Question of the Day - 24 August 2020

We can try. Admittedly, it's a bit technical and can get complicated, but here goes.  

You "middle" when you have bets at different numbers on both sides of a game and both win.

For example, you bet the Steelers -2.5 and the Ravens +3.5 and the game ends something like Steelers 27/Ravens 24. The key is having the result fall between the two numbers, otherwise you'll win one and lose the other, which results in a loss of the vigorish. If that Steelers game ends 28-24 with $110/$100 bets on both sides, you lose $110 and win $100 for a net $10 loss. Hence, you have to win better than 1-in-20 tries to profit.

It's possible to find line discrepancies simultaneously, but they typically occur when you already have a bet on one side and the market moves the line enough to bet the other way.

It's a lot easier to middle in basketball than football. The reason is that there are no "key" numbers in basketball. In football, the 3 and 7 are most important (in that order) for sides, and there are also (lesser) key numbers for totals. Middling around these numbers brings additional considerations into play and that makes it difficult to provide a compact answer (you can do well getting on either side of 3 in pro football, as demonstrated in the above Steelers-Ravens example). In basketball, none of the numbers land a lot more than others. Hence, it's much easier to lay out good general rules of thumb for this game. Pursuant to that, here are some rules for middling in basketball.

1. A 2-point middle laying no more than -110 on each bet is profitable over the long term for both sides and totals. Not surprisingly, sides have less variance than totals, so a -7/+9 play on a side is much better than an OV 134/UN 136 total. The former has a solid edge, the latter just a tiny one. Other gaming authors, including the usually very solid David Sklansky, have gotten this wrong. The reason sides are better is that there's a stronger "regression to the mean" on sides (in games where a team is over-performing, they tend to regress).

2. A 1.5-point middle on a side is usually profitable, but is more suspect at higher pointspreads. Playing -6 and +7.5 in the NBA will get the money, but -24 with +25.5 in a college basketball game is a lot more dicey.

3. Since there are no ties in basketball, games can't land on zero, so a +1 with the other team +1 is not profitable. Be careful crossing over the zero.

4. Given that you can lay a combined 20¢ with a 2-point middle and have the best of it, when betting with reduced vigorish and laying 10¢ or less, you need only one point. So -6 -100 with +7 -110 is a good middle. In the "old days," good middles used to be everywhere. Fast forward to 2010, and many middles players now must have a good sense of which way a line will move later. By example, Kansas State's two best players were once suspended right before a game with UNLV. Pros getting wind of this could max bet (even if overbetting) UNLV +4 and +3.5, then after the news broke, play some back on K State -2 -100 at post to put themselves in a favorable situation.

5. Many middlers prefer to arbitrage and lock in sure profits (or at least a riskless bet). In the example above, some pros chose to play K State -3 +170 to have a middle and an arbirtage working for them. Assuming they bet the max on both sides, they were guaranteed to either break even (on a UNLV cover) or make money (on a K State cover) and they had a chance to win big if the game landed 3 or 4.

Of course, when you get bigger spreads in middle opportunities, your results will be better. But be prepared to fade a lot of small vigorish losses along the way while you wait for a result that wins both sides.