Join Bridge Winners
All comments by Doug Bennion
You are ignoring the author of this comment. Click to temporarily show the comment.
Your best fit will be 5-2 or worse about 23%, 5-3 about 43%, 5-4 about 28%, and at least 5-5 about 6%. Bidding looks routine to me at matchpoints, and did before I looked at the sim, except maybe unfavorable.
March 19, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
5440 hands play about 5 points better when a fit has been found, so those minimums will almost always really grow up. Quick and dirty sims show responder will have at least 3 hearts more than half the time, and the balance of the time responder is a huge favorite to have at least a 4-fit with one of the minors. The best way to find one of those fits is to open.
March 18, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Transfers. X = hearts, 2 = spades, 2 = balanced 12-14, 2NT = balanced 15-17. 2m = natural NF.
March 17, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks for the article.

I wrote an early-generation bridge program so have some experience with the foibles of double-dummy play. Bear in mind I’m speaking from memory here, not having looked at a line of code in 10+ years. Also I confess to little first-hand knowledge of how the BBO bots perform, rarely playing them. I’m assuming the principles of play are largely similar to how Buff played 10 years ago.

Your first hypothesis is correct. Declarer bot has sifted through a batch of simulated deals and chosen the card which generates the best results over those deals, in this case a trump lead . Bot with Kx under AQJT… deals a bunch of hands based on his perspective and sees that the K and the x lose equally, so are equivalent holdings.

At this point I think the programmer has a choice. He can randomize the play of equals, which apparently is what the BBO bots do, or he can direct the bot to play the lowest of the equivalents, here the trump x, which is what a human would do .

Which leads to your second point. The bots apparently discard randomly from Q5 when the 4 remains out, and sometimes ‘fool’ the opps by playing the Q.

So l think your hypothesis compresses to ‘the bots play randomly from effectively ‘equivalent’ cards’.

If the programmer wanted, he could ‘humanize’ bot play by specifying it always play low from equivalents. Or if he wanted to inject a random element of falsecarding — play a card other than low 5% of the time, or whatever.

There are other ‘cosmetic fixes’ a programmer might undertake. In the example case, bot declarer might have cashing cards on the side, say AKQ in a suit. The batch of simulated hands might not reflect any ‘bad’ things happening to an initial cash of one or two of those honors, then getting around to drawing trump. So a cash would be equivalent to a trump play. That behaviour would look weird to a human — cash side ace, draw a trump, cash side king, draw another trump. I seem to recall writing code to mitigate some of that kind of goofy play sequence.

A point about the bot leads. They do lean conservative. I once built a 100K sample of double dummy leads, compiled and ranked the results. If I wasn’t so lazy, I would publish the results. Suffice to say, as you pointed out, double dummy engines ‘dislike’ leading from honors, often defaulting to xxx and xxxx. . How those bot preferences relate to human play, is problematic.
March 15, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
The bots don’t (or shouldn’t) describe their hands when they announce ‘this bid shows'. They should explain the meaning of the call, not what they actually hold. Say the auction has proceeded (1) 2 (2) 2. The same ‘this bid shows’ will be made for 2 whether a bot makes the bid, or a human. The bot examines his hand, makes his bid, then looks up some table/algorithm which purportedly explains what the bid means — 6+ spades, 8+ HCP, 2+ clubs (or whatever). Everyone at the table is entitled to know that description.

That’s a prime source of bidding error/poor judgement of course. Advancer bot happens to hold 5 spades and 5 HCP. His ‘bidding logic’ tells him that hand should bid 2, so he does. He now looks up the ‘what 2 shows’ table, and announces to the table ‘6+ spades, 8+ HCP, 2+ clubs’. In this scenario his side might get into trouble because overcaller thinks advancer is stronger and longer. Sometimes the bidding logic is problematic, sometimes the auction description likewise, sometimes both. Sometimes the bidding anomalies will impact play.
March 12, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
If the seed is the same, I don't understand how one bot can lead the A, and another the K. The difference in play might result from the K lead possibly suggesting/implying/promising the Q, leading to different sims.
March 12, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Play 2 ranges.

Proper shapes for takeout doubles are fairly rare, and a good chunk of them will be for the ‘wrong’ major, so don't bother wasting bids on them until suit is revealed. Play transfers so X is hearts, 2 is spades, then 2 is a weaker notrump range maybe 13-15, and 2NT is 16-18 or whatever suits.
March 1, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
All 5332 in 1 or 1NT really does sharpen your 1M openers and followups.
Feb. 28, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Might be useful: Google search beginning with the site name like so;

bridgewinners.com flannery worst ever convention
Feb. 23, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
There is a good chance RHO has a weak notrump, and if you also have one, it's more likely your hand than theirs (you have position), so start by overcalling 1NT with some weak NT range and a balanced hand. RHO starts 1 short, you would overcall 1NT with the likes of 2434 or 3244 or 3325 or 3424 or basically any balanced hand that doesn't fit a classic takeout X. A corollary is your takeout doubles are sharpened. Don't insist on a stopper, but your followups might include a stopper-ask.

You'll need a runout routine, and a scrambly kind of Stayman to locate those major fits you preempted yourself out of, including Moysians.
Feb. 22, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
That hand should perform very well. It has a 5-card suit and awesome spot-quality. The two T9’s are worth a point and a half.

That hand should perform very poorly. It has two very attackable weak suits. Often the best-performing hands are 4333 with all suits stopped.

Since the lead will be double dummy, I’m voting for a less-than-spectacular performance. Not a favorite to make 9 tricks.
Feb. 17, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
High if you wouldn't mind partner leading the suit should he ever get the chance, otherwise low.
Feb. 16, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
It’s an instructive hand. A little shape goes a long way. Say opener has 14 HCP. If opener is 5332 and responder has that hand, game makes ~ 38%.

Make opener unbalanced (not 5332), and percentage making shoots up to 74%.

Now make the spade 7 the club 7, so responder is 2-4-3-4, and if opener is 5332, game makes 58%

With responder 2-4-3-4 and unbalanced opener, game makes 80%.
Feb. 15, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I learned that a limit raise asks opener to proceed to game unless he is plainly minimum. I tested a few opener holdings with a double dummy solver which might shed a little light because we're just blindly stumbling around lol. The hand looks to me like a good constructive raise.

Anyway if opener has 14 HCP and is NOT 5332, 4 makes about 74% of the time. That hand is plainly not plainly minimum.

With 13 HCP and NOT 5332, 4 makes 60%. IMO that general holding also isn't plainly minimum, largely 5431 and 5422 hands.

With 12 HCP and NOT 5332, 4 makes 44%. That's bordering on too minimum; some hands you would pass, others not, and form of scoring would come into play.

The real drecks would be the 5332 holdings. 12 HCP and 5332 makes just 8% geesh (I double-checked that result).
Feb. 14, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I think it is problematic to value most hands, before you know which strain you will be playing. You hold AKQx/QJTx/xxx/xx. Partner opens a 12-14 1NT. If you find an 8+ major fit, you’re close to being a favorite to make 4M. But if you don’t find a fit, that hand is sub-par for notrump purposes — because of the two unstopped suits. You’re considerably better off with something like (using same honors) AJxx/QT9x/Qxx/Kx which has all the suits stopped.
Feb. 1, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Two 50K sims (to see ~ error margin), opener has 5+ spades and 11+ HCP. Overcaller has 11+ HCP. Spade lengths:

> 5 0.73% (0.70)
=5 4.37 (4.30)
=4 14.60 (14.93)
=3 29.22 (28.59)
=2 30.87 (30.84)
=1 16.86 (17.05)
=0 3.36 (3.60)
Jan. 30, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
For the HCP ranges specified in the post, the frequencies for 2 are about as follows, for first seat:

Flannery (5 or 6 hearts) ~0.9%

Weak (6-suiter, max 4M) ~ 1.9%

Multi (6-suiters, max 4OM) ~ 3.9%

BAL (any 5332, not 5422) 18-19 ~ 3.2%

Prec ~ 0.6%

Roman 17-20 ~ 0.2%

Roman 11-15 ~ 1.5%

Ekren (4-5 both M) ~ 5.5%
Jan. 27, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
David: Thanks for verifying. I'm getting roughly comparable matchpoint results. For example for AKxxx suit, 1NT wins/ties/loses 44/26/30%, and average IMP win 0.54.
Jan. 18, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Some 5-3 info. When opener has specifically 3 hearts (includes 4333), then when responder’s heart suit is 4+ HCP, 1NT makes the ‘normal’ 57%, 2 making zooms to 72%.

When opener has 3 hearts but won’t be 4333, 1NT down a bit at 50%, 2 up more to 76%.
Jan. 17, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I was curious, so looked at what kind of difference suit ‘quality’ makes in a 5-2 fit. Opener has 12-14 balanced, with 2 hearts. Responder has 7-9 HCP and 5 hearts in a 5332 hand. Sample size is 2000, large enough to be comfortable with the results.

Seems the suit quality has little impact on the chances of making 1NT, but a major impact on the chances of making 2.

When heart suit is xxxxx, 1NT makes 55%, 2 makes 28%

When suit is Qxxxx, 1NT makes 55%, 2 is creeping up at 35%

When suit is Kxxxx, 1NT makes 57%, 2 makes 39%

When suit is QJTxx, 1NT makes 56%, 2 makes 46%

When suit is AKxxx, 1NT makes 63%, 2 makes 46%

When heart suit has a generic 4+ HCP, 1NT makes 62%, 2 makes 45%
Jan. 17, 2018
.

Bottom Home Top