Who uses Losing Trick Count?

[perko90]

I've watched a decent amount of Expert Bridge streamers and I've noticed nobody mentions LTC after finding a fit.

 

I have some speculation why that might be: For finding games, it's not really needed. And for finding slams, they use the same technique that I find valuable - visualization - where they "try out" sample hands that they imagine are in range for their partner's bidding. Nonetheless, I still find LTC useful as another (imperfect) tool that I believe helps my judgment in certain situations.

 

Have I taken too limited of a sample of streamers? Or do experts not rely on LTC for hand evaluation?

[Douglas43]

Personal view, you might get different perspectives. It's pretty widely used in the UK. The "total points" or "add points for distribution" idea that BBO robots use is not so widely used here. I found LTC very helpful as a learner and 40 years later still use it on borderline cases.

[mw64ahw]

I always calculate a modified loosing trick count on initial valuation of a hand as I find it aids the decision making process before and after a trump fit is found.

[DavidKok]

I am a fan of (some version of) the LTC. That being said, it seems that the topic is rather polarising. I think LTC is a valuable tool, much like HCP is a valuable tool. Unfortunately most people I've run into either seem to think that LTC is awful ("You how how to count losers? Turn your head around and check how many people are using LTC, that's the number of losers in the room! Ha ha.") or divine ("My 6HCP hand had 7.5 losers so you can't disagree with me for opening it at the 1-level."), while in truth it's just one tool among many.

 

As far as I know, experts don't rely on any particular tool for hand evaluation. Rather, all simple tools are at best approximations of what experts are doing, and at worst one of many different ways of looking at a hand. I do recommend LTC to beginner, intermediate and advanced players but only with the caveats just mentioned.

 

A while back I picked up ♠x, ♥Axxx, ♦Ax, ♣ATxxxx first in hand and the auction (opponents silent) went 1♣-1♥ - I rated this hand too strong for a 2♥ rebid (and far too weak for a 3♠ splinter) so I jumped to 3♥ and partner pushed to a thin slam with a balanced 11HCP, making on a finesse. That got some comments from bystanders, and had I not known about the LTC I might have bid differently. Not exactly a prize auction, but then again if partner held an extra king slam was likely cold and we would not have gotten there over 2♥.

[paulg]

As far as I know, experts don't rely on any particular tool for hand evaluation. Rather, all simple tools are at best approximations of what experts are doing, and at worst one of many different ways of looking at a hand.

I'm of the same mind. I think most experts use elements of LTC when evaluating hands but there are always plus and minus points flying around that seem quite flexible which can make it difficult to explain.

 

I'm not surprised that expert streamers keep it as simple as possible and easy to understand for their audience.

 

 

 

[cherdano]

I can't say much about the objective merits of LTC.

I can say, however, that I've seen more very bad bidding decisions justified in bridge forums based on LTC than based on any other evaluation method.

 

There is one upside of LTC, however: at some point, beginners need to learn to bid more with shapely hands. LTC is a way to get them to do that. But the reason to bid more with shapely hands is of course only partially the playing strength, and in part also the lack of defensive strength. So learning to bid more for the wrong reasons then doesn't teach you quite exactly when to bid more.

[mikeh]

20+ years ago, in what remains the strongest partnership I’ve had, we explicitly incorporated LTC into our valuation criteria in terms of opening unbalanced hands

 

We required 3 controls (A=2, K=1) and a LTC of 7 or less if we lacked 12 hcp.

 

Now, the reason for the 3 control requirement was that we played a relay method, and the first step to a control asking relay (always after we had gone through at least two shape relays) was 0=3 and we had no simple way of showing fewer than 3.

 

We’ve begun playing again after a 20 year layoff but now play a very different method. We also open shapely hands, especially in the majors, more aggressively than we did back then. Our system notes no longer set out valuation criteria. However, as I’ve written here several times, personally I use LTC as one minor factor. If I am uncertain as to how light a hand to open, I will use LTC as the last tie breaker: if the LTC is lower than 7 and I have 3 controls, I tend to open. This means that we would open AJxxxx Axxx xx x 1S. But, even though the LTC is the same, we’d not open 1S with AJxxxx Kxxx xx x (well, we would in third seat).

 

Thus while I ‘use’ LTC, it’s probably the least important/useful of a wide range of factors.

 

LTC can (but usually doesn’t) play a role in second round action: I’ve never consciously used it beyond the second round. Once one has established a game force or slam try, one should usually know or seek to know far more specific information. Where it can come in, for me, is on my second bid once I’ve opened 1x, where x is a suit, is when I’m close to a jump raise of partner’s suit. In that regard, I agree with David’s description of jump raising with a 6421 hand with three aces, none in the singleton. Since a minimum opening hand, with shape, will usually have 7 losers, the hand is at least a full trick stronger than that. However, bear in mind that Aces are far more important than are kings, etc. x Axxx Ax Kxxxxx is the same (simple) LTC as x Axxx Ax Axxxxx but the second hand is significantly stronger and I’d not be tempted to jump with the first.

[perko90]

Thanks for the replies. I'm glad to see that LTC is seen to have some uses in expert circles.

I admit that the target audience for the streamers I watch is for a skill level around intermediate (or a bit less). So, perhaps that's the reason they don't discuss LTC. Although they sometimes touch on advanced concepts.

And I still find some nuggets even though I'm above the target audience skill range. (For ex, I discovered from a stream that 1NT-2♣; 2♦-2♠ shows an invite w/ or w/o 4 Hs and is "expert standard" instead of the traditional 1NT-2♥*; 2♠-2NT).

 

And yes, I've seen some abuses of LTC by less experienced players. The worst of which is strict LTC Ogust responses for weak 2 bids. Me: "So, you would respond 3♠ to an Ogust 2NT inquiry with xx KQxxxx x xxxx." Opp: "Yeah, for sure." Me (to self): {crazy!}.

[mycroft]

One of the issues (as Mike and I have discussed here a number of times) is that LTC is a good algorithm to fake judgement (as are all counts, sure, but HCP is treated as "the base" that people use judgement around).

 

One of the things that make experts experts is that they have good judgement and don't need algorithms to fake it. One of the nice parts of LTC is that they can shortcut explaining their judgement to non-experts using it; but it's not really what they do at the table. And streamers, who would be "narrating their thought patterns", will do that and not go to the shortcuts they might use away from the table when explaining.

 

Having said that - and saying that I am not an expert, just a decent flight A player - in one system I played "basic LTC-vs-vul based" 3- and 4-level preempts. And we carried them over, for 3M at least, to our "normal" system, because we thought that codified the aggressive style we wanted to play.

 

I do think that a good, solid, detailed "LTC and cover card"-based system - one that, possibly, used Walrus points to fake judgement - probably works as well as anything else. I don't think that any of the people I play against playing "LTC and cover cards" have a good, solid, detailed system - because it's as complicated as the good, solid, detailed HCP-based system they moved to LTC from because it would have been too hard.

[Vampyr]

I have never been a fan of LTC. All of the example hands given in this thread have aces. I am leery of a method that counts queens the same as aces.

[Vampyr]

I have never been a fan of LTC. All of the example hands given in this thread have aces. I am leery of a method that counts queens the same as aces.

[Vampyr]

I have never been a fan of LTC. All of the example hands given in this thread have aces. I am leery of a method that counts queens the same as aces.

[Vampyr]

I have never been a fan of LTC. All of the example hands given in this thread have aces. I am leery of a method that counts queens the same as aces.

[DavidKok]

I have never been a fan of LTC. All of the example hands given in this thread have aces. I am leery of a method that counts queens the same as aces.

Nobody is forcing you to stick to this.

[helene_t]

Modified ltc is not too bad when you have a 9 card fit, but with an 8 card fit it is excessive to count 2,1 and 0 card suits as 1, 2 and 3 Kings respectively.

Another issue is that you initially evaluate your hand using hcp and switching to a different scale after finding a fit is awkward.

 

So even if you like ltc you might consider rescaling it to the hcp scale.

[cherdano]

Come on guys, LTC is awful.

 

I mean, doesn't it think that AKxxxx Axx xxx x is as strong as AKxxxx xx xxxx x? That Axx is as good as QJT?

 

Yeah, even the silliest method can become good if you start with "silliest method" and then adjust with your judgement to correct for its flaws. But why not start out with something that isn't awful?

 

The whole idea that first-round losers are as bad as third-round losers is just ridiculously bad. Yes, modified LTC is better, but that introduces the problem that three people talk about evaluating a hand using LTC, and they mean three different things. Evaluation methods are also meant to allow us to communicate about hand strength, and nothing other than the tower of babylon and LTC have ever done as much damage to communication between fellow human beings.

[Douglas43]

Come on guys, LTC is awful.

 

I mean, doesn't it think that AKxxxx Axx xxx x is as strong as AKxxxx xx xxxx x? That Axx is as good as QJT?

 

Yeah, even the silliest method can become good if you start with "silliest method" and then adjust with your judgement to correct for its flaws. But why not start out with something that isn't awful?

 

The whole idea that first-round losers are as bad as third-round losers is just ridiculously bad. Yes, modified LTC is better, but that introduces the problem that three people talk about evaluating a hand using LTC, and they mean three different things. Evaluation methods are also meant to allow us to communicate about hand strength, and nothing other than the tower of babylon and LTC have ever done as much damage to communication between fellow human beings.

 

As opposed to a point count that tells you that Qxx Qxx Qxx Qxxx is equivalent to Axxxxx Axxxx x x? Anybody can rubbish an unsophisticated version of a valuation when used in isolation. I use a mix of HCP and LTC, but it's really a matter of choosing the tools that work for you.

[cherdano]

As opposed to a point count that tells you that Qxx Qxx Qxx Qxxx is equivalent to Axxxxx Axxxx x x? Anybody can rubbish an unsophisticated version of a valuation when used in isolation. I use a mix of HCP and LTC, but it's really a matter of choosing the tools that work for you.

HCP works well for balanced hands, and doesn't need drastic adjustments. LTC works well for - no hand type at all, and needs drastic adjustments to even be a modestly reliable guide.

[helene_t]

nothing other than the tower of babylon and LTC have ever done as much damage to communication between fellow human beings.

The Sumerian goddess Enki was asked to give everyone a common language so they could collaborate and build a tower. For some reason, a lot of people think that the god(s) gave people different languages, but that is a distortion of the original myth.

[nullve]

As opposed to a point count that tells you that Qxx Qxx Qxx Qxxx is equivalent to Axxxxx Axxxx x x?

The Milton Work point count?

 

In his writings on bridge during the last seven years of his life, Work adopted Bryant McCampbell's 1915 suggestion[6] for evaluating balanced hands using a point count method: Ace = 4, King = 3, Queen = 2 and Jack = 1.

Since LTC is just a disguised way of counting points for high cards and distribution, it must be better to compare it with the Goren point count.

[nullve]

Btw, it's easy to create a pair of hands where some version of LTC looks dumber than Goren points.

 

Two examples:

 

1)

 

♠Qxx ♥Qxx ♦Qxx ♣Qxxx (8 losers, 7 Goren points*)

♠AJT ♥AJT ♦AJT ♣AJT9 (8 losers, 20 Goren points*)

 

2)

 

♠ Qxxxxx ♥Qxxxx ♦x ♣x (6 losers, 7 Goren points*)

♠ AJT9xx ♥AJT9x ♦K ♣K (6 losers, 18 Goren points*)

 

* using the (Goren point) algorithm in Jens Boeck, Moderne bridge, 1980

 

Can anyone come up with a pair of hands where Goren points look dumber than some version of LTC?

Edited June 22, 2022 by nullve

[Douglas43]

In response to the request for a pair of hands which LTC handles better than Goren

 

1. ♠ AJT9xx ♥AJT9x ♦K ♣K (6 losers, 18 Goren points*)

2. ♠ AJT9xxx ♥AJT9xx ♦- ♣- (4 losers, 16 Goren points )

 

[mw64ahw]

Btw, it's easy to create a pair of hands where some version of LTC looks dumber than Goren points.

 

Two examples:

 

1)

 

♠Qxx ♥Qxx ♦Qxx ♣Qxxx (8 losers, 7 Goren points*)

♠AJT ♥AJT ♦AJT ♣AJT9 (8 losers, 20 Goren points*)

 

2)

 

♠ Qxxxxx ♥Qxxxx ♦x ♣x (6 losers, 7 Goren points*)

♠ AJT9xx ♥AJT9x ♦K ♣K (6 losers, 18 Goren points*)

 

* using the (Goren point) algorithm in Jens Boeck, Moderne bridge, 1980

 

Can anyone come up with a pair of hands where Goren points look dumber than some version of LTC?

Using a MLTC rather than LTC gets you to

1) 10 vs 5

2) 8 vs 5.5

so makes more sense

[mycroft]

"Single mode of evaluation" is bad.

 

Mollo mocks it with the Walrus; it is screamingly clear that pure and only LTC is also (usually) awful.

 

That's why nobody past novice does that.

 

We just assume this when having Walrus count conversations.

 

Why does nobody allow this for LTC, but just say that "KQxxxx xxxx xx x and AJ9xxx AT9 A98 Q are considered the same, and that's stupid"? Of course it's stupid. That's why nobody past low intermediates and LTC zealots (but I repeat myself) does that.

 

As I have said repeatedly - including here - LTC is a great "substitute for judgement" *on top of* HCP. It's a great way of explaining your judgement to those who don't have it yet - MikeH's "a 6-loser 18 count is very strongly on the weak side of 18s, I would downgrade out of GF" - makes sense to people in one sentence that wouldn't understand the 3 paragraphs of *actual judgement* Mike actually does. I think - but have not tried - LTC + cover cards could play well; "an Ace is equal to a Queen" is fine if partner's cards are therefore more likely to be Aces than Queens because you have them. After all, that's what we do with control cue bids (for AK) and spiral scan (for AKQ). I'm at the point where my judgement is good enough for me that relearning how to think isn't worth it, but I think a good, solid, detailed version of this (likely with some sort of spiral scan) would work at least as well as HCP + adjustments.

 

But no single mode of evaluation will be better than good judgement. It's just that all the people here have judgement for HCP "automatically", and laugh at LTC without judgement applied (partly because it's easy, partly because they don't "automatically" have the LTC-based judgement to apply).

[Flem72]

I use R of 22 and R of 21+ when I have spades, LTC might be a tie breaker. A few of my partners use LTC. It is amusing to me when they pass R of 22 hands and I open hands that don't pass LTC muster but are R of 21+ or 22. NO accountin' for personal taste, I says.