smerriman

The vast majority of BBO users who claim to play SAYC aren't actually playing it / have no idea what SAYC actually means. Many incorrectly assume SAYC means what you have described, so the person who wrote those notes likely has too. (There may be some interest in rubber bridge, but the fact it's not something BBO even supports - other than manually adding up the scores with friends - suggests it is a very small amount of interest only.)

I've run several large-scale tests in the past, and what is written in GIB's system notes about signalling is provably wrong. Here is one example about the opening lead signals. I can't find my post where I tested count, but I dealt it the most basic situation when teaching beginners about count (dummy has a long suit missing the ace and no entry, and the defender with the ace needs to know how many times to hold up). GIB as the other defender consistently failed to give count, playing the same card regardless of whether it had an even or odd number, making the hold up a coin flip rather than a certainty. GIB is not programmed to signal attitude at all on its first discard, so I suspect what you are seeing there is just coincidence / confirmation bias (ie, when its card matches a signal, you notice it more than when it doesn't). Edit - found the example disproving count signals.

You're right that good defense is a skill, and that it is enjoyable. But it's a partnership skill, not an individual skill. You need a partner who can provide (and understand) interpretable leads and attitude/count/suit preference signals. With GIB you don't get any of that, which is why not many enjoy defending with a robot partner. Add to that the fact that discarding one irrelevant card will make the robots take a completely different line than discarding another irrelevant card, and the results will end up being almost entirely random, rather than skill based. (That's true for declarer play too, but not to as great an extent.)

(Quick note - looks like an extra bidding diagram needs to be edited into the top of the post to make the intro more understandable.)

So if you had an agreement that your weak 2s deny a major void (but may have a minor void) - that one would be illegal because now you're strictly showing 1+ in the other major? I mean, I agree, most laws in bridge tend to be silly when you read them literally instead of what they were meant to mean (*cough* lamford). Though in this case it's just hard to think of examples where a 'two suited bid' which only promises a card or two in the second suit actually makes sense in the first place in order to make it illegal. I guess it was probably meant to solely cover the 3 card suit option like the one helene_t mentioned (but then that was made legal anyway, so I don't know.)

nullve does have a point. What is your definition of 2 suited hand? The standard definition is 54 or better, which clearly is already ruled out here. Is your definition of 2 suited that those two suits must be longer than any other suits - ie this is solely ruling out bids that deal with hand shapes like 6322? Even if so, the wording does not talk about 2 suited hands; it talks about 2 suited bids. The bridge laws don't stop me from overcalling 1NT to show the majors with 5440, even though that's a two suited bid, not a two suited hand.

It doesn't rule out an opening bid which promises at least 6 spades and fewer than four hearts. It rules out an opening bid which promises at least 6 spades and at least x hearts, for some specific 0 < x < 4. Your alert is an 'at least'; the 'fewer than 4' applies to the constant.

Ekren?

Basic GIB doubles unless favorable, when it passes. Advanced GIB simulates hands at specific times during the auction, so can come up with different results each time (based on the seed as discussed previously). That's why it passed in the originally posted case, but doubles at other times. There is always a noticable pause when GIB is simulating (as it takes time to calculate double dummy results, vs the database lookup being near instant). What I'm talking about here is that it isn't analyzing the right alternatives when a simulation is taking place.

The following hand was posted in BridgeWinners recently: [hv=https://www.bridgebase.com/tools/handviewer.html?s=ST97H7DQT7542CJ92&w=SAK8432HT43DJ3C53&n=HAKQ9865CAKQ764&v=b&d=s&a=P2S3S4SPPP]400|300[/hv] Upon testing, Advanced GIB pauses to run some simulations over 4♠.. before either passing or doubling (and double is clearly not expecting partner to take it out, since it'll pass as South too). It will never bid anything else. With a 13-card solid spade suit, advanced GIB.. opens a weak 4♠. OK, we can probably set that one aside, since there is a flag in the bidding database where GIB is only allowed to run simulations for certain later bids, to prevent it from being super slow every bid. But have South bid like this: [hv=https://www.bridgebase.com/tools/handviewer.html?d=s&n=SAKQJT98765432&a=1HP1SP1NP2CP2DP3SP3NP6SPPP]200|300[/hv] It pauses after the 3N bid to run simulations.. and bids 6♠? It appears as if GIB is only considering bids that are *defined* in the database - ie ones that basic GIB may make with some hand. Couldn't this be improved by allowing GIB to make 'undefined' bids? Assume everyone else will pass (which is how undefined bids work anyway - could even alert the bid as 'to play') and see what works best. The *only* downside is that it will take some more processing time - you need to test each simulated deal in all 5 strains, whereas the original simulation may have only involved a couple. But isn't getting to the right contract worth one bid taking a little bit more time than normal? (And aren't some double dummy analyzers optimized so that calculating all 5 results is simpler than doing 5 unrelated calculations? And if not, restricting it to suits you have some combined length in is probably fast enough already..)

At the bottom of the page, there's a blue bar saying 'Click here to show filter options'. The default is to only show posts from the last 30 days, but you can change it to all.

FYP. For the vast majority of users it works perfectly fine with no disconnects, and based on BBO replies they've never been able to replicate it themselves.

Bids of 4 of a major are dual-purpose (and not covered in your list) - rather than solely obstructing the opponents like 2 or 3 level jumps, you're bidding it in the hope that either you make the contract, OR it preempts the opponents if not. Alerting it here would be silly. To be honest, your suggestion of a 2♥ overcall would be much more of a stretch from what I'd expect..

It *somewhat* does - if RHO had passed and you bid 1♠ / 2♠, it will continue with 2♦ / 3♦ with the description changing to rebiddable diamonds and 18+ total points. Not that diamonds are good enough for that here. The main problem is that it doesn't plan for any further interference from the opps, so gets into huge trouble when that happens. And that's the main reason takeout doubles on hands like this are just plain bad. (Of course, East psyching a weak hand doesn't help). If you think this was bad, be extremely thankful that East didn't bid 3♥. If South continues with 3♠ (showing nothing more than a 6-10 count), then North will bid.. wait for it.. 4♥..

Off-shape takeout doubles is a long-standing issue with GIB. It adds 2 points for the singleton spade and deems that hand to be too strong for a simple overcall, which forces it to double. But it doesn't understand any followups after that even if the opponents hadn't preempted from there. (And won't overcall 1NT with a singleton).

There is a monumental difference between this hand and a 13 card club suit. Trying to pretend you're weak when there is only one suit the opponents could possibly sacrifice in (as opposed to 3), and you have 20 HCP (as opposed to 10 HCP), and you aren't even sure if you're bidding/making slam (if you are and they sacrifice you're holding AKQ in their trump suit).. would be well past insane.

2♣, then jump to 3♥ and see if partner can contribute a spade control.

By my calculations, A+B would say: card / probability of beating contract / average tricks taken: S3 - 31.6% - 3.98 HA - 21.8% - 3.56 H6 - 35.2% - 3.84 D7 - 26.3% - 3.77 CJ - 22.2% - 3.59 So they would indeed recommend the passive lead at MPs - but the heart at IMPs. Of the 1000 deals I generated, if the choice was solely between spade + heart, a heart was strictly necessary to beat the contract on 146, while a spade was strictly necessary on 110, so I can see a limited-simulation GIB that uses A+B's logic going either way.

It was in the OP - Kxx KJT9x QJ AQx. If the agreement for the double is really as described, I don't understand why it was made with that hand either.

You could argue that posting about GIB bugs is immaterial too since they're not going to be fixed :) But they spark interesting discussions about real bridge, whether directly or indirectly, which are never immaterial.

I suspect nige1 was looking at the results sheet for his 12 boards (where 660 happens to be the highest score), rather than the results for board 12 :)

It opens 1♦, but yes, that's definitely not something you have to worry about here :) Unrelated to GIB, the main reason Soloway is good for the two-suited-with-support hands is that it's hard to show by starting with 1♥ - you often have to go through a fourth suit forcing sequence, and by the time you've gotten around to bidding diamonds you could be at the 3 or 4 level, without having shown any shortage yet (and possibly not even clear if you have slam interest or were looking for 3NT, or whether opener is giving preference to your heart suit, or whether they have the missing honor in hearts). So Soloway actually *saves* space and gives you a lot of important info quickly. Suits like AKQJxx or KQJT9x are good for Soloway too, but in that case you have a self-sufficient suit - so don't need to treat these as two suited.

Heart suit isn't good enough? When you are showing a two suited hand with support for partner, all it promises is 2 of the top 3 honors (and could even be 5 cards). You might be mixing it up with the case where you're planning to show a 1 suited hand with hearts. Seems like thepossum got it exactly right. GIB just over counted points (it adds 3 for its void even in your suit).

GIB really passed 1NT too? Yikes.

No, GIB doesn't play crawling Stayman. 2♣ - 2♦ - 2♥ promises invitational values with 5-4 in the majors.