<<<Back to Pagetworld

An Exercise In Time Wastage

What's all this then? Allow me to explain:

My partner and I were watching Countdown one day (I TiVo it every day. Sad eh?) and she said something like "the youngest person always wins this programme". That chimed with my own personal theory that younger people tend to choose strange combinations for the numbers round (none from the top, or four from the top for example). So I created an Access database which analyses the lengths of words found by contestants and compares the performance of different ages, genders and occupations.

To answer Sarah's theory, it performs two analyses. Firstly it asks the question "Who won the game?". That's easily done and the result a pie chart called "Youngest". If the contestants are of the same approximate age, the result is ignored for the purposes of this test (more about the contestants' age later).

The next series of charts work out which contestant found the longest words. Note, that this is not necessarily the winner of the programme!
This is because of the way I have decided to count the scores. In the TV show, if one contestant gets a 7 letter word and the other gets a 6 letter word, only the first player gets 7 points. The other person gets nothing. However, when I'm calculating average score, I also count the points for the losing contestants. After all, they deserve credit for finding the words! Just because they happen to have been pitched against a better player doesn't detract from their own skill, and it's the individual's performance which interests me.

So, imagine this sequence of scores:

On the TV show Player A would win the first 3 rounds and have 19 points. Player B would only have 8 points. So the winner would be Player A.
But Player B managed to get 24 letters! Their actual total is higher, and that's what I'm counting.
I use a similar system for the numbers round. If they get it exactly they get 10 points, 7 points for getting within 5 of the target number and 5 for within 10. Losing contestants also get points according to their claim.

The Conundrum is a tricky one. There is no way to know if the losing contestant would have solved it within the time limit, so I'm assuming not. It may be that I decide to drop these points from the totals at a later date.

The mean score is calculated for each age bracket, and the result is a column chart called "Ages". A comparison of "Gender" and "Occupation" is also provided.

Finally there are a few charts designed to satisfy my curiosity about how people choose numbers for the numbers round. Essentially, contestants get to choose between zero and four large numbers to be included in their calculation. If someone chooses zero or four that makes the round very difficult. This normally means they are either very cocky or they are deliberately trying to ensure than no one gets the answer.
My theory is that younger people are more likely to do this. So the chart called "Numbers 3D" shows the percentage of times each age bracket will choose each variation of numbers. Because this is quite confusing I've also split it up into separate age brackets for "Numbers Individual".

The final chart is called "Sneaky", and analyses the age of contestants who choose zero or four top numbers.

Now, the issue of ages. They rarely state the ages of the contestants on the programme. Only in the case of child prodigies do they bother to make an issue of it. This gives me a problem, as I have to estimate the age of each contestant. I have deliberately left the age brackets quite wide to minimise the effect of errors and so far I have had few problems. In general it is possible to distinguish a 30something and a 40something but there is a possibility of confusion, especially where people lie on the boundary.

Added 17/8/2004

I have decided to end the experiment. After 236 shows (over a year's worth of programmes) a pattern has emerged, and I think I can relax a bit. It seems the youngest person DOES win most of the time. The final results are available here.