Posts: 758
Threads: 97
Joined: Apr 2011
Reputation:
53
(07-30-2013, 01:53 PM)ToreadorElder Wrote: Bidding tests are probably impractical. First, there's only a few plausible answers for any particular situation, so the choices are rather restricted. Second, a scoring system now becomes probably excessively dependent on the preferences of the test creator/scorer. Maybe you think giving meld is more important...your meld-asking bids either don't have 20+ meld, or have a notably more offensive hand (8 card run instead of 7, or a 2-suiter). The Bridge World has the monthly feature called the Master Solver's Club...they present *tricky* hands, with the auction proceeding up to this point, and it's your turn to bid the hand. They have about 30 panelists submit their answers, and frequently comments, for each problem. The comments are probably the more critical, as they expose the thinking process. That's what lets you compare and contrast, and really think through a situation. Of course, bridge bidding is much more technical and with a much richer language. Another distinction: there's a common frame of reference, Bridge World Standard. It's a system that's been developed and refined through polling experts and readers alike, so yeah, it's something of a committee project and therefore not perfect...but it *does* cover most situations. We've got nothing formal.
I agree with your concerns and have thought on the subject enough to pencil in the following methods to overcome some of the challenges in Bidding Tests and 2nd-round Play Tests:
1. Get support and contributions from our member base.
2. Restrict the situational play to a point where there is only 1 correct bid (to 99% confidence, some cases will be impossible to achieve 100%).
3. Situations will display some or all considerations a player might have:
- Game Score
- Seat Position
- Hand's Bidding History
- Cards Dealt
- Scorechart
- Cards Played, if any (for 2nd-round Play)
- ...and more as necessary to build a solid scenario
4. Random generations may not be viable, set sequences will likely need to be supplied.
Every aspect of a given Test will have to have a solid justification, so if something is marked as "incorrect" then a reason is supplied to the player. I'm very much just brainstorming this, so don't expect me to start rolling these Tests out anytime soon. I just want to plant a seed in the minds of our members. I am sure I won't do this alone!
It's unbelievable how much you don't know about the game you've been playing all your life. -- Mickey Mantle
Posts: 834
Threads: 112
Joined: Jul 2012
Reputation:
80
I'll strongly suggest 1 thing at a time, by which I mean, don't even *contemplate* play problems until such time as we can reasonably capture the issues of bidding problems.
And yes, the test hands have to be constructed. The latest Bridge World bidding contest has a funny side note from the moderator, about submitting problems that only have 2 viable answers...such hands are rare, because they don't really illustrate much. But there's only a few hand types in pinochle, that reasonably have 3 options...and that, only when 51 for aces is available. Offhand, I can think of 1 such:
              
You're in first seat. Options: 50, 51, 54. But note I'm thinking 65 is not an option...but many people bid it.
Now...rather than a test, what might be *quite* useful, is a POLL. Given that hand and a situation, what should be done?
Posts: 33
Threads: 5
Joined: Jul 2013
Reputation:
6
I would open with 54, but I was taught that the first bid is meld.
(07-31-2013, 04:38 PM)ToreadorElder Wrote: (snip)
But there's only a few hand types in pinochle, that reasonably have 3 options...and that, only when 51 for aces is available. Offhand, I can think of 1 such:
You're in first seat. Options: 50, 51, 54. But note I'm thinking 65 is not an option...but many people bid it.
Now...rather than a test, what might be *quite* useful, is a POLL. Given that hand and a situation, what should be done?
Posts: 758
Threads: 97
Joined: Apr 2011
Reputation:
53
(07-31-2013, 07:10 PM)tony ennis Wrote: I would open with 54, but I was taught that the first bid is meld.
... I can already hear TE's keyboard clicking loudly.
(07-31-2013, 04:38 PM)ToreadorElder Wrote: Now...rather than a test, what might be *quite* useful, is a POLL. Given that hand and a situation, what should be done?
For now, I think this is an excellent suggestion. For anyone who wants to post a 20-card hand and ask "How would you play it?", I would like to see the hand and situation displayed as a POLL with reasonable answers and some silly ones to keep it interesting. Posting can still be done at the Poll in case commenting is necessary.
*Polls should be left infinitely open (I think that is 0 as a deadline) so that new members can always go and vote. Rak, maybe a weekly Hand Poll?
*PowerPinochle Polls are by default limited to 10 options. If anyone needs more, I think I can raise the limit in the Admin settings. Just let me know.
*I don't know if the visibility of the Poll Results can be initially hidden and then switched to public, investigation is necessary.
I'll have to confirm with rakbeater the priorities of my to-do list, but I think the next training and test programs will be trick-taker counting. I may do only trick counting, or jump straight to the 3-part counting/analysis referred to as Standard Hand Assessment Notation in the Help! Opening With A Biddable Hand And Aces Around! thread. The one that talks about Meld - Length of Trump - Estimated Trick Count. I don't have time to do either at the moment, so I'll give it some thought at a later date.
It's unbelievable how much you don't know about the game you've been playing all your life. -- Mickey Mantle
Posts: 834
Threads: 112
Joined: Jul 2012
Reputation:
80
Heh...yeah, I don't think I need to comment about what I think about meld first always.
But the hand I posted has no clear-cut answer, which is why it's a good question. Personally I'd never start with 54, in any situation; I'd either bid 50 or 51. Unless the auction has reached 75 by my next turn...possible but unlikely...I can show the meld by jumping, if that appears to be my best choice at that point. But the 54 opener is constructive, and this is largely a two-way hand.
On the polls, I *doubt* there would be any bidding problem that admits 10 different, non-ludicrous answers.
Trick-taking programs sharply define the hand type on which you can ask. Trick taking potential is ONLY an issue when you play the hand...therefore, the hand has to have a marriage, and preferably, something close to a trump suit. (My requirement to make a save bid is TTKQJJ.) It doesn't have to have a big trump suit, or meld...partner opens with 54, you bid 55, partner passes as opponents pass at their turns. But, this isn't a case where counting your tricks matters. I can think of a few situations where it matters:
1. Making the meld-ask bid...do you have the tricks to ask for meld? 35-7-6 is probably the target...but recognize that some people will open on 6 card runs, or out of fear with something like
   
This hand type may have to consider bidding after partner starts with a meld bid, so it's a useful one.
2. General, medium to strong hands...7-8 trump with extra tricks.
3. The really big hands...the big 1 suiters (9+ trump, no side suit longer than 5), the big 2 suiters.
A hand type you can't include: one where any potential trump suit is 5 cards or less. There's one obvious case where I have to bid this to play: parter bids double aces. But I'm not counting MY tricks, I'm counting on his. I can offhand construct 1 other case where I'd bid this to play...double bidder out, partner started by giving meld, and I can't give meld. But this is also an easy trick-counting problem...your aces == your tricks.
But the problem really is...by whose standard? We've had these discussions; my estimations are often higher because I estimate the expected number of tricks. rak prefers to consider more negative hands. Then there's estimating the play...there may be notably different trick expectations depending on your line of play. Say your hand is:
 
There are 3 plausible lines:
1. at trick 1.
2. , , then at trick 3.
3. , , at trick 3
In my experience: line 3 beats line 2 by 1 to 2 tricks overall. The (and the 7th diamond) may become tricks. Partner can ruff diamonds. An opponent with a dangerous trump holding may be forced to ruff, reducing my trump losers. Line 1 is a gamble; it might work great, and it can fall apart (oh gee, watch those diamond aces get ruffed away.) The hand has 4 or 5 tricks from spades, plus 2 diamond tricks for 6 or 7, following line 2. Line 3's minimum is that same 6, and offers chances for 8 or 9.
Posts: 642
Threads: 111
Joined: Apr 2011
Reputation:
10
This is late to the conversation, but another variable is your opposition. Since I have been playing in the beginners lounge for our pinochle nights, I can attest to the huge difference in defense played against you compared to the higher level players, and I never played at TE's rating level, so I can only assume it is even tougher. Tricks you can count at the beginner level, don't always pan out at the higher levels, because opponents are generally better defenders.
Posts: 834
Threads: 112
Joined: Jul 2012
Reputation:
80
It *can* be tougher, yes. At low levels, say South is declarer and cashes aces...no reason to think he's bagging. He leads to North, who cashes aces. North exits. What do you see? East and West start cashing their aces, all over the place. NO! 
But one of the things that's necessary, is you have to change your declarer play to a degree.
|
