[numbers] Our Countdown Teaser - verifier and the UK rules

Great to see our suggestion of the Countdown Puzzle to be taken up by Matt for this month and to see it causing a stir - and yes, it's not a trivial one.

For interest I thought it worth posting details of how the game is played as it deviates a little here, although it shouldn't affect the implementation of solution algorithms.

The Puzzle is based on 6 numbers but not all numbers are available in the domain. Numbers are laid on a table for selection by a contestant, and with one row of large numbers (25, 50, 75, 100) and the rest being small. (1-9). There ay be repeated numbers in the selection.

Typically one large and the rest small is chosen.

A computer of course produces a target number, however contestants do not have to use all numbers in arriving at an answer, even though there is always at least one solution that is based on all numbers.

For verification purposes the site at http://www.williamtp.com/numbers.html may be useful (there is no source available)

Good luck!

Nick

Here in Spain there is also a show with a very very similar puzzle. With 6 numbers from 1-9,10,25,50,75 and 100 (repetitions allowed), they have to get the closest possible to the target number.

However, the rules specify that the result of an operation (or sub-operation) must be integer. E.g., the following whould not be acceptable by those rules:

(100-(((75/50)+1)/25))*10=999

I guess that this doesn't apply to this contest. Does it, Matt?

No, it does not.

hmmm. It would be nice to see the code of http://www.williamtp.com/numbers.html after the contest results are published. Maybe an official codewalker could try to ask him .
Nice contest anyway. I'm not advanced enough in math and logics to do this, i'm just 16, but i am looking out for the results. Good luck to everyone!

I e-mailed William T-P, asking for the strategy of his script, but he was somewhat evasive:



I didn't ask specifically for the source at that point... that would have been cheating.

Hi !
Sure that is a good website: it is really fast to find solutions...
However, the solutions found are NOT distinct all the time as they are said to be...
So congratulations to the guy that developed it, but it can still be improved a little bit ;)

/Flood

The solver is probably written in faster language like C. With a little optimization even a brute force method would be quick.