Footy 30 May 2005 02:18 pm

Footy tipping - round 10

Well, it took me nine weeks, but I finally noticed that, if I wanted this little experiment to at least look scientific, there was something missing: a control group. Which, in this case, translates to a “pseudo-strategy” that would select random winners for all games. In this way, I’d be able to compare the results of my simple strategies to blind luck.

Fortunately, this little flaw is easy to fix. I wrote a program to randomly select eight results (”home” or “away”, with a 50-50 chance); ran it a few million times to ensure that I was indeed getting a random spread of results; and then ran it nine times and compared the results retroactively with the first nine rounds of matches. Then I ran it once more to generate results for this week.

The results were… interesting. When compared to the other three strategies, random tipping is achieving at least as many correct guesses as the worst-performing of the others, and is returning more money. Most of this is due to managing to get seven correct results in round 9, where most of those were upsets. So, it would seem that none of the other strategies is that much better, or even as good as, random guessing.

As for this week, we got mixed results: “money” and “home” got five correct guesses each (both had exact the same tips, by the way), and “ladder” and “random” got three each. This was the second week in a row when all three strategies, other than the control group, returned net financial losses.

And here are the numbers:

Round 10 Ladder Money Home Random
Correct tips 3 5 5 3
Accuracy 37.5% 62.5% 62.5% 37.5%
$ result $-5.70 $-2.50 $-2.50 $-15.45
 
Cumulative Ladder Money Home Random
Correct tips 45 46 47 45
Accuracy 56.25% 57.50% 58.75% 56.25%
$ result $17.60 $-55.65 $31.60 $55.25

Links:

Subscribe to the comments through RSS Feed

Leave a Reply