where I couldn't do better, which immediately ruled out winning. I actually think my suggestion would have considerably more than just a slight effect, but understand the complexities of implementing it. (Though, to be fair, MP where everyone plays different boards is pretty ideosyncratic too! BBO scoring is very unique compared to any other tournament.)

I'd say the most important thing to note is that it shouldn't (just) be about points. Given it takes up a lot of bidding room, you have to know you're going to be able to exchange the right information in time. As above, there are various ways to play them, but most involve certain shapes - some strong hands you definitely don't want to jump shift with.

Let's play bridge robots = basic robots = no simulations during bidding, so that's not the case here. The description of 3♣ is actually compatible with GIB's hand, so it's hard to know when GIB chooses to bid it and when it chooses to pass. But if you want GIB to take into account things like cards being more valuable over the doubler, you'd need the advanced bots, where it will run simulations of possible opponents hands and use them to decide whether it's worth pressing for game or not.

You might like to read this post, where I prove that there are indeed times where GIB simply leads randomly, against the documentation and not even possibly based on simulations. This is my favorite example - GIB is programmed to lead high from a doubleton, unless a simulation tells it another card works better. I wonder what hands it simulated to achieve this lead: [hv=http://www.bridgebase.com/tools/handviewer.html?bbo=y&lin=pn%7Cmauroem%2CRobot%2CRobot%2CRobot%7Cst%7C%7Cmd%7C4SAT932HAKT3DQT2CK%2CSQ4HQJ985D64CJ954%2CS875H64DAKJ973CQT%2CSKJ6H72D85CA87632%7Csv%7CN%7Cah%7CBoard%202%7Cmb%7CP%7Cmb%7C1S%7Can%7CMajor%20suit%20opening%20--%205%2B%20%21S%3B%2011-21%20HCP%3B%2012-22%20total%20points%7Cmb%7CP%7Cmb%7C1N%7Can%7CForcing%20one%20notrump%20--%203-%20%21S%3B%206%2B%20HCP%3B%2012-%20total%20points%7Cmb%7CP%7Cmb%7C2H%7Can%7CNew%20suit%20--%204%2B%20%21H%3B%205%2B%20%21S%3B%2011%2B%20HCP%3B%2012-18%20total%20points%7Cmb%7CP%7Cmb%7C3S%7Can%7CLimit%20major%20raise%20--%203%20%21S%3B%2010-12%20total%20points%7Cmb%7CP%7Cmb%7C4S%7Can%7C4%2B%20%21H%3B%205%2B%20%21S%3B%2014%2B%20HCP%3B%2015-18%20total%20points%7Cmb%7CP%7Cmb%7CP%7Cmb%7CP%7Cpc%7CD4%7Cpc%7CD3%7Cpc%7CD8%7Cpc%7CDT%7Cpc%7CSA%7Cpc%7CS4%7Cpc%7CS5%7Cpc%7CS6%7Cpc%7CS2%7Cpc%7CSQ%7Cpc%7CS7%7Cpc%7CSK%7Cpc%7CH2%7Cpc%7CHA%7Cpc%7CH5%7Cpc%7CH4%7Cpc%7CDQ%7Cpc%7CD6%7Cpc%7CD7%7Cpc%7CD5%7Cpc%7CD2%7Cpc%7CC9%7Cpc%7CDJ%7Cpc%7CSJ%7Cpc%7CCA%7Cpc%7CCK%7Cpc%7CC4%7Cpc%7CCT%7Cpc%7CH7%7Cpc%7CHK%7Cpc%7CH8%7Cpc%7CH6%7Cpc%7CH3%7Cpc%7CH9%7Cpc%7CS8%7Cpc%7CC2%7Cpc%7CDA%7Cpc%7CC3%7Cpc%7CHT%7Cpc%7CC5%7Cpc%7CCQ%7Cpc%7CC6%7Cpc%7CS3%7Cpc%7CCJ%7Cpc%7CST%7Cpc%7CHQ%7Cpc%7CD9%7Cpc%7CC8%7Cpc%7CS9%7Cpc%7CHJ%7Cpc%7CDK%7Cpc%7CC7%7C]400|300[/hv]

Yup, they should, on average - but statistically some are guaranteed to get more than average, and some are going to get less than average. Weighting them as described above removes this variation completely.

Yeah, thus my earlier comment about best-combined-hands being the best of both worlds, if that were possible one day.

Yeah. Just most of the time it comes down to random guesses - with one guess being the difference between 0% and 100% on the board.

At the table, your partner does everything he can to help you - opening lead conventions, signals, etc. With GIB, your partner leads and plays based on the assumption that you can magically guess which suit he wants you to switch to without any signals.

You're right, of course - in reality, there is simply no way to compare A and B based on the hands they received. However, I would rather receive a lower score because *I* had the potential to do better - vs receiving a lower score because *others* had the potential to do better.

MP does that perfectly - if everyone else plays the same boards. Your score isn't affected compared to the players who played the same board - but it is affected compared to those who didn't.

See the blue box at the top of here.

Suitplay is a nice program that answers questions like this for you. It says you can make 2 tricks 87.5% of the time, starting by leading low from J98x and covering whatever the next opponent plays.

All of your spade raises work exactly the same as if RHO didn't double - 2♠ is minimum, 3♠ is invitational, 4♠ is max. Your partner just has the additional information that you're not raising with a shapely 3 card support hand, which (without the interference) you would otherwise sometimes have to do.

To basic GIB the hand above is equivalent to any flat 12 point hand. It doesn't seem to take into account extra trump length sadly, though I guess advanced GIB could.

Also, I just realised the weighting formula is extremely trivial. It's how many people didn't tie your score. In MPs, if 50% of people do better than you, and 50% of people do worse, it proves you are a 50% player. If 10% of people do better than you, and 10% of people do worse, it proves you are somewhere between a 10% and a 90% player. MPs assigns a score of 50% - too small for good players, and too large for lesser players. In my proposal, the first score would count towards a full 50% - the second score would only count as 0.2 of a board to counterbalance this.