To study all NFL mid-game lightning delays in the 21st century, determine how influential they are in determining a game's outcome, and explain the correlation.
As a Philadelphia Eagles fan, I was excited to watch the Eagles face the Cowboys on September 4, 2025. During the 3rd quarter, there was a lightning delay that lasted for a few hours, so I turned off the game. I checked back on the score hours later and saw that the Eagles won 24-20. That, on its own, had little significance towards this project. As a native Floridian, I was used to high school football games getting canceled and not watching UCF Football due to lightning delays, so the one instance in the Cowboys-Eagles game held little significance. That Sunday, on September 7, two more games experienced mid-game lightning delays: the Panthers-Jaguars game and the Titans-Broncos game. It was then I noticed that all three home teams won the game and that all three winning teams had leads at the time of delay. While national media at the time was debating about whether or not changes to the NFL's Lightning Delay rules were necessary, I dug deeper, trying to see if all three home teams leading at the time of delay and winning the game was a coincidence or indicative of a greater trend.
I found 21 instances across 19 games of a mid-game delay due to lightning from the 2000 season up until Week 5 of the 2025 season. I went through the Game Book
I did not consider any pre-game delays or mid-game delays due to power outages. While mid-game lightning delays give teams extra time to evaluate the ongoing game and make adjustments, like a second halftime, pre-game lightning delays don’t give teams that same opportunity since they occur before the game is played. Teams gain a few extra hours at most to plan after having the whole week to plan. In the grand scheme, those few hours make next-to-no effect. For power outages, fans on the field, or other distractions, while they do disrupt play, they don’t push teams back into the locker room, providing no chance for a team to make major adjustments in the same way. Mid-game lightning delays are more interesting because of the strategic element they bring. They give teams a chance to catch their breath and coaches a chance to change the game plan in a way that can affect the game differently compared to just during halftime.
I tracked 13 different attributes for every instance: the season, the week in the season, home and away teams for each game, the game clock at the time of delay, the reason for delay, the score at the time of delay, the final score, whether the home team won, whether the team in the lead won, the spread (taken from Pro Football Reference’s “Vegas Line”), whether the favorite covered the spread, and whether the favorite won.
1. The main takeaway from the table is that teams in the lead at the time of delay won 17 out of 18 times (the three ties at the time of delay were not included in that calculation). While it may seem obvious that a team in the lead will win the game, the fascinating thing was that teams in the lead at the time of delay won 94.44% of the time no matter when during the game the delay occurred.
2. Having home-field advantage did not have any meaningful effect on whether the home team won games featuring mid-game lightning delays. Greg Auman of Fox Sports reported that in the 2024 regular season, home teams won 53.3% of the time. While the percentage has fluctuated little-by-little over the years, according to Bradley Locker of Pro Football Focus, the highest home winning percentage since 2000 by half-decade was 2000-2004, where home teams won 57.3% of the time. My data matches these expectations. Across all 21 instances, the home teams went 12-9, or won 57.14% of the time. Even when measuring only the 19 games played, home teams went 11-8, or won 57.89% of the time. The data shows that being a home team in these situations does not give that team an advantage.
3. Being a favorite does not correlate with winning in games featuring mid-game lightning delays. I found that favorites in these games went 11-10, or won 52.38% of the time. According to Jeff Hochman of CBS Sports, from 2018-2024, NFL money line favorites maintained a 66.6% win rate looking at BetMGM data. I have an explanation for why the win rate for favorites in games with mid-game lightning delays is less than what we would normally see. I believe that the underdog feels emboldened by their play by the time of delay. Out of the 10 underdogs who won these types of games, every one of them was either winning or tied at the time of delay. The underdog may experience a morale boost for hanging with the favorite. From a game-strategy perspective, playing ahead makes it easier for the team to adjust their playstyle to run out the clock and maintain their lead, not needing to take as many risks as if they were behind.
4. The favorite tends to not cover the spread. I found that favorites in these games were an abysmal 5-16 against the spread, or covered the spread 23.81% of the time. According to David Purdum on ESPN, during the 2024 season, favorites covered the spread in 53.9% of games, best since 2017 and good for seventh-best in the Super Bowl era. It is harder to explain this discrepancy. While ATS records are typically worse than straight-up records when looking at favorites, NFL season average ATS odds for favorites are usually around even odds. Further research is needed to analyze that discrepancy.
Note: The 2000 Cardinals-Giants game took place at old Giants Stadium, replaced by MetLife Stadium, which hasn’t experienced a mid-game lightning delay.
Spatial Analysis:
The biggest takeaway from these maps lies in Map 3. In that map, the more red a team is, the worse record they have in games with mid-game lightning delays. The more blue a team is, the better. From Map 3, we can see that most teams that perform well in these scenarios are teams in areas with a low National Risk Index for Lightning. Excluding the Jaguars, Dolphins, and Bears, teams in the higher risk areas that experienced more than one instance of a mid-game lightning delay have a combined record of 2-14. Meanwhile, teams in lower risk areas that have experienced more than one instance of a mid-game lighting delay are a combined 9-2.
The Bucs and Dolphins experienced the most mid-game lightning delays at home, at five and three respectively. The Bucs even suffered mid-game lightning delays in back-to-back weeks in Weeks 3 and 4 of the 2016 season.
Despite their experience with mid-game lightning delays, the Bucs went 1-4 in Tampa Bay and 2-4 overall, including two home losses to the Rams.
In fact, out of teams that played more than one game with a mid-game lightning delay, there were only three undefeated teams: the Rams, the Bills, and the Jaguars, all 2-0. The Rams and the Bills were the away team in both of their victories.
On the opposite side, there were three winless teams out of all teams that played more than one game with mid-game lightning delays: the Cowboys (0-2), the Titans (0-2), and the Panthers (0-3).
Explaining the Correlation:
My theory of why teams leading at the time of delay are 17-1 centers around more time to rest and strategic planning by coaches of the leading team.
Since players have an extra "halftime" to rest and coaches to make offensive and defensive adjustments meant to maintain their lead. If the players exit the delay in the same energy-state as the beginning of the game, but are leading by four or seven points and with little time left in the game, the losing team at the time of delay has a harder road to climb.
The coaching side of the benefits are more easily seen when the delay occurs late-game. When a team is up ten points at the time of delay late-game, their best running backs and offensive linemen are more easily prepared to run the ball to eat as much game-clock as possible. Meanwhile, the losing team would be more inclined to pass more and thus, more one-dimensional. Stacking man coverage or blitzing every time would be what a leading defensive coordinator would draw up, and those schemes lead to more interceptions, fumbles (by way of the quarterback), and incompletions.
Under these circumstances, it is possible for a losing team to complete a comeback, but having all players of the leading team rested for an extended amount of time before play resumes from the delay and having only one real means of making the comeback leads to an uphill climb for the losing team, regardless whether they are the favorite or home team.
Regarding league's transition to closed stadiums: Weather plays an equalizer role in football that isn't seen in many other sports. The weather can cause players to fumble, miss catches, or slip more often, which all result in a more exciting game to watch. Even the hidden benefit teams receive when leading in the event of a mid-game lightning delay can result in a difference in playcalling that can spice up a game. More than half of games with mid-game lightning delays (12 out of 19) were one-score games, with 9 of those games being decided by six points or fewer. That’s a recipe for entertaining games. With more stadiums moving indoors, we risk losing part of football's identity. Many iconic football games were made so by the weather conditions, for example such as the “Pickle Juice Game” or the multiple “Snow Bowls” in NFL history, especially with the 2024 NFL Divisional Round matchup between the Eagles and Rams in recent memory. Closed stadiums and stadiums with retractable roofs, while they improve the overall fan experience, comes at the expense of the whimsy and chaos that weather conditions can provide an NFL game. Certain teams harness the weather to their advantage, such as Lambeau Field in Green Bay being notoriously cold (and the Packers performing well historically in low temperature conditions) and at Hard Rock Stadium in Miami where the Dolphins players are in the shaded part of the stadium while the away team is constantly exposed to the harsh Florida sun. If the NFL was committed to raising entertainment value of the modern NFL, they should encourage teams playing in open-air stadiums.
Regarding sports betting: It is undeniable that the above statistic of teams leading at the time of a mid-game delay due to lightning are 17-1 can be used to place sports bets on the outright winner of a game. However, I must state that this is a correlation, not proof of causation. The found correlation and collected statistics were means to try to explain the past, not predict the future. Without performing statistical regression, I cannot be confident in proving that this relationship exists enough to make predictions based on past data, especially when money is on the line. For that reason, I do not encourage using these numbers to bet on and I do not take responsibility in any gambling decisions made using these numbers.