[kosti]

Not having much of a mathematical background, I find myself stuck in a dead end in my research....maybe not a dead end but actually an end with multiple forks in it. Hoping for some guidance here from the pros.

I have worked on an over/under strategy for soccer for weeks now. I have back tested a year with tens of thousands of results. The hit rate is surprisingly strong BUT this is just on outcome, and NOT outcome and odds. I've only been researching the odds since I started testing (no back testing includes odds as I wasn't able to find them). I know this is a huge issue and I need to start addressing and researching it before I put serious money into this. I have no doubt that some of my favourable results would have had odds of 1.60 (-165) and have even seen it as bad as 1.45 (-220)...this was only once in an entire month but I'm sure there are multiple of those in my back testing.

I know that working a 'system' or not considering odds when gathering data is stupid and will end up in bankrupcy...but my question is this: What sort of odds do I start betting at? How do I determine the odds where this would become profittable?

Although I have a mathematical sort of mind, I don't have any extensive mathematical or statistical knowledge (no university courses in math etc.).

No idea if the answer is clear, or whether the question is even answerable. But I am at the point of not knowing where to go from here. If anyone could give me a push in the right direction I would greatly appreciate it.

Thanks
Peter

[allin1]

in order to determine the odds where it would become profitable you have to come up with you own probability for that result-event

in other words: every line (odds from bookmakers) has it's implied probability. if you believe the probability for that result is bigger than the implied probability of the odds the Bookmaker is offering then you take it because you think there is value.

if we are talking about major markets with high limits (major leagues, world cup, euros, champions league) then the closing lines are the ones you want to watch and compare those and their implied probability to your estimations. if you are near those then you might have something. if not, then you are probably treating this in a very simple approach that can fool you into thinking it can be used to make money.

I am an amateur when it comes to maths in gambling so I hope I am not wrong. If I am, hopefully someone that knows better will correct me.

there is a lot of useful information on this forum if you spend some time doing research, you can learn a lot from here

start with ganchrow's threads regarding kelly

[allin1]

[Dark Horse]

Using SBR's odds conversion tool, translate the winning percentage of your model into a line. That line is where you play even. Then add an overlay of maybe 2% (or whatever you prefer) and determine the corresponding line.

[VLR100]

If you're looking to increase your proficiency in math you could do a lot worse than checking out www.khanacademy.org
If you have the prerequisites the probability and statistics sections would be quite useful. Take it slow through the topics and try and think about how they could be applied to sports and odds.

Not trying to put you down, but you're not going to get very far unless you have at least a passing knowledge of the basics of statistics and probability.

[indio]

Without really understanding your system of making selections, it makes it a little tough, but I'll try to keep it simple.

You say you tested the veracity of your system with 10,000+ results, so you have a solid volume to combat variance. What is the win rate of your selections? Whatever it is, simply demand a price that's better than that, and keep your risk amounts around 2% of your bankroll.

So if your win rate is 57%, you should only be betting selections where you have an advantage. In this case selections at -125 or better will give you an implied edge of 2.6% +. If your win rate is 60%, you need -140 or better, etc..

You should have an estimation of how many bets you will be able to make in a given year, and with simple binomial distributions, you should have an understanding of your chances of obtaining the profit you desire.

Of course, all of this is dependent on the veracity of your system, but if it's based on over 10,000 results, it's worth trying. Don't be afraid of failure so give it a try, but don't discount the possibility of failure, and don't let it ruin you.

The other variable is where you will be making these wagers? Be prepared that if you start winning consistently, many shops will limit you or close your account, so make sure you're aware of that variable.

[kosti]

Thank you very much for the help guys.

Logically I thought about matching my win rate % to the corresponding odds. My only concern would be that my back testing would give me an inflated number since it would have included -165 odds as well as +150 and everything else in between. What if a high percentage of my win rate were at terrible odds and most of the losses were at odds where I had implied edge? No real way of determining this at this point in time, but I will go back to see the results for the last month (where I do have the odds info) and see what I can come up with. A month isn't enough results, but will at least give me an idea.

Thanks for the links and suggestions everyone!

[HUY]

Not having much of a mathematical background, I find myself stuck in a dead end in my research....maybe not a dead end but actually an end with multiple forks in it. Hoping for some guidance here from the pros.

I have worked on an over/under strategy for soccer for weeks now. I have back tested a year with tens of thousands of results. The hit rate is surprisingly strong BUT this is just on outcome, and NOT outcome and odds. I've only been researching the odds since I started testing (no back testing includes odds as I wasn't able to find them). I know this is a huge issue and I need to start addressing and researching it before I put serious money into this. I have no doubt that some of my favourable results would have had odds of 1.60 (-165) and have even seen it as bad as 1.45 (-220)...this was only once in an entire month but I'm sure there are multiple of those in my back testing.

I know that working a 'system' or not considering odds when gathering data is stupid and will end up in bankrupcy...but my question is this: What sort of odds do I start betting at? How do I determine the odds where this would become profittable?

Although I have a mathematical sort of mind, I don't have any extensive mathematical or statistical knowledge (no university courses in math etc.).

No idea if the answer is clear, or whether the question is even answerable. But I am at the point of not knowing where to go from here. If anyone could give me a push in the right direction I would greatly appreciate it.

Thanks
Peter

I don't know what's so complicated about it. Just get the odds and see how your system does against them. You'll most likely find that your system is useless.

[allin1]

kosti I recommend that you fully understand these things first:

- implied probability and no vig lines
http://www.onlinebetting.com/remove-vig/
http://www.sportsbookreview.com/betting-tools/
http://www.sportsbookreview.com/betting-tools/odds-converter/

- kelly criterion
http://www.sportsbookreview.com/betting-tools/kelly-calculator/

- market efficiency and beating the closing line

here are some videos to get you started