How I Turned A Losing Slate Into A Winning Slate with Betscope

The best part of finally launching Betscope is not having to talk about the concepts that power it in the abstract any more- there are actual products with actual use cases on how it can drive positive sports betting returns. So today, I thought I’d walk through a use case that helped me find some winning bets last night.

I typically start my sports betting process with Betscope not with any particular game I want to target, but rather with looking at the entire slate of games and assessing how many markets have price discrepancies. This is the foundation of the market scanning approach to betting, which starts not with a specific belief in a game I want to bet on, but rather identifying where prices are weakest in a game. Here’s what the slate looked like yesterday afternoon when I opened it up:

As a reminder, each color in the games dropdown represents a different kind of price discrepancy for a market in that game:

There are markets with price discrepancies in each game, but the Hornets/Celtics game jumped out at me immediately with how many it had, including a very valuable arbitrage opportunity. I knew that game had a lot of discrepancies to capitalize on, so I clicked through to the Celtics-Hornets game to assess where I wanted to attack the market.

As it turned out, all of the markets with pricing discrepancies were in prop bets, not an uncommon occurrence. Here’s how the layout looked at that time:

A reminder of what each of these numbers are: these are the implied totals from the market consensus lines for each prop bet. We get these numbers by finding the mathematical distribution that corresponds to the consensus lines. This distribution fitting process is critical for assessing price discrepancies, which is why we take such pains in getting it right. 

Every box that has a colored border has a price discrepancy across two or more sportsbooks, and is indicative of a market that I should be targeting. Ideally, I would have my own projections for every single number and I could fill in my own for each of these boxes, but that’s largely impractical. Fortunately, Betscope has a great feature, the correlations engine, where all I need to do is input a single bet I think might be different than the market numbers, and the correlations engine will explain how that view affects every other bet. (On my appearance on the Trill Withers Show yesterday, he called it the “butterfly effect”, which sums it up perfectly.)

Not having a whole lot of strong opinions on how the game would go, I looked at some other opinions on the game and other bets people might be targeting. I found a reasonable Action Network article indicating the author was targeting Robert Williams’ points+rebounds market:

Coincidentally, this was one of the markets that had a low synthetic hold, as identified by Betscope’s analytics that finds price discrepancies: 

I adjusted Williams’ projected points and rebounds totals, then clicked the “Calculate Stats” button to activate the correlations engine. In short, the correlations engine figures out what happens in a world where my new projected totals for Williams’ points + rebounds are correct and how all the other stats are different as well. Here’s what those updated stats look like:

And here are the bets the ROI calculator, which runs all of the updated stats against every price at every sportsbook, recommended I make:

These recommended bets come from the principle of attacking your correlated beliefs through markets with pricing discrepancies, as this is one of the best ways to improve your long-term win rate at sports betting. The first recommended bet, coincidentally, was already the bet indicated by the article; the fact that I was able to target it with low synthetic hold was an added bonus. The ROI calculator also highlights the importance of how much finding the best price possible is to maximizing your returns: even though Fanduel and Sugarhouse both had profitable opportunities, the expected ROI of the Fanduel line was much higher than Sugarhouse. But it’s the other two bets that were particularly interesting to me as well, because these were angles not considered in the original analysis or the lines that I changed. P.J. Washington’s points prop highlighted a split market price inefficiency, where the market consensus was between two over/under numbers, but the price was particularly good on one of the numbers:

I was okay betting over 10.5 at +104 because as the correlations engine indicated, in a world where Williams’ points + rebounds were higher, Washington’s point total would also be a little higher as well. The combination of the price inefficiency plus the correlated outcome being in the right direction was enough for me to pull the trigger.

The other market was Al Horford’s rebounds, which had such a large price discrepancy across markets, there was an outright arbitrage situation, where I could have locked in a profit by betting both sides: 

Even though Horford’s correlated outcomes for rebounds was slightly negative with my adjustments, I knew the price I was getting on over 6.5 rebounds at PointsBet was too good not to bet. As others have pointed out, an arbitrage situation usually occurs when there is one line that is so mispriced, you have to look for a reason not to bet it. The ROI calculator figured out which side of the arb I should be on, and I fired knowing it was too big of a mispriced line not to attack it.

So how did the results go? The WIlliams rebounds bet didn’t hit, but the Washington points and the Horford rebounds did, resulting in a 2-1 record for this slate:

If I hadn’t discovered the additional bets I wasn’t even looking for, it would have been an 0-1 night from the recommended bet from the article, but incorporating correlated outcomes and mispriced lines turned what would have been a loss into a win.

One of the impressions a lot of people have had in the first week of launch is that they need super strong opinions to use as inputs into Betscope’s tools, like a handicapping model or super detailed analysis, in order to leverage it the most. I walked through these steps to illustrate how that’s not the case: once you spot a game with enough price discrepancies, even running only semi-strong opinions through the correlations engine is enough to highlight profitable bets. As long as you’re attacking your opinions through markets with price discrepancies and exploiting them, you have more than a fighting chance to expect positive returns on your bets. I encourage you to go try out your own hunches, knowing that firing them into correlated markets with mispriced lines is a great way to boost your winnings.