Help
i have a frustration trying to master gib card play and while 99% is my own fault,,as i have started to improve i find in some instances with identical bidding and card play gib will open a different new suit . 8 players will get say a club lead and the contract will make and 4/5 others (usually smaller number) will get say a heart suit lead and contract will fail.
is this just a limitation in gib or a slightly random factor that mimics real life on borderline situations.. i presume the speed at which u play different cards is not a factor . just curious
Page 1 of 1
gib leads
#2
- Group: Advanced Members
- Posts: 4,151
- Joined: 2003-May-14
Posted 2013-March-03, 14:31
For main bridge club hands (not robot tourneys), there are at least 3 versions of gib in play. Advanced bots vs. std bots; adv bots have a lot more thinking time. Also people running old downloadable BBO (as opposed to flash client in web browser), run even older version of gib bidding db, and have access to variable thinking time controls. These things account for the variation, given same human bids/card selection.
#3
- mmm Duvel
-
- Group: Advanced Members
- Posts: 10,728
- Joined: 2003-July-30
- Gender:Male
- Location:Belgium
- Interests:Duvel, Whisky
Posted 2013-March-05, 03:51
GIB bases his decision on random deals which the auction can represent. Some deals will suggest one lead, other deals will suggest another lead, and the average result will be important. When there's one lead which is much better than the rest, GIB has an easy decision and it will be very rare that the random set will result in another lead. However, when 2 leads are very close (for example if the exact odds of lead A is 81% successful and lead B is 82% successful), then a relatively small set of deals will have a small deviation and the 'inferior' lead may become GIB's choice.
The more random deals GIB analyzes, the more accurate (= exact odds) the result will be. The odds however don't always give GIB the best result on a single hand.
So in our example lead B will be chosen most of the time, lead A some of the time. When lead A defeats the contract and lead B lets the contract win, then the minority will get the preferable lead. If it's the other way around, then the majority will get the preferable lead. In the long run, you'll get some good leads and some bad leads, so it should even out.
The more random deals GIB analyzes, the more accurate (= exact odds) the result will be. The odds however don't always give GIB the best result on a single hand.
So in our example lead B will be chosen most of the time, lead A some of the time. When lead A defeats the contract and lead B lets the contract win, then the minority will get the preferable lead. If it's the other way around, then the majority will get the preferable lead. In the long run, you'll get some good leads and some bad leads, so it should even out.
"It may be rude to leave to go to the bathroom, but it's downright stupid to sit there and piss yourself" - blackshoe
Page 1 of 1