This article is always being updated. Version 1 (Sep 2006) is philosophy
Part 2 (Sep 2006) contains PCF opening bids and outline responses
Part 3 (Oct 2006) looks at 1 major and its philosophy
Part 4 (Apr 2007) looks at interference over 1C and counter-measures

Part 1

Reese Precision (RP) spawned Old fashioned Cambridge (OFCP) as a result of the work done by Granville, Fleet et al in Cambridge in the 70's. OFCP is a definitive version of precision from those times. I'd agree with Geraint that original RP is out of date, as indeed is OFCP, and there are a number of changes made from RP to get to OFCP and a few more that proddy and I play (Proddy/ChienFou - PCF). Most of our changes are no more than modernisation in line with current thinking. Our original base was RP with ideas from many sources, particularly OFCP

I'm going to discuss the ground rules for PCF, and try to set out how we plugged the gaps in RP.

Fundamentally the limited opener allows for very light openers, probably a full point lighter than 4cM, and as much as 2 points lighter than 2/1GF. If you play genetic precision (GP) then you realise the main reason is so you *can/must* open light

In RP 1NT was 13-15. This is awful. It meant one had to open 1C with 16 bal, and 1C 1N was not GF. It meant that 1D, 1NT was messy. The obvious thing to do was to widen the NT range to 6 points and have a 1NT ladder.  Modern precision has a 14-16 NT (Rigal, Precision in the 90's - PITN) and 1D, 1NT is 11-13. Players who like a variable NT can play 1NT 11-13, and 1D, 1NT 14-16. interchangeably with PITN. This achieves 2 things. It gets the NT ladder into shape, and it allows 1C (bal) to be 17, so that all positive reponses to 1C are now GF. This shifts the problem onto the 1D opener, which can now be balanced out-of-range, or D, or clubs. With negative doubles this is manageable. The nett effect is that we now have a structure for light openers and full openers, with ranges 11-13, 14-16 for balanced nands, and 10-12, 13-15 for unbalanced hands. this was never evident in RP, and only partially resolved in OFCP. It is fundamental in PCF.

1H 1N! forcing caused horrible problems with the Flannery hand (4522 minimum). Without question the greatest advance of OFCP was to invert the meaning of 1S and 1NT over 1H. This inversion of NT and the next suit up where partner is known to have 5 or more cards in a suit is a foundation of PCF.

1H 1S!? 1N! promises the minimum 45, and therefore the 1S response denies 5 spades, but may hold 4. thus the auction 1H 1S 1N 2S is a 4-4 fit. 1H 1N shows 5+ spades. There is almost no downside to this method. The argument that 1H 1S 1N natural is a good sequence doesn't hold water. A 5 card major is ok for a suit even facing a singleton, and secondly the 1S response can be 4 or 5 cards. With the inversion we *know* the major lengths.

Part 2

We started by playing 1NT as 10-12 all seats, but found that it put too much weight on 1D. After a few years, probably early 90's, we saw an article suggesting that in a 5cM system one should play a weak NT in seats 1&2, and stronger NT with 4cM in seats 3&4, irrespective of vulnerability. It is hugely aggressive, a GP trait. I play this style with all my partners. You'll see it also in TimWM's and my EHAA system.

So we have defined the 1NT ladder as follows:
11-13 1NT seat 1&2; 1D,1NT 3&4
14-16 1D,1NT seat 1&2; 1NT 3&4
17-19 1C,1NT; we put some 20's through this
20-21 1C,2NT
22-23 2NT We messed around with 2NT
but 22-23 is low enough frequency that we don't mind it. I'll take a free swipe at the Benjaminised Acol players who play 2NT 19-20. Do you really think that 2NT taking away 9 natural bids is helpful? Additionally it is such a stupid contract when partner passes that you should seek to lower its frequency not raise it.

At the beginning we made a few obvious changes to opening bids. Why open 2D with a 4414 opener? You might as well open 2H, no space is lost and the downside of passing with a weak 6 card D suit is small. This allows you to open 2D with a wk 2 in H or the Roman 4441, 17-24. The responses are fairly obvious, based on the original Blue team club stuff. The original 3C bid can only be in the system because they'd had a bad result with a 2C opener. Play 3C as a modern pre-empt. 1C became 16+ unbalanced, 17+ balanced.

This leads us to the Opening bid structure we still use today. I've also indicated the reponse structures. It changes a bit in 3rd seat:
1C 16+unbalanced, 17+balanced; reponses 1D negative (0-7ish), 1H,1S,1N,2C,2D natural GF, higher bids PRE, 2N 4441 shape GF
1D out-of-range NT, D, or D+C. responses natural approach force, 2C and 2D both F2N, WJS majors
1H/1S 10-15, 5cM (Granville's inversion of 1S/1N over 1H), 2/1 F2N, 2N GF fit, fit-jump, PRE raises
1NT, by seat, 5cM/singleton permitted. *Under discussion is changing responses to weak NT in line with the EHAA method.
2C, 10-15, 6C or 5C and 4cM (GP). Responses: 2D asking relay, 2H/S nat NF, 2N descriptive relay puppet 3C, 3C const.
2D, H wk2 or 4441 17-24, responses 2H NF relay, 2S forcing relay. very complex continuations for 4441's
2H, 4414 10(11)-15(16), some shapes go through 2C, responses: 2N shape and strength relay, other calls, aggressive limit.
2S, wk 2. feature style development over the weak 2's rather than Ogust. We could change this.
2NT, 22-23, 5C puppet Stayman.

In 3rd seat we freely open v light 4cM if planning to pass the response, a 3rd seat rebid promises a full opener. We don't play the inversion and 1NT is semi-forcing. We use Flannery 2H as we can open 4cM and this solves the inversion problem. We generally pass any response with the 4cM openers, as they're (9)10-13, and partner hasn't 11-13 NT'd
We revert to 1st seat method in 4th seat (except 1NT is now 14-16). We can still pass responses though.
we don't play Drury, though Lawrence "Passed hand bidding" recommends it. I think it's a style thing.

Part 3

Moving on to 1 of a major. Our primary objectives here are
1) to pre-empt opponents when the hand is not ours
2) to land well fitting 21 point games.

In most 5cM methods with F1N, the single raise is constructive and this makes sense when the opener is unlimited. In GP this is not the case; so in PCF with a 6 count facing a 1H opener, no game is likely, and we just raise. We do this freely with a doubleton from about 5 up to about an 8 or 9 count, and with 3 cards we just raise with almost any hand up to about a 7 count. Partner doesn't even expect this to make, but who cares, he won't take another bid. Opponents don't know whether we have a 7 or 8 card fit, nor whether our side has 22 or 15, and you don't get doubled in 2 major often, particularly at imps.

For the constructive raise we have to go via the relay. eg. 1H 1S! 2C! 2H!. 2C is better minor, and 2H is alertable as it's constructive not preference, about 7-10 with 3 pieces. The jump raise is pre-emptive too, shows a 4-card fit. It probably won't make. 2NT is the 4 card GF raise, and relay followed by the 3 level is a flat 3-card high card raise (about 11). 2/1's are F2N, and we play the 2NT rebid as "I shouldn't have opened", so the suit rebid is neutral but not a pile of tat. I'm not sure we're right here, but drifting 1 off in 2N, when there never was a chance of game (11 facing 11) is the price we pay. It pays to keep the 2/1's up to strength; about 10 with a fit, 11 with no fit. You've got the relay if you're desperate. More recently we've taken to opening the real minimum 5332's with 1NT in first and second and this helps a lot. It's better in our opinion (and I thank TimWM for this idea) to put as many junk openers through 1NT as possible and have better definition in suit auctions.

We play fit-jumps, promising values for the 3-level. A card in the JS suit will be enough for game. Sometimes the opponents can cash 4 quick winners, we don't mind this at all, they've got to find the lead.

Let's look at the thin games: If we can find an 8 card fit at the 2-level then we play Extended Reverse Romex (ERR). Every bid between 2 Maj and 3 Maj is working overtime. Initiator will be 1 loser better than a minimum. This has a huge upside. Example sequences: 1H 1S! 1N! 2S; 1S 1N! 2D! 2S; 1H 2C 2D 2H; etc. In all these sequences the next bid up says "I want to make a short suit game try" (SSGT), and a new suit is a long suit game try (LSGT). We tend to make SSGT on singletons, and LSGT on H(H/x)xx.
Note that if H are trump then a 2N bid is a LSGT in spades. We don't need the natural 2N with limit openers.
Let's look at some sequences: 1H 1S! 2C! 2H! 2S! 2N! 3H! !! 1S relay; 2C better minor; 2H constructive raise; 2S "SSGT coming up"; 2N, "will co-operate"; 3H - SSGT in the non-available suit (ie spades). Opener's hand is x KQxxx AJx Axxx, and here he wants to know whether there any wasted values in spades. Partner accepts with xxxx ATx QTxx xx and these two hands would both be minimum. It's not a great game, but both hands could be better, and the auction would go 1H 2H in most prec partnerships, and the same in Acol.
If you don't necessarily want to co-operate in a SSGT then bid the first suit you *would* co-operate in. This suggests good trumps in those sequences where you will have killed off the trump ask, and we use it sparingly.
Opener holding x AQJxx xxx AQxx would probaly bid 1H 1S! 2C! 2H! 3C! as a club LSGT looking for the CK.
There's a lovely twiddle here too. If you make a SSGT in your second suit, then it becomes a trump ask, so we can bid 1H 1S! 2C! 2H! 2S! 2N! 3C! as asking for decent trumps. Opener will be x xxxxx AKx AQJx or something like that. Partner will know that the losers are in trumps and can evaluate on that. KQJ tight would be a huge holding.

The Wolfarth auction on the board can now be decoded. I leave it as an exercise for the reader :) PCF is a *lot* of imps up with this stuff. For the record the auction was: 1H 1S 2C 2H 2S 2N 3D 3N 4H and gained a games swing when precision opponents at the other table in the same seats bid 1H 2H AP and commented "flat board"

Part 4

Having played against the "gives" on BCL! who also play GP they asked me to outline our methods for interfering over a GP 1C opener, and our own counter-measures to these defenses. Again we have a philosophy; we assume we don't have game and bid to the level of our fit on the basis that, in principle, this is a competitive auction.

Over 1C
X is hearts contructive
1D is spades constructive
1H is Coolour
1S is Rank
1N is Shape. We bid on almost any hand and only need 4-4. Paradox(*) responses
2 of a suit is destructive with the suit, or more constructive with the other 3, Paradox(*)
2N is any bid 2-suiter, and almost GF.
3 of a suit is a WASP (Wild-arsed s****y pre-empt)

Paradox is where you bid what you don't like - think of responding to a multi for an example.
For the Crash hands we bid the better of the pair we don't like at a TNT level, subject to being able to convert to the other pair if partner takes out to them by bidding the cheaper. We'll raise defensively holding 5, assuming partner has 4
For the 1 or 3-suiters we raise to the level where we can tolerate preference on a TNT basis. An example will help. Partner overcalls 2D and we have 3D and 5 spades. We bid 3D (assumed 9 card fit) and if partner now bids 3H ("sorry, not D's") we bid 3S as that is a 9 card fit.

When the opponents interfere over our strong club we have a few gadgets. Positives show 3K (K=Neapolitan Controls; A=2, K=1), as you need 9K for game and it helps partner assess whether to double if opponents keep competing. We pass with 0-4 and double with hands 5+ denying 3K (This is borrowed from some Blue Club ideas). A cue of a known suit is a DAB for NT and natural NT bids show a stopper. We don't worry too much about trying to extract a penalty below 2NT, but go for the throat at the 3-level.