Nearly Two in Five NHL Goals Come Within Six Seconds of Zone Entry

Goals come within seconds of entry, or much later

Since 2023-24 the NHL has published a short clip for nearly every goal: the tracked position of the puck and every player, ten frames a second, the data behind our goal replays page. We ran three seasons of them, 23,393 goals scored with a goalie in net, through one set of rules that finds when the puck last crossed the blue line, where the scorer let the shot go (the release point), and who touched it and how. We wanted to know two things: how a goal gets there, and whether the way it gets there belongs to the team or to the player who scores it.

First, the clock. Of the 23,393 goals, the clip shows when the puck came into the zone for 22,865. For the other 528 it starts too late to tell, so the timing below uses the 22,865. About two in five, 39.1%, went in within six seconds of the puck crossing the blue line. One in five, 19.7%, went in within three seconds. The single most common moment was two to three seconds after entry. The other 60.9% came after the puck had been in the zone six seconds or longer. For most of those, the puck was already in the zone when the clip begins, usually about eight seconds before the shot, so the clip never shows it coming in. We still know they were slow, because the clip shows the puck in the zone for six seconds or more. The 528 we could not time are the ones whose clip shows less than six seconds in the zone and no entry.

Almost nothing goes in faster than a second and a half: only 73 of the 22,865 timed goals did. The puck has 64 feet to cover from the blue line to the goal line. The fastest in the sample belongs to Timothy Liljegren, then a Toronto defenceman, at Philadelphia on March 14, 2024: 0.9 seconds from the blue line to the back of the net, on a wrist shot from 45 feet.

The fast goals bunch up at two to three seconds. That is a rush: carry it in, make one move or one pass, shoot. Edmonton’s only goal in that Game 7, by Mattias Janmark from a Cody Ceci pass, took 2.3 seconds too. After four seconds the bars flatten, and the long tail is the set play, the cycle and the scramble in front. The clip cannot say how long those possessions ran, only that they ran six seconds or longer.

That split between fast and slow barely moves from year to year. The share of goals inside six seconds was 38.8% in 2023-24, 39.2% in 2024-25 and 39.3% in 2025-26. We held back the first 1,066 goals of this season as a check and set none of the headline numbers with them. They came in at 40.2%.

The extra man is the slowest scorer in hockey

That clock runs at a different speed depending on who has the extra man. The power play looks like the easy goal. By the clock it is the slowest one. At 5-on-5, 45.5% of goals came within six seconds of entry. On the power play it was 11.6%. With an extra skater a team sets up, moves the puck and waits for a lane.

The power play is also where the one-timer lives. It is Leon Draisaitl’s shot: the pass arrives and he hits it in the same motion, without stopping the puck, so it leaves as fast as it came and in a new direction. 25.8% of power-play goals were one-timers, about double the 5-on-5 rate of 13.0%. Power-play goals also came after the most passes. We count the passes the clip shows in the attack that ended in the goal. A power-play goal had 1.84 of them on average, against 0.98 at 5-on-5. More power-play goals had a pass across the slot before the shot, too: 35.7% of power-play goals, against 19.3% at 5-on-5. Our pass count misses some passes, so the true counts are higher, as one of Leon Draisaitl’s biggest goals shows further down.

Shorthanded goals run the other way. 89.7% came inside six seconds and 78.4% off the rush. Nobody in the sample scored more of them than Sam Reinhart, with 11, and nine of his came within four seconds of entry. In Game 2 of the 2025 Final, Brad Marchand scored one for Florida on what TSN called “a short-handed breakaway,” 2.2 seconds after the puck crossed the line. Three-on-three overtime sits between the extremes, with 58.0% of its goals off the rush.

Turnovers, rushes and the net front make most goals

We also sorted every goal by the kind of play that made it. We put each goal into one kind of play, using the same fixed rules every season (they are in the method below). Four kinds of play made nearly three goals in four (72.7%). The biggest is the turnover in the zone. A defender has the puck in his own end, an attacker takes it off him, and the puck is in the net within eight seconds. Then come the odd-man rush, the rebound and the deflection. There are two kinds of rush. Reinhart’s Game 7 winner was an odd-man rush: by the tracking, Florida had more attackers coming than Edmonton had defenders back. In an ordinary rush, marked “even numbers” in the chart, the numbers are level. The rarer kinds include the dump and chase, where a team shoots the puck deep and races after it.

The mix barely moves from one season to the next

Season to season, the mix hardly changes. Across the three full seasons, no kind of play's share of goals moved by more than 1.7 percentage points.

Rushes and the net front far outscore the set play

Put the two kinds of rush together and 30.6% of goals came off the rush. Rebounds and deflections, the goals made in front of the net, add another 30.2%. The classic set play, where a team settles into the zone and works the puck around the boards and the blue line until a lane opens (the cycle, or a settled possession), made only 8.9% of all goals. At 5-on-5 it was 5.7%. On the power play it rose to 19.4%. The point shot, a long shot from a defenceman up near the blue line (the point), is rarer still. Where one play type ends and the next begins is a judgment call, but the rules are the same every season, and the mix barely moved. Between any two full seasons, no kind of play’s share of all goals changed by more than 1.7 percentage points, fewer than two goals in every hundred.

Marchand finished Game 2 with another rush. At 8:05 of the second overtime, he took a pass from Anton Lundell and, in TSN’s words, “was able to sneak a backhand shot through Oilers goalie Stuart Skinner’s legs.” The tracking reads an odd-man rush, 2.7 seconds from the blue line to the goal, released 6 feet out. Florida won 5-4.

A third of goals are scored from inside 10 feet

Rebounds and tips are goals made in front of the net, and the net front is where a big share of all goals come from. Verhaeghe’s Game 7 tip left his stick about 2 feet from the goal line. Reinhart’s winner left his 36 feet out. We took that distance for every goal, from the spot where the scorer last touched the puck to the centre of the goal line. A third of all goals came from inside 10 feet, roughly the crease and the doorstep, and only 10.1% from 40 feet or more. Line every goal up from closest to farthest and the one in the middle, the median goal, was released 16.0 feet out: half came from closer, half from farther.

No regular scorer lives closer to the net than Zach Hyman. Half his goals in the sample were released within 7.4 feet, the shortest mark of the 51 players with 80 or more goals; for the league as a whole it is 16.0 feet. Of his 124, 79 came from inside 10 feet, the most of anyone, and he has the most deflection goals, 33. “I’ve always been trying to figure out how to score goals, and that’s the way I score goals,” he told NHL.com’s Derek Van Diest in December 2023. A little later he put it more simply: “If you go to the net enough, you’re going to figure it out.”

Distance and play type go together. Rebounds (typically 7.9 feet out) and deflections (8.3 feet) are the short goals. Ordinary rushes were typically released 23.0 feet out (odd-man rushes 19.6). Of the goals released 15 feet out or farther, about a fifth were one-timers. Inside 10 feet, fewer than one in ten were.

A team’s one-timer habit barely lasts a season

So far this is a portrait of the league: how fast goals come and what kind of play makes them. The second question is whose habit that is. On a leaderboard, teams look as if they have a style. Montreal scored 23.4% of its 2025-26 goals on one-timers and Calgary 8.9%. But a team scores only 165 to 345 goals a season, so a few lucky one-timers move its share a lot.

Picture 32 teams that all had exactly the same habit. Over three seasons, about 575 to 945 goals each, luck alone would still scatter their one-timer shares, typically by about 1.4 points either side of the average. (Each point is one goal in every hundred.) The real teams scatter by about 2.2 points. If luck explained everything, the real teams would be spread no wider than the imaginary ones. They are spread wider, so some of the difference is real. You can’t just subtract one number from the other, because luck and habit don’t stack up like plain numbers. Worked out properly (the arithmetic is in the method below), a typical team’s real habit puts it about 1.7 points above or below the average over those three seasons.

That is a small habit, and one season hides it. A single season has far fewer goals, and luck alone can move a team’s share by two or three points either way, as the chart below notes. That is more than the habit itself. Over three seasons the luck partly evens out, and the small real difference shows through.

Season to season, team one-timer shares bounce around. The three full seasons give two pairs to line each team up against itself the next year: 2023-24 against 2024-25, and 2024-25 against 2025-26. In the first pair, a team’s share told you nothing about its next one; the second showed some link: teams high one season tended to be high the next, with plenty of exceptions. A third look, 2025-26 against the first weeks of this season, also showed nothing, but each team has only 20 to 47 goals so far this season, too few to count for much. Rush share holds a little better: some link in the first pair, a weaker one in the second that luck alone could produce, and nothing so far this season. Rosters and systems change between seasons too, so a team’s habit can change with them.

The fairest test keeps the roster fixed. Take the same three seasons and split each team’s games into two piles, odd-numbered games and even-numbered games, like dealing a deck into two hands. Same players, same coach, same seasons. If a team really has a one-timer habit, both piles should show it. The two piles do line up, but loosely. If every team had the same habit, the piles would usually match less closely than this. So a few teams really do lean on the one-timer. The gap is small, though, and one season’s team ranking is mostly luck.

The style belongs to the scorer, not the team

Run the same tests on the men holding the sticks and the answer flips. Take the 51 regulars, the players with 80 or more goals over the three seasons. After taking luck out, a typical regular’s one-timer share sits about 8 points above or below the average, roughly five times as far as a typical team. Now deal each player’s games into the same two piles. That test needs enough goals in each pile, so it uses a different, larger group: the 86 players with at least 30 goals in each pile. A player’s one-timer share in one pile closely matches his share in the other. That match is far too close to be luck.

Three of the extremes belong to one team, Edmonton. Leon Draisaitl scored the most goals in the sample, 146, and the most one-timers by a distance: 62, or 42.5% against 16.0% league-wide. Of those, 41 came on the power play, where the next-highest total was Steven Stamkos’s 30. Season by season his one-timer share was 53%, 33% and 43%. Over the three seasons combined, only three of the 51 regulars scored more than 40% of their goals on one-timers: Draisaitl, Nikita Kucherov and Steven Stamkos. Next came Bo Horvat, at 31.5%.

His captain scores another way. Connor McDavid had the highest rush share of the 51, 45.5% (49%, 42% and 45% by season), and the most goals within six seconds of entry, 53. Their winger, Zach Hyman, is the crease scorer we met above. Sidney Crosby sits at the other end from McDavid. He scored only 14.0% of his goals off the rush, the lowest of the 51, and 38.7% from rebounds and tips. Chris Kreider has the highest share of goals off rebounds and tips of the 51, at 58.3%.

The net-front habits repeat too. A player’s share of goals off rebounds and tips, and his share from inside 10 feet, match across the two piles about as closely as his one-timers do. The rush is the weakest habit: it repeats, but less. These tests show the habits are real, but they can’t pin down any one player’s exact rate. Draisaitl’s 43% still has some luck in it. And a scorer’s share includes the job his coach gives him. Of Draisaitl’s 62 one-timers, 41 came on the power play, so part of his style is the spot Edmonton gives him there.

Playoff goals lean slower, a little

One more test of the league picture: does the clock change when the games matter most? The playoffs tilt it a little toward the slow goal. Playoff goals came off the rush less often, 28.1% against 30.8% in the regular season, and were a bit less likely to come within six seconds of entry (36.5% against 39.3%). The gap points the same way in each of the three playoffs. With only about 450 playoff goals a year, though, it is just a little bigger than luck alone usually produces, so it could still be luck. And playoff games pit better teams against each other. Better teams may simply score differently, and then the gap comes from who is playing. Call it a lean, and no more.

In all three seasons, playoff goals came off the rush less often and took longer after entry

Draisaitl’s biggest goal shows what the clips miss

The slow goals are also where the clips see least. Draisaitl’s overtime winner in Game 1 of the 2025 Final was one of those slow goals, and his trademark: a power-play one-timer, by our tracking, with 31 seconds left in overtime. NHL.com’s recap traced the puck “from Evan Bouchard to Ryan Nugent-Hopkins to Corey Perry to Connor McDavid and finally across to Draisaitl.” Draisaitl called it “Four good plays in a row.”

That is one goal. The team and player habits above are built from these same labels, so it matters how often they are right. We checked them against the NHL’s own scoresheet and play-by-play for every goal. The list below gives what the checks found, and the other things the clips cannot do.

• The clips catch nearly every goal. 25,347 of the 25,597 goals in 2023-24 to 2025-26, not counting shootouts, have a clip (99.0%). 2023-24 is the thinnest season, at 98.0%.
• When the clip names a setup man, he is usually the right one. He is the player the NHL credited with the first assist (usually the last teammate to pass before the goal) 90.2% of the time, about nine times in ten. But on 38.0% of assisted goals, about two in five, our rules find no teammate passing to the scorer in the clip. The pass may have come during a line change, in a scramble, or before the clip started. So our pass counts miss passes, and the true counts can only be higher.
• Tips are hard to see. Only 39.4% of the shots the NHL records as tip-ins or deflections read as a tip in the tracking, because the tracking follows a player’s body, not his stick. That is why our deflection label also uses the NHL’s own shot type.
• Rebounds mostly agree. Our rebound flag matches the NHL’s play-by-play, its official log of every event, on 86.1% of goals. Most goals are not rebounds, so two sources would agree on most goals even by chance, just by both saying “not a rebound”. That makes the match look better than it is: they still disagree on about one goal in seven.
• Shot speed is rough. At ten frames a second, a puck crossing 15 feet shows up in only two or three frames, so any one goal’s shot speed can be well off. We do not lean on it.
• Empty-net goals are out. With nobody in net, the entry clock and the shot distance mean something else, so the 1,583 empty-net goals are left out of every figure.

Continues below

• The clips are short and new. Each runs about 12 seconds in 2023-24 and 14 seconds after, and usually starts only about eight seconds before the shot, so long possessions get cut off, and the clip begins mid-play. They start in 2023-24, and three seasons cannot show a long-run trend.
• Every label is ours. The rules are our own, not the NHL’s, and beyond spot checks like Draisaitl’s goal there is no independent record of the right answer.
• The clips exist only for goals. They show how goals went in, not how often a kind of chance works: Hyman’s crease goals say nothing about how often his crease chances go in.
• A player’s shares include his role. The odd-and-even test cannot tell a player’s talent apart from how his coach uses him.

The last Cup’s MVP scored from the doorstep

The last Cup in the sample ended the way this piece began, with both speeds of goal. On June 14, 2026, in Las Vegas, Taylor Hall opened the scoring in Game 6 of the Final the fast way, 2.4 seconds after the puck crossed the Vegas blue line. Carolina won 3-0 and took the Stanley Cup.

The Conn Smythe Trophy went to Jordan Staal, the Hurricanes captain, who at 37 became the oldest playoff MVP since the trophy was first awarded in 1965, NHL.com noted. He scored six goals in the Final, most of them the way Hyman scores. Five of them have a clip our rules could read cleanly. One came off a rush in Game 1, from 30 feet. The other four came with the puck already in the zone when the clip begins, and all four were released within 11 feet of the goal line. Across his playoff run, three of his cleanly tracked goals were tip-ins.

Two Junes earlier, Reinhart had won Florida a Cup with a shot from 36 feet, 2.3 seconds after the puck crossed the blue line. Staal won his from in close, on pucks that had been in the zone longer than the clip could count.

Show the work

Tests written down

Not in advance (exploratory)

Results computed

October 27, 2026

Every goal with an NHL EDGE tracking clip, 2023-24 to 2025-26 regular season and playoffs, labelled by one fixed set of rules (analyzer ga-1.0.0) and joined to the NHL play-by-play. Of 25,347 tracked goals we set aside 1,583 empty-net goals and 371 penalty shots and unreadable clips, leaving 23,393 goals into a guarded net. Entry is the last time the puck crossed the attacking blue line before the shot; when the puck is already in the zone at the first frame, the goal counts as 6+ seconds if the clip shows at least 6 seconds of zone time. Play types are assigned in a fixed order: a deflection (a fast puck redirected with a small turn, or a shot the NHL records as a tip-in or deflection), then a rebound (a fast puck reached the goalie within 3 seconds before the shot), then off a faceoff (within 5 seconds of a draw), dump and chase, a rush (shot within 6 seconds of entry, odd-man when attackers outnumber defenders goal-side of the puck), a turnover in the zone (the attackers took the puck off a defender in the zone in the 8 seconds before the shot), a point shot (from 40+ feet, near the blue line), and otherwise a cycle or settled play. Each figure is checked season by season, and the first 1,066 goals of 2026-27 are a holdout never used for the headline numbers.

Continues below

Team and scorer style use the same two tests: the observed spread of shares against the spread chance alone would give at each team’s or player’s goal count (pooled three seasons), and split-half correlation between odd- and even-numbered games. Scorers are credited from the play-by-play; the spread test uses the 51 players with 80+ goals, the split-half players with 30+ goals in each half (86). Named goals come from the same clips joined to the play-by-play; Staal’s Game 5 goal in the 2026 Final has no clean release read and is not in the sample. “Into a guarded net” means with a goalie in net. The 22,865 goals in the entry-time chart are the 23,393 less the 528 whose clip shows under 6 seconds of zone time and no entry, so their entry time cannot be read. The median is the middle goal: half are closer (or faster), half farther. Shooter speed is the tracked speed of the scorer over the half-second before release, and defender distance is the nearest defender to the shooter at the release frame; both are geometry from the clip and have not been checked against any official figure. Team spread, with the arithmetic: pooled over three seasons, team one-timer shares spread by 2.2 points (one standard deviation); luck alone, at each team’s goal count, would spread them by 1.4.

Continues below

Luck and real differences add like the two short sides of a right triangle, not like plain numbers, so the real spread is the square root of 2.2² minus 1.4², about 1.7 points, not 2.2 minus 1.4. A correlation (r) measures how closely two lists move together: 1 is lockstep, 0 is no link; with 32 teams, chance alone lands within about ±0.4 of zero. Team one-timer share year to year: r = -0.07, 0.42, -0.16 for 2023-24 to 2024-25, 2024-25 to 2025-26, and 2025-26 to the start of 2026-27 (each team has only 20 to 47 goals so far in 2026-27). Rush share: 0.48, 0.27, -0.01. Two seasons combined (2023-24 and 2024-25) against 2025-26, the team chart: r = 0.44, 32 teams. Split-half, a team’s one-timer share in odd-numbered games against even-numbered games of the same three seasons (same roster on both sides): r = 0.48. Splitting by odd and even game numbers is a fair test because it deals games into two piles that share players, coach and seasons, so only luck differs between them. Scorers: among the 51 players with 80+ goals, one-timer shares spread by 9.3 points where chance alone would give 4.0, leaving about 8.4 real. Split-half one-timer r = 0.67 across 86 players with 30+ goals in each half; chance alone would land within about ±0.2 of zero. Rebound-or-tip share repeats at r = 0.71 on the same split, release inside 10 feet at 0.67, rush share 0.51.

Continues below

These correlations say the habits are real; they don’t say how big any one player’s true rate is. The primary assister is the player the NHL credits with the first assist; the setup-man check compares him with the primary assister our rules predict from the tracked passes, on the assisted goals where the rules name one. The rebound flag's chance agreement (about 70%) is what the two sources would share if each flagged rebounds at its own rate with no link between them. The playoff lean (rush share 28.1% against 30.8%) is about two standard errors on the pooled gap, with about 450 playoff goals a year. Why Draisaitl’s 2025 Final Game 1 winner reads as a turnover with no passes: the rules give the puck, frame by frame, to the nearest skater; a change between teammates is a pass and a change between teams is a turnover. The final attacking chain starts after the last moment a defender had the puck, and passes are counted only inside that chain. A turnover in the zone is a change from a defender to an attacker in the zone within 8 seconds of the release, and in the fixed order it outranks the point shot and the cycle. So a moment with a Florida skater nearest the puck late in that passing play ends the chain there, leaves no passes in it, and labels the goal a turnover in the zone.

Continues below

Frequently asked questions

More in The Clock

• Verdict: Data story

  With Both Goalies In, NHL Third Periods Now Score 13% Less Than Second Periods

• Verdict: Data story

  Scoring First Wins Two NHL Games in Three Again, Now That Ties Are Gone

• Verdict: Data story

  NHL Teams Down a Goal Late Still Tie About 40% of the Time

All How Hockey Changed research →

Keep reading

• How Hockey Changed: NHL data stories →One chart-led story a day about how the NHL game, its players and its rules have changed.
• NHL goal replays →Last night’s goals drawn from NHL EDGE tracking, with entry, passes and shot speed.
• The power play is disappearing →Power-play chances halved since 2005-06, and the scoring rebound came at even strength.
• Shots on goal fell. Shooting didn’t. →NHL games have eight fewer shots on goal than in 2017-18, yet scoring rose.

Sources and method

1. NHL EDGE: puck and player tracking (goal clips 2023-24 on). Tracking-derived via NHL EDGE.
2. ESPN, Greg Wyshynski: Panthers edge Oilers in Game 7, win first Stanley Cup title (June 24, 2024)
3. NHL.com, Dan Rosen: Draisaitl goal in OT lifts Oilers past Panthers in Game 1 of Stanley Cup Final (June 5, 2025)
4. TSN / The Canadian Press: Marchand scores double OT winner as Panthers even Cup Final in Game 2 (June 6, 2025)
5. NHL.com, Derek Van Diest: Hyman still thriving as net-front presence for Oilers (Dec. 7, 2023)
6. NHL.com: EDGE stats behind Staal winning Conn Smythe Trophy with Hurricanes (June 15, 2026)
7. NHL.com: NHL EDGE stats leaders for 2026 Stanley Cup Final (June 16, 2026)
8. AP photos via WJAR: Hurricanes blank Golden Knights 3-0 in Game 6 and win Stanley Cup (June 14, 2026)
9. NHL stats (NHL.com), every season since 1917-18
10. NHL API reference (community documentation of the stats endpoints)

Research and analysis, not betting advice. Past results do not guarantee future returns; bet only what you can afford to lose.