[svengolly]

4 card end position declarer in Hearts. As South, I lead the spade J which is known to be high. W (dummy) is irrelevant. Robot pard (N) ruffs my J low. Declarer (E) follows suit. I think OK, N has all trump left. N then returns a low trump, which E takes. E plays a high trump, N follows and I'm like WTH? If N doesn't ruff and simply pitches the D, declarer is off 2 because I get my J of spades AND N gets one trump (tenace). It was like the bot figured out the ONLY way to not score a single trick on the 4-card ending.

This is how I remember the position:

---

Is there a rational reason for GIB to do this? Am I missing something (I'm no expert)?

[smerriman]

The "usual" reason for this is when something unusual happens in the bidding (where humans don't have what they promised). If assumptions based on the bidding make the actual layout impossible, GIB may well have determined that ruffing was no worse than any other play. GIB has an annoying habit of taking bidding too literally - especially when there are other 100% lines that don't rely on bidding - but at the same time, that leads to good plays in other situations.

For example, if you've promised an extra spade that East actually has, then it makes no difference whether GIB ruffs or not.

You'd need to show the full hand to be sure (should be available after logging in via My BBO - Hands and results).

[svengolly]

smerriman, on 2017-December-21, 19:02, said:

You'd need to show the full hand to be sure (should be available after logging in via My BBO - Hands and results).

You're right of course. I searched for it but I played a slew of tournaments the day this happened and really didn't think to record the hand. Couldn't figure out which it was.

Regardless, I almost never make weird bids with GIB. I've learned my lesson on that account.

[johnu]

svengolly, on 2017-December-21, 18:32, said:

4 card end position declarer in Hearts. As South, I lead the spade J which is known to be high. W (dummy) is irrelevant. Robot pard (N) ruffs my J low. Declarer (E) follows suit. I think OK, N has all trump left. N then returns a low trump, which E takes. E plays a high trump, N follows and I'm like WTH? If N doesn't ruff and simply pitches the D, declarer is off 2 because I get my J of spades AND N gets one trump (tenace). It was like the bot figured out the ONLY way to not score a single trick on the 4-card ending.

You must mean GIB found a way to take 1 trick in the 4 card ending, not zero tricks. GIB could have scored a 2nd trump trick by leading a diamond and endplaying East instead of a trump, but what was the original hand?

I can say that ruffing with the trump 3 is a clear mistake. If East has a spade, there is no point in ruffing if the jack is high.
Again, without complete hands, can't fully analyze this, but if East can overruff, GIB could ruff with the 6 if East had J4 which may gain a trick.

[steve2005]

Ruffing doesn't matter. When the play doesn't matter Gib doesn't necessarily make the obvious play.. Leading trump on the second trick is the blunder.
Gib should just leads it's ♦ then declarer must give defense a second trick in trump.
Gib usually gets these endings right. Your post is wrong: defense did score 1 trick in the 4-card ending not zero.

What type of tournament/game was this? Sometimes Gib plays worse depending on event.

[svengolly]

steve2005, on 2017-December-22, 07:45, said:

Your post is wrong: defense did score 1 trick in the 4-card ending not zero.

What type of tournament/game was this? Sometimes Gib plays worse depending on event.

Sorry, yes defense did score a trick. I said that Robot ruffed my good trick and took it returning a low trump. I erred in saying 0 tricks.

It was a standard .25 robot tourney.