Do NFL Coaches Keep Covering? ATS Records Tested

An NFL head coach's record against the spread in one season predicts almost nothing about the next. Across 643 consecutive coach-seasons from 1999 through 2025, the correlation was -0.031. Coaches who covered 65 percent or more went 642-626 the following year, 50.63 percent. Winning percentage carried over at 0.296. Even 25-year career leaders look close to what chance alone would produce.
An NFL head coach's record against the spread in one season tells you almost nothing about the next one. We tested 643 pairs of back-to-back seasons by the same coach from 1999 through 2025, and the correlation between year one and year two ATS was -0.031, which is effectively zero. Coaches who covered 65 percent or more of their games went 642-626 against the spread the following season, 50.63 percent. Winning, by contrast, carries over: the year-to-year correlation for straight-up winning percentage was 0.296. The Best Bet on Sports has built a verified $367,520+ in profit across all sportsbooks over more than twenty years, and this study says a lot about how to judge anyone's betting record, including ours.
Coaching ATS records get quoted every week. "He's 12-4 against the spread as an underdog." "His teams always cover after a loss." It sounds like handicapping, and it's easy to find. The question is whether any of it holds up when the next season starts.
We tested it three ways: one season to the next, career record to the next season, and the career leaderboard itself.
How Did We Test Coach ATS Records?
We used the nflverse schedule file, which includes the closing spread, the final score and both head coaches for every NFL game since 1999. We kept regular-season games only, from 1999 through 2025, and graded each coach's side against the closing number. Pushes were dropped.
For the season-to-season test, we kept a coach-season only if he coached at least 10 graded games, and paired it with his next season only if it came the very next year. That gave us 643 pairs. We left out 2026, partly because the season isn't over and partly because the schedule file's 2026 coach columns still lag some offseason hires.
Does a Great ATS Season Carry Over?
No. Here's what happened when we split every coach-season into five equal groups by ATS record and looked at the following year:
| ATS group | Pairs | ATS that season | ATS next season | Win % that season | Win % next season |
|---|---|---|---|---|---|
| Bottom fifth | 130 | 35.54% | 51.06% | 40.16% | 48.60% |
| Second | 179 | 46.97% | 51.12% | 49.42% | 51.70% |
| Middle | 135 | 55.10% | 48.51% | 57.05% | 51.51% |
| Fourth | 104 | 60.81% | 49.45% | 63.67% | 52.43% |
| Top fifth | 95 | 69.92% | 50.49% | 68.82% | 58.24% |
Look at the two right-hand columns side by side. The best ATS coaches won 68.82 percent of their games that season and 58.24 percent the next, so a big chunk of their winning carried over. Their cover rate went from 69.92 percent to 50.49 percent, so none of the covering did.
That's the whole story in one table. Good teams stay good. The market just prices them as good, so their ATS edge vanishes.
The extremes tell the same story. In 81 seasons where a coach covered 65 percent or more, his teams averaged 70.86 percent ATS. The next year they went 642-626, 50.63 percent. In 51 seasons where a coach covered 35 percent or less, his teams went 329-319 the following year, 50.77 percent. The worst ATS coaches did slightly better the next season than the best ones.
What About a Whole Career Record?
One season is 16 or 17 games, which is a small sample. A fair objection is that a career record should be more reliable. So we tested that too.
For every coach with at least three prior seasons and 40 graded games, we took his career ATS record entering a season and compared it with what he did that season. That gave us 460 coach-seasons:
| Career ATS entering the season | Coach-seasons | Average career ATS | ATS that season | Games |
|---|---|---|---|---|
| Below 47% | 57 | 43.95% | 50.43% | 809 |
| 47% to 50% | 96 | 49.00% | 48.68% | 1,442 |
| 50% to 53% | 103 | 51.58% | 50.15% | 1,613 |
| 53% to 56% | 93 | 54.38% | 50.66% | 1,431 |
| 56% or better | 111 | 59.75% | 50.55% | 1,739 |
Coaches who came into a season with a career cover rate near 60 percent covered 50.55 percent that year over 1,739 games. Coaches with career records below 47 percent covered 50.43 percent. The correlation was 0.074. A bigger sample didn't rescue the trend.
We also ran one more check that's even harsher. Within a single season, we split each coach's games into odd and even weeks and compared the two halves. The correlation was 0.005 across 899 coach-seasons. A coach's ATS record in his odd-numbered games doesn't even predict his record in the even-numbered games of the same season. That's what noise looks like.
Why Doesn't Covering Carry Over When Winning Does?
Because the spread already includes the winning. A point spread is the market's estimate of the margin, built from everything known about both teams, including how good the coach is. If Bill Belichick's teams are better prepared, the line goes up to reflect it. Once the line is right, the coach's skill is in the number, and what's left over is mostly random.
There's a second thing. A season of ATS results is about as spread out as coin flips would be. The standard deviation of coach-season ATS records was 11.89 percentage points. For 16 coin flips, it's 12.5. About 18 percent of coach-seasons (with at least 14 graded games) finished at 62.5 percent or better, and a coin flipped 16 times lands 10 or more heads about 23 percent of the time. A 10-6 ATS season is not rare. It's ordinary variance wearing a headset.
We found the same thing at the team level when we tested what a 4-0 ATS start predicts, and when we looked at whether first-year NFL head coaches cover. Records against a sharp closing number regress hard.
Are There Any Coaches Who Really Beat the Spread?
Here's the career leaderboard for every coach with at least 150 graded regular-season games from 1999 through 2025:
| Coach | ATS record | Cover rate | Seasons in sample |
|---|---|---|---|
| Bill Belichick | 215-162 | 57.03% | 24 |
| Mike McCarthy | 156-125 | 55.52% | 18 |
| Sean Payton | 158-130 | 54.86% | 18 |
| Tony Dungy | 84-70 | 54.55% | 10 |
| Mike Tomlin | 163-140 | 53.80% | 19 |
| Marvin Lewis | 132-114 | 53.66% | 16 |
| Andy Reid | 226-201 | 52.93% | 27 |
| Mike Holmgren (lowest of 19) | 71-83 | 46.10% | 10 |
Belichick's 57 percent over 377 games is a genuinely unusual number. But 19 coaches made this list, and only two of them finished at 55 percent or better. We simulated 20,000 leagues of 19 coaches flipping coins for the same number of games each. On average, 1.26 of them reached 55 percent by luck alone. Two is about what chance produces.
Splitting the top careers in half shows the fade:
- Bill Belichick: 112-76 (59.57%) from 2000 through 2011, then 103-86 (54.50%) from 2012 through 2023. As an underdog, 34-16 (68.00%) in the first half and 21-22 (48.84%) in the second.
- Mike McCarthy: 92-64 (58.97%) through 2015, then 64-61 (51.20%) after.
- Andy Reid: 54.34% through 2012, then 51.44% after.
There's one result we can't explain away. Mike Tomlin as an underdog went 23-15 (60.53%) from 2007 through 2016 and 45-22 (67.16%) from 2017 through 2025. It held in both halves and actually got stronger after it became one of the most quoted trends in football. Maybe it's real. But when you test dozens of coaches in several situations, you should expect one or two trends to survive both halves by luck. One survivor out of a crowd isn't proof. It's a reason to keep watching.
What Does This Mean When You Pay for Picks?
This is the part that matters if you're looking for NFL picks against the spread or comparing football handicappers.
A handicapper's season record is exactly as noisy as a coach's season record. Sixteen weeks of ATS picks can run 62 percent or 38 percent with no skill involved at all. If a coach who covered 70 percent of his games one year went back to 50 percent the next, a tout who went 70 percent on 50 picks will probably do the same. A hot record is the easiest thing in this business to sell and the least useful thing to buy.
Here's what to ask any sports handicapping service instead:
1. How many graded bets is the record built on? Our guide to reading a sports betting track record explains why hundreds of bets is a minimum and thousands is better. 2. Are the bets posted before the game, at a price you could actually get? A 60 percent record at numbers that disappeared in a minute isn't a record you can follow. 3. Has a sportsbook ever limited the account? Books don't limit losing bettors. They limit people who take their money, and they do it based on years of data, not a hot month.
That third test is how The Best Bet on Sports is built. We've been limited on all six major U.S. sportsbooks (FanDuel, DraftKings, Caesars, BetMGM, Fanatics and ESPN BET) for winning too much at live betting. A sportsbook risk department is a much harsher judge than a highlight reel. You can check our results and compare us with other expert sports picks services before you decide.
Where This Analysis Stops
Three limits worth stating. First, coaching staffs change around the head coach. A coordinator leaving can change a team more than the head coach does, and our data only tracks the head coach. Second, we used closing lines. A coach's teams might have beaten the opening number more often if sharp money moved the line toward them during the week, and this study can't see that. Third, these are regular-season games only. Playoff ATS records are even smaller samples and even noisier.
None of that changes the main finding. Against a closing NFL line, past covering didn't predict future covering, at the season level or the career level.
Get Tonight's Live Picks
Want live in-game NFL calls, built for the moments when the live line is weaker than the closing number, sent to your phone during the game?
The Best Bet on Sports is the only live betting handicapping service limited on all six major U.S. sportsbooks (FanDuel, DraftKings, Caesars, BetMGM, Fanatics, ESPN BET) for winning too much during in-game action. Verified profit: $367,520+. Live calls go out through Email and Discord during games, and every bet in the $10,000 to $100,000 challenge is posted as it happens.
ā Join The Best Bet on Sports community ā newsletter, live stream and Discord ā Try a free live pick first ā reserve your spot for tonight's pick
Frequently Asked Questions
Do NFL coaches with good ATS records keep covering?
No. Across 643 back-to-back coach-seasons from 1999 through 2025, the correlation between one season's ATS record and the next was -0.031. Coaches who covered 65 percent or more went 642-626 the following season, 50.63 percent.
Which NFL coach has the best career record against the spread?
Among coaches with at least 150 graded regular-season games from 1999 through 2025, Bill Belichick has the best record at 215-162, 57.03 percent. Mike McCarthy is second at 55.52 percent and Sean Payton third at 54.86 percent.
Is Bill Belichick's ATS record proof that coaches can beat the spread?
It's the strongest case, but not proof. Belichick covered 59.57 percent from 2000 through 2011 and 54.50 percent from 2012 through 2023. With 19 coaches in the sample, chance alone would produce about 1.26 coaches at 55 percent or better, and two did.
Does a coach's winning record carry over to the next season?
Yes. The year-to-year correlation for straight-up winning percentage was 0.296. Coaches in the top fifth by ATS won 68.82 percent of their games that season and 58.24 percent the next. The market prices that winning into the spread, which is why covering doesn't carry over.
Is Mike Tomlin a good bet as an underdog?
His record is the one trend in this study that held up. Tomlin's teams went 23-15 ATS as underdogs from 2007 through 2016 and 45-22 from 2017 through 2025. Even so, when many coaches and situations are tested, one or two trends will survive by chance, so it's a reason to watch rather than proof.
How many picks do you need to judge a handicapper?
Far more than one season. A coach's 16-game ATS record has about the same spread as 16 coin flips, and a handicapper's record works the same way. Hundreds of graded bets at prices you could actually get is a minimum, and sportsbook limits are a stronger signal than any short record.
Why are NFL closing lines so hard to beat?
By kickoff, the closing line includes injuries, weather, betting from sharp bettors and everything known about both coaches. That's why past ATS records don't predict future ones. Live lines move much faster and with less information, which is where the edges are larger.
The Best Bet on Sports Ā· 1.5M followers on TikTok
Sports Betting TikTok is the live betting account behind The Best Bet on Sports, covering NFL, NCAAF, NBA, NCAAB, MLB and WNBA markets in real time. The account has been limited on all six major U.S. sportsbooks ā FanDuel, DraftKings, Caesars, BetMGM, Fanatics and ESPN BET ā for winning on live betting, with $367,520 in verified profit since 2005. View full profile ā
Past results do not guarantee future performance. Must be 21 or older to wager.
Related Articles
NFL Over/Under After a Shootout: 27 Years Tested
NFL Double-Digit Favorites Against the Spread: 27 Years Tested
NFL Christmas Day Against the Spread: Favorites Are 9-19-1
4-0 Against the Spread: What a Hot Start Predicts
Top NFL Handicappers Review and Rating: What Separates the Best from the Rest
NFL Total Sacks Props, Tested Against the Total
Get real live betting picks
From $39.95/mo Ā· Cancel anytime Ā· Email + Discord delivery
ā Get The Picks
Choose your package
Tap any tier below to select Ā· Subscribe in one tap
š Tap any package to select it
ā You're selecting
25 Units Ā· $99.95/month
25 unit picks per month. Sweet spot for most bettors ā plenty of action without going all-in.
- ā25 live betting selections/month
- āEmail + Discord delivery
- āAll sports (NFL, NCAAF, NBA, NCAAB, MLB)
- āMix of 1, 2, 3, 5-unit rated plays
- āCancel anytime
Secure PayPal Ā· Cancel anytime Ā· No hidden fees Ā· 21+ Ā· +$367K verified
+$367,520 verified Ā· Limited on 6 sportsbooks Ā· 20 years documented Ā· See Proof
Join Our Newsletter
Get free expert sports picks and analysis delivered weekly.