How many different ways are there to score a touchdown in the NFL? I found myself asking this question one day. I think the answer is a staggering twenty, or even twenty-three, on a fairly intuitive level of granularity.
I've organized them below, grouping them intofour coarser-grained categories. I've also added comments and links where appropriate.
I was not guided by statistical conventions, like those found in the NFL's guide for statisticians. This is most apparent in my distinguishing various scenarios that, in stats, are all recorded as touchdowns from fumbles.
Instead, I considered the rules of the NFL themselves, determining which scenarios they treat as distinct.
If I have erred anywhere, I have erred on the side of differentiation, on the side of a finer-grained taxonomy.
Following the list of NFL touchdown types, I address some "types" of touchdowns that might be includable, but which I have not included as separate types.
Official boxscores distinguish between a fumble return and a "fumble recovery in end zone". I would consider the latter to be a 0 yard fumble return touchdown.
Just like in the case of fumble returns, official boxscores distinguish between an interception return and an "interception in end zone". I would consider the latter to be a 0 yard interception return touchdown. This has occurred, by the way; here is an example from 2013, in a game between San Diego and Washington.
After the ball goes at least 10 yards on any kickoff, the kicking team can recover the ball.
This is what the kicking team aims to do on an onside kick, which is a kind of kickoff play, not a separate scenario demarcated by the rules.
However, under current NFL rules, the kicking team cannot advance the ball; they can only take possession of it where it is recovered, and, if the kick makes it to the receiving team's endzone untouched by the receiving team, the result is a touchback.
It is still possible for the kicking team to score a touchdown by recovering their kickoff, but the updated rules now require that, first, the receiving team touch the ball before it enters their endzone.
Remarkably, a touchdown of this kind did happen prior to rule change concerning touchbacks; here is the example from 2016. On this kickoff, the player for the Bills appears to anticipate the new rule, and backs away from the ball once it rolls into his team's endzone. A time traveller, perhaps?
At one time, it was even possible for the kicking team to recover the ball and advance it. The only other instance of this kind of touchdown recorded in the annals of the NFL belongs to this period.
In 1921, the second year of the NFL's existence (when it was still known as the APFA, the American Professional Football Association), future Hall-of-Famer Jimmy Conzelman of the Rock Island Independents scored on a kickoff recovery, according to Pro Football Research; see the boxscore for their game against the Chicago Cardinals (today's Arizona Cardinals).
Indeed, further research reveals that this was recorded as an "onside kick" recovery; see the scoring information reproduced directly from contemporary newspaper accounts here; the information from the relevant game appears on page 2, and the source is listed as the Chicago Tribune.
In fact, you can read the original article from the October 17, 1921 edition of the Trib for free here. The relevant excerpt:
...the [Independents] registered another score by virtue of a questionable onside kick. One official and the Cardinals players claimed Conzelman was not behind the oval [i.e. football] when it was kicked.
The referee...ruled the play legal with the result [that] the touchdown was allowed, and this proved to be the winning play, as neither team scored in the second half.
It seems that, in college football, kicking teams were barred from advancing their own kickoffs after the 1948 season. It would be interesting to know if the NFL rule was changed in response to the rule change in college football, or if it was changed even earlier.
Don't confuse this kind of a touchdown with a touchdown scored by the receiving team when the kicking team attempts an onside kick, like in this example from the highest scoring fourth quarter in NFL history. That is just an uncommonly short kickoff return touchdown.
Likewise, do not confuse this kind of touchdown with a different kind touchdown scored by the kicking team on the kickoff. For example, the kicking team can score by recovering and/or advancing a fumble by the receiving team, like one team did as recently as the 2023 season, but this is just a fumble return touchdown.
If a player on the receiving team touches the punted ball without taking possession of it, the punting team can recover, but not advance, the ball.
If a muffed punt finds its way into the receiving team's endzone, the punting team can recover it for a touchdown.
This has happened at least once, in 2020, in a game between the Carolina Panthers and the Washington Football Team.
The official boxscore for this game (which can be found in the gamebook, accessible here) records this as a fumble recovery in the endzone.
This is an incorrect use of the term "fumble", according to the definition of fumble given in the NFL Rulebook, because the punt returner for Washington never had possession of the ball.
Missed field goals can be returned for a touchdown by the receiving team, i.e. the defense.
This has happened several times in the NFL, but is still rare.
My favorite example is Antonio Cromartie's return for the San Diego Chargers in 2007.
There's a lot to mention about this touchdown: it was just a hair short of a strictly-speaking impossible 110 yards, it occurred in the same game that Vikings' running back and occasional local news interviewee Adrian Peterson ran for a still-record 296 yards, and the kicker was Ryan Longwell, a great example (when his kicks were good from a great distance) of nominative determinism.
Somewhat frustratingly, these returns are not counted towards a player's touchdowns on "combined kick returns", which is the relevant official category in the annual NFL Record and Fact Book. (The 2023 edition is freely downloadable at the NFL's website here.)
This stat-side decision is salient in at least one case. In Devin Hester's rookie season for the Chicago Bears (2006), he returned a missed field goal for a touchdown against the New York Giants.
Since this touchdown does not count towards Hester's combined kick returns, the rookie record in that category is 5, not 6. At the time, it was also the (rookie or nonrookie) single-season record, but Hester would have a combined 6 touchdowns the very next season on kickoff or punt returns, resulting in him breaking his own single-season record set the previous year, instead of tying it.
This is similar to #8, muffed punt recovery.
When the ball is placekicked or dropkicked from scrimmage, and the ball is beyond the line of scrimmage, the kicking team can only recover if it is touched, beyond the line of scrimmage, by the receiving team, i.e. the defense.
I don't think that such a recovery for a touchdown has ever been recorded.
A famous example of a recovery in this situation is the game-losing blunder, one of the most infamous ever, by Leon Lett, who was already famous for his much less consequential blunder in the previous year's Super Bowl.
The Dallas Cowboys blocked a potential game-winning field goal by the Miami Dolphins on Thanksgiving Day, with the ball lying beyond the line of scrimmage, and Leon Lett attempted to recover the ball, which was tactically unnecessary. Leon Lett touched the ball, but couldn't recover, and Miami did. NFL Films ranked it as the "number 2 Thanksgiving moment of all time". It was certainly a Thanksgiving moment of all time.
Now, this was a snow game, with the lines on the field difficult to see, and I've always thought that, in reality, it could have been the case that the Miami player recovered, i.e. gained possession of, the ball in the Cowboys' endzone, which would have been a touchdown of this kind. But the ball was ruled by the officials to have been recovered just short of the endzone, leading to a second attempt by Miami at a game-winning field goal, which was successful.
The holder on both of those field goals, by the way? Doug Pederson, current head coach of the Jacksonville Jaguars.
A safety kick is the free kick that follows a safety. It can be punted, unlike a kickoff, and usually is, because a kicking tee cannot be used on safety kick, unlike in the case of a kickoff. It can also be recovered by the kicking team, just like a kickoff.
Multiple safety kicks have been returned for touchdowns. Here is an example from 2009. In this example, Jacoby Jones of the Houston Texans burns the Oakland Raiders with a 95-yard return. Two weeks later, the Raiders' special teams unit would enlist the help of a pigeon.
Based on all scoring information available on databases like Pro Football Reference, a safety kick has never been recovered for a touchdown in NFL history.
These are the most obscure kinds of touchdowns, by far.
Every time a player completes a fair catch or has been awarded one by a penalty, his team actually has the option to attempt a fair-catch kick, which is, basically, a kickoff with which the kicking team can score a field goal.
The fair-catch kick isn't the en passant of NFL football; it's more like the underpromoting-to-a-piece-of-the-opponent's-color of NFL football.
There are only 26 known examples of fair-catch kicks in the entire history of the NFL; the most recent was in 2019.
These kicks can be returned for touchdowns, just like kickoffs. They can also be recovered by the kicking team for touchdowns, just like kickoffs, though they cannot be recovered anywhere on the field—not just in the endzone—unless the receiving team first touches the ball, unlike kickoffs. (This difference is moot in the context of recoveries in the endzone, however.)
Interestingly, Bill Belichick, wary of the possibility of a fair-catch kick return touchdown, did not have his Patriots attempt a 75-yard field goal via fair-catch kick at the end of Super Bowl LI.
A fumbling team can recover and advance their own fumble, even to score a touchdown, in most circumstances.
The most famous touchdown in this category is the "Holy Roller", a last-second game-winning play by the Oakland Raiders against the San Diego Chargers.
This play was controversial, and led to rule changes that prevent a situation like it from recurring. First, intentional forward fumbles are considered forward passes. Second, the only player on the fumbling team who can advance an offensive fumble on fourth down or a fumble after the two-minute warning is the fumbling player. The other players on the fumbling team can recover the ball, but not advance it. In fact, the result is a strange pseudo-recovering, where the ball is placed at the spot of the fumble or the spot of the recovery, whichever is less generous to the fumbling team.
If the ball is behind the line of scrimmage after a punt is blocked, the punting team can recover and advance the ball.
As far as I know, this has never resulted in a touchdown in NFL history.
If the ball is behind the line of scrimmage after a field goal attempt is blocked, the kicking team can recover and advance the ball.
In one of the greatest plays in NFL history that one almost never hears about, the Green Bay Packers scored in this fashion to win a game in 1980. In overtime. In Green Bay. Against their biggest rivals, the Chicago Bears. And it was the kicker who ran it in for a touchdown. Oh, and it was on opening day, too.
Yes, the referee can award a touchdown to a team as a result of a "palpably unfair act". This is implied by the arithmetically progressive rule 12-3-4 governing palpably unfair acts, which states: "[t]he Referee may award a score".
The scenario everyone thinks of is that of a player running onto the field to tackle an opposing player before he scores a touchdown. While a player has run onto the field during a play and participated in it in recent years, that was quite different; he merely helped block for a teammate on an interception return (and no penalty was called).
When it comes to this rule, the real closest call in recent memory actually involves a coach, Pittsburgh Steeler's Mike Tomlin. Tomlin was standing with one foot on the football field, and the Baltimore Ravens' kick returner, Jacoby Jones, who was running a kickoff back down the sideline, moved slightly towards the middle of the field to avoid running into him. Jacoby Jones was tackled almost immediately after this. Tomlin very well could have prevented a kickoff return touchdown, and his smile afterwards is less than exculpatory.
In my opinion, a penalty touchdown should have been awarded here, but it is hard to imagine the referee ever doing so. Instead, the NFL fined Tomlin $100,000. At the time, the only example of a head coach being fined more was Bill Belichick's fine of $500,000 as a result of the "Spygate" scandal.
In fact, the rules require the referee to award a touchdown in at least one circumstance, that of "successive or repeated fouls to prevent a score". This scenario is handled by the nearby rule 12-3-2, which states: "[i]f the violation is repeated after a warning, the score involved is awarded to the offensive team."
There is at least one scenario where 12-3-2 could realistically be enforced in an NFL game. Legendary defensive coach Buddy Ryan included a "Polish goalline defense" in one of his playbooks (see this archived blogpost, which includes a working link to the playbook in question).
I'll let Ryan's description in the playbook explain it:
Three extra linebackers go into the game.
Situation: the opponent is inside [our] 5 yard line going in to score. There [are] less than 15 seconds left. We want to stop their offense from scoring and[,] in the process, we want to run the clock down to where they have enough time for just one play. So, we will stop them, get penalized half the distance to the goal [for too many men on the field], but leave them with enough time to run one play. We will then go back to our regular goalline defense and stop them to win the game.
I'm not thinking specifically of intentionally committing a foul for too many men on the field. I don't think that's terribly realistic.
More realistic are repeated intentional pass interference fouls. On my interpretation of this rule, it would take at least 3. Once, then again with a warning, then a third time to meet the conditions of the rule.
Some readers might be confused by the inclusion of Leon Lett's blunder as an almost-example of a "field goal recovery" touchdown.
The field goal attempt in question was blocked, they would emphaize, and was touched by the defense on their side of the line of scrimmage, allowing the kicking team to recover.
It is conceivable, they would correctly claim, though highly, highly unlikely, that an unblocked field goal attempt could be muffed by a returner and recovered by the kicking team for a touchdown.
This is the true analogue, among field goal attempts, of a muffed punt recovery touchdown, they would say.
On the contrary, I am fairly confident that my categorization is correct here, and that this scenario presents no new touchdown type.
I can find nothing in the NFL rules that distinguishes the following two scenarios from each other:
This to say that, as far as the rules are concerned, when the ball lies beyond the line following a blocked field goal attempt, it might as well have not been blocked at all.
There is an inverse potential objection regarding my handling of punts.
Consider a Leon-Lett-type play that occurs on a punt, as opposed to a field goal attempt.
That is, the punt is blocked, but the ball still goes beyond the line of scrimmage. It is then touched by a member of the receiving team, and is finally recovered by the kicking team beyond the line. If this recovery occurs in the endzone, it is a touchdown.
This is actually not as unlikely as it sounds, because some blocked (read: tipped) punts can travel quite far beyond the line, behaving like mediocre, unblocked punts. A recovery following a touching by the receiving team on such a play would be no more extraordinary than one following a muffing of an unblocked punt.
In fact, there is an example in recent memory of a punting team getting a first down in this scenario, though, in this example, the blocked punt does not travel very far.
The verdict here is the same as in the case of unblocked field goal muffs: the rules view this scenario as equivalent to that of a muffed, unblocked punt.
Indeed, blocked scrimmage kicks hardly exist in the NFL rules as things to be legislated, as things that become entangled in permutations with other rules-theoretic entities. What really matters is simply whether the scrimmage kick is currently behind the line of scrimmage.
The possibility of passing the ball backwards, i.e. performing a so-called "lateral", leads to at least four questions for the taxonomy.
First, what if, say, the ball is successfully passed backwards to a teammate by a player who has intercepted the ball, and the teammate scores a touchdown?
In cases like these, I am happy to say that the play "inherits" the type of touchdown it would have been if a touchdown had been scored without any laterals. In the example above, the result would be an interception return touchdown.
Second, what if, say, the ball is unsuccessfully passed backwards to a teammate on a hook-and-ladder play following the initial catch, such that the ball hits the ground before being possessed by the intended receiver of the lateral, but the ball is nevertheless recovered by the offense and a touchdown is scored?
In this case, some would categorize this as a touchdown scored as a result of a fumbling team advancing their own fumble (included in the "Other" meta-category above).
But, the NFL rules clearly imply that, in such a case, no fumble has occurred. The NFL rules (8-7-3) define a fumble as "any act, other than a pass or kick, which results in a loss of player possession". Note that "pass" here means "forward or backward pass".
Furthermore, the NFL rules (8-7-1) also imply that, when a backward pass hits the ground, it is still a backward pass: "Players of either team may advance after catching a backward pass, or recovering a backward pass after it touches the ground."
So, instead, I think this case should be handled just as the previous case was handled, with the play "inheriting" the touchdown type of the sequence of play prior to any backward pass(es).
In contrast, unsuccessful handoffs, what the NFL rules also call "handling the ball", are fumbles. The rules 3-15(b) and 8-7-4(b) explicitly state that a muffed handoff, whether backward or forward, is a fumble.
Touchdowns scored therefrom should be classified accordingly. For example, in the 2006 AFC Championship game, when an offensive lineman for the New England Patriots recovered a fumbled handoff in the Indianapolis Colts' endzone for a touchdown, that was an own fumble advancement touchdown according to the typology here. Incidentally, in that game, three touchdowns were scored by offensive linemen: two own fumble advancement touchdowns and one receiving touchdown on a tackle-eligible play.
Third, what if an opponent gains possession following an unrecovered backward pass, and scores a touchdown?
Here, there is genuine trouble for the taxonomy. It seems like "backward pass return" should indeed be an additional type of touchdown. This is acute given the taxonomy's distinguishing between muffed punts and fumbles. For consistency's sake, a distinction should also be made between mishandled or unhandled backward passes and fumbles.
Frankly, I'm just too lazy to go back and renumber the touchdown types above, because I'd want to include it in the "Defensive" meta-category. Here's a recent, spectacular example of such a touchdown as an apology and recognition of this touchdown type's plausible legitimacy.
Including this type would bring the total number of touchdown types to 21, by the way.
Fourth, how should "fumbled" snaps be handled?
The NFL rules clearly imply that it is impossible to fumble a snap, because, in such an interaction, no offensive player ever has possession of the ball, and a fumble is a loss of possession. Note that the center cannot have possession of the ball, in any relevant sense, because it is the completion of the snap that puts the ball in play.
Indeed, the snap is defined as a backward pass in rule 3-32, and the note to rule 8-7-1 reminds the reader that "[a] direct snap from center to a player in the backfield, a muffed hand-to-hand snap, or a snap that is untouched by any player are backward passes, and the ball remains alive" (emphasis added).
So, it seems that a defensive player scoring a touchdown by returning or recovering a loose snap—to introduce a general term—has scored a backwards pass return touchdown, the same type just introduced above. Here is an example.
A case could be made in the spirit of fine-grain division that, because snaps are clearly a distinct kind of backward pass, this should be an additional touchdown type.
This would bring the total to 22.
The idea would have to be that, much like how the NFL rules define free kicks as kickoffs and safety kicks, the NFL rules define backward passes as snaps and...non-snap backward passes? Here, the analogy breaks down. Snaps are a particular kind of backward pass. Unlike kickoffs/safety kicks, they are not essential "building blocks" of the concept (backward pass) in question.
Nevertheless, we still have to consider a loose snap that is recovered by the offense (or, more generally, the "snapping team"). I will address this in a later subsection.
I said above that a play with a backward pass inherits the type of play for the purposes of classifying a touchdown, if one is scored from the play.
If so, what of a touchdown scored by the player who receives the snap, and runs into the endzone without any intervening complications, like a forward pass or fumble?
Intuitively, that is a rushing touchdown, belonging to the first type listed above. But how did it inherit the state of "rushing", if "rushing" is not the default state?
The answer is that rushing is the default state, because every scrimmage play begins with a snap, and a snap is a backward pass, and "rushing" really refers to plays with backward passes, and nothing more exotic.
In other words, it's not that the default state is defined to be "rushing", it's that "rushing" happens to be the default state, because rushing plays are plays only with backward passes, and that is exactly what a scrimmage play always is, initially.
This certainly matches our intuitions when the backward passes are all successful. Snap to backward handoff to touchdown? Rushing. Snap to "pitch", i.e. planned backward pass, to touchdown? Rushing. Snap directly to touchdown? Rushing.
When the non-snap backward pass is unsuccesful, it becomes less intuitive, and one is inclined to conceptualize it as a fumble, but the rules are clear that it's not. Accordingly, touchdowns on such plays would be rushing touchdowns in the above typology.
Consider now the following scenario. The snap is muffed, rolls forward into the opponent's endzone, and is recovered by a member of the offense for the touchdown.
It might seem like the logical conclusion that, in the typology developed here, such a touchdown is a monstrosity, an apparent impossibility: a 0-yard rushing touchdown, a backward pass recovered forward.
Even ignoring how odd that would look in a boxscore, it seems markedly unintuitive to classify such a touchdown as a rushing touchdown. It should either be an own fumble advanced touchdown or belong to its own type.
Justification for it belonging to its own type might come from the rules governing when forward passes are legal. Rule 8-1-2(1c) outlaws forward passes when the ball has already crossed the line of scrimmage, even if the pass itself is made behind the line of scrimmage.
The suggestion is this. When a loose snap is recovered by the offense while or after the ball has crossed the line of scrimmage and a touchdown is scored from this, the result should not be counted as a rushing play, because something "special" has happened, i.e. the ball has crossed the line of scrimmage without ever being possessed by the offense, and play is therefore no longer in the default state./p>
We should probably view the advancement of a backward pass that is recovered while or after the loose ball has crossed the line of scrimmage in the same way, to be consistent, so as not to introduce a distinction between snaps and non-snap backward passes.
Touchdowns resulting from these plays would be true backward pass advancement touchdowns. I don't know if any such touchdown has ever been scored in the NFL.
In conclusion, the number of ways to score a touchdown in the NFL is at least 20, and might be as high as 23.