Author Topic: How to know if you have a potentially profitable strategy  (Read 4366 times)

Tags:
  • Nerd
  • Élite
  • Posts: 462
  • Gender: Male
  • I think I could be on to something here!
*
I always run ideas through test mode, but I have been burnt a couple of times when turning loose a seemingly profitable strategy.

So I started wondering how will i know how likely a strategy is to be successful? Should I stick it out, maybe it's a rough patch?

After a bit of googling I came across a formula to calculate the probability of success based on your test data. The strategy must already be profitable. This measure should be used as a confidence factor.

The formula is not a holy grail as deviations occur (see standard deviation), however it seems very

99.9% confidence factor is ideal, but 95% and starting with small stakes initially is also OK.


Examples calculation:
Backing/Laying at 11.00 (10/1) you need 73 bets to achieve 99.9% confidence that the results achieved can be sustained when going live.

Backing/Laying at 15.00 (14/1) you need 101 bets to achieve 99.9% confidence that the results achieved can be sustained when going live.


Note: on lower odds, the number of bets required to achieve 99.9% also decreases. I would recommend a minimum of 25 bets regardless.

Using Microsoft Excel [Not tried on Google sheets]

Type the average odds of all your selection in A1
Type the total number of selection in B1
Paste below formula in C1
=(1-POWER(1-(1/A1),B1))*100
Fortune favors the brave!

  • All members
  • Posts: 13
One of the best things to use is called an Archie score.  Here is a link to an explanation of it: http://false-favourites.co.uk/blog/archie-score-how-to-check-your-selections-true-chance-of-winning/

An Archie score looks at the probability distribution of the difference between the square of the expected number of wins and the actual number of wins, which is a Chi Square distribution.

It is easy to calculate.  The only variables are the number of bets, number of wins and the expected number of wins.  The expected number of wins is calculated by adding up the inverse of each bet price (using decimal odds).  So if your 3 bets were at 2, 3.5 and 6, the expected number of wins is 1/2 + 1/3.5 +1/6 = 0.95

What it calculates is the probability that your results are due to chance.  An Archie score > 4 implies that there is a less than 5% chance that your results are due to luck.

  • Élite
  • Posts: 110
*
In my opinion testing strategies in "Test Mode" only makes  sense for non in-running strategies. If you can afford, it's always better to test strategies with small amounts in real mode.

Yes, there is expected winners, there is Archie Score but be aware that even the best and most profitable strategies will have its losing runs. The best strategy can start with losing 30 or 40 points and this is the difficult moment where you have to mentally prepared to stick to the strategy. Not easy at all....

A good strategy must be based on some grounds and logic. So it's more about whether you believe in your strategy or not.

Also don't forget that markets are dynamic, i.e. what works today does not necessary in three months time. For example if you had bet on all away teams on all Premier League games in the 2015/16 season when the odds were 4.0 and bigger you would have won 70 points (after 5% commission). Do you think that the same system will work for this season?

One good starting point is to bet against anything that is hype at the moment. At the end of the day you have to find some value. In my opinion, value can be found at over/ under bet runners/players/markets, meaning going against the market is more likely to be profitable in the long rung then to go with the market.

That is my very personal opinion....

 

Please note, BetFair is seems to be currently OFFLINE