Dylan DeMelo — 2023–24 stats and season rank | IceSheet
Dylan DeMelo
Winnipeg Jets·#2·Defenseman · Shoots right
Age 33 · 6′1″ · 194 lb · London, ON, CAN
Season grade
Withheld
Incomplete data
Performance / 60−0.0119/82 GP
Games82T-2nd in NHL
Goals30.04/GP▼0.04T-526th in NHL
Assists280.34/GP▲0.06T-151st in NHL
Points310.38/GP▲0.02…
5v5 xG share…
Avg TOI21:44…
/
Loading season standing and profile…
Season comparisons
Loading peer comparisons and model evidence…
Optional recent-history profile
Rolling Radar
Loading recent-history profile…
Loading model evidence…
Loading defensive evidence…
Shot locations
Loading canonical shot events…
Usage by situation
Loading usage…
Loading special teams…
T-263rd in NHL
58th in NHL
Not qualified · 2023–24 defensemen
Playing-time minimum met. Rank withheld because the app’s scoring data are incomplete. The app has checked scoring results for 19 of 42 appearances (45.2%); season standings require at least 90%. Scoring data cover less than 90% of recorded workload or could not be verified.
vs 0 qualified defensemen
Percentiles appear once this player qualifies for season standings. The score breakdown is under How this is calculated.
How this is calculated
Score breakdown
Component
Weight
Scaled
Contribution
Offense
25%
0.292
+0.073
Defense
25%
−0.115
−0.029
Chance creation
20%
−0.431
−0.086
Finishing
10%
0.107
+0.011
Assisting
20%
0.120
+0.024
Performance rate
−0.007
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 defensemen 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: 17 covered appearances, 337.8 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: 6.8 min excluded; official TOI 415.6 min.
Coverage: partial. 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: 46.1 / 100. The fixed exposure baseline is 50; this is separate from the season peer grade.
Rolling Radar
Recent NHL history, carried across seasons. The radar compares this broader profile with a fixed position-group reference.
Rolling profile is currently unavailable.
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.
Five-on-five · time together from shift charts (1,526:59 of his 5v5 ice time) · xG share together and apart
Teammate
Pos
TOI together
xG share together
GF–GA
CF%
Him without
Teammate without
Josh Morrissey
D
1,194:48
51.5%
48–35
50.9%
50.7%
59.3%
Mark Scheifele
C
501:02
47.9%
11–16
47.6%
53.5%
50.9%
Mason Appleton
C
463:23
52.8%
30–8
50.6%
50.7%
54.2%
Adam Lowry
C
449:25
54.3%
25–8
52.6%
50.2%
56.5%
Kyle Connor
L
424:13
46.7%
9–17
45.5%
53.6%
51.8%
Alex Iafallo
L
393:13
54.6%
17–9
50.4%
50.3%
52.5%
Nikolaj Ehlers
L
389:33
56.7%
8–15
56.2%
49.5%
51.3%
Nino Niederreiter
R
367:07
53.4%
27–9
52.5%
50.7%
57.0%
Vladislav Namestnikov
C
309:04
55.1%
12–11
51.3%
50.5%
53.1%
Cole Perfetti
C
256:07
56.3%
3–9
53.2%
50.6%
53.3%
Morgan Barron
C
246:44
47.1%
10–8
47.4%
52.1%
52.7%
Gabriel Vilardi
C
238:29
44.0%
2–11
46.8%
52.9%
52.5%
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%.
2023–24 playoffs
Official playoff totals · no peer ranks, since playoff samples are too small for a fair comparison
Games
5
22:24 per game
Goals
0
0 on the power play
Assists
1
official
Points
1
0 power-play points
Plus-minus
-1
even strength and shorthanded
Shots
5
0.0% shooting
Playoff games
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: 1.19 Peer range: 0.00 – 1.84 / 60
MeanDylan DeMelo
Lower Higher (better)
Player 1.19 · mean 0.85. 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
Metric2023–24ReferenceDiff
0.040.08-0.04
0.340.28+0.06
0.380.36+0.02
1.011.24-0.23
Reference: 2022–23 · 75 games.
Each game against his own season
Points in each game, coloured by how good the game wasTrailing 10 games, per gameSeason 0.38
