The cousin of arbitrage: instead of locking a small profit, you accept a small cost for a shot at winning both bets.
A middle is arbitrage's more interesting cousin. Instead of locking a small profit, you accept a small cost for a shot at winning both bets at once.
Two books disagree about a number. You take both sides at their own numbers, and if the result lands in the gap between them, both tickets win.
You risk $10 for a $200 payoff on a margin of 5 or 6, which happens in roughly 6% of NFL games.
That is the trade in general: a small, certain cost against an occasional large win. At those numbers you need the middle to hit about once in twenty attempts to break even, and a 5-or-6 margin comes in comfortably more often than that.
Middles live and die on key numbers. A gap spanning 3 and 7 in football is worth several times a gap of the same width spanning 8 and 9, because those two margins are where games actually land.
The half point matters here more than anywhere. A middle from +7 to -4 includes 5 and 6. A middle from +6.5 to -4.5 includes 5 and 6 as well but pushes on nothing — and a middle from +7 to -3 includes 4, 5 and 6 *and* pushes on an exact 3 or 7, which is a free extra outcome on both ends.
| Sides | Wins both on | Pushes on |
|---|---|---|
| +7 / -4 | 5, 6 | 7 (one side) |
| +7.5 / -3.5 | 4, 5, 6, 7 | nothing |
| +3.5 / -2.5 | 3 | nothing |
| +3 / -3 | nothing | 3 (both sides) |
The second row is the one to want: four winning margins including 7, and no pushes to muddy it. The last row is a low-hold with no middle in it at all.
A low-hold bet takes both sides at prices adding to just over 100%. There is no middle and no profit — the outcome is a small guaranteed loss, often under 1%.
Pointless on its own. Useful for three things: clearing a rollover requirement cheaply, converting a promotion into cash, and moving a balance between books without paying real vig for it.
Live middles across pick'em platforms and sportsbooks, with the gap and the cost shown. Open the middles board.