Teasers are the most misunderstood bet type in sports betting — most recreational bettors use them wrong, but used correctly at key numbers, they represent one of the few bet types with a credible positive expected value.
A teaser lets you adjust the point spread in your favor by 6, 6.5, or 7 points (in the NFL) in exchange for a reduced payout. The critical insight is that not all teaser legs have equal value.
How Teasers Work
A standard 6-point NFL teaser combines two spread bets, each moved 6 points in your favor, for a combined payout of roughly -120 (risk $120 to win $100). A three-team teaser pays better but requires all three legs to win.
Example: If the Chiefs are -8.5, a teaser moves them to -2.5. If the Raiders are +1.5, a teaser moves them to +7.5. You're now backing the Chiefs to win by 3+ and the Raiders to lose by 7 or fewer.
Why the Key Numbers Matter
NFL games cluster at specific margins: roughly 15% of games end with a 3-point margin and 9% end with exactly 7. A teaser that crosses through both 3 and 7 (e.g., moving a -7.5 team to -1.5) captures both key numbers. This is what handicappers call a "Wong Teaser" — named after gambling author Stanford Wong, who first systematically identified the edge.
A teaser leg crossing both 3 and 7 has an estimated win rate of roughly 75%, compared to the ~52% needed to break even on a standard -110 spread bet. This makes two-team, 6-point NFL teasers crossing key numbers one of the more credible edges in mainstream sports betting.
When Teasers Don't Work
Teasers that don't cross key numbers, college football teasers (lower key number clustering), and basketball teasers (no strong key number effect) do not offer the same mathematical benefit. Using teasers outside of NFL key number situations dramatically increases the house edge.
Tracking Teaser Performance
Oddible lets you log teasers with all legs, track your key-number crossing rate, and see if your teaser strategy is actually producing results.
Track your teasers and measure your key number edge with Oddible →
