How to read these odds
Each table lists the home team's last digit as the row and the away team's last digit as the column, matching the orientation of the heatmaps above. A pair of 7-0 means the home team's score ended in 7 and the away team's ended in 0. The share is the percentage of games that landed on that pair at the end of the period; the counts are exact. The denominator is 4,484 games for every period, including the partial 2026 season and the postseason. A random square on a 100-square board has a 1.00 percent chance before the draw, so the best square's share divided by 1.00 is its edge over a random square. Scores are cumulative at the end of each period, and the final includes overtime.
End of Q1
The first quarter is the most concentrated period in the data. A 0-0 score is the single most common result: 712 of 4,484 games ended the first quarter scoreless on both sides, and 0-0 alone accounts for 15.88 percent of first-quarter outcomes. The 0, 3 and 7 digits dominate the rest of the table. The bottom ten pairs did not occur at all in the sample, which is a useful reminder of how many combinations a football score can avoid in twelve minutes.
Top 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 0 | 0 | 712 | 15.88% |
| 2 | 7 | 0 | 615 | 13.71% |
| 3 | 0 | 7 | 484 | 10.79% |
| 4 | 0 | 3 | 376 | 8.38% |
| 5 | 3 | 0 | 368 | 8.21% |
| 6 | 7 | 7 | 335 | 7.47% |
| 7 | 7 | 3 | 268 | 5.98% |
| 8 | 3 | 7 | 229 | 5.11% |
| 9 | 3 | 3 | 172 | 3.84% |
| 10 | 4 | 0 | 169 | 3.77% |
Bottom 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 9 | 8 | 0 | 0.00% |
| 2 | 9 | 5 | 0 | 0.00% |
| 3 | 9 | 4 | 0 | 0.00% |
| 4 | 9 | 2 | 0 | 0.00% |
| 5 | 9 | 1 | 0 | 0.00% |
| 6 | 8 | 9 | 0 | 0.00% |
| 7 | 8 | 8 | 0 | 0.00% |
| 8 | 8 | 6 | 0 | 0.00% |
| 9 | 8 | 5 | 0 | 0.00% |
| 10 | 8 | 4 | 0 | 0.00% |
Halftime (end of Q2)
By halftime the 0-0 share has fallen to 6.56 percent, but zero and seven still lead the table. The best square, home 0 and away 0, is 6.6 times the 1 percent baseline. The spread between the top pair and the tenth pair is much smaller than in the first quarter, and the bottom of the table is a mix of pairs that never occurred and pairs that appeared once or twice.
Top 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 0 | 0 | 294 | 6.56% |
| 2 | 7 | 0 | 290 | 6.47% |
| 3 | 0 | 7 | 236 | 5.26% |
| 4 | 0 | 3 | 233 | 5.20% |
| 5 | 3 | 0 | 215 | 4.79% |
| 6 | 7 | 7 | 213 | 4.75% |
| 7 | 7 | 3 | 194 | 4.33% |
| 8 | 4 | 0 | 172 | 3.84% |
| 9 | 3 | 7 | 158 | 3.52% |
| 10 | 3 | 3 | 149 | 3.32% |
Bottom 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 9 | 8 | 0 | 0.00% |
| 2 | 8 | 9 | 0 | 0.00% |
| 3 | 8 | 8 | 0 | 0.00% |
| 4 | 8 | 2 | 0 | 0.00% |
| 5 | 5 | 8 | 0 | 0.00% |
| 6 | 5 | 5 | 0 | 0.00% |
| 7 | 5 | 2 | 0 | 0.00% |
| 8 | 2 | 8 | 0 | 0.00% |
| 9 | 2 | 5 | 0 | 0.00% |
| 10 | 9 | 5 | 1 | 0.02% |
End of Q3
The third quarter is where the table starts to flatten. The best square is home 0 and away 7 at 4.33 percent, only 4.3 times the baseline, and the difference between the first and tenth rows is about 1.5 percentage points. The 0, 3, 4 and 7 digits are still prominent, but the rare pairs are no longer all zero: 2-2 did not occur, while 9-8 and 5-2 appeared once each.
Top 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 0 | 7 | 194 | 4.33% |
| 2 | 7 | 0 | 193 | 4.30% |
| 3 | 0 | 0 | 190 | 4.24% |
| 4 | 3 | 0 | 170 | 3.79% |
| 5 | 7 | 7 | 161 | 3.59% |
| 6 | 0 | 3 | 159 | 3.55% |
| 7 | 4 | 7 | 159 | 3.55% |
| 8 | 3 | 7 | 142 | 3.17% |
| 9 | 4 | 0 | 133 | 2.97% |
| 10 | 3 | 3 | 127 | 2.83% |
Bottom 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 2 | 2 | 0 | 0.00% |
| 2 | 9 | 8 | 1 | 0.02% |
| 3 | 5 | 2 | 1 | 0.02% |
| 4 | 9 | 5 | 3 | 0.07% |
| 5 | 8 | 2 | 3 | 0.07% |
| 6 | 5 | 5 | 3 | 0.07% |
| 7 | 2 | 5 | 3 | 0.07% |
| 8 | 5 | 9 | 4 | 0.09% |
| 9 | 2 | 8 | 4 | 0.09% |
| 10 | 8 | 9 | 5 | 0.11% |
Final (including overtime)
The final table is the flattest of the four. The best pair, home 7 and away 0, wins 3.52 percent of games, 3.5 times the 1 percent baseline, and the tenth row is still above 2 percent. The bottom of the table is more informative here: 2-2 appeared three times in 4,484 games, 5-5 six times and 9-8 nine times. The final prize is the most widely distributed, which is why a 20/20/20/40 split feels different from a first-quarter prize.
Top 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 7 | 0 | 158 | 3.52% |
| 2 | 0 | 7 | 150 | 3.35% |
| 3 | 0 | 3 | 141 | 3.15% |
| 4 | 7 | 4 | 138 | 3.08% |
| 5 | 3 | 0 | 137 | 3.06% |
| 6 | 4 | 7 | 135 | 3.01% |
| 7 | 3 | 7 | 114 | 2.54% |
| 8 | 4 | 0 | 105 | 2.34% |
| 9 | 0 | 4 | 103 | 2.30% |
| 10 | 6 | 0 | 93 | 2.07% |
Bottom 10 pairs
| Rank | Home digit | Away digit | Games | Share |
|---|---|---|---|---|
| 1 | 2 | 2 | 3 | 0.07% |
| 2 | 5 | 5 | 6 | 0.13% |
| 3 | 9 | 8 | 9 | 0.20% |
| 4 | 5 | 2 | 10 | 0.22% |
| 5 | 8 | 8 | 11 | 0.24% |
| 6 | 9 | 5 | 12 | 0.27% |
| 7 | 5 | 9 | 12 | 0.27% |
| 8 | 2 | 5 | 14 | 0.31% |
| 9 | 8 | 9 | 15 | 0.34% |
| 10 | 8 | 2 | 16 | 0.36% |
Best square vs a random square
The summary table puts the four periods side by side. The multiple is the best pair's share divided by the 1.00 percent baseline for a random square on a 100-square board.
| Period | Best pair (home-away) | Share | Games | Multiple of 1% |
|---|---|---|---|---|
| End of Q1 | 0-0 | 15.88% | 712 | 15.9x |
| Halftime | 0-0 | 6.56% | 294 | 6.6x |
| End of Q3 | 0-7 | 4.33% | 194 | 4.3x |
| Final (incl. OT) | 7-0 | 3.52% | 158 | 3.5x |
The edge shrinks as the game goes on. Early scores are concentrated on 0, 3 and 7; by the final, more combinations have had time to appear. That is why a first-quarter prize often lands on a 0-0 or 7-0 square, while the final prize is more spread out. The same pattern shows up in the Super Bowl history on the results page, though the Super Bowl sample is only 60 games.
How to use this in a pool
You cannot choose a square after the draw in a fair pool, so these tables are context, not a strategy for a blind draw. They are useful when you assign squares after the digits are visible, such as an auction, a draft or a charity board where players bid on known pairs. In those formats, price the 0-0 and 7-0 squares higher for early periods and the 7-0, 0-7 and 0-3 pairs higher for the final. They also help when someone asks why the 0-0 square keeps winning early. For a 25-square board, each square covers four pairs, which smooths the extremes. For a reverse-winner side prize, the same frequencies apply with the digits swapped.
Data and method
Source: the site's NFL digit dataset, built from nflverse play-by-play (regular season and postseason, 2010 through 2026). The sample is 4,484 games; the 2026 season is partial at the time the file was built. Quarter scores are the cumulative home and away scores at the last play of each quarter, and the final score includes overtime. A pair is the home team's last digit followed by the away team's last digit. Percentages are rounded to two decimals; counts are exact. The same dataset powers the best-numbers page and the heatmaps above. Not affiliated with or endorsed by the NFL.