Chicago (7-2) at Detroit (7-2), Philadelphia (2-7) at Dallas (4-6) and Kansas City (7-2) at Buffalo (8-1) play on Thursday, November 26, with Green Bay at the Rams the night before and Denver at Pittsburgh on Black Friday. Records through week 10, where each team sits on our power ratings, and what 73 Thanksgiving games since 1999 show: hosts went 33-40, and the holiday's games have been less close than the rest of their week.
By C. B. Zakarian · Published October 18, 2026
Three games on Thursday, November 26, 2026: Chicago at Detroit at 1:00 p.m. Eastern, Philadelphia at Dallas at 4:30 p.m. and Kansas City at Buffalo at 8:20 p.m. The week also has Green Bay at the Rams on Wednesday night, November 25, and Denver at Pittsburgh on Friday afternoon, November 27. History is not kind to the hosts: from 1999 to 2025, home teams went 33-40 on Thanksgiving, and the reason is mostly who the hosts were. In 40 of the 73 games the visitor had the better record.
Related: 2026 standings · the 2026 schedule's quirks · NFL key dates 2026-27
Schedule from the nflverse schedule file, re-read November 16, 2026 and unchanged since October 18, when it was checked against the ESPN scoreboard. Records through Sunday of week 10. History from the nflverse game log, 1999-2025 regular seasons.
All five games belong to week 12. Kickoff times are Eastern. We re-read the schedule on November 16 and nothing has moved since October 18: same teams, kickoffs and stadiums.
| Day | Kickoff (ET) | Game | Stadium | Records | Days of rest |
|---|---|---|---|---|---|
| Wed., Nov. 25 | 8:00 p.m. | Green Bay at LA Rams | SoFi Stadium | 6-4 / 6-4 | 10 / 10 |
| Thu., Nov. 26 | 1:00 p.m. | Chicago at Detroit | Ford Field | 7-2 / 7-2 | 4 / 4 |
| Thu., Nov. 26 | 4:30 p.m. | Philadelphia at Dallas | AT&T Stadium | 2-7 / 4-6 | 4 / 4 |
| Thu., Nov. 26 | 8:20 p.m. | Kansas City at Buffalo | Highmark Stadium | 7-2 / 8-1 | 4 / 4 |
| Fri., Nov. 27 | 3:00 p.m. | Denver at Pittsburgh | Acrisure Stadium | 7-2 / 6-3 | 5 / 5 |
Records are visitor first, through Sunday of week 10: 149 games, the last of them on Sunday, November 15. Monday night's game, the Chargers at Baltimore, is not in them. Week 11 is still to come before the holiday.
Every team playing on Thanksgiving Day comes in on four days' rest, since all six play the Sunday before. The Wednesday game is the opposite: Green Bay and the Rams both come in on 10 days. The schedule-quirks page explains how the league built that week.
For each team: its record and point differential, its last three results, where it sits on our NFL power rankings, its quarterback on our QB power rankings, and its chance to make the playoffs and to win its division in the site's own season simulation. The two rating pages, as published today, count games through week 9 (136 games) and add week 10 once all of its games are in; the records and the chances already include Sunday's games.
| Team | Record | Pt. diff. | Last three (oldest first) | Power rank | Quarterback (QB power rank) | Playoff chance | Division chance |
|---|---|---|---|---|---|---|---|
| Green Bay | 6-4 | +5 | W 27-19 vs CAR, W 45-27 at NE, L 10-30 vs MIN | 10 (+3.1) | Jordan Love (5) | 26% | 1% |
| LA Rams | 6-4 | +61 | W 24-9 vs LAC, W 20-13 at WAS, W 34-6 at ARI | 5 (+4.5) | Matthew Stafford (13) | 43% | 4% |
| Chicago | 7-2 | +57 | W 30-24 vs NE, W 34-17 at SEA, W 27-24 vs TB | 7 (+3.4) | Caleb Williams (11) | 84% | 21% |
| Detroit | 7-2 | +58 | W 19-15 vs MIN, W 19-9 at MIA, W 35-10 vs NE | 9 (+3.3) | Jared Goff (8) | 92% | 35% |
| Philadelphia | 2-7 | −73 | L 21-38 vs DAL, L 10-27 at WAS, L 11-19 vs NYG | 23 (−3.3) | Joe Milton III (44 of all 61) | 1% | 1% |
| Dallas | 4-6 | +4 | L 30-33 vs ARI, W 27-23 at IND, L 40-46 vs SF | 17 (−0.8) | Dak Prescott (6) | 17% | 16% |
| Kansas City | 7-2 | +81 | L 24-28 at DEN, W 35-12 vs NYJ, W 35-13 at ATL | 8 (+3.3) | Patrick Mahomes (9) | 96% | 24% |
| Buffalo | 8-1 | +88 | W 36-10 vs BAL, W 23-6 at MIN, W 31-12 at NYJ | 2 (+6.7) | Josh Allen (2) | >99% | >99% |
| Denver | 7-2 | +12 | W 30-16 at ARI, W 28-24 vs KC, W 26-21 at CAR | 11 (+2.9) | Bo Nix (16) | >99% | 76% |
| Pittsburgh | 6-3 | +16 | W 16-10 at NO, W 27-6 vs CLE, L 18-38 at CIN | 14 (+2.2) | Aaron Rodgers (27) | 85% | 14% |
Power rank: the NFLAnalytic Power Rating, in points per game better than an average team on a neutral field (rank of 32). QB power rank: rank among that page's 34 qualifiers, the quarterbacks with a regular workload (61 have dropped back this season); the quarterback listed is the one with the most dropbacks in his team's latest game on that page. Chances: 20,000 simulated seasons from the 149 games played, the same run as the playoff chances page. Chicago, Philadelphia and Denver had week 10 off.
Chicago at Detroit is the game of the afternoon by record: both are 7-2, and so is Minnesota, so three NFC North teams share the same record through week 10, with Green Bay a game and a half back at 6-4. It is the teams' first meeting of 2026. Chicago has won four in a row and Detroit five, the latest 35-10 over New England in Munich on Sunday, and their point differentials are a point apart (+57 and +58). Our power rating has them close too, Chicago No. 7 and Detroit No. 9, with Jared Goff No. 8 and Caleb Williams No. 11 on the QB power page. The simulation gives Detroit a 92% chance to make the playoffs and Chicago 84%, and the division is wide open: Detroit wins the North in 35% of runs, Chicago in 21%, and Minnesota in 43%, more often than either.
Philadelphia at Dallas matches two teams below .500. Philadelphia is 2-7, has lost seven in a row and has a 1% chance; Dallas is 4-6 with a 17% chance, nearly all of it through the division (16%), where it trails the 5-3-1 Giants. Dallas already won the first meeting, 38-21 in Philadelphia on October 26, and lost 46-40 to San Francisco on Sunday. Dak Prescott is No. 6 on the QB power page. Philadelphia's quarterback has changed twice: Jalen Hurts led the team in dropbacks in weeks 1-5 (No. 15 on the QB page), Tanner McKee in weeks 6-8 (No. 33) and Joe Milton III in week 9, his only game, which leaves him in the table of quarterbacks below the workload bar at No. 44 of 61.
Kansas City at Buffalo has the best combined record of the three holiday games, 15-3. Buffalo is 8-1, has won five straight, is No. 2 on our power rating with Josh Allen No. 2 on the QB page, and wins the AFC East in more than 99% of simulated seasons, three and a half games clear of New England. Kansas City is 7-2 and tied with Denver atop the AFC West after the two split their meetings, 31-10 to Kansas City in September and 28-24 to Denver on November 1. Kansas City is No. 8 on the power rating, Patrick Mahomes No. 9, with a 96% playoff chance and a 24% division chance.
Green Bay at the Rams on Wednesday night brings two 6-4 teams going opposite ways: the Rams have won four straight, the last 34-6 at Arizona, and Green Bay lost 30-10 at home to Minnesota on Sunday. The Rams' point differential is +61 to Green Bay's +5, and the power rating has them No. 5 and No. 10. Both are chasing wild cards more than division titles (Rams 43% to make the playoffs, Packers 26%); San Francisco, 8-1, leads the NFC West.
Denver at Pittsburgh is the Black Friday game, below. Denver is 7-2 with five straight wins and a 76% chance to win the AFC West; Pittsburgh is 6-3, second in the AFC North behind Cincinnati after losing 38-18 there on Sunday.
Records, ratings and chances do not always agree, and the reasons are on the pages they come from: the power rating counts margins, efficiency and opponents rather than wins (Denver is 7-2 with a +12 point differential and sits No. 11), while the chances come from the Elo ratings that drive our game predictions and play out the rest of the schedule. The playoff picker lets you set any of these results yourself.
From 1999 to 2025 the NFL played 73 regular-season games on Thanksgiving: two a year through 2005, three a year since 2006, and two in 2020, when the third game, Baltimore at Pittsburgh, was played the following Wednesday, December 2. Detroit and Dallas hosted every year. The home teams won 33 and lost 40, a .452 share. In the other 6,828 regular-season games played at a home team's stadium over the same seasons, hosts won .564.
Most of that gap comes from the hosts themselves. Detroit went 8-19 at home on Thanksgiving, Dallas 15-12, and the third game's hosts 10-9. Detroit had the worse record of the two teams, counting each side's other games that season, in 19 of its 27 Thanksgiving games. Across all 73, the visitor had the better record 40 times and the host 30 times; in 3 the two were level.
Split that way, the holiday mostly disappears. When the host had the better record, it won 21 of 30 on Thanksgiving (.700), close to the .713 such hosts won in other games. When the visitor had the better record, the host won 10 of 40 (.250), against .412 in other games. That second gap is about two standard errors on 40 games, which is enough to notice and not enough to call a Thanksgiving effect. A host that is the weaker team has done worse than usual on the holiday, but 40 games cannot separate that from chance.
Two other things stand out in the 73 games. They score a little more: 46.0 points per game, against 44.2 across all regular-season games from 1999 to 2025. And 35 of them were division games, which this year's afternoon pair continues.
The split above sorts each game by one fact, which team had the better record. A finer check gives every Thanksgiving game an expected result from both teams' records in their other games that season, with the usual home edge (the same log5 calculation as the rest section below), and adds them up by host.
| Host | Games | Home wins | Expected from the records | Gap |
|---|---|---|---|---|
| Detroit | 27 | 8 | 10.5 | −2.5 |
| Dallas | 27 | 15 | 16.1 | −1.1 |
| Third game (since 2006) | 19 | 10 | 10.1 | −0.1 |
| All Thanksgiving hosts | 73 | 33 | 36.8 | −3.8 |
Two thirds of the shortfall, 2.5 of the 3.8 wins, is Detroit's. The third game's hosts have won almost exactly what their records said they would, and Dallas has come close. The calculation does not flatter the hosts in general: run on the other 6,828 home games of 1999-2025, it expected 3,688.7 home wins and the hosts won 3,849, so it slightly underrates home teams, and Detroit's gap would be a little larger against a better-calibrated version. Even so, 2.5 wins on 27 games is about 1.2 standard deviations of what chance alone produces, so it is a lean, not a curse.
Comparing Thanksgiving with the whole season mixes in September and January. A cleaner comparison is the 354 other games played in the same week of the same seasons, 1999-2025, which shares the time of year, the weather and the injury toll.
| Games | Games | Points per game | Home win share | Decided by 8 or fewer | Decided by 17 or more |
|---|---|---|---|---|---|
| Thanksgiving, all | 73 | 46.0 | .452 | 31 (42%) | 29 (40%) |
| Same week, other games | 354 | 43.9 | .555 | 200 (56%) | 91 (26%) |
| Detroit's game | 27 | 47.8 | .296 | 14 (52%) | 10 (37%) |
| Dallas's game | 27 | 47.5 | .556 | 13 (48%) | 11 (41%) |
| Third game (since 2006) | 19 | 41.2 | .526 | 4 (21%) | 8 (42%) |
Against its own week, Thanksgiving still scores a little more, 2.1 points per game, but the bigger difference is in the margins. Only 42% of Thanksgiving games were decided by eight points or fewer, against 56% of the other games that week, and 40% were decided by 17 or more, against 26%. On 73 games those gaps are about 2.4 and 2.7 standard errors, larger than the home-record gap (1.8), so the holiday has produced more lopsided games than its week, though 73 games is still one small sample. The night game, played since 2006, is the extreme: 4 of its 19 games were within one score. The home win share on the holiday is .452 against .555 for the rest of the week, which is the host gap the sections above trace mostly to who the hosts were.
A Thanksgiving team gives up three days of rest before the game and usually gets them back after it: most play their next game ten or more days later, against an opponent coming off a normal week. Of the 146 next games after a Thanksgiving game from 1999 to 2025, 108 came with a rest edge of three days or more.
The edge has not shown up in the results. The Thanksgiving teams went 69-76-1 in their next game. From each team's record in its season's other games, with the usual home edge, they would have been expected to win 70.3; they won 69.5. In the 108 games with a rest edge of three or more days they won 51.5 against 52.2 expected. The extra rest after the holiday has bought nothing this record can see. Across the whole schedule, the rest and scheduling page found that a big rest gap is worth about a field goal, so 108 games are a small sample against that, not a refutation of it.
For 2026, five of the six Thanksgiving Day teams get that edge next: Chicago at home to Jacksonville, Detroit at Atlanta, Philadelphia at Arizona and Buffalo at New England on Sunday, December 6, each with 10 days against 7, and Dallas at Seattle on Monday night, December 7, with 11 against 8. Kansas City does not. It plays again on Thursday, December 3, at the Rams, on 7 days against 8.
For each of the 146 Thanksgiving team-games of 1999-2025, we took the team's record going into the holiday, the result, and whether it reached that season's playoffs (12 teams a year through 2019, 14 since 2020).
| Record going in | Won on Thanksgiving: made playoffs | Lost on Thanksgiving: made playoffs |
|---|---|---|
| Above .500 | 39 of 49 (80%) | 11 of 26 (42%) |
| .500 or below | 8 of 24 (33%) | 1 of 47 (2%) |
| All | 47 of 73 (64%) | 12 of 73 (16%) |
The holiday result has lined up closely with the postseason. Part of that is the result telling us who was good, since better teams win more of these games, so this is not a measure of what one win is worth. It is still a plain record of what has happened: of the 47 teams that came in at .500 or worse and lost, one made the playoffs, Green Bay in 2013, which was 5-5-1 and lost 40-10 in Detroit. Of this year's six Thursday teams, four are above .500 through week 10; Dallas (4-6) and Philadelphia (2-7) are in the .500-or-below row, where 46 of 47 Thanksgiving losers missed the playoffs.
None of this history decides a 2026 game. The records and ratings above say much more about these matchups than the holiday does, and one more week of games will move them before Thursday.
The league has played one game on the day after Thanksgiving in each of the last three seasons, every one at 3:00 p.m. Eastern with both teams on five days' rest. Before 2023 the nflverse game log, which starts in 1999, has none.
| Season | Result | Records going in | Points | Rest of that week | Made the playoffs |
|---|---|---|---|---|---|
| 2023 | Miami 34, NY Jets 13 (at New York) | 7-3 / 4-6 | 47 | 38.7 | Miami |
| 2024 | Kansas City 19, Las Vegas 17 (at Kansas City) | 10-1 / 2-9 | 36 | 50.2 | Kansas City |
| 2025 | Chicago 24, Philadelphia 15 (at Philadelphia) | 8-3 / 8-3 | 39 | 40.7 | both |
Records are winner first, going into the game; "rest of that week" is the points per game in the week's other games, excluding Thanksgiving (12 games each year). The team with the better record won both games where the records differed, and in 2025 two 8-3 teams met and the visitor won. The three games averaged 40.7 points, below the 43.2 of the other games in those weeks, but three games show no pattern of their own. Every Black Friday winner made the playoffs.
For 2026, Denver at Pittsburgh is the fourth Black Friday game and, like 2025's, not a division game. Both teams come off a Sunday game, so both get five days. Denver is 7-2, No. 11 on our power rating, with Bo Nix No. 16 on the QB power page and a chance to make the playoffs above 99%; Pittsburgh is 6-3, No. 14, with Aaron Rodgers No. 27 and an 85% chance.
Thanksgiving = the fourth Thursday of November; every regular-season game on that date, 1999-2025 (73 games)
Black Friday = the day after; every regular-season game on that date, 1999-2025 (3 games, 2023-2025)
better record = winning percentage in the team's other regular-season games that season (the game itself excluded,
ties count half)
other games = every other regular-season game at the home team's stadium, 1999-2025 (6,828; neutral sites left out)
same week = the other regular-season games of the same season and week as a Thanksgiving game (354)
one score = decided by 8 points or fewer
going in = the team's record in its games before the holiday that season
next game = each Thanksgiving team's following regular-season game (146)
expected wins = log5 from the two teams' records in their other games, plus a home edge of .56 when one team is at home
2026 records = 149 games through Sunday, November 15 (week 10 without Monday night)
chances = 20,000 simulated seasons, the site's playoff simulation through the same 149 games
Our check script re-reads the 2026 schedule, compares it with October's two sources, reads the records, the two rating pages and the simulation, finds every Thanksgiving and Black Friday game, computes each table above and asserts every number on this page.
Sources: the nflverse schedule and game log, re-read 8:05 am ET on November 16, 2026 (its 1999-2025 games are identical to the copy the October version used), and the ESPN public scoreboard for November 25-27, 2026, read October 18. Ratings from our power rankings and QB power rankings as of November 16, chances from our playoff chances page. Records are through Pittsburgh at Cincinnati on Sunday night, November 15.
Kept in this browser only.
Pick the games left and see both conferences’ seeds under the NFL’s tie-breakers.
Want the code behind these metrics? Work through the 43-chapter NFL analytics tutorial.
Browse tutorials Free tools