[hrothgar]

akhare, on Sep 26 2006, 01:40 AM, said:

awm, on Sep 25 2006, 04:39 PM, said:

In addition, I see a lot of methods opposite the 8-12 opening which are based on either a relay approach or a two-over-one game-force approach, either of which works well for the occasional slam hand and not so well for the more common game/no game type or find the best partial type decisions. The followup methods advocated in MALEX seem to be roughly of this form, as do the followups advocated in most forms of MOSCITO and in Josh Sher's TOSR system and in DrTodd's strong pass methods.

Well, I don't know if you are referring to something specific here, but the system that DrTodd and I play gives us plenty of chances to settle down in a part score.

Same here: After a MOSCITO 1♥ opening showing 4+ Spades

1♠ = relays, game invite+
1n = Natural, non-forcing
2♣ = 5+ CLubs, non foricng
2♦ = 5+ Diamonds, non-forcing
2♥ = 5+ hearts, non-forcing
2♠ = Value raise
2N = Limit raise +
3♣ = Fit jump (6 clubs and 3 Spades) non-forcing
3♦ = fit jump
3♥ = fit jump
3♠ = value raise

[glen]

hrothgar, on Sep 25 2006, 09:11 PM, said:

I would go so far as to argue that its easier to discover partner's strength when he holds a skewed distribution as in the second example.  A 11-15 HCP opening range is inherently tighter than a 8-12 HCP opening range.

There is another factor that makes the 11-15 more effective (in finding out the strength) - responder can often just assume 11-13, knowing that opener can bid again/compete with 14-15. A problem with 8-12 is usually both the bottom and top of the range needs to keep quiet afterward. Also there is not as much use of bidding space - for example you open 1♠, partner bids 1NT (say, like a forcing NT), and now with 8-12 you hardly want to jump to the three level except for the rare hand types (some of which you would have been better off to open a preempt).

[Impact]

As someone who has played variations of FP for 25 years (and my regular partner fro the last 15 has kept every hand record and bidding sequence we have had!!) let me weigh in with a few observations:-

1. Playing imps a vul fert is a BIG loser. It is inherently unsound.

2. One famous hand when we had a 54 Major fit and 17 HCP we went for 1100 at the one level: if you had seen the hand it could have been a James Bond design against the system but that hand was genuinely bizarre and not a reason to change the system of itself;

3. ferts n/v are potentially big winners as they force opponents to guess outside their comfort zone (or even to switch systems) when it is their hand: we play 2C opening n/v as 0-5 for 17 years, and 1H n/v opening as either 6-9 unsuitable for single-suited/2-suited pre-empt OR singlesuited 6+C 10-15.

4. Defending against ferts is an artform and the goalposts move: if they open vul with a fert your first option should be penalties but when the vul is unfavourable to the defending side, the odds change dramatically...

5. Having played all sorts of ranges for "mainstream bids" , I confirm Richard's analysis of 8-12 versus 10/11-15: 8-12 while frequently occurring is a BAD range for taking median positions, whereas 10-15 is a good one.

6. Often overlooked is the judgement developed in responding to mainstream openings is largely lost when you switch to 8-12 and it really takes huge adjustments to make this work whereas the minor adjustment for the other is practical.

7. 8-12 openings push the opponents (particularly international class) into more potentially makeable games....pressure can rebound, and too frequently provide the road map when an early couple of Honour cards are located...

8. Some systems which are largely banned at and for pairs events (owing to the rarity and few boards played against each pair) were developed for pairs originally - and are less effective at teams/imps (eg Regres and a few of the other Polish systems)

9. Relay systems are very good for imps - but the concentration of strength which may be vital for pairs decisions are lost in relays - particularly at the partscore/Moysian marginal level (as you tend to show shape regardless unless branching out...). Having different information will create swings, but even having more information when it is of a different kind is NO GUARANTEE of a better result (eg the typical relay nightmare of finding doubleton opposite doubleton and avoiding NT only to find it was a very strong doubleton in partner's hand, or worse still AQ tight opposite singleton -cf concentration of strength but there are compensations)

10. My methods differentiate and revolve around the S suit as I am a strong believer in the boss suit: I like to differentiate between 4S, 5+S, pre-empts and hands which hold enormous playing strength but middling defence to open a strong bid. The Malex style is overly locked into precise length and ignores many practicalities of modern bidding.

[DrTodd13]

The following numbers are for a period in time in which we played our FP system both vul and non-vul. I don't differentiate in these numbers. These represent 1644 times that we opened the bidding. I'm lazy so I'll figure you can interpret the numbers. If they are too cryptic let me know and I'll explain further. I have a bunch more numbers than this if there are any specific queries. One thing to notice is that our 1N opening is a surprising loser. Opening pass is doing better than 1♣. 1♠ isn't that frequent but seems to be a big winner. Overall, we seem to have gained the most from the opening pass.

Total - opening | frequency | average IMPs | freq * avg
P | 0.20 | 0.531548 | 0.108637
1C | 0.15 | 0.446250 | 0.065146
1D | 0.11 | 0.782065 | 0.087530
1H | 0.28 | 0.268478 | 0.075122
1S | 0.03 | 1.330233 | 0.034793
1N | 0.09 | -0.086667 | -0.007908
2C | 0.03 | 0.116667 | 0.002981
2D | 0.05 | 0.851685 | 0.046107
2H | 0.02 | 0.588000 | 0.008942
2S | 0.01 | -1.635000 | -0.019891
2N | 0.01 | 1.776470 | 0.018370
3C | 0.01 | 0.890909 | 0.005961
3D | 0.01 | 0.653846 | 0.005170
3H | 0.00 | -2.375000 | -0.005779
3S | 0.00 | -1.160000 | -0.003528
3N | 0.00 | nan | 0.000000
4C | 0.00 | 3.300000 | 0.002007
4D | 0.00 | nan | 0.000000
4H | 0.00 | 0.000000 | 0.000000
4S | 0.00 | 0.000000 | 0.000000
4N | 0.00 | 0.000000 | 0.000000
5C | 0.00 | 8.300000 | 0.005049

Todd

[hrothgar]

DrTodd13, on Sep 26 2006, 05:47 AM, said:

The following numbers are for a period in time in which we played our FP system both vul and non-vul.  I don't differentiate in these numbers.  These represent 1644 times that we opened the bidding.  I'm lazy so I'll figure you can interpret the numbers.  If they are too cryptic let me know and I'll explain further.  I have a bunch more numbers than this if there are any specific queries.

Todd

Its late and I'm sleeping.

Are these numbers raw or have you normalized them versus your average score?
Normalized would be easier to look at...

[glen]

DrTodd13, on Sep 26 2006, 02:47 AM, said:

The following numbers...

So everything worked, except for 1NT, 2♠, 3♥ a few higher bids?

[DrTodd13]

Everything gained IMPs except for 1N, 2♠ and 3♥ but the frequency of those are pretty low so the std deviation is pretty large. The numbers are raw and not normalized to average IMPs won but I think you have the information there to normalize the results. One flawed pair playing a system against sometimes crazy opps hardly is an accurate measure of worth so take it for what it is worth.

[glen]

so almost everything worked and "hardly an accurate measure of worth" ... this would seem to prove not too much

[awm]

Well DrTodd's average is about 0.43 imps per board playing this system. This seems to indicate:

(1) The 1NT opening is faring remarkably badly. Perhaps this is due to playing weak NTs at vulnerable?

(2) The 1♥ fert is not really doing very well. While it's mildly positive, it's scoring a good bit less than DrTodd's overall average. The times I've watched him play, his opponents often seem to have accidents defending the fert. I'd bet that against well-prepared opponents the 1♥ fert would do quite badly. Not clear how much worse this is at vulnerable rather than NV.

(3) The 1♠ andn 2NT openings are big winners. My guess is that bids which show real length in a minor suit and take up a lot of space are generally good bids. The 1♠ bid is also potentially hard to defend (much easier to have an accident against this than against 1♣/♦ showing a major).

(4) Wilkosz is of course a net winner.

(5) The 1♦ bid showing spades seems to do dramatically better than the 1♣ bid showing hearts. Perhaps this could be due to the impossibility of holding a longer secondary major? I'm not totally sure how the openings work in this system.

(6) Most of DrTodd's other openings more or less conform to his overall average. Of course, the overall average is some combination of the relative skill of DrTodd and partner to opponents, the advantage gained due to the methods, and the advantage gained due to opponents unfamiliarity with the methods/lack of a good defense.

[akhare]

Impact, on Sep 25 2006, 07:21 PM, said:

As someone who has played variations of FP for 25 years (and my regular partner fro the last 15 has kept every hand record and bidding sequence we have had!!) let me weigh in with a few observations:-

Wow -- can you please a summary and/or details preferably of your methods when you get a chance?

It would be especially interesting to see the methods that favour the ♠ suit.

For what it's worth, I did a rough simulation of a 1♥ fert w/ only condition being that the hand be balanced and have 0-9 points. We could scramble to 7 tricks in 1♥/1♠ only about 46% of the time and interestingly enough could make 7 tricks in a minor about 50% of the time. This would seem to indicate that there's at least a 1 trick penalty available at the 1♥/1♠ level at least 54% of the time.

In any case, the analysis isn't worth much because it doesn't take into account what the opps could have made, but it's pretty easy to see how it can rapidly go downhill when we are vul.

[kes]

Impact, on Sep 25 2006, 07:21 PM, said:

The Malex style . . .and ignores many practicalities of modern bidding.

You may be rigth - but would you please go a little bit more in details .

[hrothgar]

awm, on Sep 26 2006, 08:09 AM, said:

Well DrTodd's average is about 0.43 imps per board playing this system. This seems to indicate:

(1) The 1NT opening is faring remarkably badly. Perhaps this is due to playing weak NTs at vulnerable?

One other thought:

Its been a while since I saw Dr Todd play, however, I seem to recall that he played a fairly standard weak 1NT opening. Simply put, his opponents have much more experience with this method than any of the other stuff that he is playing. It wouldn't surprise me if they are able to score better against it.

[akhare]

hrothgar, on Sep 26 2006, 09:36 AM, said:

awm, on Sep 26 2006, 08:09 AM, said:

Well DrTodd's average is about 0.43 imps per board playing this system. This seems to indicate:

(1) The 1NT opening is faring remarkably badly. Perhaps this is due to playing weak NTs at vulnerable?

One other thought:

Its been a while since I saw Dr Todd play, however, I seem to recall that he played a fairly standard weak 1NT opening.

Well, "fairly standard" if you are the kamikaze NT type -- the range was actually 9-13 at any vul and denies a 4 card major unless 4333.

It's probably a little better than it sounds vul. because it's 1st and 2nd seat only and some quacky hands and / or hands w/o 2 controls are systemically downgraded to a fert opening...