[Cthulhu D]

mikeh, on 2014-November-26, 18:51, said:

I don't have any familiarity with the assumed fit style, altho I am aware of it and the frequency-based arguments that suggest that most of the time one has a playable fit.

So maybe my dislike of the method is more due to the innate conservatism that grows in most people as we age, but for a good imps event I don't want to play with team-mates who on a completely routine 'nothing' hand can go for 800 or 1100 at the 2-level into the opps' game, even if it only happens 10% of the time.

Mps: sure...frequency of gain is the main parameter. Imps, not so much

As for justifying the methods as 'perfectly reasonable bridge', colour me unimpressed.

I am capable of mediocrity at any time with any method: I don't need to adopt inherently flawed methods to attain that level of bridge. I would far rather pursue the less attainable, but far more enjoyable, goal of 'reasonably perfect bridge'

I see your point, but I question 'inherently flawed'

I think the worst case disaster potential estimate for any assumed fit method has to be 0.9% - and that's assuming that 100% of the time you blunder into a no fit auction that you get doubled and go for 800 and it's always a part score deal which is just not true. Lots of Moysians are going to play for 1 off, against a part score their way, and that's assuming they get you every time. I think it's closer to 0.1% as a practical matter - for a disaster anyway.

The breakdown of the issues with the misfitting hands are more like:

Hand type  How likely are they to let you off the hook when you have no fit?             How bad is the damage? 
Part score It depends, if you're showing minor+major often they will bid a 5 card major. Potential part score swing, usually modest, unlikely to get a big X
Game       Most serious danger zone, often they can get you for 800 or 1100 vs a game.   This hurts if Vul, if NV it's usually quite tame (3 off vs a making 3NT is usually fine)
Slam       Very, hard to double 2D when you know you might be making 6C                  Not usually that bad. 

A 3 imp loss is possible on lots of boards, the worst are usually when you open in 1st and partner has 55 minors, whereas you would have got out in 2C playing standard, you are totally railed in this. It's rarely a disaster, but it can a part score swing and 5 imps hurts.

This assessment of the is conservative and fits with my practical experience - my biggest penalties have been going for 800 and 1100 against a tight vul game that the field wasn't always making in a swiss pairs (a loss), and a cold NV game (a big loss) in ~2.5 years of playing the methods (so about 3000 boards, perhaps more, but I play TWO assumed fit preempts not one with one partner). You could play quite a few team games and never notice.

But that can happen playing a weak NT! You always take risks making any bid, and while some styles have more penalty risk, you have to take some risk.

[WesleyC]

Even I think opening this hand 2S is pretty gross. However in this match we were playing against a strong seniors pair who are excellent card players, but less impressive in the bidding (specifically too conservative in competitive auctions).

As an example, in our previous encounter, I opened 2S first seat W/R on a similar hand and after 2S (P) 3S, RHO Passed with a 1543 13 count and scored +150. Earlier this year they found a 3NT balance over my 3H opening on a sharp 23 count and scored up 720.

Regarding MikeH's question about what to do with [KQTxxx xx KTx xx] - 1st seat favourable I would happily open 3S.

Having a partnership agreement of wide ranging preempts when favourable makes a lot of sense. Missing a game costs your side less, while the opponents missing a game costs them more. So from a mathematical point of view it is certainly the best time to take a randomizing action.

[WesleyC]

Regarding the Lead problem:

I felt like the situation called for an aggressive lead so at the table I swung the Jh. The full layout was:

---

Not a success...

However I subsequently put together a simulation of the situation:

West is:
14-18 HCP, semibalanced without a 7cm or 6cM, with an appropriate NT stopper based on spade length.
I also removed hands with a doubleton spade and 4+ hearts.

East is:
8-15 HCP, without 4+H, unless 4333.
No very long minors in a strong hand

South is:
No spade support unless very weak, however I didn't remove any hands with a long suits (because i struggled to come up with a foolproof way to work out which hands to remove).

The results over 10000 simulations:
Lead Beat3NT Tricks(average)
-------------------------
JH: 1240 10.385
4C: 1027 10.419
TS: 693 10.597
KD: 799 10.872

In practice, the heart lead would be less successful than this (due to partner having bid hearts on some of those layouts) but it does feel like probably the best shot.

[Cthulhu D]

WesleyC, on 2014-November-26, 20:47, said:

Regarding MikeH's question about what to do with [KQTxxx xx KTx xx] - 1st seat favourable I would happily open 3S.

Having a partnership agreement of wide ranging preempts when favourable makes a lot of sense. Missing a game costs your side less, while the opponents missing a game costs them more. So from a mathematical point of view it is certainly the best time to take a randomizing action.

Incidentally, JLall has a good article on this: http://justinlall.co...on-bridge-pros/

Paragraph 8 of the post and onwards. Mikeh is expressing the client's view. It's impossible to know who is right though. Obviously as a hyperactive bidder I think my style is correct, or atleast more fun for the same cost, but you have to pick your spots.

[gnasher]

WesleyC, on 2014-November-26, 22:06, said:

South is:
No spade support unless very weak

If you're going to open 2♠ open this hand, it seems very unwise to routinely raise with three-card support.

[whereagles]

lol.. ♥J best theoretical lead, but in this case it's the only to give it away

typical unlucky expert

[mikeh]

Cthulhu D, on 2014-November-27, 01:02, said:

Incidentally, JLall has a good article on this: http://justinlall.co...on-bridge-pros/

Paragraph 8 of the post and onwards. Mikeh is expressing the client's view. It's impossible to know who is right though. Obviously as a hyperactive bidder I think my style is correct, or atleast more fun for the same cost, but you have to pick your spots.

I think that is unfair, because my take on Justin's post is quite different from yours. Yes, he is arguing that NA pros tend to be more conservative in the system they adopt, and that higher variance methods may be more effective, but he isn't, as I read his article, arguing that ALL high variance actions are 'good' or that all high variance actions are equal. I have no doubt but that his main thesis is correct, but to go from that and argue that 1098xx is a weak 2 bid is way out in left field. To argue that 4=4 weak hands should be opened with a 2 bid is almost as far out. The reality he was discussing did not involve and generally at the higher levels of bridge does not involve the more extreme forms of pre-emption. I suggest that the evidence demonstrates that while NA pros tend to be conservative pre-emptors, the hyper-aggressive players tend to show up even less, and with much less success, at the higher levels of bridge.

If your ideas really had the merit you think they do, then why don't we see them prominent in the methods chosen by the best in the world? Too much variance is not a good thing unless the methods on balance show a real benefit, after accounting for all of the systemic implications of the methods. I think the evidence suggests that the hyoer-methods don't work well enough.

Incidentally, I am not and never have been a 'client', and have never tried to be a pro, altho I have played on pro teams on occasion. When I did so, the opinion of the client meant nothing to me, since I have never looked at bridge as a source of income....when I played on those teams it was because of the opportunity to play with some extraordinary players.

I don't mind when a good teammate makes a reasoned, aggressive action that didn't work out. What I won't do is play an important event with a pair at the other table whose methods carry within them a virtual assurance that disaster will strike in a way that won't be offset by enough gains. My take is that a style that calls for a weak 2♠ in first seat on the OP hand is ridiculously disaster prone and, if there is proper disclosure, unlikely to discomfit good opps. So I don't see the upside as significant and the downside is obvious.

[mikeh]

BTW, is there not a chance that had N passed, the opps would have reached 4♥, which appears to be doomed?

[whereagles]

mikeh, on 2014-November-27, 03:29, said:

BTW, is there not a chance that had N passed, the opps would have reached 4♥, which appears to be doomed?

hmmm.. the trump pattern suggests a Devil's coup, but I don't think it can pulled off without "assistance".

[Cthulhu D]

mikeh, on 2014-November-27, 03:27, said:

I think that is unfair, because my take on Justin's post is quite different from yours. Yes, he is arguing that NA pros tend to be more conservative in the system they adopt, and that higher variance methods may be more effective, but he isn't, as I read his article, arguing that ALL high variance actions are 'good' or that all high variance actions are equal. I have no doubt but that his main thesis is correct, but to go from that and argue that 1098xx is a weak 2 bid is way out in left field. To argue that 4=4 weak hands should be opened with a 2 bid is almost as far out.

I'm not suggesting that ALL high variance actions may be good - but that's not what we're discussing. You flat out stated that you are unwilling to explore whether high variance methods may be good because regardless of practical frequency, the risk of disasters offset any gain for you (i.e. you have a low risk appetite, which is reasonable).

The argument about 'well, if it was good, why don't pros play it' - four observations

A) Pros have stated that learning two systems is too much, you need to play the same system all the time for your 'serious' partnership.
B) ACBL and WBF restrictions ensure that a lot of methods are banned in some or all competitions you play. For example, assumed fit pre-empts are banned in ACBL midchart competition. Every professional pair plays these events from what I can tell.
C) The Ekren's assumed fit preempt was certainly popularised by Helgemo (or atleast he brought to the Bermuda bowl, which is where it was indirectly exposed to me)... so they do.
D) JLall makes this very good point: If you are a top pair playing against weaker pairs (which, if you are in the top 10 pairs in the world, is basically all the time), you don't want to play high variance methods! This ties into both A, B AND the risk aversion issue.

There is another issue as well - Fred has commented that when playing with team mates who didn't play standard weak 2 preempts he felt as though he was under huge pressure when one came up at his table because his opponents wouldn't get the same auction and this was a possible swing board. Playing Ekrens or whatever has a cost, you cannot play a different possible weak bid - and this creates the same pressure!

The combination of these five factors mean that imho, looking at top players who regularly play in the US to see if these methods are good or bad is not useful, because they are unable to play these methods. (You can make a similar analysis to quickly see why no-one would play forcing pass even if it was massively plus, because you can never play it..)

This makes the entire thing a complete crapshoot. We really have not substantive evidence whether, say, Ekrens 2H is a good idea (nor how much variance it actually has). The only way to see if it's good or not is, realistically, to program Jack or wbridge 5 to play the convention, then deal and play 5000+ hands to see if it is plus or minus imps.

Edit: Another intresting note - I think it's generally hard to get new methods adopted, see: total failure of transfer responses to 1♣ to become popular with American pairs, and even the general lack of uptake in Europe.

This post has been edited by Cthulhu D: 2014-November-27, 05:02

[Jinksy]

Cthulhu D, on 2014-November-26, 18:26, said:

     Style One        Style Two
2C   Strong only      Weak 2 diamonds or strong
2D   Weak 2 Diamonds  Weak with diamonds and a major
2H   Weak 2 Hearts    Weak 2 hearts
2S   Weak 2 Spades    Weak 2 Spades

Style Two is behind Style One on Strong 2C auctions (arguably, and not very much, though the odd slam bidding that starts 2C-2NT-?? Damnit, I've got the strong hand and he's inviting to game opposite a weak 2? is difficult to bid), you're behind on weak 2D auctions because you must cater for the strong hand when you are weak with diamonds, but you're ahead on the 6% of hands when you open the assumed fit 2D. BUT, that Frelling 2D opening has a bunch of disaster potential.

I feel it's perfectly reasonable bridge though.

I haven't played Style 2, but out of curiosity, is 2N over 2♣ the generic invitational bid? If so, why not switch it (or something slightly more nuanced) for 2♥? Then when opener has the strong hand and responder makes a space-eating bid, you at least know his primary suit.

[Cthulhu D]

Because we use 2H for the double negative.

More seriously that is very workable, it depends what you want to do with your continuations when 2C is strong. We like the 2H super negative (it's basically zero cost because if you have a QJ or less and partner does actually have the weak option he can just pass on the basis that they have atleast 27 HCP and probably slam, and realistically if you are that weak partner is gonna be strong or someone has bid already) because that gives us nice absolutely GF auctions after 2C-2D with kokish and stuff. We've decide to sacrifice some accuracy and preemption when pard has the weak 2D.

I'm not sure if it's right, just where we have landed and inertia has truly set in. I have no idea where the right balance is (I think for example that maybe 2C-3C should show a very bad hand with 5+ diamonds for responder a la bramars' transfer negative, then if opener is weak and fourth seat bids opener can take preemptive action, and if opener is strong he knows what responders suit is and that his hand is awful).

I've looked at that idea (2H as the invite) or something similar to manage 2C as weak diamonds, weak spades or strong, but it doesn't save that much space in the end.

The other option is make the invite 2S but I work about wrong siding 3NT - most of the information about the method is in Dutch so I unsure about what the general experience is.

Highly recommend it though if you play an artificial 2D and a strong 2C.

[Cthulhu D]

Edit: Double post.

[Cthulhu D]

For some reason cannot edit: if 2C is weak diamonds only 2H as inv+ saves a ton of space. If 2C is weak in either pointy suit or strong it saves less space.

[cherdano]

mikeh, on 2014-November-27, 03:29, said:

BTW, is there not a chance that had N passed, the opps would have reached 4♥, which appears to be doomed?

Given that the 2♠ opening apparently shows a balanced-ish hand, I am not sure East should act differently over a 1NT opening than over a 2NT overcall

[Fluffy]

whereagles, on 2014-November-27, 02:47, said:

lol.. ♥J best theoretical lead, but in this case it's the only to give it away

typical unlucky expert

I wonder if I would defeat on ♦K lead ducked around, sure it seems I should, but I am far from certain, there was a hand in the cavendish where most declarers made 3NT after ♠Q lead, declarer ducking from ♠AK102, there was no way to make unless you got 3 tricks from spades.

EDIT: There is no way to defeat after a ♦K lead taken by the ace.

[gnasher]

mikeh, on 2014-November-27, 03:29, said:

BTW, is there not a chance that had N passed, the opps would have reached 4♥, which appears to be doomed?

3NT was doomed if you just lead your longest suit.

I dislike the red-suit leads almost as much as Mike dislikes the 2♠ opening. Are we really that desperate? Given the methods, partner may well have three spades and not have raised, so it's possible that we'll set the spades up and get in with ♦K to cash them. And passivity may well be right anyway. The opponents haven't shown extra strength, and quite often they're stretching on this type of auction. If 3NT depends on a guess as to who has ♥J, declarer is certain to get it wrong unless we lead it.

[Zzjurt]

WesleyC, on 2014-November-26, 22:06, said:

The results over 10000 simulations:
Lead Beat3NT Tricks(average)
-------------------------
JH: 1240 10.385
4C: 1027 10.419
TS: 693 10.597
KD: 799 10.872

In practice, the heart lead would be less successful than this (due to partner having bid hearts on some of those layouts) but it does feel like probably the best shot.

I did the same simulation in Jack6.10 with these results
♠10: -461.7
♣7: -470.4
♥J: -484.9
♦K: -580,6

So ♦K is clearly the worst lead and ♠10 best but close to ♣7 and ♥J

[WesleyC]

Zzjurt, on 2014-November-30, 07:37, said:

I did the same simulation in Jack6.10 with these results
♠10: -461.7
♣7: -470.4
♥J: -484.9
♦K: -580,6

So ♦K is clearly the worst lead and ♠10 best but close to ♣7 and ♥J

Can you tell me more about how you did the simulation and what those numbers mean?