In the modern world of bridge we receive hand records at the end of the session to remind us — and partner — of exactly what it was that we held when we got that disastrous score. The hand records at my local club include double dummy analysis, which tells us all exactly how many tricks might be made in each denomination given the best possible defence and declarer play. Now obviously the results of these analyses aren't always realistic: you don't drop singleton kings and doubleton queens every time they happen to be off side. You don't magically lead unsupported aces to unblock partner's unbid suit to defeat 3NT. And you don't always get the best possible defence or declarer play, as the following two hands demonstrate.

It's a well known fact that if you put me in a slam there is every chance that I'll go off in it, even if it's cold. So naturally when I bid myself into an allegedly cold slam the poor bridge alarm bells should start ringing. The full hand was as follows:

2 is weak with both majors. There are no specific agreements as to what the rest of the bids might mean.

Now according to the double dummy analysis this is cold for twelve tricks in no-trumps. This is probably a piece of good fortune for me as declarer, because it doesn't really look like there should be a whole lot of play to it. In fact, so convinced was I that I couldn't possibly make the contract that I didn't really spend any time thinking about it. So I cashed the top two spades after winning the diamond lead. Now, when the J popped out doubleton I realised that there was play for this contract. However, I should have played ace-another heart before playing the spades thereby either setting up the hearts or taking two heart tricks should the king have been ducked. Correction: had I been thinking at all that might have occurred to me. What I actually did was play a low spade from dummy, automatically, on the queen, thereby killing dummy entirely and making no tricks over there at all! Nice job! I now drift two off and confound the double dummy analysis by two tricks: 10 tricks made when 12 were there.

On the following deal, the double dummy analysis suggests that East-West can make precisely no contracts in any denomination. Mr Harker begged to differ.

I confess to being a little bit surprised that, double dummy, East can't find his way to at least seven tricks in spades and I rather suspect that in practice this is the sort of hand that will make seven or eight tricks at the table. Let's take a look at the actual play of the hand:

South leads the K which goes to the Ace in dummy. Declarer tries a spade to the king and ace. Now South plays the Q, which holds, followed by the J. Dummy plays low on this as well and North's ace takes the trick — setting up three heart tricks in dummy. North decides that with a source of tricks now established for declarer, trumps should be drawn (to help him cash them), and leads a spade. Geraint draws the trumps and cashes out for ten tricks. Err, well done partner! Ten tricks made when only six should be there.

So there you have it: double dummy analysis is shown to not give an accurate assessment of the number of tricks that can be made on any given hand. Sometimes — when people like me are declaring — the double dummy overestimates the tricks won, regularly by two or more. At other times, it's possible to make as many as four extra tricks over and above what the double dummy analysis suggests.

Conclusion: when poorbridge is afoot, the analysis is worthless!

Last week, in this article we wondered if there were any other poor bridge hands involving the 9 8 6 5 4 suit combination. Well, Robin Barker to the rescue! He pointed us in the direction of an amusing hand from the Canberra Bridge Club Bulletin where slam is made on a 5 count. Scroll down to page 5 — it's called 33 Points Schmoints. There must be something to that combo after all!