[1175]

This hand had a very suboptimal bid (and reminds me of another question):

---

One South managed to bring in some twelfth-dimension bidding:

---

The 6♠ bid by East seems a bit light on HCP (according to the description) as does the (correct) 4♥ bid for that matter, but aside from that, what tricks did South expect to take after North showed five clubs? Making, for -2470, and -10.1 IMPs. I have to ask here why West bid hearts before diamonds (only on this sequence - if South bids either Pass or 3♣, then West bids diamonds).

One South put himself in an incredible winning position,... and then threw it all away:

---

The redouble looks as random as any bid I have seen for a while, but then South compounded his problems by ruffing the third diamond low, allowing East to overruff with the ♣6. What should have ended up as -800 became -2200 and -5.4 IMPS (actually 0.5 IMPs better than -2210, though).

At seven tables, North-South never entered the auction (other than a double), and East-West quickly wrapped up a grand slam (-2210 and -5.9 IMPs):

---

At six tables, South made the rather pedestrian 3♣ bid, which kept East-West out of a grand slam (-1460 and +6.9 IMPs).

---

I want to go on a bit of a tangent here. A huge percentage of my big IMP losses in the duplicate IMPs comes from sacrifices not taken. Almost all of said sacrifices come from unilateral actions by the carbon-based unit at the table. The Robot almost never sacrifices, even when it seems "obvious." On this auction, I think that North should bid 5♣, which will not stop this freight train, but might push East-West into 6♦ instead. How would the bidding proceed after a 5♣ bid (or maybe even a 6♣ bid by North)? To what degree does the free Robot consider sacrifices? Does the advanced Robot handle them better?

I have already learned that the Robot does not deal with 4-level preempts well, and on this hand, mine caused West to make a bid that, shall I say, does not express the full values of its hand:

---

Obviously, North-South have no tricks against spades, either, but -710 returned a healthy +15.4 IMPs. Of course, I want to ask the obvious questions: Why did the free Robot settle on 4♠, and what would the advanced Robot have done? To continue with the thoughts on sacrifices, clearly North should not (and did not) raise clubs here, when it appears that East-West might have settled in a contract well below the trick-taking capabilities of the East-West hands.

[smerriman]

1175, on 2024-June-29, 03:00, said:

I have to ask here why West bid hearts before diamonds (only on this sequence - if South bids either Pass or 3♣, then West bids diamonds).

It simply has a rule that after an overall, it should prioritise 2♥ bids over a new minor. It doesn't have such a rule in other situations.

1175, on 2024-June-29, 03:00, said:

On this auction, I think that North should bid 5♣, which will not stop this freight train, but might push East-West into 6♦ instead. How would the bidding proceed after a 5♣ bid (or maybe even a 6♣ bid by North)? To what degree does the free Robot consider sacrifices? Does the advanced Robot handle them better?

GIB is programmed to raise partner's preempt here with suitably weak hands (limited in both points and controls), but also requires a singleton, which it does not have. Also, the old version has one rule when favorable, and another rule when unfavorable (with exactly the same conditions), but totally skips equal vulnerability, likely by accident. The current version of GIB appears to be the same, so will bid 5♣ if you move a diamond to the heart suit, unless at equal vulnerability when it passes. The proper robot isn't going to do any better here; this isn't a situation to simulate in.

1175, on 2024-June-29, 03:00, said:

Of course, I want to ask the obvious questions: Why did the free Robot settle on 4♠, and what would the advanced Robot have done?

The free robot is only permitted to leap to slam if it knows it has a fit in that suit along with the appropriate strength. Likewise, it can't bid 4NT here without a fit, and the highest matching remaining rule is falling back to partner's suit with a 7 card fit.

The old version of the proper robot leaps to either 6♦ (corrected in this case to 6♠), 6♥ (passed), or 6♠ (passed), depending on its mood. It's not allowed to consider grand and it didn't think 4NT was Blackwood while the current version does, so it's possible the current version of GIB would try that, or a 5♣ cue raise, though they may be excluded based on whatever mysterious logic says some bits are permissible and others aren't. One big issue with the proper robot is that in considering potential bids it can make, it extrapolates the bidding based on the free robot's rules. This is the primary time the free robot's rules are important (when it's not about what it should bid *now*, but what it would bid *later*). Since the free robot is never going to bid grand by itself, the advanced robot is generally unable to tell that going slow may give it a chance to *simulate* at a later round and come up with grand once it has more information. jdonn, a BBO developer who used to post on the forum, say it was one of his wishes to alter GIB to force it to choose the slow option. But then he disappeared in 2018, never to be heard from again, and that was the end of GIB development as we knew it.