[PhilKing]

jerdonald, on 2012-November-03, 19:55, said:

BBO Forum,
There are a lot of good points made in response to my post
but no one has answered the original question. How often do
these type of hands occur?

PhilKing says WJS hands are very rare and Bergen is falling
out of favor. Are we going back to Strong(18+) jump shifts?

jerdonald

In England, almost all pairs in the Premier League who play 5-card majors use natural invitational jump shifts (call it 9-11 with a decent 6-card suit if you like). Looking through the US teams trials, it seems to be the most popular treatment as well.

Bergen raises can be handled by playing 2NT as 10+ 4+ and 3M as 7-9 4+. I don't believe the very weak raise is worth much as a preempt, nor is it very common. Responding 1NT with very weak hands with 4+ card support can be just as effective in a weird way.

[Vampyr]

PhilKing, on 2012-November-04, 09:20, said:

In England, almost all pairs in the Premier League who play 5-card majors use natural invitational jump shifts (call it 9-11 with a decent 6-card suit if you like). Looking through the US teams trials, it seems to be the most popular treatment as well.

I think that a lot of these people are playing 2/1GF. It would be interesting to know the most popular treatment for pairs who don't play that.

[Flem72]

PhilKing, on 2012-November-04, 09:20, said:In England, almost all pairs in the Premier League who play 5-card majors use natural invitational jump shifts (call it 9-11 with a decent 6-card suit if you like). Looking through the US teams trials, it seems to be the most popular treatment as well.

Vampyr, on 2012-November-04, 10:04, said:

I think that a lot of these people are playing 2/1GF. It would be interesting to know the most popular treatment for pairs who don't play that.

Also interesting to get frequencies for these....

[PhilKing]

Flem72, on 2012-November-04, 10:15, said:

PhilKing, on 2012-November-04, 09:20, said:In England, almost all pairs in the Premier League who play 5-card majors use natural invitational jump shifts (call it 9-11 with a decent 6-card suit if you like). Looking through the US teams trials, it seems to be the most popular treatment as well.




Also interesting to get frequencies for these....

The sample size is too small unless you include division 2.

Quite a few of the 4CM 14-16/15-17 NT brigade use IJS's as well (Hacketts definitely, and most of DeBotton team I think, but Price/Simpson play strong).

[fromageGB]

Vampyr, on 2012-November-04, 06:43, said:

I think w**k might have been referring to bidding the suit 1/1 and then repeating it, rather than jumping and then repeating it.

Thanks. But we are comparing WJOs with Bergen raises, which are (as I understand) only after major opens, and comparing different uses of the 3m bids.

[Flem72]

PhilKing, on 2012-November-04, 10:41, said:

The sample size is too small unless you include division 2.

yes, but I meant simulation-type frequencies.

[PhilKing]

Flem72, on 2012-November-04, 11:35, said:

yes, but I meant simulation-type frequencies.

Frequency is overrated.

[blackshoe]

fromageGB, on 2012-November-04, 04:34, said:

I think the OP question has been answered, but don't forget, Bergen type raises are in themselves just as pre-emptive as WJOs. As they occur far more often, that is a big plus. Moreover, it must be remembered that Bergen type raises are not just for reaching the right slam or game, they are very useful in enabling the correct action when 4th seat does enter the bidding. As partner has described both his length and his pretty exact strength, opener is well placed to make a penalty X, or knows it is correct to pass or to compete in the expectation of going off one as a sacrifice, or in the expectation of making. The Bergen raise has given him better judgement.

Contrary to blackshoe's opinion, I think it is a great asset in the partscore battle.

If by "it" you mean Bergen Raises, you have misunderstood my position.

[fromageGB]

blackshoe, on 2012-November-04, 13:29, said:

If by "it" you mean Bergen Raises, you have misunderstood my position.

I think you meant your preference is to gear the system to game/slam bidding rather than part-score, and I inferred you thought Bergen was better for the former. I play much more matchpoint than IMPs, so I lean to the partscore/game, and find Bergen is good for that.

[blackshoe]

Yes, Bergen is good for the partscore fight. My point was, I think, that the complete Bergen Raise complex is also good for game/slam bidding, while WJSs are only good for partscore battles (and in some cases for deterring opponents from bidding their game),

[Zelandakh]

The Hackett CC is pretty poor in places and does not mention jump shifts over 1M openings at all. Perhaps it should since one might get the wrong idea from the WJS in the 1m openings section. What I can see is that they play some form of Swiss over the 1M openings instead of Bergen. The Hackett influence is also pretty strong in England and many pairs seem to have copied them. My observations of their system and style has been that they are extremely efficient bunny killers but lose somewhat when playing against the top players. Perhaps Justin or Mike have some experiences of playing against them they can share.

On the actual topic, I strongly prefer more immediate raise options over 1M openings than any of the natural jump shift possibilities. If I have to play natural then my preference would be for weak (up to 7 or 8, not the super-weak style being proposed here) or intermediate but it does somewhat depend on the rest of the system in use.

[relknes]

People have mentioned Strong Jump Shifts, Bergen Raises, Fit Showing Jumps (54+), Weak Jump Shifts (0-5), and slightly less Weak Jump Shifts (4-8). I hope it is not off topic if I bring in "Mini-splinters" where a jump shift shows inv+ values, 4+ support, and a singleton in the bid suit. I am a fan of these since it allows opener to re-evaluate based on distribution of values. They seem a lot more useful for evaluating how the hands fit when they come up, though I am not sure exactly what their frequency is.

[MickyB]

Zelandakh, on 2012-November-05, 02:40, said:

The Hackett CC is pretty poor in places and does not mention jump shifts over 1M openings at all. Perhaps it should since one might get the wrong idea from the WJS in the 1m openings section. What I can see is that they play some form of Swiss over the 1M openings instead of Bergen. The Hackett influence is also pretty strong in England and many pairs seem to have copied them. My observations of their system and style has been that they are extremely efficient bunny killers but lose somewhat when playing against the top players. Perhaps Justin or Mike have some experiences of playing against them they can share.

On the actual topic, I strongly prefer more immediate raise options over 1M openings than any of the natural jump shift possibilities. If I have to play natural then my preference would be for weak (up to 7 or 8, not the super-weak style being proposed here) or intermediate but it does somewhat depend on the rest of the system in use.

The card I've found from 2004 says "Intermediate jump shifts at 3-level, weak at 2-level" under System Summary. It includes a form of Swiss instead of direct splinters, which go through 1M:3M+1, but I think they now play weak splinters directly and stronger splinters through 3M+1.

I don't think many pairs have copied the Hackett style, there are a few in Manc who play the whole system but it doesn't lend itself well to copying parts of. It does have a slight bunny-bashing bent, it's certainly felt to be worse than 5cM for slam-bidding.

I prefer Bergen-style to INV jumps partly because INV jumps don't actually work that well when they come up - after 1H:2C [GF except rebid] we find 6-2 or 4-4 major fits if present, and if not we rebid 3C invitational.

[Zelandakh]

MickyB, on 2012-November-05, 05:24, said:

The card I've found from 2004 says "Intermediate jump shifts at 3-level, weak at 2-level" under System Summary. It includes a form of Swiss instead of direct splinters, which go through 1M:3M+1, but I think they now play weak splinters directly and stronger splinters through 3M+1.

Ah, thank you for this Micky! I have found it now too on the 2010 card. This is the most recent I have and still includes the line that 1♠ - 3NT is any splinter, although the formatting seems to have removed this for a 1♥ opening somehow.

On the subject of mini-splinters, it is possible to have your mini-splinters and still keep the Bergen-style raises. Here's a structure I posted many times already...

Over 1♥
======
2♠ = mini-splinter or strong splinter (2NT asks)
2NT = GF raise
3♣ = limit raise
3♦ = mixed raise
3♥ = preemptive raise
3♠ = void splinter
3NT = splinter with spade singleton
4m = singleton splinters
4♥ = preemptive

Over 1♠
======
2NT = mini-splinter or strong splinter (3♣ asks)
3♣ = GF raise
3♦ = limit raise
3♥ = mixed raise
3♠ = preemptive raise
3NT = void splinter
4m/4♥ = singleton splinters
4♠ = preemptive

[PhilKing]

MickyB, on 2012-November-05, 05:24, said:

I don't think many pairs have copied the Hackett style, there are a few in Manc who play the whole system but it doesn't lend itself well to copying parts of.

Nor will they.

Armstrong/Davies played a fairly similar style. I remember a set of 16 boards where they opened 1M three times where a Hackett passed at the other table ...

The London pairs that played strong and four (Townsend/Gold, Price/Simpson) were playing a rather different system influenced by their rubber bridge background, but only Price/Simpson retained the simple approach.

[PhilKing]

Zelandakh, on 2012-November-05, 05:44, said:

3♠ = void splinter
3NT = splinter with spade singleton
4m = singleton splinters

This is a pretty serious leak - play it the other way round.

Singletons are more frequent, so should go through 3♠ for concealment purposes when opener has a unilateral sign-off. Also there is a small extra technical reason for this way - opener can show two voids with the steps between relay and game.

When responder's undisclosed feature is a void, opener is less likely to have a unilateral sign-off and ends up asking anyway leaking information by asking anyway.

[Zelandakh]

I have actually thought about this a little and am well aware that expert opinion is to play the singleton hidden and the void direct. That method prevents unnecessary information leakage. On the other hand, it loses one step on singleton hands, which are, as you say, more common. So the decision is whether to sometimes hide Dummy's singleton, or to be one step lower for slam investigations. At the level I play, the most common response to a splinter bid, pretty much any splinter bid, is to lead a trump. That makes, for me, the information leakage less important than the better slam auctions opposite a singleton. If I ever play at a higher level and find this varies then I will change it.

[PhilKing]

Zelandakh, on 2012-November-05, 07:59, said:

I have actually thought about this a little and am well aware that expert opinion is to play the singleton hidden and the void direct. That method prevents unnecessary information leakage. On the other hand, it loses one step on singleton hands, which are, as you say, more common. So the decision is whether to sometimes hide Dummy's singleton, or to be one step lower for slam investigations. At the level I play, the most common response to a splinter bid, pretty much any splinter bid, is to lead a trump. That makes, for me, the information leakage less important than the better slam auctions opposite a singleton. If I ever play at a higher level and find this varies then I will change it.

It's not just revealing the splinter - quite a few decent pairs play a double of, say, 1♠-4♣ as asking for a specific red lead at 3 of the 4 vulnerabilities. But I guess that's not common in the big scheme of things.

Also, after the direct splinters, the lowest one is the only one where you can easily RKCB below 4M, so the loss is not great. For instance, after 1♥-4♣ I prefer 4♦ as a train, so I just lose the buck-pass. And after 1♥-4♦ you have not really gained a step unless you play split range splinters (which I really don't like).

[blackshoe]

relknes, on 2012-November-05, 04:19, said:

People have mentioned Strong Jump Shifts, Bergen Raises, Fit Showing Jumps (54+), Weak Jump Shifts (0-5), and slightly less Weak Jump Shifts (4-8). I hope it is not off topic if I bring in "Mini-splinters" where a jump shift shows inv+ values, 4+ support, and a singleton in the bid suit. I am a fan of these since it allows opener to re-evaluate based on distribution of values. They seem a lot more useful for evaluating how the hands fit when they come up, though I am not sure exactly what their frequency is.

I don't know the frequency either, but Romex included these for a while. They were removed for two reasons: low frequency, and when they did come up, often they helped the opponents more than the Romex pair.

[Zelandakh]

Doubling of the splinter bid to ask for a lead in a different suit is something I have not come across yet - it is something that might tip the balance. I almost never ask for key cards below 4M, generally preferring to use this space to check for controls. After 1♠ - 4♣, both 4♦ and 4♥ are control-asking. After the middle splinter, I have been experimenting with giving up Kickback for something similar, that is (after 1♥ - 4♣) 4♦ and 4♠ as control-asking and 4NT RKCB. I need a bit of work/time before I decide for/against this. 1♥ - 4♦ is the problem case and I have not got a good solution for this. I do play split-range though as you can see from #34. The strong/in-between splinters are contained in the 2M+1 response along with mini-splinters.

On the direct comparison, 1♠ - 3NT; 4♣, now 4♦ = diamond void; 4♥ = heart void; 4♠ = club void. Compared with the direct splinters, diamonds and hearts are the same but clubs loses 3 steps. This is a simpler way of viewing it than 1 step for each. Those 3 steps are pretty important imho. The other subtle (rare) point is this - if you splinter with a slam drive and a club void, you can get away with rebidding 4NT here and saving a step over the equivalent XRKCB auction. You cannot really take advantage of this playing it the other way round.

All-in-all, I agree that there are clear and obvious advantages of playing direct splinters as void-showing. But there are also disadvantages and it is not clear that this is the better approach. At intermediate level my feeling is that direct showing a singleton are better. I cannot say that of expert level and indeed I suspect that the reverse may be true. But I will worry about that another time - it is a very simple change for anyone that prefers it the other way round after all!