Bands show position percentiles; axis labels show current rolling values.
Dashed outline: peer average in the fixed reference. Outside = above average; inside = below average. All axes point outward for stronger results.
Player value: 2.65 Peer range: 1.68 – 4.20 xGF/60
MeanDylan Cozens
Lower Higher (better)
Player 2.65 · mean 2.54. Higher raw values rank higher.
Rolling profile against a fixed reference group
Fixed rolling-profile reference · 430 players. Values use the same history and smoothing as the radar.
History, sample sizes & calculation
Higher radar percentiles mean stronger recorded results against the fixed reference, including older seasons. These are separate from this season’s league percentile. Each player has a different amount of covered history. Weighted minutes and attempts give recent games more influence. “Prior” is the share contributed by a frozen league baseline: it is higher when a player has less evidence, and differs by metric because each uses its own smoothing strength. It is not a confidence score.
On-ice offense · xGF/60
1182.2 weighted minutes · 14% prior
On-ice defense · xGA/60
1182.2 weighted minutes · 25% prior
Chance generation · ixG/60
1182.2 weighted minutes · 14% prior
Shot quality · ixG/attempt
215.6 weighted attempts · 10% prior
Finishing · goals above expected/60
215.6 weighted attempts · 16% prior
History through Oct 17, 2026. Recent games count more than older games. History carries across seasons and stays unchanged while inactive. On-ice results are shared; finishing describes results rather than established talent.
Recent scoring against his own season
Recent scoring against his own season
Loading recent games…
Loading splits…
Loading linemates…
Loading playoff games…
Loading game log…
Loading career ledger…
Every metric name opens its definition; the full list lives in Metrics.
Traditional statistics cover all situations. Supporting five-on-five analyses use completed games, raw adjustment and exposure from xG model v4. The two samples can differ.
An em dash means unavailable, not zero. A partial season remains partial. Where a panel reports evidence coverage, its rates use only the covered games. History never substitutes a different model, and a missing appearance interrupts a rolling model window.
Shot maps use the same shot model as Game Center and show which model version scored recent shots beside the rink. Older seasons can use earlier model versions. High danger means a model scoring probability of at least 0.20 on an unblocked attempt; for goalies, medium danger is 0.08 to 0.20 and low is below 0.08.
Special-team buckets include the wider recorded man-advantage and man-down states, read from the player’s own team; for a goalie, penalty kill means facing a power play. Their ranks need 100 recorded minutes and at least 30 qualified peers; below 100 minutes, penalty-kill percentiles are provisional. Playoff values have no ranks or percentiles. Team share is player ice time divided by team ice time in that situation, over the games the player dressed. Zone starts describe faceoff-started five-on-five shifts and exclude on-the-fly starts.
Penalties count events, including one for a double minor; bench and served-by roles are excluded, and victimless infractions have no drawn role. EPV added sums changes in the probability that the possessing team scores next; it is not standings points or an isolated player grade. Rolling windows never cross seasons or season types, and a missed game is not a zero-performance appearance. Career rates use summed counts and exposure, never an average of season rates; a missing contributing season withholds its career metric, and Performance begins in 2023–24.
Elite 0.94 individual expected goals per 60 (16th of 220)
Elite 0.94 individual expected goals per 60 (16th of 220)
58.6%7th of 220
Elite 58.6% on-ice expected goal share at five-on-five (7th of 220)
Elite 58.6% on-ice expected goal share at five-on-five (7th of 220)
Elite 2.01 on-ice expected goals against per 60 (9th of 220)
Elite 2.01 on-ice expected goals against per 60 (9th of 220)
Defensive metrics
Five-on-five on-ice suppression; individual events in all situations.
Suppression
Player
Percentile
Rank
48.08
6 / 22098th
1.33
71 / 20466th
0.058
59 / 22074th
Puck management
Player
Percentile
Rank
0.82
145 / 22034th
2.28
58 / 22074th
0.00
T-126 / 22043rd
Physical style
Player
Percentile
Rank
1.03
208 / 2205th
9.23
36 / 22084th
Context · not graded
Player value: 48.08 Peer range: 44.90 – 72.11 / 60
MeanDylan Cozens
Lower (better)Higher
Player 48.08 · mean 56.94. Lower raw values are better.
On-ice attempts against per 60, including blocks.
Evidence coverage: high danger 74 of 82 games.
Discipline & puck management
Recorded season event counts · all situations
Penaltiesnet 0
Taken 24Drawn 24
Puck eventsnet -34
Giveaways 53Takeaways 19
Blocks
24
all situations
Hits
215
all situations
Strong 6.67 power-play points per 60 (11th of 92)
Strong 6.67 power-play points per 60 (11th of 92)
Power play
261:02 recorded ice time · peers need 100 minutes
Share of team PP
60.0%
in games dressed
PP points
29
goals + assists
Unit xGF / 60
8.51
team without: 7.16
Power play rates
Player
Percentile
Rank
6.67
11 / 9289th
4.60
10 / 8790th
13.10
28 / 9270th
1.90
47 / 9249th
8.51
56 / 9240th
11.20
8 / 9292nd
Player value: 6.67 Peer range: 0.96 – 10.95 / 60
MeanDylan Cozens
Lower Higher (better)
Player 6.67 · mean 4.79. Higher raw values rank higher.
Recorded goals plus assists per 60 in this scope; widened power-play and penalty-kill scopes.
Evidence coverage: goals 79 of 81 games.
Penalty kill
7:09 recorded ice time · peers need 100 minutes
Share of team PK
1.8%
in games dressed
SH points
0
goals + assists
Unit xGA / 60
5.58
team without: 7.27
Penalty kill rates
Player
Percentile
Rank
5.58
—
8.55
—
83.92
—
0.00
—
4.55
—
No rank yet
Player value: 5.58 Peer range: 5.58 – 11.51 / 60
MeanDylan Cozens
Lower (better)Higher
Player 5.58 · mean 8.60. Lower raw values are better. Provisional PK percentile: under 100 recorded minutes. Smaller samples are less reliable; official ranks need 100 minutes.
Raw on-ice expected goals against per 60; the Season metrics sample.
Provisional PK percentile: under 100 recorded minutes. Smaller samples are less reliable; official ranks need 100 minutes.
Evidence coverage: goals 24 of 25 games · high danger 24 of 25 games.
17:02 TOI per game (12:51 at five-on-five · 3:11 PP · 0:05 PK)
17:02 TOI per game (12:51 at five-on-five · 3:11 PP · 0:05 PK)
Results vs expected
Five-on-five · what happened beside the chances created and allowed
Chances
Player
Percentile
Rank
2.85
36 / 22084th
2.01
9 / 22096th
58.6
7 / 22097th
Goals
Player
Percentile
Rank
2.38
93 / 17447th
2.57
92 / 17447th
48.1
89 / 17449th
Luck & variance
Player
Percentile
Rank
8.2
119 / 17432nd
88.5
152 / 17413th
96.71
150 / 17414th
Context · not graded
Player value: 2.85 Peer range: 1.49 – 3.89 / 60
MeanDylan Cozens
Lower Higher (better)
Player 2.85 · mean 2.41. Higher raw values rank higher.
Raw on-ice expected goals for per 60; the Season metrics sample.
5v5 goals
38–41
74 of 82 games
5v5 expected
50.0–35.3
xG for–against
EPV added
+8.41
82 attributed games · all situations
On ice vs off ice
Ottawa's share with Dylan Cozens on the ice against the same team without him.
Share
On ice and off ice
On-minus-off gap in percentage points
xGF%
On ice 58.6%; off ice 53.8%.
53.858.6
+4.8
CF%
On ice 56.8%; off ice 51.4%.
51.456.8
+5.4
GF%
On ice 48.1%; off ice 52.6%.
52.648.1
-4.5
On iceOff ice
Scale 40–65% · marker at 50% · right column is the on-minus-off gap in points
Steady 8 points in the last 10 games (0.80 per game vs 0.72 this season)
Steady 8 points in the last 10 games (0.80 per game vs 0.72 this season)
0.66 points per game at home · 0.78 away
0.66 points per game at home · 0.78 away
82 games · most recent vs TOR, W 3–1: 1 G, 0 A
82 games · most recent vs TOR, W 3–1: 1 G, 0 A
Dylan Cozens
294 attempts27.33 xG28 G
Goal28Shot on goal177Missed shot89Blocked shot0Larger dot = higher xG
Shooting by zone
Zone
Att
G
xG
Goal rate / exp.
G−xG
High slot
127
16
12.6
12.6% / 9.9%
+3.4
Point
33
2
1.8
6.1% / 5.4%
+0.2
Below goal line
1
0
0.0
0.0% / 1.9%
0.0
Outer slot
20
1
1.2
5.0% / 5.8%
−0.2
Left circle
9
0
0.2
0.0% / 2.2%
−0.2
Right circle
10
0
0.2
0.0% / 2.2%
−0.2
Inner slot
94
9
11.3
9.6% / 12.1%
−2.3
All zones
294
28
27.3
9.5% / 9.3%
+0.7
Sorted by goals minus expected goals. Goal rate is goals per unblocked attempt against the expected rate. Hover a row to light up those attempts on the rink.
Finishing evidence. Scoring 9.5% on chances against 9.3% expected, tracking the model across 7 zones.
Shot types294 attempts
Type
Share
Sh%
G−xG
Wrist
47.3%
16.2%
+2.8
Snap
22.1%
7.7%
−3.1
Backhand
12.2%
15.4%
+0.8
Tip-in
8.2%
18.8%
+1.1
Slap
5.1%
7.1%
−0.6
Deflected
2.7%
16.7%
+0.2
Other
2.4%
—
−0.5
Share is context, not a grade. Sh% counts shots on goal only.
High-danger chancesmodel probability ≥ 0.20
Attempts5.4% of the selected attempts
16
Goals
3
Expected goals
4.7
Conversion
18.8%29.6% expected
Shot profilecontext · not graded
On-net rate205 of 294 attempts on goal
69.7%
Slot share221 of 294 from the inner or high slot
75.2%
Average distancefrom the centre of the net
25.0 ft
Rebound attempts / 6026 this season · all situations
1.12
Usage by situation
Ice time, share of team ice time, scoring and scoring rate, with where shifts start.
Scope
TOI/GP
Team share
G
A
P
P/60
Five-on-five
12:51
26.8%
13
15
28
1.59
Power play
3:11
60.0%
13
16
29
6.67
Penalty kill
0:05
1.8%
0
0
0
0.00
Other
0:55
—
2
0
2
1.58
All situations
17:02
28.1%
28
31
59
2.53
Where the 59 points come from · All situations
Goals28 · 1.20/60
Primary assists14 · 0.60/60
Secondary assists17 · 0.73/60
28.8% of points are secondary assists. Peer average 25.1%. Primary points per 60 (1.80).
Zone starts — offensive of OZ + DZ
Offensive —Neutral —Defensive —
Shifts
23.0 a game · 0:45 average
Zone starts: five-on-five · Shifts: all situations
Faceoffs
All situations · zones are read from his side of the draw · values only; no peer ranks yet
Win %
53.2%
435 of 818 draws
Draws per game
9.98
82 games
Offensive zone
53.3%
306 draws
Defensive zone
55.2%
172 draws
Power play
56.1%
107 draws
Every game · rest counts days off before the game; the playoff field is the standings' top three per division plus two wild cards
Split
GP
G
A
P
P/GP
5v5 xG share
Team record
Venue
Home
41
14
13
27
0.66
62.1%
23–12–6
Away
41
14
18
32
0.78
55.6%
21–15–5
Rest
Back-to-back
14
8
9
17
1.21
66.1%
9–3–2
1 day of rest
50
14
18
32
0.64
57.1%
26–20–4
2+ days of rest
17
5
4
9
0.53
57.2%
8–4–5
Opponent
vs Eastern
50
18
24
42
0.84
61.2%
27–15–8
vs Western
32
10
7
17
0.53
54.6%
17–12–3
vs playoff teams
39
12
14
26
0.67
55.9%
19–12–8
vs non-playoff teams
43
16
17
33
0.77
61.1%
25–15–3
Recent
Last 10
10
4
4
8
0.80
58.1%
6–3–1
By period and score state
Split
G
A
P
5v5 xG share
5v5 CF%
5v5 GF–GA
Period
1st period
11
12
23
—
—
—
2nd period
9
8
17
—
—
—
3rd period
8
11
19
—
—
—
Overtime
0
0
0
—
—
—
Score state · 5v5 on ice
Leading
—
—
—
59.0%
53.8%
23–18
Tied
—
—
—
61.9%
57.6%
16–11
Trailing
—
—
—
53.8%
59.9%
4–14
From play-by-play; shootouts excluded. Score state is his team's, before each event.
Official totals; expected-goal columns are five-on-five.
Elite90+Strong70–89Average30–69Weak10–29Poor<10Bar length is the peer percentile, so longer is always better · tick at the 50th
Player value: 1.59 Peer range: 0.47 – 3.42 / 60
MeanDylan Cozens
Lower Higher (better)
Player 1.59 · mean 1.55. Higher raw values rank higher.
Recorded five-on-five goals plus assists per 60 minutes.
How to read these comparisons
Player values are recorded season results. Peer averages are arithmetic means of qualified players in the selected group; each row uses the same peers for its average and rank. Ranks use unrounded values, with tied players sharing a rank and the next rank skipped. Lower xG against ranks better. Bar colours are the five performance tiers; context rows are blue and not graded. Differences from the average describe the season, not isolated player impact.
Goals above expected is the raw goals-minus-xG rate, without the shrinkage used in the older finishing comparison. On-ice measures include teammates, opponents and deployment. These recorded metrics describe different quantities from the adjusted Season score components beside them.
Skaters need 20 appearances and 300 five-on-five minutes; quality and finishing also need 75 unblocked attempts. Early in a season these minimums scale down with games played so far. At least 30 peers must qualify. Playoff ranks are withheld. A reviewed axis whose own evidence fails is withheld and shown with its reason rather than ranked. These workload rules do not express statistical confidence.
Position uses the NHL listed primary position available now; it does not establish historical position or actual deployment. Broad groups are forwards, defence and goalies. This version uses the current model’s covered games only.
Season vs reference
Per game · all situations
Metric2025–26ReferenceDiff
0.340.20+0.15
0.380.380.00
0.720.57+0.15
2.502.29+0.21
Reference: 2024–25 · 82 games.
Each game against his own season
Points in each game, coloured by how good the game wasTrailing 10 games, per gameSeason 0.72
Points
8
4 G · 4 A
Team record
6–3–1
in his games
5v5 xG share
58.1%
on ice
5v5 ixG
2.0
individual expected goals
Avg TOI
16:54
all situations
Tiles cover the last 10 games.
Season grade
A
#50 of 400 qualified F88th percentile
Performance / 600.1682/82 GP
Elite defense with strong offense and chance creation — highlighted by 2.01 on-ice expected goals against per 60 (9th of 220).
7 NHL seasons: 453 games, 274 points
7 NHL seasons: 453 games, 274 points
Totals cover all situations; rates are five-on-five.
Expected goals against while the player is on the ice (fewer is better)
Chance creation
Expected goals from the player’s own shots
Finishing
Goals minus expected goals on the player’s own shots
Assisting
Primary assists plus half of secondary assists
Each is measured against expectation, per 60 rated minutes: plus is above, minus below.
How this is calculated
Score breakdown
Component
Weight
Scaled
Contribution
Offense
25%
0.209
+0.052
Defense
25%
0.286
+0.071
Chance creation
20%
0.217
+0.043
Finishing
10%
0.053
+0.005
Assisting
20%
−0.038
−0.008
Performance rate
0.165
Per 60 rated minutes, over the scored appearances. Each component is scaled separately, then weighted; the contributions add to the performance rate at full precision, so rounded figures may differ.
Scoring
Season grade is the qualified peer percentile on the same letter bands as Game Grade — A+ ≥88, A ≥78, A− ≥68, B+ ≥58, B ≥50, B− ≥42, C+ ≥35, C ≥28, C− ≥22, D ≥15, F below 15 — and needs qualification plus at least 30 eligible peers. The letter describes this season’s performance, not permanent ability.
Comparisons
Rank and component percentiles compare this season’s qualified forwards on unrounded rates over every checked appearance; Game Grade compares one appearance with frozen historical games.
Component percentiles are never averaged into the score. Recent history is optional, from up to three completed prior seasons; small samples are provisional.
Coverage
Qualification: 29 covered appearances, 492.0 rated minutes, and 90% appearance and workload coverage. Minimums advance with league progress; valid low-sample results stay visible.
Rated minutes exclude play with a goalie pulled: 45.2 min excluded; official TOI 1397.3 min.
Coverage: complete. Updated .
Method
Five weighted components add to the performance rate. Five-on-five offense and defense account for recorded teammates, opponents and game state, so shared on-ice results are not isolated individual ability. Finishing is scored against the shot model’s historical-shooter expectation and assisting counts primary assists plus half of secondary assists; both use the strength baseline, chance creation its positional baseline.
Legacy rating
Legacy season score: 82.8 / 100. The fixed exposure baseline is 50; this is separate from the season peer grade.
Most often with Ridly Greig: 503 five-on-five minutes, 57.6% xG share together
Most often with Ridly Greig: 503 five-on-five minutes, 57.6% xG share together
Five-on-five · time together from shift charts (1,101:05 of his 5v5 ice time) · xG share together and apart
Teammate
Pos
TOI together
xG share together
GF–GA
CF%
Him without
Teammate without
Ridly Greig
C
503:19
57.6%
24–17
56.5%
59.7%
45.8%
Brady Tkachuk
L
488:25
61.5%
22–17
59.6%
56.1%
56.8%
Drake Batherson
R
382:04
59.6%
11–13
57.1%
58.2%
55.4%
Jordan Spence
D
350:32
67.5%
15–12
65.6%
53.8%
59.7%
Tyler Kleven
D
306:17
62.9%
11–8
60.8%
57.0%
51.1%
Nick Jensen
D
304:43
53.0%
17–18
49.9%
60.9%
51.4%
Thomas Chabot
D
303:18
54.4%
18–20
52.7%
60.5%
55.0%
Artem Zub
D
293:43
55.5%
11–9
55.8%
59.8%
52.4%
Jake Sanderson
D
289:23
57.4%
7–9
59.8%
59.0%
56.5%
David Perron
L
256:48
54.8%
6–12
51.6%
59.6%
61.2%
Nikolas Matinpalo
D
199:00
55.7%
5–7
48.9%
59.3%
49.1%
Fabian Zetterlund
L
184:30
61.8%
9–8
58.1%
58.0%
55.7%
Without columns are five-on-five xG share apart: him on the ice without that teammate, and the teammate without him (with this club). Bars are centred on an even 50%.
