Real data. Everything on this page is computed from actual games.
Spare Line Change
Numbers from
Hits and takeaways
final

PHI2

Philadelphia Flyers

NJD3

New Jersey Devils

Thursday, October 1

Final in overtime · Prudential Center

The short version

What happened

Hughes wins it in overtime

Rasmus Ristolainen opened the scoring midway through the first. Twice the Devils answered a Flyers lead. Jack Hughes won it in overtime.

The score was fair. Replay this game’s chances a hundred times and the Devils win nine in ten. The Flyers tried a line they hadn’t used before: Matvei Michkov, Noah Cates and Tyson Foerster.

What it means

  • Keep an eye onWhether the Flyers keep that line together at home against Carolina on Saturday.
  • The previewBoth of the preview’s calls held.

Our pre-game read

The win chance our model worked out before the game, kept off the site until now.

We gave Philadelphia only a 36% chance, and it went the way we feared.

Two even teams on neutral ice50%
Home ice−4 points46%
Recent chances+2 points48%
Who's been playing−4 points44%
Back-to-back−8 points36%
Our chance for Philadelphia the morning of the game36%
Tonight's lineup−5 points
At puck drop, once the lineups were known30%

Recent chances: how much better each team’s chances have been than their opponents’ lately, with the latest games counting most. Who’s been playing: whether the players who dressed last time were stronger or weaker than the ones behind those recent games, so a star who is out or back counts straight away. Back-to-back: playing for the second night in a row. Tonight’s lineup: any change in who dressed since the last game. Only earlier games went into it, and the preview hinted at it in words without showing the number.

We gave New Jersey a 64% chance, and New Jersey delivered.

Two even teams on neutral ice50%
Home ice+4 points54%
Recent chances−2 points52%
Who's been playing+4 points56%
Back-to-back+8 points64%
Our chance for New Jersey the morning of the game64%
Tonight's lineup+5 points
At puck drop, once the lineups were known70%

Recent chances: how much better each team’s chances have been than their opponents’ lately, with the latest games counting most. Who’s been playing: whether the players who dressed last time were stronger or weaker than the ones behind those recent games, so a star who is out or back counts straight away. Back-to-back: playing for the second night in a row. Tonight’s lineup: any change in who dressed since the last game. Only earlier games went into it, and the preview hinted at it in words without showing the number.

What the preview said, and what happened

Of the 2 calls the preview made, 2 held up, 0 partly did, and 0 didn’t happen.

  1. Before the game

    The Flyers roll four lines; the Devils lean on their top six

    On depth, Philadelphia is one of the league’s best and New Jersey is one of the league’s worst. That’s about as far apart as two teams get.

    The call The Flyers spread their forward ice time deeper down the lineup than the Devils.

    Depth: Philadelphia 3rd, New Jersey 31st

    Held up

    On bottom-six share of forward ice time, Philadelphia had 42% and New Jersey 39%. Coming in, Philadelphia was at 43% and New Jersey at 40%.

    PHI 43% · NJD 40%
    Coming in
    PHI 42% · NJD 39%
    In this game

  2. Before the game

    The Flyers hit; the Devils mostly don’t

    On physicality, Philadelphia is one of the league’s best and New Jersey is one of the league’s worst. That’s about as far apart as two teams get.

    The call The Flyers hit more than the Devils in the close stretches of the game.

    Physicality: Philadelphia 4th, New Jersey 31st

    Held up

    On hit score in close games (arena-adjusted, 100 is league average), Philadelphia had hit score 124 and New Jersey hit score 88. Coming in, Philadelphia was at hit score 116 and New Jersey at hit score 85.

    PHI hit score 116 · NJD hit score 85
    Coming in
    PHI hit score 124 · NJD hit score 88
    In this game

Each call is checked by a rule set before the game. Held up: the team reached their usual level. Partly: they fell short of their usual level but still beat the league average. When a call says one team will come out ahead, held up means they won by at least half their usual gap, and partly means they won by less. One game is a small sample, so a miss doesn’t mean the preview was wrong about the team.

Read the full preview as it stood at puck drop.

This season so far: 94 held up, 12 partly, 64 didn’t happen, out of 170 graded calls. See the track record.

Run it back

We replayed this game’s chances 100 times. Here is how often each team came out ahead.

New Jersey earned this one

Played out 100 times with these chances, New Jersey wins 90.

10%
Philadelphia: 6 wins, plus half of the 8 ties
90%
New Jersey: 85 wins, plus half of the 8 ties

Philadelphia had 28 chances worth 1.2 expected goals. New Jersey had 54 chances worth 3.9 expected goals.

How it works: every shot in regulation that was on target or missed the net goes in as often as our expected-goals model says it should, and the game is played out 100 times that way. A rebound scramble counts as one chance, and empty-net shots are left out. A replay tied after regulation counts as half a win each, since overtime and shootouts are close to coin flips.

It doesn’t know about goalies, finishing skill or passes before a shot. It also favours the team that was chasing, because a team with a lead tends to sit back. Over the last three seasons, teams it gave 80% or more won about two games in three.

What the preview didn’t see coming

The parts of this game furthest from how these teams usually play, biggest first.

Broke the pattern: a team did the opposite of their usual habit. Further than usual: a usual habit, taken much further. Closer than usual: a usual strength or weakness that mostly disappeared. Out of character: a usually average team, far from average. Each is measured against the teams’ numbers before the game.

The numbers

How the game went, beyond the score.

PhiladelphiaNew Jersey
Goals23
Shots on goal2544
Expected goals1.84.5
Shot attempts at 5-on-53472
Power plays0 of 10 of 4

Swing

How far each player swung tonight’s result, for his team or against it.

The biggest swings: L. Hughes +1.2, D. Hamilton +1.1 and J. Bratt +0.8.

Showhighlighting

The number is roughly in goals: what he added to his team tonight, or cost it. Tap a player to see where it came from. The label beside it is his report card from “Did they do their job?” below; L1 to L4 and P1 to P3 are tonight’s lines and pairs.

How Swing is worked outHide how Swing is worked out

Scoresheet: each goal is shared 40% to the scorer, 35% to the first assist and 25% to the second, plus a quarter of the expected goals of his own shots. A penalty drawn adds 0.16 and one taken costs 0.16, what a power play is worth on average; each faceoff won or lost moves it by 0.006.

5-on-5 counts most: expected goals for minus against with him on, half compared with league average (adjusted for score and venue) and half compared with his own team when he was on the bench, plus a smaller amount for actual goals. A player on a losing team can score well if his line held its own. Special teams: power-play and penalty-kill time compared with a league-average unit.

Company kept (counted inside 5-on-5): from every player’s 5-on-5 impact over the two seasons before, a player out with strong linemates against weak opponents gives a little back, and the reverse earns a little. Goalies: goals saved above expected. Empty-net shots and shootouts are left out. One game is a small sample: this says who swung this result, not who is better.

Did they do their job?

What each player is relied on for or better than most at, and whether he delivered.

Show
  • T. SanheimMixed night

    A goal, 1 assist. A good scoresheet, but his usual jobs went quiet.

    Defense
    25:47
    G
    1
    A
    1
    +/−
    0
    xGF–xGA
    0.5–1.3
    Shots
    3
    Hits
    0
    Blocks
    3
    • ✕Kills penalties Relied onexpected goals per 60 allowed on the penalty kill7.7typical 6.5
    • ≈Creating chances at 5-on-5 Better than mostexpected goals per 60 with him on at 5-on-51.6typical 2.2
  • C. YorkRough night

    No points, and 83% of the 5-on-5 chances went the other way with him on.

    Defense
    24:01
    G
    0
    A
    0
    +/−
    −1
    xGF–xGA
    0.2–1.1
    Shots
    1
    Hits
    2
    Blocks
    3
    • ✕Power play Relied onexpected goals per 60 on the power play0.0typical 7.1
    • ✕Kills penalties Relied onexpected goals per 60 allowed on the penalty kill7.1typical 6.3
    • ≈Preventing chances at 5-on-5 Better than mostexpected goals per 60 allowed with him on at 5-on-53.6typical 2.1
  • R. RistolainenMixed night

    A goal from a player who scores in about 1 game in 26, so it counts for more.

    Defense
    22:44
    G
    1
    A
    0
    +/−
    +1
    xGF–xGA
    0.5–1.2
    Shots
    4
    Hits
    2
    Blocks
    3
    • ✓Kills penalties Relied onexpected goals per 60 allowed on the penalty kill3.3typical 6.2
    • ✕Preventing chances at 5-on-5 Better than mostexpected goals per 60 allowed with him on at 5-on-53.7typical 2.0
  • C. DvorakRough night

    No points, and 72% of the 5-on-5 chances went the other way with him on.

    Center
    20:45
    G
    0
    A
    0
    +/−
    0
    xGF–xGA
    0.4–0.9
    Shots
    1
    Hits
    1
    Blocks
    2
    • ✓Kills penalties Relied onexpected goals per 60 allowed on the penalty kill5.7typical 6.8
    • ✕Faceoffs Better than mostfaceoffs won8 of 18typical 56%
    • ✕Getting to dangerous spots Better than mostexpected goals from his own shots at 5-on-50.04typical 0.16
  • J. DrysdaleRough night

    No points, and 84% of the 5-on-5 chances went the other way with him on.

    Defense
    20:39
    G
    0
    A
    0
    +/−
    −1
    xGF–xGA
    0.2–1.2
    Shots
    1
    Hits
    0
    Blocks
    6
    • ✕Power play Relied onexpected goals per 60 on the power play3.6typical 6.9
  • T. FoersterRough night

    No points, and 73% of the 5-on-5 chances went the other way with him on.

    Right wing
    18:50
    G
    0
    A
    0
    +/−
    −1
    xGF–xGA
    0.3–0.9
    Shots
    2
    Hits
    1
    Blocks
    1
    • ✕Power play Relied onexpected goals per 60 on the power play3.8typical 7.1
    • ✕Kills penalties Relied onexpected goals per 60 allowed on the penalty kill7.6typical 6.3
    • ✕Preventing chances at 5-on-5 Better than mostexpected goals per 60 allowed with him on at 5-on-53.7typical 2.1
Show 11 more Philadelphia playersHide the other Philadelphia players
  • T. ZegrasRough night

    No points, and 87% of the 5-on-5 chances went the other way with him on.

    Center
    17:12
    G
    0
    A
    0
    +/−
    0
    xGF–xGA
    0.2–1.1
    Shots
    1
    Hits
    0
    Blocks
    0
    • ✕Power play Relied onexpected goals per 60 on the power play3.2typical 7.0
    • ✕Creating chances at 5-on-5 Better than mostexpected goals per 60 with him on at 5-on-50.7typical 2.5
  • S. CouturierOff night
    Center
    17:02
    G
    0
    A
    1
    +/−
    +1
    xGF–xGA
    0.3–0.6
    Shots
    2
    Hits
    1
    Blocks
    0
    • ✕Kills penalties Relied onexpected goals per 60 allowed on the penalty kill8.2typical 6.6
    • ≈Faceoffs Better than mostfaceoffs won7 of 13typical 55%
    • ✕Creating chances at 5-on-5 Better than mostexpected goals per 60 with him on at 5-on-51.1typical 2.2
  • T. KonecnyDid his job
    Right wing
    16:59
    G
    0
    A
    1
    +/−
    −1
    xGF–xGA
    0.3–0.7
    Shots
    1
    Hits
    4
    Blocks
    2
    • ✓Hitting Better than mosthits4typical 2
    • –Power play Relied onToo little time to judge
  • N. AcciariDid his job
    Center
    16:36
    G
    0
    A
    0
    +/−
    0
    xGF–xGA
    0.2–0.8
    Shots
    0
    Hits
    1
    Blocks
    2
    • ✓Kills penalties Relied onexpected goals per 60 allowed on the penalty kill5.5typical 6.6
    • ✓Blocking shots Better than mostblocked shots2typical 1
    • ≈Faceoffs Better than mostfaceoffs won7 of 13typical 55%
  • O. TippettMixed night

    No points, and 79% of the 5-on-5 chances went the other way with him on, though he still did his usual jobs.

    Right wing
    16:05
    G
    0
    A
    0
    +/−
    0
    xGF–xGA
    0.3–1.1
    Shots
    2
    Hits
    4
    Blocks
    1
    • ✓Hitting Better than mosthits4typical 3
    • ✕Preventing chances at 5-on-5 Better than mostexpected goals per 60 allowed with him on at 5-on-54.9typical 2.1
    • ✓Shooting Better than mostshot attempts at 5-on-54typical 4
  • N. CatesOff night
    Left wing
    15:12
    G
    0
    A
    0
    +/−
    −1
    xGF–xGA
    0.3–0.4
    Shots
    2
    Hits
    0
    Blocks
    1
    • ≈Kills penalties Relied onexpected goals per 60 allowed on the penalty kill5.9typical 5.6
    • ≈Preventing chances at 5-on-5 Better than mostexpected goals per 60 allowed with him on at 5-on-52.3typical 2.0
    • ✕Faceoffs Relied onfaceoffs won6 of 13typical 47%
  • N. SeelerDid his job
    Defense
    14:43
    G
    0
    A
    0
    +/−
    0
    xGF–xGA
    0.3–0.6
    Shots
    1
    Hits
    1
    Blocks
    5
    • ✓Kills penalties Relied onexpected goals per 60 allowed on the penalty kill5.1typical 6.1
    • ✓Blocking shots Better than mostblocked shots5typical 1
    • ✓Hitting Better than mosthits1typical 1
  • S. BenoitDid his job
    Defense
    13:29
    G
    0
    A
    0
    +/−
    0
    xGF–xGA
    0.3–0.7
    Shots
    1
    Hits
    2
    Blocks
    0
    • ✓Hitting Better than mosthits2typical 1
    • –Kills penalties Relied onToo little time to judge
  • M. MichkovOff night
    Right wing
    13:24
    G
    0
    A
    0
    +/−
    −1
    xGF–xGA
    0.3–0.3
    Shots
    1
    Hits
    0
    Blocks
    2
    • ✕Power play Relied onexpected goals per 60 on the power play3.2typical 7.0
  • P. MartoneRough night

    No points, and 81% of the 5-on-5 chances went the other way with him on.

    Right wing
    12:27
    G
    0
    A
    0
    +/−
    +1
    xGF–xGA
    0.2–0.9
    Shots
    1
    Hits
    0
    Blocks
    0
    • ✕Power play Relied onexpected goals per 60 on the power play3.2typical 6.8
    • ✕Preventing chances at 5-on-5 Better than mostexpected goals per 60 allowed with him on at 5-on-55.1typical 1.9
  • C. GrundstromDid his job
    Right wing
    8:30
    G
    0
    A
    0
    +/−
    +1
    xGF–xGA
    0.3–0.4
    Shots
    1
    Hits
    2
    Blocks
    1
    • ✓Hitting Better than mosthits2typical 2
    • ≈Preventing chances at 5-on-5 Better than mostexpected goals per 60 allowed with him on at 5-on-53.2typical 2.2
How the cards are gradedHide how the cards are graded

✓ a typical night for him in the ice time he got (the level he reaches in about half his games). ≈ beat an average player at his position, a typical power play or kill, or his own team with him off the ice. ✕ neither. Power-play and penalty-kill results count half, since five players share them.

Huge night: a hat trick or four or more points. Big night: two or more points, or a goal from a player who scores in fewer than one game in five on top of doing his jobs. Rough night: no points while the other team had 70% or more of the 5-on-5 chances with him on (xGF–xGA). Plus-minus never sets the level. A shorthanded goal counts as two points. Points never fully cover for jobs that went badly: with his jobs off, a huge night reads big and a big night reads mixed.

Judged only from games before this one; 5-on-5 rates adjusted for score and venue. One game is a small sample: even a strong player misses a typical night about half the time. Both teams: 36 ✓, 10 ≈, 27 ✕.

Somewhere new

Things they aren’t known for, far beyond their usual pace.

A night this big would happen less than one game in fifty at his usual pace, and is at least twice what an average player at his position would manage.

In net

Each goalie’s night: the goals he stopped beyond what the shots were worth, and the dangerous and routine saves.

  • J. WollBig night

    41 saves on 44 shots, +1.5 goals saved above expected.

    Goalie
    62:49
    Shots
    44
    Saves
    41
    Sv%
    .932
    xGA
    4.5
    GSAx
    +1.5
    Dangerous
    4/6
    • ✓Saving more than expected goals saved above expected+1.5typical +0.2
    • ✓Dangerous shots saves on shots with a 15%+ chance4 of 6typically saves 4 of 6
    • ✓Routine shots saves on shots under a 15% chance37 of 38typically saves 35 of 38
  • J. AllenOff night
    Goalie
    63:23
    Shots
    25
    Saves
    23
    Sv%
    .920
    xGA
    1.8
    GSAx
    −0.2
    Dangerous
    1/1
    • ✕Saving more than expected goals saved above expected-0.2typical +0.3
    • –Dangerous shots Too few shots to judge
    • ≈Routine shots saves on shots under a 15% chance22 of 24typically saves 23 of 24

Every goalie has the same three jobs, judged against his own earlier starts (✓) and an average goalie (≈). Huge night: a shutout on 25 or more shots, or 3+ goals saved above expected. Big night: any other shutout, or 1.5+. Rough night: 2+ goals worse than expected, or pulled early after three or more.

Empty-net shots are left out. Expected-goal models can’t see passes before a shot, so a goalie beaten by cross-ice passes looks worse here than he really was.

Team pages: Philadelphia Flyers · New Jersey Devils