Edges

Edges in sports betting are wide-ranging. They can be anything from an extremely complex model to noticing that a player went out clubbing last night. I think the hardest part about finding an edge is knowing what a real edge is. Going from zero edges to one is far harder than going from one to ten. There is a snowball effect to knowing what winning looks like and how to repeat it.

Most of my edge personally comes from being able to price NBA and WNBA sides, totals, and props well enough to win in that market. That’s one type of edge—one that will typically last longer but is always being adjusted and requires constant work. The edges we’re talking about in this newsletter are the more discrete ones: you know something that the sportsbooks don’t and make money off that knowledge.

I’m going to go through five of these edges I’ve had over time, most being dead edges with a couple still somewhat existing. You’ll often hear winning sports bettors talk about how easy it is to win—not necessarily win big, but to win at some scale. I think that’s mostly true, but really, the hardest part of sports betting is figuring out how to learn the right things. It’s a bit of a paradox: anyone who is giving away a current “edge” or writing about how to win is probably losing, and anyone with a significant edge is trying to protect that information. I hope reading about how I found and thought about edges I used in the past will be valuable to people trying to become winning bettors.

Lots of these edges also came when I was much more inexperienced as a bettor, and if I had the same opportunities today, I would definitely do some things differently.

Edge 1: PJ Tucker and James Harden on the Sixers

I believe I have talked about this edge before on a podcast, but it is one of my favorites of all time. This is what I would categorize as a pure “ball-knower” edge and a great example of how having more basketball knowledge helps your betting ability (which somehow many disagree with). The 2022-23 season for PJ Tucker was one of the most unique in NBA history. He started all 75 games he played in, averaged 3.5 points per game, and had the lowest usage rate in the league at 6.4%. He was out there to rebound, defend, stand in the corner on offense, and let Joel and Harden do their thing.

Tucker received an average of 11 passes per game and dribbled the ball only 0.5 times per touch. Sixty percent of Tucker’s shot attempts that season were corner threes, and 100% of his three-pointers were assisted. Generally, the only ways he’d score were from a corner three assisted by Harden or Joel or a putback on an offensive rebound. James Harden missed 14 games that season, and when he was out, the books would increase Tucker’s points and three-pointer props. This is typical—if a high-usage starter goes out, every other starter's props will go up, and that is correct almost every time. But in this case, it wasn’t. Tucker’s role was so unique and dependent that his props should have gone down considerably with no Harden.

Every game Harden was out that year, I took PJ under points and under threes. I don’t have the exact stats on how those 14 games did, but in 589 minutes that season without Harden, Tucker averaged 2.5 points per 36 minutes, compared to 6.21 points per 36 minutes with Harden. He also shot 2.02 three-point attempts per 36 without Harden and 2.98 per 36 with him. When you have significant basketball knowledge, you can take advantage of edge cases that are very hard for a traditional algorithm to price, and I believe this was a great example of that.

Edge 2: Women’s NCAA

The next two edges both took place on the same exchange that my partner and I exploited while building our bankroll. The Harden and PJ Tucker edge required both basketball and betting knowledge while being confident in my understanding of both. This edge was much, much dumber than that.

We had an exchange that offered Pinnacle lines for many smaller-level sports, and they used Pinnacle’s lines at the book. The kicker was they would offer $4K a pop with reloads on women’s college basketball, while Pinnacle was only taking $50. We would max bet Pinnacle four times one way, with no knowledge of the teams or games, assuming the price was fair enough. Pinnacle would move the total around 10 points, then we would just max click the exchange as many times as possible.

This edge was quite obvious, and I’m not going to pretend I’m a genius for betting it, but I think there is value in realizing that edges don’t have to be complicated.

Edge 3: Our Version of Top-Down

The exchange we were using was a crypto exchange, and one of the "selling points" was that all bets were on the blockchain, meaning all bets were public. We could track every bet that was made and by which bettor.

We started tracking every bettor on the website and profiling sharp tendencies. Once we had enough info to feel confident someone was sharp, we would get a notification of the bet they placed and immediately jam every account we had on that bet. Of course, the bettor was generally placing these bets at the exchange to keep them quiet, and we would move the market, driving the bettor away. Again, I would do some things differently now, but this was still a very profitable endeavor for us.

The best bettors for us were small-time sharps who didn’t realize how big an edge they had. They’d bet a small amount; we’d get a notification and go hit it everywhere, and the bettor wouldn’t mind at all.

Edge 4: Position Players Pitching

As MLB teams got smarter, they realized there was no reason to burn bullpen arms in unwinnable games. As a result, teams started putting position players in at a far greater rate in blowouts. In 2017, there were 36 instances of position players pitching; by 2022, that number had jumped to 132.

Obviously, position players are not very good at pitching, and the live algorithms for books struggled with these adjustments. We tracked teams, identified thresholds for when they would put in a position player, and monitored bullpen usage. During blowouts, we’d refresh Twitter feverishly to see if any beat reporters mentioned a position player coming in. Once confident, we’d hammer the live over and spread.

There were even times I’d buy tickets to Jays games in the 6th or 7th inning, hustle to Rogers Centre, and check out the bullpen myself. Sadly, in 2023, MLB implemented rules to make position-player pitching less common and more predictable, and books started pulling live markets when these scenarios became likely.

Edge 5: Full Cashouts

This is the edge I’d say is the most alive right now—and probably the one that will get some people riled up for me publishing it. The basic idea is to take full advantage of any site that offers full cashouts, and there are a lot of ways to make this work in your favor. When we were starting out and felt like we had a small edge but weren’t sure, we’d place bets overnight based on what we thought was an edge. If the bet moved in our favor, we’d let it ride. But if it stayed neutral, we’d cash out without risking a dime. Of course, it’s not risk-free—bets can go against you—but it definitely helped boost our ROI by cashing out on any neutral (-4.5% EV) bets. Another way we’d attack this angle is with injury info, especially when a star player is 95%+ likely to play (or not). When a player is that likely to be in or out, the prop lines usually won’t move if the news goes as expected. We’d bet longshot SGPs or ladder individual player stats that depend on that player. If the player went in/out as expected, we’d cash out, but one in 20 there’d be that surprise injury direction—and we’d have a 200-1 bet that was fair 20-1. Lastly, some sites will get jittery if they see a lot of action on certain markets, especially with super long futures, and especially when they notice sharp accounts betting them. We’d place a 200-1 bet on a player for an award from a fresh account, and then get others with limited accounts to max bet whatever their limit was. The line would shift to 20-1 almost immediately, and the site would offer 9x cashouts on a bet we didn’t have an opinion on.

Final Thoughts

The last thing I’ll say is that betting edges—especially ones like this—can be risky. I’ve got the benefit of hindsight now, so I know for sure that these edges were real, but that’s not always the case. It’s easy to convince yourself you’ve got an edge when you don’t, and that’s how you end up going bust, so proceed at your own risk. I recently got a text from a sharp bettor who was pricing NBA half-markets. His logic for why teams would play slower or faster in specific halves made perfect sense, but he missed one key factor (which I’m not sharing here) that’s crucial when pricing NBA first and second halves. Because of that, all his analysis was essentially useless, and he was making some bad bets. Sometimes you think you’ve got the nuts, but then you get kicked in the nuts instead. Be careful out there.