My first attempts at a system

[AndrewKent]

Hi folks, I'm currently experimenting with different ideas for systems. As I am a data analyst for a living (and a VBA programmer) I have to admit I am enjoying the challanege of putting a system together, dealing with the data and seeing if I can make a little bit of success from nothing. I found an interesting Over/Under system that I wanted to share with you:

This is an Under/Over strategy 

I'm afraid it look's a bit simple, but it has been quite profitable for me. 
As mentioned it's an under/over strategy and it's based only on statistics. 

Let's imagine that there's a match between team A and team B. 
(A is the home team). First of all I will check what is team 
A's average of goals scored and conceded at home. Let's say 
that team A scores 2.1 and concedes 1.20 - total of 3.30 
goals average in team A's home games. 

Then I will check what team B's average of goals scored and 
conceived in their away games is. 
Let's say that team B scores 1.75 avg. and concedes 1.15 - 
a total of 2.90 goals avg. in team B's away games. 

Then I take the two smallest numbers 1.15 + 1.20 = 2.35 - 
minimum goals in this match. In the same manner you find the 
maximum which is 3.85. The avg. of the min/max is 3.10 which 
is much more the line of 2.50, so I will bet over in this game. 

If I get an avg. of 2.70 or more I will bet over with high 
stakes. If it's between 2.50 and 2.70 I will decide by myself 
if the OVER bet is worth a try, but it will definitely be 
with small stakes. 
If I get an average between 2.30 and 2.50 I will decide if the 
UNDER bet is worth a try and it will definitely be with 
small stakes. 

If I get an average of 2.30 or lower I will bet under with 
high stakes. 

Of course, sometimes I don't necessarily bet on the 2.50 line. 
I might also bet on lines of 3.00, 2.00 and others. 

I have already increased my bank by 50% in one month using this 
betting strategy and I will certainly continue increasing 
my betting bank further. 

now I have put this system together with my source of data (from www.football-data.co.uk) and although I can see there can be a profit made here, it seems to be very very small indeed (about 6% ROI) can anyone suggest any factors that I might have overlooked. I thought about the following: 1) Only analysing data where Team A / Team B are playing in the same division 2) Looking to find the Over / Under % for the league as a whole and working from there Cheers, Andy

[Dices]

Re: My first attempts at a system Can you associate in your analysis the "relationship" between this average figure and the offered odds? I mean can you define "value" in your possible bets? An average of 3.1 goals is hight but what if the odds at this case is 1.4? If you divide 3.1/1.4 you will get a figure and based on that you place bets only if it's above/below .... (and maybe varying stakes) I'm only proposing an approach. 6% ROI (yield) is not small. How many is your bets?

[Norfolk Enchance]

Re: My first attempts at a system Agree with Dices about defining value. The odds should determine whether you bet or not, alongside your estimations for average no. of goals and probability of over/under whatever lines are offered. 6% is not very very small for over/under betting. It’s actually pretty good for the kind of odds over/under betting deals with. Especially if the system generates a large number of bets.

[merlin]

Re: My first attempts at a system ..... Hi Andrew and welcome to the site. I know just what you need in 2) above, League percentage for Overs and Unders. Go to:- www.h2hstats.com and click on Soccer, then Leagues, and your Unders are three columns from the right hand side of the screen, with Overs next to it. Hope this helps and good luck. Cheers.;)

[AndrewKent]

Re: My first attempts at a system Hi guys, Thanks for the input! In truth I didn't know what was a good or bad ROI but I thought 6% was low. Sadly from the sample set of games I analysed the volume was low. I will continue to work on this. The good thing is I have a source of data that I've built to automatically update. That at least is a start. Cheers, Andy