Tuesday May 27th 2008
Posted: Tue May 27, 2008 3:03 pm
R13 FEROCITY.
A group for contestants and lovers of the Channel 4 game show 'Countdown'.
http://www.c4countdown.co.uk/
Well done though for getting 2 wins .lelb72 wrote:He whopped me, i had no chance!Oh well
You actually scored quite well then, if you didn't pay any attention to the letters when deciding your wordslelb72 wrote:As you say he was just very clever. If you watch him he studies the letters first before deciding a word. I cant and didnt do that.
Hmm, I really should be revising, but all I can think about now is devising a strategy to play optimally when you don't know what the letters selection is. If I fail I'm holding you directly responsible.Corby wrote:You actually scored quite well then, if you didn't pay any attention to the letters when deciding your wordslelb72 wrote:As you say he was just very clever. If you watch him he studies the letters first before deciding a word. I cant and didnt do that.
Who's worried?Richard Brittain wrote:This guy is really rather good. If I was Jason Cullen, I'd be worried. If I was David O'Donnell, I'd even be slightly worried.
This intrigued me, I have to say. There's not a strategy that you could employ to win the game, but what would be the word that would leave you with the least egg on your face? For a start, most times your word's not even going to be within the selection, and even if it is, it's probably not going to win the round, unless you get lucky and your opponent declares something invalid.Paul Howe wrote:Hmm, I really should be revising, but all I can think about now is devising a strategy to play optimally when you don't know what the letters selection is. If I fail I'm holding you directly responsible.
Hah, not really! I'm not sure you can do better than a slight modification of what you suggested, which is to build a table comparing how each word would score against every other word in the dictionary, on average (so this hypothetical table would have N^2 entries, as opposed to just a list of N words) and then pick the word that scores best against all other words. Assuming you knew the number of vowels and consonants in each round, you'd need three separate tables. My intuition is that five vowels would result in slightly higher scoring, but it doesn't matter as both players would play the same word. Kind of a boring game really.jimbentley wrote:
I realise that Paul was probably thinking along more subtle and
nuanced lines in terms of an optimal strategy, but it's best to start
with a brute force approximation, I reckon.