Burke: Betting in-game or second half college basketball totals


Reader Stewart C. brought a very interesting segment from A Numbers Game to my attention and asked if I could write about the topic that was discussed. On Friday, Todd Wishnev joined Gill Alexander to talk about betting college basketball totals from an in-game standpoint and it was a fascinating discussion, but one that may have been tricky to grasp on air.

Before I get into the topic at hand, I want to tell everybody that Burke’s Betting Blurbs will be moving to the free newsletter (the earlier one) starting tomorrow. The goal is to reach even more people with these concepts and topics, so it will be part of the other daily email going forward.


Back to the Wishnev topic… Todd spoke about a shortcut formula to calculating the number of possessions that have been played to a certain point in the game and how that creates a betting opportunity with the live total. You only have to calculate one team, as the other one should roughly be the same, but you can do both just to make sure.

The formula is: Field Goal Attempts %plussign% (Free Throw Attempts * 0.5) %plussign% (Turnovers – Offensive Rebounds). The halftime example Todd used on air is that if there have been 30 FG attempts, 10 FT attempts, 8 TO and 5 OReb, you would have 30 %plussign% 5 %plussign% 3, so that would be a pace of 38 possessions, which is a very fast pace. A very slow first-half pace would be 28 or fewer possessions. The “average” will be somewhere around 32 or 33.

Full-game total lines are built on the expectation of pace. A game with a total of 125 implies a very slow pace, likely around 62 or 63 possessions. If a game is moving faster than that, 125 likely won't be enough for the total and the live betting line may not adjust properly.

Todd said on the segment that the median number of possessions in a game is 67.5, but you will see slow-paced teams that play to 63 or fewer possessions or fast-paced teams that play to 73 or more possessions. Games trending below or ahead of their expected pace create in-game totals betting opportunities because they’re being played slower or faster than the expectation, thus creating the chance for the game to be lower-scoring or higher-scoring than the line.

In-game algorithms won’t properly account for this. They are based on the pregame total and the pace that was expected. Furthermore, there are typically more possessions in the second half than the first half because of late-game fouling, teams trying to play faster to catch up, etc. If a game is ahead of its projected pace in the first half, you can reasonably assume that it will likely stay that way.

There are exceptions to this rule, like a game with really high shooting percentages that are likely to fall off. Games with a lot of turnovers create a “fake fast game”, as Wishnev noted. But, the idea here is that the live betting algorithms incorporated by the sportsbooks are limited in their scope of knowing what is actually happening in the game. The algorithms are mostly formulaic (purely math-based), so you can find edges in this market if you are studying live box scores.

I texted with Gill last night and he told me that he’s hoping to elaborate more on this with Wishnev tentatively scheduled for Monday’s edition of ANG. I’m also filling in on Prime Time Action on Monday and Tuesday and Gill and I will try to expand more on the topic as well, especially with games taking place as we’re on the air.

Thanks to Stewart C. for bringing this to my attention and I encourage readers and subscribers to reach out if they ever have any questions like this about something said on air or something written about here at VSiN.com.