[bluecalm]

Ok, so in 3/5 problem holdings are T9x and 98x depending if you choose to lead a 9 from the latter.
In 2/4 the only problem holding is JTx and higher (but again, it's a problem for 3/5 too).
So it seems that 2/4 is better in this respect (confusing 2 and 3 less often). On the other hand 3/5 is better in a sense that it gives faster information and sometimes you know the count by the first round of the suit.
This trade off looks like it favours 3/5 intuitively but I have no experience playing it so it's just a guess.

[gnasher]

bluecalm, on 2012-June-28, 09:00, said:

Ok, so in 3/5 problem holdings are T9x and 98x depending if you choose to lead a 9 from the latter.
In 2/4 the only problem holding is JTx and higher (but again, it's a problem for 3/5 too).
So it seems that 2/4 is better in this respect (confusing 2 and 3 less often). On the other hand 3/5 is better in a sense that it gives faster information and sometimes you know the count by the first round of the suit.
This trade off looks like it favours 3/5 intuitively but I have no experience playing it so it's just a guess.

Isn't this completely symmetrical? In both methods you know the count after the second card. On the first card, if 982 is a problem for 3/5, 932 is a problem for 2/4.

[bluecalm]

Quote

Isn't this completely symmetrical? In both methods you know the count after the second card.

You don't know count by the 2nd round with 98x and T9x because it still could be 9x or Tx while in 2/4 you lead 8 and 9 respectively and then follow with low making it clear you still have the 9 or T.

[blackshoe]

"What we've got here is failure to communicate." - Strother Martin, as The Captain, in Cool Hand Luke.

[gnasher]

bluecalm, on 2012-June-28, 10:30, said:

You don't know count by the 2nd round with 98x and T9x because it still could be 9x or Tx while in 2/4 you lead 8 and 9 respectively and then follow with low making it clear you still have the 9 or T.

Playing 3/5, with 982 you lead the 2 and follow with a higher one, and partner knows what your length is. By trick two, you're in the same situation as the 2/4 players.

You could agree to lead the 2 from 1092 as well, with the same result. People usually agree to lead the 10 from 109x, but that's because they choose to prioritise information about the honour holding over information about count. However, they could if they chose, lead the 2 and then the 9, with exactly the same information conveyed by trick two. If they did this, they would be worse off than the Polish 2/4 players at trick one, but level with them by trick 2.

There is a corresponding gain for 3/5 players with 1032: they know the count at trick one, whereas 2/4 players don't. By trick 2, everyone is in the same position.

The 3/5 players gain because:
- They can more easily distinguish between two- and three-card suits at trick one
- With Hx partner knows what the honour is at trick one
- With Hx they may hold the lead to repeat a finesse through dummy.
The 3/5 players lose because:
- They can more easily distinguish between three- and four-card suits at trick one
- Leading high from Hx may cost a trick
- With a three-card holding, partner has more clue at trick one about what the highest one is.

When partner has bid the suit, leading high from Hx is usually a good thing rather than not. With three or four cards, you will often have raised anyway. Hence I think that even if you lead Polish 2/4 normally, you should switch to 3/5 in partner's unsupported suit.

I agree it's not clear which is best when partner hasn't bid the suit.

[Heavily edited]

This post has been edited by gnasher: 2012-June-29, 02:01

[bluecalm]

My point is that it' better to lead a high card from 98x and from T9x and you are paying significant prize for leading low from those configurations. In 2/4 you systematically lead high (well, 2nd one but the trick tacking effect is the same).

Quote

With Hx partner knows what the honour is at trick one

Low from xx is up to 9x, from Tx and higher we (polish people and others playing 2/4) lead a H.
Notice that T9x isn't problematic because we lead a 9 from that.
That, combined with above mentioned point, is why I referred to 98x and T9x as problem hands for 3/5 leads - you either don't give count information even by trick 2 or you have to make inferior play of leading low from those configurations.

Quote

They can more easily distinguish between three- and four-card suits at trick one

I think you mean 2/4 players. Yeah, that's true. I have little experience playing 3/5 so I completely forgot about 3 and 4 problem which is non-existent in 2/4 world.

Quote

Leading high from Hx may cost a trick

You keep mentioning Hx. We lead high from Hx, it's much much better to lead H in almost all situations regardless of the agreements. This play just takes more tricks on average. I really don't have any idea why you think anybody would lead low from Hx. The same goes for your last point.

[gnasher]

Sorry, I hadn't realise that you led high from Hx. There was no need to tell me twice, though.

In Poznan two years ago I must have read the leads section on about 50 Polish convention cards without ever noticing this.

What do you lead from 109 doubleton?

[bluecalm]

Quote

There was no need to tell me twice, though.

Yeah, sorry for that.

Quote

What do you lead from 109 doubleton?

A T, to distinguish it from T9x+.
The only configurations which are uncomfortable for "2/4" rule are HTx when sometimes people feel the need to lead low and T8xx+/T9xx+ when sometimes you face the dilemma of leading 2nd (standard) or 4th (might be good especially vs nt).

[mgoetze]

I like the rule that T is an honor vs NT and not an honor vs suits. Thus against NT, I lead

T9
T8xx
HTx
etc.

whereas against suits I lead

T9
T8xx
HTx
etc.

I can see why some might be tempted to lead HTx instead but it's not for me.

[dake50]

Consider.... xx
...... xxx :: AKJ10x
...... Qxx
My count lead from 3 is lowest. Partner wins A, returns J
to put declarer to an immediate guess. Known in NT contracts,
but may be testing win Q now against discard on Q later.
On other similar layouts top of nothing takes away these
put-to-a-guess.
I like count certainly shown on opening lead.
Top of xx? xxx? xxxx? xxxxx? Which?