I was planning on writing up a big post outlining every single formula you need to know for sports betting, but instead I made an open source website SportsBettingCalcs.com that does all the work for you. The javascript and little bit of python should be simple enough to follow and wikipedia and khan academy can do a much better job of explaining the formulas than I ever could. What a day we live in! So much free information out there. I left some relevant links at the bottom of the post if you’re interested in the formulas behind the site.

Let’s suppose you want to make a living sports betting. How many bets will you need to make to earn $100k per year? How large a bankroll do you need? How much should you bet? Which book should you bet with? If you can manage a 10% ROI (unlikely) you would only have to bet $1M/year ($100k / 0.10 = $1M). A more realistic goal is a 2-5% ROI which means you’ll need to wager between $2M to $5M. In order to bet $3M/year you need to make

-300 bets of $10k a piece or
-600 bets of $5k a piece or
-3000 bets of $1k a piece

There’s just no way you’re going to find 3000 +EV bets per year. That’s nearly 10 per day. So betting small in volume is not going to earn you much. Even if you gain a big edge in a small market like esports 300 bets of $100 a piece at 10% ROI would earn you $3,000. Yeah it’s decent, but the amount of work and effort required to achieve this would probably be better served working a plain old job with much less risk. Sports betting for most people is a hobby and should be treated as such by 99% of people.

Now you understand why there are so few people making a living betting sports. In order to make it full time you’ll need to bet a large amount(4 figures) on hundreds of games per year with a positive expectation. Obviously the amount you bet will not matter if you have no edge. Increasing your wager size will only increase the amount of money you will lose per bet if you have no edge. This is very important and obvious, but worth repeating. Knowing you have a winning strategy is your #1 priority before risking ANY amount of money wagering.

There’s very few people that have the bankroll, risk tolerance and skills to making living betting sports. If they do they’ll probably make more with less risk as a consultant or tout. Don’t expect to do this full time anytime soon. In order to make $100k/year you are going to have to risk something like 5% of your expected yearly earnings per bet. That’s incredibly risky. If you do manage to find a big edge in a sport, it’s unlikely you’ll be able to bet very large because the limits will be low. For the major sports, the limits are higher but potential edges are lower. For a sport like the NFL it’s easy to bet $10k+ per game before close but there’s only 256 games per year. There’s thousands of college basketball games per year, but limits are much lower. If your goal is to maximize $ earned I would recommend focusing on a sport that balances the amount of $ you can bet with the potential amount of edge you can gain. More realistically though your goal should be to maximize your enjoyment and minimize your losses with your likely losing hobby so I’d say focus on a sport you like and forget about everything else!

So say we’re shooting for a 3% ROI. A realistic goal. What kind of winrate do we need to have a 3% ROI? Using the ROI Calculator and typing in 1.9091 for odds (-110 is typical for a 50/50 wager) and 3% ROI and clicking calculate we get that we need to win about 54% of our bets. So we need to build a system that can go 54-46 on average for every 100 bets.

Let’s suppose we download a database of our sport of choosing, get all the historical closing lines and try to find an angle. Let’s say after much digging we find that on night games where there’s a full moon the home team covers the spread 40 times and loses to the spread 30 times. Great! Winning strategy, let’s bet it right?

Hold on a second. First, the angle has to make sense. You should be able to explain why the full moon favors the home team. Remember if you watch the roulette wheel long enough in the casino you may start to think you’re seeing winning patterns there too. Second, you need to be sure the results are significant. Is going 40-30 good enough to bet? Try typing it into the T-test calculator. This will perform a one sided t-test to see if the results are >52.38% (enough to cover the vig on a typical -110 odds bet). You will find that is less than one standard deviation away from the mean and not at all significant. What about a tout saying they are 12-5 or 100-70? Which is more significant? Type them in and find out for yourself 🙂

A more realistic example is say we find that betting large home underdog moneylines in MLB are underpriced.

Does this angle make sense? This makes intuitive sense because the public likes betting favorites so the bookies may shade their line towards underdogs. So say for example we think that betting home dogs of +170 are +EV.

What kind of winrate do we need?
Using the Odds Converter and typing in +170 in the US Odds and clicking calculate we find that we only need to win 37% of the time or more to make a profit on these wagers.

Are our findings significant? In the past season let’s say we see that they actually went 127-173. Using the t-test calculator once again and using 127 for wins, 173 for losses and a population mean of 37% we find that the results are a winrate of 42% with a p-value of 0.03 which may be significant.

What’s our expected ROI? Using the odds converter we find that +170 equates to 2.7 decimal odds. Using the ROI calculator with 2.7 decimal odds, 42% winrate and solving for ROI we get an expected ROI of 13.4%.

How much should we bet? In general I’d say set aside a sports betting bankroll you are willing to lose and bet 1-2% of it per bet. Expected ROIs are just guesses. Just because an angle did well in the past does not mean it will do well in the future or at a bare minimum the ROI for the angle will not go down significantly. If you’d like to use the kelly criterion calculator I think 1/4 kelly is somewhat reasonable. So for our example, kelly multiple is 0.25, odds are 2.7, winrate is 42% and we get 1.97% recommended betsize.

Finally, let’s say after the first month of the season we’ve gone 18-42 getting an average of 2.7 odds on our bets. This is pretty bad, but is it bad enough to quit? Try out the Confidence Interval Calculator We expected to win 42% and need to win 37% to profit remember? 18-42 is a 30% winrate and the 95% confidence interval is 18.2% – 41.8%. So there’s still a decent chance we have a winning bet strategy, but it’s unlikely we are winning at the 42% chance we expected. It may be time to quit or re-evaluate or at bare minimum lower your betsize.

One last note… Always line shop at multiple books and always bet the larger number. For american odds +111 is better than +110. +101 is better than -101. -101 is better than -103. For decimal odds 2.101 is better than 2.09, 1.9 is better than 1.87. If you shop long enough you may even find two lines that when combined have negative vig on the Vig Free Calculator A few tips for when/if you do.

1. Make sure the book dealing the bad line will not scam you
2. Always bet the bad line before you hedge with the widely available line
3. Locking in the profit ie Hedging is -EV. In general you’ll make more just betting the rogue line but if you bet more than you’re comfortable with letting ride it’s a smart play to hedge

Kelly Criterion on Wikipedia

Leave a Reply

You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>