His repeatable signature. A step only appears if its reliability score (sample size × match coverage) reaches at least Medium confidence — pattern mining, not anecdotes.
RECEIVE
Right Half-Space, Attacking third — between the lines — mostly in the attacking third (60% of his 35 passes)
54% completion · 35× in 4/4 matches
↓
TURN / BEAT MAN
face forward in the pocket
75% won · 4× in 2/4 matches
↓
FINISH
arrive and shoot from central zones
1 goals · 1.29 xG · 9× in 3/4 matches
Signature: Receive → Turn / Beat Man → Finish — the only steps proven repeatable this season (3 of 4 in the full pattern). Not enough evidence yet for the rest of the chain below.
Not a pattern: create — 0× (insufficient confidence, hidden — not enough evidence yet) — omitted from the signature rather than presented as one.
In possession
In possession
What he does with the ball: build-up, progression, creation, carrying and finishing.
0 key passes (0.00/90), 0 of them assists — where his chance creation specifically comes from, 18 zones (6×3), showing his share of key passes originating there.
Chance creation — key pass origin zone share
Filter:
chance creation (0)assist (0)% of the selected group's key passes originating in each zone
03
Where does he become vulnerable?
17 possessions lost. Directness has a cost — the question is where. Attacking-third losses are cheap; deep ones are dangerous.
8
attacking third (cheap)
6
middle third
3
own third (dangerous)
Vision AI 47% of losses are in the attacking third — he gambles where a turnover costs little.
Mistake danger (proxy): 3.81 danger-index/90 · 0.287 avg/turnover. Weights each loss by threat conceded at that location. True xT-conceded needs the opponent possession model; positional proxy.
24 defensive actions across the season (18.8/90). Where he wins the ball, and how — the core question for a defender. Aerial duels are not recorded for this competition, so aerial dominance is not measured here.
click a card to isolate that action type · own goal left, attacking →
3
tackles · 33% won
0
interceptions · 0.0/90
8
recoveries · 6.3/90
3
clearances · 2.3/90
1
blocks
52
avg action height
7
won high (att 3rd)
10
middle third
7
own third
Vision AI A front-foot / high defender. He wins the ball mainly by stepping in — tackles, interceptions & recoveries outweigh clearances. 11 proactive ball-wins (8.6/90), tackle success 33%. He defends high, stepping out to intercept — suits a high line.
Defensive work in the attacking half
Ball-wins in the attacking half only (own half vs. attacking half, a simple 2-way split, not the usual thirds) — the clearest single signal of how much he presses and recovers the ball high up the pitch.
14
attacking-half actions · 10.96/90
Team
Team context
How much of the team runs through him, and what changes without him.
Hull City's output per match with Joe Gelhardt (4 matches) versus without him (4 matches). Insufficient evidence — the without-sample is only 4 matches; treat every Δ below as noise until more games exist. A proper on-off model needs many more minutes.
WITH (4)
WITHOUT (4)
Δ
goals for /match
1.00
1.25
-0.25
xG for /match
1.19
0.00
+1.19
goals against /match
1.50
0.50
+1.00
points /match
0.50
2.00
-1.50
06
His Share of the Team
Of everything Hull City produced this season, the categories where the biggest share ran through Joe Gelhardt specifically — his share of the team total, not a per-90 rate.
category
share
him / team
expected goals (xG)
7%
1.3 / 17.9
shots
5%
9 / 175
successful take-ons
2%
3 / 136
Recruitment
Style and verdict
Style profile, closest comparables, and the summary.
Joe Gelhardt is an attacking mid: 0.00 key passes /90, 0.8 progressive passes, 1 goals from 1.29 xG. A between-the-lines creator. [Off-ball movement into pockets needs tracking.]
Based on 4 matches (115 minutes) — an early read, treat with caution until more minutes are logged. He profiles as an attacking mid — best used in a system built around that identity.
✨ This report is interactive
Click any stat card to filter its map, hover a dot for the exact match, and use the dropdowns to slice passes zone-to-zone. Nothing here is static.
