NHL penalties explained, from a minor to a match penalty

Every punishment in ice hockey is measured in the same currency, which is time without a player. There is no free kick, no penalty yardage, no possession handed over. A referee's arm goes up, somebody skates to a box, and for the next two minutes his team plays a man short against opponents who have spent all week practising exactly this.

That single design choice explains most of what is strange about NHL penalties. It explains why a trip in the neutral zone and a slash in the corner cost precisely the same. It explains why teams commit fouls on purpose in some situations and never in others. And it explains why the sport needed to invent so many exceptions to the two-minute rule, because a punishment measured in time is blunt, and the rule book has spent a century sanding down the corners.

What follows is the whole structure: what each class of penalty costs, what ends it, what happens when both teams offend at once, and the parallel disciplinary system that operates after the final horn and answers to nobody in stripes.

Why hockey punishes by subtraction rather than by giving something away

Most team sports punish a foul by handing the other side territory or possession. Hockey cannot really do either. Possession changes hands constantly and is worth little on its own. Territory is meaningless on a surface where the puck can cross the whole rink in under two seconds.

So hockey took the one thing that is genuinely scarce, which is bodies on the ice, and priced everything in that. Five skaters against four is an enormous edge, and the sport has built an entire coaching discipline around exploiting it, covered in the piece on what a power play unit is actually trying to do. The mirror discipline, killing the penalty off, is treated separately in the guide to the systems a short-handed team uses.

The consequence is that every penalty is a swing rather than a deduction. The offending team does not merely lose something. The other team gains a two-minute window in which its best five players can be on the ice together against a defensive unit chosen for its capacity to suffer.

Minor2min

Double minor4min

Major5min

Match penalty5min

Misconduct10min

Durations fixed by the NHL rule book. A game misconduct has no duration because it is an ejection; the offending team is not reduced unless a separate penalty is attached to it.

Show the numbers

What each class of penalty costs in minutes

Item	Value
Minor	2min
Double minor	4min
Major	5min
Match penalty	5min
Misconduct	10min

The chart hides the most important distinction on it, which is that only the first four of those five actually cost the team anything. A misconduct is the longest sentence in the list and the cheapest one to serve.

The minor, the double minor, and what makes each of them end

The minor is the workhorse. Two minutes for tripping, hooking, holding, slashing, interference, cross-checking, roughing, delay of game, too many men, and the several dozen other things a hockey player does when losing a footrace.

Two features of the minor matter more than its length.

The first is that it ends early if the team on the power play scores. One goal, one penalty wiped out, player released from the box the instant the puck crosses the line. That is why a two-minute power play is worth at most one goal, and it is a deliberate cap on how much a single foul can cost. If a team is already two men short and a goal goes in, only the first of the two penalties expires, so the side stays a man down.

The second is that the minor is not scaled to the damage done. A trip that stops a nothing rush behind the net and a trip that kills a two-on-one both cost two minutes. The rule book accepts that unfairness because the alternative, grading penalties by how much they cost the other team, would require a referee to assess a counterfactual in real time. The penalty shot, discussed below, is the one place hockey does grade a foul by its consequence, and the conditions for it are narrow precisely because the judgement is hard.

The double minor is four minutes served as two consecutive twos, and it exists for two quite different reasons. The first is a genuinely doubled offence, most often high-sticking that draws blood, where the outcome rather than the act determines the length. The second is a way of expressing that an act was worse than routine without reaching the threshold of a major.

The mechanics of a double minor catch people out. It is two penalties, not one long one. If the power play scores during the first two minutes, that half ends and the second two minutes begins immediately with the player still in the box. So a team can concede a goal and remain a man short, which looks like a bug and is not. A four-minute high stick can, and occasionally does, produce two goals.

The bench minor is the same two minutes assessed against a team rather than a person, for things a team does collectively: too many men on the ice, abuse of officials from the bench, delaying the game. Any player on the ice at the time may serve it, chosen by the coach, and by convention that is a defenceman who has just been off for a while.

Goaltenders never serve their own penalties. A teammate on the ice at the time goes instead. This is practical rather than indulgent, since a team with no goaltender for two minutes would be facing an unplayable situation, but it does mean a goaltender can accumulate minors at no personal cost beyond the pay he is fined for repeat offences.

The major runs its full five minutes, whatever the score

A major is five minutes, and the crucial difference from a minor is that scoring does not end it. The team on the power play can score twice, three times, four times, and the offender stays in the box until the clock reads zero.

That is why a five-minute major is not two and a half minors. It is a categorically different punishment, and the gap between the expected damage of a minor and of a major is much larger than the ratio of their lengths.

Majors are assessed for the serious physical fouls, the ones with real injury potential: boarding, charging, cross-checking, elbowing, head contact, slew-footing, kneeing and the rest, where the referee judges the act severe enough to warrant it. Fighting is a major as well, though it is usually a coincidental one and therefore costs neither team anything on the ice.

Two further consequences ride along with a major that people forget. A player assessed a major for a physical foul frequently receives a game misconduct alongside it, ejecting him. And every major is reported to the league's player safety department automatically, which starts a separate process that can end in a suspension long after the game is over.

• 3Minimum skaters a team may have on the ice
• 2Penalties that can run against one team at once
• 5Minutes a major runs regardless of goals scored
• 10Minutes a misconduct costs with no power play attached

All four are set by the NHL rule book rather than observed from any game.

The first two of those are worth dwelling on, because they set a floor on how bad things can get. A team never plays with fewer than three skaters. If a third penalty is assessed while two are already running, the third one is queued and its clock does not start until one of the earlier penalties expires. A five-on-three is the worst position in hockey, and by rule it is also the worst position available.

Misconducts punish the man and spare the team

A misconduct is ten minutes in the box for the player and nothing at all for his team. He is replaced on the ice immediately, the sides stay even, and he returns at the first stoppage after his ten minutes have run.

This looks like a soft option and is more useful than it appears. It gives officials a way to remove a specific troublemaker from a situation that is heating up without handing the other team a power play it has not earned. Referees use it for abuse, for refusing to stop when told, for continuing an argument after being warned, and for the general category of conduct that is a problem of behaviour rather than of play.

A game misconduct is an ejection. The player goes to the dressing room and takes no further part. The team is not reduced unless another penalty is attached, which it usually is. Like a major, a game misconduct is automatically reported to the league.

The design here is neat once you see it. Hockey has separated two questions that most sports answer with the same call: how much did this foul cost the opponent, and how much does this player deserve? The minutes-in-the-box answer the first question. The misconduct answers the second, and it answers it without contaminating the game situation.

The match penalty, the only call that turns on intent

A match penalty is the heaviest in-game punishment available. It is assessed for deliberately injuring an opponent, or attempting to. The player is ejected for the remainder of the game, a teammate serves five minutes in his place, and the incident goes straight to the league for review.

What separates a match penalty from a five-minute major for the same physical act is not the severity of the outcome. It is the referee's judgement about what the player was trying to do. Every other call in the rule book can be made from the act itself. This one requires a conclusion about a state of mind, arrived at in a couple of seconds by an official who was thirty feet away.

That makes match penalties rare, contentious, and disproportionately likely to be revisited afterwards. It also makes them honest about something the rest of the rule book hides, which is that the difference between a hard hockey play and an assault is often entirely a matter of intent, and somebody has to decide.

Coincidental penalties, and why two wrongs make four-on-four

When both teams are penalised on the same stoppage, the rule is to cancel as many matching penalties as possible and apply only the difference.

Matching minors at even strength take both players off and leave four skaters a side. Nobody has a power play. The sides are equally reduced, and four-on-four is a genuinely different game with more ice, more odd-man rushes and more chances at both ends, which is why it is sometimes worth a coach's while to be down a man in that particular way.

Matching majors, the usual outcome of a fight, are different in one important respect. Both players sit for five minutes, but both are substituted for immediately, so the teams play five-on-five throughout. The fighters stay in the box until the first stoppage after their time expires. Nothing about the game state changes. Two players have effectively removed each other from the next five minutes of hockey and the score does not care.

The cancelling is done by matching penalties of the same class, and any remainder is served. A major against one team and a minor against the other leaves three minutes of power play for the side that took the minor, because the five and the two do not simply subtract to a clean number of a single type. If one team is already short-handed when coincidental minors are assessed, the players serving them cannot come back the moment their time runs out; they wait for the next stoppage, so that the return does not accidentally create an advantage nobody was awarded.

• Ravens
• Foresters

Constructed example, not a real game. Minute 0 is a minor to the Ravens; a second Ravens minor follows twenty seconds later; the first expires at two minutes; matching minors are then assessed to both sides.

Show the numbers

Skaters on the ice through an invented penalty sequence

Item	Ravens	Foresters
Even strength	5	5
First minor	4	5
Second minor	3	5
First minor expires	4	5
Second minor expires	5	5
Matching minors	4	4

Read the trough in the middle of that. Twenty seconds of carelessness turned a two-minute penalty kill into a stretch at three skaters against five, and the second penalty's clock did not even begin until the first had run. Then, once everything expires, a pair of matching minors drops both teams to four without giving anyone an advantage. Same rule book, three completely different games inside four minutes.

The delayed penalty, and the sixth attacker nobody defends

If a player commits a foul while his team does not have the puck, the referee does not blow the whistle. He raises an arm and lets play continue until the offending team gains possession and control, at which point the whistle goes and the penalty is assessed.

The logic is straightforward. Stopping play immediately would punish the team that was fouled by ending its attack, so the rule lets the attack finish first.

What the rule produces is one of the most peculiar sights in sport. The team about to receive the power play pulls its goaltender and sends on a sixth attacker, because it cannot concede in that window: the moment the other side touches the puck with control, the whistle goes. Six attackers, five defenders, an empty net at the far end and no risk attached to it.

There is one risk, and it is the sort of thing that happens twice a decade. The empty net is not protected by the delayed penalty rule against the attacking team itself. If a player on the team with the extra attacker puts the puck into his own net, it counts, and the crowd's reaction is worth waiting for.

The other detail worth knowing is what happens if the goal goes in at the right end. A minor penalty is wiped out entirely if the non-offending team scores during the delayed call. The player never serves it. He skates off the ice having committed a foul that cost his team a goal but not a second in the box. A major is assessed regardless, because a major is punishing the act rather than pricing the disadvantage.

1. 1A foul is committed by the team without the puckThe referee raises his arm and holds it there. No whistle. Play continues exactly as it was, and the offending team is now playing under a sentence it cannot escape by defending well.
2. 2The attacking team pulls its goaltenderThe bench sends on a sixth skater as the goaltender races off. There is no defensive risk in the ordinary sense, because play will stop the moment the other side controls the puck.
3. 3The attacking team keeps possessionSix attackers against five defenders, with the puck cycled rather than shot from distance, because a rebound that reaches an opponent ends the whole thing.
4. 4Either a goal is scored, or the puck is controlled by the offending teamThose are the only two ways the sequence ends. There is no clock on it and no limit to how long it can last.
5. 5If a goal is scored, a minor is cancelledThe penalty is never served and the offender leaves the ice with the whistle. A major, or any penalty carrying an ejection, still stands.
6. 6If the offending team gains control, the whistle goesThe penalty is assessed from that moment, the goaltender returns, and the power play begins.
7. 7The face-off location is set by rulePlay restarts in the offending team's defensive zone, so the team with the advantage begins its power play with possession in the right end of the rink.

The delayed penalty sequence set out in the NHL rule book. No particular game is described.

When a penalty shot replaces the two minutes

Sometimes two minutes is not the right price, because the foul destroyed a goal that was very likely to be scored. Hockey's answer is the penalty shot: one skater, one goaltender, from centre ice.

The classic case has four conditions, and all four must hold. The infraction happened in the neutral or attacking zone. The attacking player was fouled from behind. A reasonable scoring chance was thereby denied. And there was no opposing player between the puck carrier and the goaltender at the time. Miss any one of those and it is an ordinary minor.

The list runs wider than that. A penalty shot is also awarded when a defending player throws his stick or any object at the puck carrier on a breakaway, when a player who should not be on the ice interferes with a breakaway, and when a player other than the goaltender picks up or covers the puck in the goal crease. Deliberately displacing the net in the closing stage of a game, where a penalty could not be fully served, produces one as well.

There is a further step beyond the penalty shot that almost nobody has seen. If the offending team's goaltender has already been pulled and an attacker with no opposition between him and the empty net is fouled, the referee does not award a penalty shot. He awards the goal.

The penalty shot replaces the minor rather than accompanying it. The team does not go short-handed as well. That is the whole point: the foul is being priced at roughly what it destroyed, rather than at the standard two minutes. If the shot is scored, play resumes at centre ice. If it is not, the offending team has escaped with nothing at all, which is a harder outcome than a penalty kill and is why coaches are less delighted by penalty shot calls than crowds are.

Obstruction fouls, physical fouls, and the pendulum between them

Penalties fall into two families that behave completely differently.

Obstruction fouls are hooking, holding, interference, tripping and the grey area around them. They are committed by a player who has been beaten and is trying to reduce the cost. They are almost never dangerous. They are also, and this is the important part, almost always available: on any given shift there are several moments where a stick is on a hand or a body is in a lane, and whether they are called is a matter of where the threshold sits that night.

Physical fouls are boarding, charging, elbowing, head contact, kneeing and the rest. They are about danger rather than advantage, they are far less frequent, and the threshold for calling them moves much less, because the consequences of not calling them are visible on a stretcher.

The enforcement history of the sport is essentially the history of the obstruction threshold moving. Hockey has gone through periods where clutching and grabbing were tolerated to the point that skill players could be neutralised by a competent journeyman with a long stick, and periods, most famously the crackdown that followed the 2004-05 lockout, where the league instructed officials to call obstruction strictly and the game changed within a season. Skating became worth more. Defencemen who could not move were exposed. Scoring rose.

The rule book barely changed in any of those swings. The instruction to officials did. That is worth holding on to, because it means the sport people watch is set as much by an enforcement standard as by the words in the book, and enforcement standards are decisions made by a small number of people in a room.

The 2025-26 rule book made this explicit in a way it never had before. The interference rule previously said that a minor penalty "shall be imposed"; it now says one "can be imposed". The league also added a provision to the glossary allowing officials to apply common sense in situations the rules do not expressly cover. Neither change alters a single definition of a foul. Both formally acknowledge that judgement is part of the job, which everybody knew and the book had been pretending otherwise.

The same rule book made the one substantive change of that cycle, to high-sticking. The old test was whether the stick was carried above the height of the opponent's shoulders. The new one asks whether the stick made contact with an opponent's neck, face or head. That is a shift from a rule about how a stick is carried to a rule about what a stick does, and it moves high-sticking a little further from the obstruction family and a little closer to the physical one.

Player safety is a separate court with different rules

Everything above happens on the ice, in real time, decided by officials who cannot pause the game. Running alongside it is a completely separate track, and confusing the two is the source of most of the arguments about hockey discipline.

The league's department of player safety reviews incidents after the fact. It works from video, at leisure, with camera angles the referee never had. It can fine a player, suspend him, or do nothing, and it can do any of those regardless of what happened on the ice. A play that drew no penalty at all can produce a suspension. A play that drew a five-minute major and an ejection can produce nothing further.

That is not an inconsistency. The two systems are answering different questions. The referee is asking how the game should be adjusted right now for something he thinks he saw. The department is asking what this act deserves, having watched it thirty times.

Its published reasoning follows a reasonably predictable shape: whether the contact was legal, what part of the body was the main point of contact, whether the injured player put himself in a vulnerable position, whether the hit was avoidable, and whether the offender has been here before. Repeat offender status is defined in the collective agreement rather than by mood, and it matters, because it moves the starting point for the next incident.

Two structural limits are worth knowing. Fines are capped by the collective agreement at a small proportion of a player's salary, with an absolute ceiling, and both figures are published in the agreement rather than set case by case. Because the cap is low relative to modern salaries, fines are a marker of disapproval rather than a deterrent for anyone earning well. And a suspended player forfeits pay for the games missed, which for a repeat offender on a large contract is by far the heaviest financial consequence in the system. The way that money interacts with a club's books is a wrinkle in itself, and it sits alongside the rest of the machinery described in the piece on how the NHL salary cap actually works.

Suspensions above a certain length trigger an in-person hearing rather than a telephone one, and longer suspensions carry an appeal route that runs first to the commissioner and then, beyond a threshold, to a neutral arbitrator. That structure exists because a system where the prosecutor is also the final appeal is not one any players' association would sign.

The instigator rule, and the argument that will not die

Fighting occupies its own strange corner of the rule book. A fight is a five-minute major, usually coincidental, usually costing neither side anything on the ice.

The instigator rule adds a two-minute minor and a ten-minute misconduct for the player judged to have started it. The criteria are set out in the rule book and are behavioural rather than physical: who travelled the distance, who dropped the gloves first, who threw the first punch, who was verbally threatening, and whether the fight looks like retribution for something earlier in the game or the season.

In the last five minutes of regulation, or in overtime, the misconduct becomes a game misconduct, and the offending player's coach can be fined and suspended for it. That escalation is the giveaway. The rule is not really aimed at fighting. It is aimed at the deliberate late-game fight designed to send a message when the result is already settled, and at the coaching decision behind it.

The argument about the rule has run for decades and both sides of it are coherent. The case for it is that a sport should not allow a player to be assaulted at no cost to the assailant's team, and that removing the free retaliation removed a category of violence that had nothing to do with hockey. The case against it is that policing used to be done on the ice, that taking away the deterrent left skill players more exposed to stickwork and cheap hits, not less, and that the league then had to build an entire department to do a job the players had been doing themselves.

Nobody has settled it, and the honest answer is that the evidence is genuinely mixed and heavily confounded by everything else that changed in the sport across the same period.

Diving, embellishment, and the fine that grows

Because penalties are so valuable, the incentive to manufacture them is obvious, and hockey's answer is unusual among major sports.

Embellishment is a minor penalty in its own right, and it is frequently assessed alongside the foul it was designed to exaggerate. Both players go to the box, the teams play four-on-four, and the message is that the referee saw a foul and saw a performance.

Beyond the in-game penalty sits a graduated schedule of fines that escalates with each incident across a season. What makes it distinctive is that the escalation reaches the head coach as well as the player. Keep diving and your coach starts paying, which turns a matter of individual conduct into a matter the dressing room has an interest in policing.

Whether it works is debatable, and the same objection applies here as in every sport that has tried it. Punishing embellishment requires a judgement about whether a player went down more easily than the contact warranted, which is a question about intent again, made at speed, about an athlete who was genuinely hit. Officials in every sport that has legislated against simulation report the same difficulty. What hockey adds, by fining the coach, is a party who can do something about it during the week.

Why the standard moves, and how to watch for it

Anybody who watches a full season notices that the same act is not called the same way in October and in May, and the explanation is not that officials are inconsistent. It is that several forces push the threshold around in ways that are entirely predictable once you know they exist.

At the start of a season the league issues points of emphasis, telling officials to tighten up on a particular family of fouls. Calls in that family spike immediately. Players adjust within a few weeks, coaches complain about the flow of the game, and by midwinter the rate has usually drifted back towards where it was. That cycle repeats most years.

In the playoffs the pressure runs the other way. Nothing in the rule book changes, but the cost of a marginal call rises enormously, and the natural human response of an official in a tied elimination game is to require more before raising an arm. Players know this and adjust what they are willing to try. The result is a version of the sport with more contact allowed, which people describe approvingly as playoff hockey and disapprovingly as the rule book being suspended, and both descriptions are of the same thing.

There is a third force, less discussed. Referees manage games as well as officiate them. A team that has taken three penalties in a row is, in practice, slightly more likely to get the next marginal call in its favour, not because anyone is levelling the score deliberately but because a human being making judgement calls under pressure is aware of what the penalty count looks like. Whether that is a flaw or an inevitable feature of employing people is one of the oldest questions in officiating, and it is examined more generally in the piece on how officials arrive at decisions under pressure.

So here is what to watch, now that the structure is visible.

Watch the referee's arm rather than the puck when a foul happens away from play, because a raised arm with no whistle means a delayed penalty and the goaltender is about to sprint for the bench. Watch whether a call is two minutes or four, because a double minor means the power play can concede a goal and still be a man up. Watch whether a major came with a game misconduct, because that combination means the league is going to look at it during the week. And watch the penalty count when a game has gone twenty minutes without a whistle, because the threshold has already moved and the next borderline hook is unlikely to be called either.

The rest of the sport's rule book rewards the same attention, and the ice hockey archive is the place to start. The obvious next stop is the rule governing contact with the goaltender, which is where hockey's tolerance for judgement calls is tested hardest, and where a two-minute penalty can turn into a goal being taken off the board.