
finesse157
Members-
Posts
33 -
Joined
-
Last visited
-
Days Won
1
finesse157 last won the day on January 3 2022
finesse157 had the most liked content!
finesse157's Achievements

(2/13)
2
Reputation
-
I thought that, but that would have given south 5+ clubs. I gave the simulator the trump void hand, and a few hands with 5/5 S/C. My simulation still gives 60+% for the ruff.
-
I think I am missing something here, but why would you not give a ruff when all the Diamonds are known https://www.bridgebase.com/tools/handviewer.html?lin=st||pn|,~~v3fakebot,~~v3fakebot,~~v3fakebot|md|1SAQJ973HJ6DJCK632,S6HKQT874DKTCA954,SHA953D976542CJT8,SKT8542H2DAQ83CQ7|sv|e|rh||ah|Board%203|mb|1S|an|Major%20suit%20opening%20--%205+%20!S;%2011-21%20HCP;%2012-22%20total%20points|mb|2H|an|Two-level%20overcall%20--%205+%20!H;%2010+%20HCP;%2011-18%20total%20points|mb|P|mb|2N|an|11-13%20HCP;%20stop%20in%20!S|mb|P|mb|4H|an|13+%20HCP;%20strong%20rebiddable%20!H;%2014-17%20total%20points|mb|P|mb|P|mb|P|pc|D5|pc|DQ|pc|DJ|pc|DT|pc|H2|pc|H6|pc|HK|pc|HA|pc|CJ|pc|CQ|pc|CK|pc|CA|pc|HQ|pc|H5|pc|S2|pc|HJ|pc|HT|pc|H3|pc|S4|pc|S3|pc|DK|pc|D2|pc|D3|pc|S7|pc|H7|pc|H9|pc|S5|pc|S9|pc|CT|pc|C7|pc|C2|pc|C5|pc|C8|pc|ST|pc|C3|pc|C4|pc|D9|pc|DA|pc|SJ|pc|S6|pc|D8|pc|C6|pc|C9|pc|D7|pc|S8|pc|SQ|pc|H8|pc|D6|pc|H4|pc|D4|pc|SK|pc|SA|
-
The "Free Acol Individual Daylong (MP)" tournament seems to disappear depending on which login I use. Other people I talk to have a similar issue when using specific accounts I can see the tournament if I use the Director account for my club, but today I cannot see it under my own account. Other people have reported the same issue and it always seems to be on their same account.
-
P-P-1N-X 2D-P-2H-3C P-3N-P-P X Let's assume for now that the 3C bid description is wrong and that the bid should be 17+HCP as per Acol. From that bidding, you would assume that the 3N bidder had 6-8 HCP and a stopper in H. So making game seems reasonable. So, it was a surprise when lefty doubled. It turns out that P has JTxxx-Jxx-xxx-xx. What I think is happening here is that lefty (Argine) knows that the description and implementation of the 3N bid is wrong. So, is this an example of Argine(West) fielding a psyche by Argine(North) https://tinyurl.com/2h4j8xg2
-
Is there a problem with the Acol room in the free dailies. I cannot see the room to play the free daily. Using my club login though, I can see the room.
-
finesse157 started following Free Acol room missing
-
This is an open question for the GiB developers... Should some hands be removed from a score where there are enough reports of the AI going AWOL? TL;DR; We all agree that the robots are only human and will get things wrong sometimes. Chances are that this will have a negative impact on your score and your sanity (see any post about Argine!) If we consider 2 types of error. One where the AI has been duped (for example where you upgrade a hand and it counts the points wrong). The other is where the programming database is wrong (Argine) or the AI blunders (Argine and occasionally GiB) In the second case, if enough people who played the hand reported it, could it be removed from the general score. Even if enough people did not report it maybe anything where 1/2 the scores are -7 down could be removed.
-
I would whole heartedly disagree with the "bid very well" comment. Playing in the free daily, Argine has a WTF ratio of about 1 in 8 hands. Anything involving a Q bid is going to be a disaster. The definitions are way off for too many sequences. However worst of all is the hand valuation. In response to a 1S opener, both of these were "worth" 3S... Jxxx-xx-Ax-xxxxx AQxxx-x-KJxx-xx
-
Without going into specific hands, the GiB Acol bidding system has several fundamental flaws. I’m not sure that all the problems are down to the descriptions of the bids, but something really basic is going wrong. From worst first Forcing bids being passed regularly Inability to value distributional hands in response to NT openings Bidding to game after intervention and not bidding the same game without intervention NB the intervention is totally irrelevant in most cases Frequent discard issues when the distribution is 100% known Picking a 50 line when there is a 100% line. This may be down to it’s understanding of the bidding/play
-
I have a hand in my hand history that I would like to put in a BBO Forum post. I can export the tinyurl, but how do I save the hand in a LIN file so I can get it on the forum?
-
Another situation where Gib "knows" that it is making exactly no matter what card it plays. Gib does not use any kind of fuzzy logic. You have shown 55?? and therefore must have it. Introducing fuzzy logic into a bidding/play system is possible but I think beyond the current design of Gib. In fact, I have implemented a simple system that uses FL for the last few tricks and this is currently beating Gib at about 57% over 10000 hands.
-
GAME FORCE with 8+ pts and a misfit-Really?!
finesse157 replied to steve2005's topic in GIB Robot Discussion
It's not a bad contract a priori. I'd be happy in this at IMPS. What is strange is the 3H is forcing. The choice of force or pass. Not my cup of tea I'm afraid. -
High-level bidding is particularly difficult for robots - this sequence may not be fully described in the bidding DB. It is also possible that Gib is passing the buck. It "knows" that you have a fit, and therefore it "knows" that you will do the right thing and bid 4S.
-
X to show a strong hand (and TOX) is reasonably common convention. What has gone wrong here is the Total Point valuation causing Gib to overvalue the hand. TPV is a common failing of the Gib bidding system where shortages and solo honours can be overvalued. In addition, the use of TPV before you have found a fit can cause even wilder point valuation as you are effectively giving value to shortage before you know that the shortage can be fully efficient. The most common effect of the failure of TPV is in mismatched NT contracts - high TPV but low trick-taking possibilities.
-
The first Pinned topic on this forum quotes. "For example, GIB never miscounts a suit. It never misses a spot card. It never forgets its own system. It never fails to notice… well, anything. Overall, GIB plays better than the average player on BBO – in fact, this has been proven, as when GIB plays in BBO tournaments that humans played in as a test, it consistently performs well above average." I have seen many cases where this is not true, or needs caveating. "GIB never miscounts a suit" Yes, it does. In the case where the end-play needs a count of the partners suit, it can get the discards wrong" "It never misses a spot card." Again, that is true, but in defence it does not always play the obvious card for partner. For example, it "knows" that its has QJT7 and that all cards are "equivalent". However, it may play the 7 making it hard/impossible to catch this. Also, it can miss the spots if you false card "It never forgets its own system" O yes it does, and with a regular (~10%) occurrence. Frequently it makes bids far outside of the HCP/TP in the hand. Even worse, it can frequently q-bid cards that it does not have - usually the 2nd q-bid. So, the question is, how to avoid these errors, and what to do when it happens. As far as I can tell - the only solution is really, really reign in your competitive bids and this spoils the game somewhat.
-
Do you adjust? It's a matter of....
finesse157 replied to finesse157's topic in General Bridge Discussion (not BBO-specific)
This is the perfect answer, thank you. I should have stated that the game was online so the 2D self-alert was late and came after I had bid, and at the same time that the 2H bid was made. It was at this point that I called the Director (ASAP). So 21B1a applies exactly, but how should the director have handled it? FYI. With the correct alert, I pass and bid 2N (leb) over the 2H bid. With the director encouraging us to continue (and tell him if we were damaged), I did not have the chance to take the bid back. TBH also, I direct online on the EBU occasionally, and I might have made the same decision. Personally I would give a score/Av or score/Av+. However, I was not looking for an adjustment, rather to learn what went wrong :)