smerriman

As you know, the bot only works double-dummy.. double dummy, GIBs play at trick 3 was irrelevant; your play at trick 6 was not.

The short answer to your questions is that if you're playing with friends, get the host to choose the 'Permission required to play' setting, which prevents timeouts altogether. This will sound like I'm picking on you, but I'm honestly wondering how you managed to look for this, and not find it, given it has been literally asked and answered over 20 times in the last couple of months. Clearly something is wrong with BBO's system in that people aren't able to find things; where exactly did you look? This is just a brief sample of the requests, there are many others: Casual players getting kicked out of game due to delay Adjusting response time BBO time out Getting booted and resuming play Avoiding robots in table and team games Ejection by computer Took to long to play and got kicked out Time available to bid When we use history to review the hand the opening bidder on the next hand is ejected for being slow Uninvited robot Bidding/play timing Log off during play Pause in play Automatic boot Table disruptions

To me bids after 2♠ would be control bids, so I wouldn't have bid 3♣. No idea how to find clubs, but the bidding might have gone something like.. 1S - 2C 2D - 2S 3D (denying ♣ control) - 3H (showing ♣+♥ controls) 3S - 4C 4N - 5D 5H - 5S but I'm still not able to count 12 tricks; South might have 4 clubs, or the ♥Q instead of the ♦Q, so I'm probably giving up; I'm sure there is a better solution. I always find slams hard to bid when side queens are important. Aside - what would 5N mean after the 5♠ queen denial show in this sequence? Pick a slam I guess, so not appropriate here. Edit - interesting; after some research I've discovered three level bids shouldn't be control bids here. Learn something every day..

No, GIB plays both IMPs and MPs. The general logic is the same, but when it calculates the outcome for possible simulations, it affects how they are scored. This therefore changes what it plays based on the scoring. Accordingly, I don't see anything wrong with deal 16; obviously if it were IMPs, taking the Ace would be correct, but at MPs the potential overtricks are important. Had you rented an advanced bot or a basic bot? Deal 14 is silly, but makes sense if it were just a basic bot. At the point it leads from hand, double dummy tells it all trumps are equivalent (all making +1), so it doesn't distinguish them. Only Advanced GIB has a single-dummy version built in which is able to realise that it does matter.

There's also the issues of paying for accommodation, which is starting to become a talking point even in our own country.. and getting people from the hotel through an airport to the plane without any contact at all doesn't sound like an easy task (or even keeping them contactless in an airport hotel to begin with).

I don't think there is a "right or wrong" answer when it comes to these types of hands. Good 6-5 hands are far more powerful than just HCP, so reverses can be made lighter than a strict HCP requirement. But with just an 11-count and two singletons, I would have bid 1♠ as well. Advice from some more respected sources: Larry Cohen suggests 1♠ when holding AJxxx AQxxxx x x. and Bridge World Standard says "With a minimum-range five-six hand, open in the higher and shorter suit only when the long suits are adjacent." Which therefore advises 1♠ here as well.

I think someone needs to tell the hosts to move to the 'Casual' server. There are 6 online currently, but none of them on the casual server, so they show up as offline as soon as you move. https://www.bridgebase.com/help/v2help/hosts.html

If 2♠ promises an opening hand, despite having passed originally, she's not going to be bidding 2♠ very often..

When you're invisible, your name at the top right has a white background. When visible, it's a darker grey/green.

Is the moral of this story that even though GIB's bidding is bad, humans can still be worse?

HCP of 100 hands from the anonymous MP option: 4,11,13,5,12,11,9,11,10,0,11,13,7,6,10,19,9,10,12,13,15,13,2,7,5,14,6,5,14,14,23,15,4,11,9,8,9,10,6,7,8,10,10,12,9,10,12,6,2,11,5,11,19,3,11,11,11,13,8,7,23,13,9,18,5,13,16,18,6,9,11,3,21,7,7,5,8,5,12,6,20,10,14,9,5,10,8,10,3,12,11,13,9,7,9,4,4,7,14,13 Average is 9.89 HCP. Perhaps you've played best-hand in other formats before, and have forgotten the rarity of actually having hands that good when playing with randomly dealt cards. And the chances of being dealt 79 or less HCP over 10 boards isn't "extremely unlikely" - it's about 1 in 18. If I play 10 boards on 12 separate occasions, there's a 50/50 chance I would also be able to say I was "once" dealt 79 or less HCP. And far more so if you were fudging the numbers and only averaged out 10 consecutive boards that you noticed were a poor run. Indeed, in boards 89 - 98 of the above I magically averaged 7.9 HCP. But there was about an 80% chance I could find a run that bad.

Where exactly are you playing without signing in? It's been demonstrated time and time again that there are no biases in the main dealing algorithm, though the most common Solitaire options (Just Play Bridge and Bridge 4) are in fact best-hand, meaning the deals are rotated to give you the best hand, so clearly you're not using that. I'm not sure how the Matchpoints and IMPs options work; I suspect it's something like the old weekly free tournaments, where you'll see exactly the same set of deals until the next one is rotated in a week later. So perhaps you're just playing the same set of deals, which happened to be one lacking in HCP, which is obviously going to happen with random dealing. (Edit - it's not - appears to be a steady stream of hands). Or perhaps there is indeed a bug in the format you're using, since I've only ever analysed deals in the main software. But would need to know what that is to test.

I think it's the other way around, to be honest. Making changes to GIB appears so complicated that they can't afford to spend any time on it, which is why nobody works on it anymore. And that there probably is a long term plan to switch to Argine, in which case there is no reason at all to spend anything more on GIB.

Sorry, but while there is nothing wrong with the specific bids mentioned in this thread, GIB's bidding system in general is monumentally broken. While it's true a lot of people don't know the system, those that have played with GIB regularly know the vast plethora of issues with it which have been around unfixed for many years.

GIBs responses are perfectly normal in this situation. 4NT is definitely Blackwood; 4♣ is a standard splinter, a new suit is a game try (which could turn into a slam try), etc. But perhaps you were playing Solitaire, where there have been reports of connection issues causing part of the auction to get lost, rather than related to GIB in any way.

Well, not really; with 9 (or 30) more breakthroughs of the same type, you're not going to reduce the rate any more than with 1 breakthrough. Or in fact, it'd probably make it worse, if you plan to give everyone 10 different drugs at the same time..

I think you misunderstood his response. Many people play that 1♦ by responder denies a 4 card major (unless holding 5+ diamonds and a game forcing hand). Opener then does *neither* of your two options, but instead bids 1NT to show a balanced hand - 1♥ and 1♠ would both show unbalanced hands. (You never miss the 4-4 major fit since responder can show it next time if holding the game forcing version).

While it's hard to know the exact simulations GIB uses (eg it wouldn't bid 2♥ with *every* single hand that matches the description above), I've done a few rough simulations and 2♥x goes down a surprisingly large proportion of the time. So even though it seems insane to humans, and may well still be insane, I'm no longer too surprised that a limited set of simulations could well tell GIB to pass.

Obviously if you play transfers or have a mixed raise available, they are better bids. But with "standard" bidding, you don't. You can show 3 and 4 card limit raises (standard for the former is redouble then a spade raise), and preemptive raises, but only have one bid available for a single raise, be it 3 or 4 cards.

I consider 3♠ as weaker than 2♠, which is the bid I would have made. You can still compete to 3♠ next time if partner has no game interest.

Most of the time it means it is a private / club-based tournament, so the TD uploads a 'web list' of club members who can participate (or make it restricted to their 'friends'). Enemies is more likely to be on an 'exclude' list, so people who have upset the TD in the past can't join.

That is true about every single bid in bridge.

I think spotlight7 might have been taking a leaf out of johnu's book with sarcasm, given 2♥x is making +2. Basic GIB responds 3♣ - perhaps the advanced bot fluked a set of simulations where passing works out better, but it really just seems like a stuff up on the robot's end.

Yeah, bidding freely to the 5 level shows a ridiculous number of points with GIB. More often comes up when you bid unusual 2NT and then try to sacrifice in 5 of the minor, which shows about 28 points I believe.