Middle Bet Calculator

The Middle Bet Calculator shows three cash-flows for two opposing (or overlapping) positions: only A wins, only B wins, and both win — the middle. Defaults: Stake A $110 at 1.91, Stake B $100 at 1.91. Profit if only A: 110 × (1.91 − 1) − 100 = 110 × 0.91 − 100 = $0.10. Profit if only B: 100 × 0.91 − 110 = −$19.00. Profit if the middle hits (both tickets paid): 110 × 0.91 + 100 × 0.91 = $191.10. The highlighted row is that $191.10.

Those defaults are a classic slightly skewed two-way: $110 on A almost Dutchs the $100 on B at the same 1.91, so a one-sided A win is roughly flat (+$0.10) while a one-sided B win drops $19.00. The middle is the whole object — for example a spread of +3.5 versus −3 at −110/−110 style decimals, where a 3-point win pays both sides. Totals middles (over 220.5 and under 221.5) use the same three columns. Type the two stakes and two decimals you actually hold or plan to hold.

If you equalized the two-way instead of overweighting A, you would stake about $100 / $100 at 1.91 / 1.91: then only-A = 91 − 100 = −$9.00, only-B = −$9.00, middle = $182.00. The default $110 / $100 pair buys a near-zero A-only result at the cost of a worse B-only result, which is how people shade toward the side they prefer if the middle misses. The calculator does not judge the shade; it prints $0.10, −$19.00, and $191.10.

Decimal 1.91 is the usual −110 American converted: 100/110 + 1 = 1.909…, displayed 1.91. If your actual prices are 1.95 and 1.87, type those. Profit if only A becomes 110 × 0.95 − 100 = $4.50; only B becomes 100 × 0.87 − 110 = −$23.00; middle = 104.50 + 87.00 = $191.50. Middles do not require equal odds; they require a scoring interval in which both contracts can be right at once. The engine never sees the spread or the total; it only sees the two tickets’ odds and stakes.

Read the three rows: Profit If Middle Hits $191.10, Profit If Only A $0.10, Profit If Only B −$19.00. If both “only” columns are positive you have an arbitrage, not a middle. A true middle has at least one negative one-sided column and a large both-win column. Educational three-state ledger. Not a recommendation to buy +3.5 and −3, and not a prediction that the landing number sits in the gap.

A middle is a pair of bets that can both win on the same event because the lines differ. Point-spread middles (Cowboys +3.5 at one book, opponent −3 at another) pay both sides when the margin is exactly 3. Totals middles pay both when the combined score sits between the two numbers. First-half / full-game splits, alternate lines, and some player-prop thresholds produce the same geometry. The calculator does not need to know which geometry you are in; the cash-flows are always “A only / B only / both.”

US sportsbooks in the 1980s–2000s made middles famous because −110 / −110 on off-by-a-half lines created cheap convexity: small one-sided losses, rare but large double wins. The default $110 / $100 at 1.91 / 1.91 is that −110 world with a $10 shade. Expected value is p_middle × 191.10 + p_Aonly × 0.10 + p_Bonly × (−19.00). Without p_middle the $191.10 is a headline, not an EV. Books now shade alternate lines specifically to starve that p_middle.

Middles are not Dutching. Dutching equalizes several backs that cannot win together. Middles require the possibility that they do win together. They are also not a two-way hedge: a hedge wants the complementary side of the same line, which cannot both win. If you hedge the same −3, you will never see the $191.10 column. You would see the partial-hedge columns instead. Mixing the vocabulary is how people accidentally close a middle they meant to keep open.

The −$19.00 B-only result is the price of the $0.10 A-only shade plus the juice. At fair 2.00 / 2.00 with $100 / $100, one-sided results would be $0 / $0 and the middle $200 — but fair 2.00 two-ways do not exist on a single book, and two books posting 2.00 / 2.00 on overlapping lines is an arb-plus-middle chimera that gets limited instantly. The 1.91 juice is why one-sided columns go negative even when stakes are balanced.

Limitations: no push column (a spread push is a fourth state: stake returned on that ticket). If A can push while B wins, add that cash-flow by hand: you get stake A back and B’s profit. The engine’s “only A” assumes A pays and B loses, not that A pushes. No correlation with other bets, no closing-line value, no probability. Related tools: arbitrage calculator when both one-sided profits are positive; partial hedge when you want to close rather than keep the gap. This page keeps the gap. Educational three-state arithmetic, not advice to hunt 3-point football middles.

A middle is a three-state ledger for two tickets that can both be right. Profit if only the first ticket wins is that ticket's net payout minus the second stake. Profit if only the second wins is the second net payout minus the first stake. Profit if both win is the sum of the two net payouts. Those three identities are the whole calculation. Defaults are 110 dollars at 1.91 and 100 dollars at 1.91, the usual conversion of minus-110 American juice with a ten-dollar shade toward the first side.

Net on the first ticket is 110 times 0.91, which is 100.10 dollars. Net on the second is 91.00 dollars. Only-first profit is 100.10 minus 100, namely 0.10 dollars. Only-second profit is 91.00 minus 110, namely minus 19.00 dollars. Both-win profit is 100.10 plus 91.00, namely 191.10 dollars. Checksum: only-first plus only-second plus both stakes equals the middle, because 0.10 plus minus 19.00 plus 110 plus 100 equals 191.10. Display matches two-decimal currency.

Balanced juice at equal stakes of 100 and equal odds 1.91 would print only-first and only-second of minus 9.00 dollars and a middle of 182.00. Overweighting the first side by ten dollars buys a near-flat only-first column at the cost of a worse only-second. An arbitrage, by contrast, needs both one-sided columns non-negative, which is the inverse-sum test that the two inverses sum to at most one. A true middle has at least one negative one-sided column and a large both-win column. Complementary moneylines cannot both cash; plus 3.5 and minus 3 can.

United States books in the late twentieth century made spread and totals middles famous because minus-110 against minus-110 on off-by-a-half lines created cheap convexity: small one-sided losses, rare large double wins. Expected value would need the three state probabilities, which are not inputs. A few percent on the 191.10 dollar column can justify the minus 19.00, or not. Books now shade alternate lines to starve that both-win chance. Mixing the vocabulary with a hedge is how people accidentally close a gap they meant to keep open.

Assumptions: three states only, no push, no void, no dead-heat. A spread push is a fourth cash-flow: stake returned on that ticket. If the first can push while the second wins, you get the first stake back plus the second profit; the only-first column assumes the first pays and the second loses. Decimal 1.91 is 100/110 plus one, about 1.909, displayed 1.91. Type the prices you actually hold. The three profits are a ledger, not a forecast of the landing margin.