Goalkeeper PerformanceGK SUITE · 6 DIMENSIONS

Season window:1 Jul – 6 Oct 2026Keepers:4Matches:12Saves:96
← Menu
DEMO · SYNTHETIC
GOALKEEPERS
Window
Comp
GK PROFILEsix-dimension percentile fingerprint vs GK peer group
1
KIM S.
Available
Shot-stoppingPSxG · Goals Prevented · save map
⊞ Shot-stopping model
SaveGoal concededpost-shot xG weighted
Distributionbuild-up · launch · under pressure
⊞ Distribution plan
Sweeping & positioningactions outside box · distance from goal
⊞ Sweeper role
Aerial & claimingcrosses · claims · command of area
⊞ Cross management
GK training loadGK-LAB · dives · GK Load · power · impact — ported & extended from Panetolikos
⊞ Load → readiness
DIVE VOLUME · LEFT vs RIGHT
GK LOAD · LAST 8 SESSIONS
SESSION DETAIL
SessionDives LDives RGK LoadAvg Power (W)Dive Impact (g)Distance (km)HML (m)
GK physical testingreaction · dive reach · explosiveness · agility
⊞ GK testing battery
GK-specific battery feeds the Athlete Profile (Layer-1). Reaction & dive reach are GK-only; CMJ, 505 COD and agility share the outfield thresholds (asymmetry >10% flag, LSI ≥90% for RTP).
Benchmarkingpercentile vs GK peer group · white line = elite norm
⊞ Benchmarks & thresholds
Readiness (today)HRV · sRPE · sleep · wellness · ACWR
flags tests taken in a red state
Korea GK Performance Suite · Layer feeds Daily Report & Player Status · DEMO — synthetic data, no real player values.

Shot-stopping model

How we read save quality — beyond raw Save %.

Post-Shot xG (PSxG) is the probability each on-target shot becomes a goal, given placement, power and context. Goals Prevented = PSxG − Goals Against. Positive = the keeper saved more than an average GK would from those shots.

1This keeper
—
2Save map read
Cluster the goals conceded by zone: top-corner goals are low-blame (high PSxG); central / near-post goals at mid-height are coachable. Weight review by PSxG, not by count.
3Session focus
Reactive-save blocks for the high-PSxG zones that still leaked; footwork & set-position drills for the low-PSxG (coachable) concessions. Re-shoot the save map every camp.

Elite reference: Save % ~75–80%, Goals Prevented positive across a season. Small samples (one camp) are noisy — read the trend.

Distribution plan

Build-up role tuned to the game model & the opponent press.

Two profiles: a build-up keeper (high short %, high pass completion, plays under pressure) vs a launch keeper (high long %, wins second balls). Match the profile to the game plan, not the other way around.

1This keeper's profile
—
2Vs a high press
If completion under pressure drops below ~70%, add a pressure-release pattern (third-man, split CBs wider, one-touch to the pivot) and rehearse the "brave short vs safe long" decision rule pre-match.
3Training
Rondo-exit passing under live pressure, weighted long-ball targeting drills to the channels, and goal-kick routines mapped to the opponent's press triggers.

Elite reference: overall pass completion ~80–90%; long-ball completion ~45–60% depending on role.

Sweeper role & positioning

Defensive actions outside the box tie the keeper to the line height.

Sweeper actions /90 and the average distance of defensive actions from goal quantify how proactively the keeper defends space behind a high line. A higher line demands a higher starting position and quicker reading of through-balls.

1This keeper
—
2Line-height link
Set the keeper's default height with the defensive line: high block → 16–20 m off the line, ready to sweep; low block → drop and protect the frame. Rehearse the trigger to sprint vs hold.
3Training
Through-ball reading + first-step sprint drills, decision games (claim / clear / hold), and pass-under-pressure once the ball is swept — the sweep only counts if the next action retains possession.

Cross & aerial management

Command of the area — win %, claims, punch decisions.

Aerial win %, claims /90 and cross-stopping % measure command of the box. Punch vs catch is a decision: catch when protected, punch far and wide (25–30 yards) when crowded.

1This keeper
—
2Set-piece defending
Map starting position vs delivery type (in-swing / out-swing / short). Drill the take-off timing and the call ("keeper!") so the back line clears the zone.
3Training
Crossed-ball repetitions under contact, punch-distance targeting, and claim-in-traffic reps. Track high-claim success separately from low crosses.

Load → readiness

GK dive load is high-impact and asymmetric — manage it like sprint load.

Dives are the GK's high-impact actions. Track total dive volume, L/R balance and dive-impact (g), and blend with ACWR & wellness before prescribing the next session.

1This keeper's status
—
2L/R balance
A persistent dive-side imbalance >15% biases hip/adductor load to one side — mirror the dive drills and check the ForceFrame ADD:ABB ratio in the Athlete Profile.
3Today's dose
Green readiness → full reactive block. Amber → cap high-impact dives, keep handling & footwork. Red → technical / distribution only, no maximal dives.

ACWR guardrail 0.8–1.3 · dive-impact trend flags accumulating landing stress.

GK physical testing battery

GK-specific tests + shared outfield thresholds.

The GK battery adds reaction time, dive reach (L/R) and explosive push to the outfield tests. Everything writes to the Athlete Profile so benchmarking, asymmetry flags and RTP criteria apply the same way.

TestMeasuresGood (sr.)
Reaction timeSimple/choice RT (ms)<190 ms
Dive reach L/RLateral reach (cm)>270 · <10% asym
Explosive pushLateral push-off (cm)>40
CMJLower-body power (cm)>40
505 / COD deficitChange of direction (s)<2.30 · <0.20 deficit
T-test / agilityMulti-dir agility (s)<9.5

Asymmetry >10% flag (>15% strong) across dive reach, CMJ single-leg, 505 L/R. LSI ≥90–95% before return to full training after injury.

Benchmarks & thresholds

Reference bands baked into the tile colours (green = good/elite · amber = watch · red = flag). Bands are the defensible cross-league professional reference; values vary by league & season, so wire the live percentile bars from your data provider when real data lands.
SHOT-STOPPING
MetricGood / eliteWatch / flagNote
Save %73–78% (elite)<67%league avg ~69%; team-dependent
Goals Prevented (PSxG−GA)+4–5 / season ≈ elitenegativenoisy <1.5–2 seasons (~200 SoT)
PSxG / shot on targetcontext ~0.19—0.21 = harder shots faced
Penalty save %~18–25% realistic—career/large-sample only
DISTRIBUTION
MetricGood / eliteWatch / flagNote
Pass completion %build-up 85–89%<78% for a possession keeperread vs launch% & team plan
Short pass %93–97%<90%technical flag if low
Long pass % · Launch %style-dependent—real long 35–60%; launch <30 build-up / 35–40 direct
Under pressure>78%<70%possession-role flag
SWEEPING & AERIAL
MetricGood / eliteWatch / flagNote
Def. actions OOB /90 (#OPA)≥1.15 (sweeper)≤0.70 (deep)FBref/StatsBomb; league-adjust
Avg action distance≥15.0 m≤12 mFBref AvgDist (native yards → m)
Cross-stop % (Stp%)~11–14%≤6%crosses-into-box stopped — NOT an aerial-duel %
Claim accuracy (xClaim)≥1.20<0.85claims ÷ expected (best quality metric)
PHYSICAL TESTING
MetricGood / eliteWatch / flagNote
Reaction time<220 ms excl.>250 msdevice-specific; GK>outfield — build in-house norms
Dive reach / explosive pushin-house norm—no validated field test — proxy w/ lateral jump + CMJ
CMJ≈42 cm (pro GK)<38 cmforce-plate
505 · COD deficit<2.20 s · <0.20 s>2.35 s · >0.30 selite male footballer
Asymmetry (VALD)<10%>15% strong flagNordic ecc <~4.0 N/kg risk · ADD:ABB ≈0.9–1.1 · LSI ≥90–95% RTP · ACWR 0.8–1.3

Caveats: single-camp Goals Prevented is noisy — read the trend. Stp% is crosses-stopped, not an outfield aerial-duel %, so single-digit values are normal. Raw GK "aerial-win %" is unreliable — xClaim / Stp% are used instead. GK reaction & dive-reach have no published cut-offs — set in-house percentile norms per device.

Sources: Opta/The Analyst · StatsBomb/FBref (definitions & leaders) · Driblab (xClaim) · Analytics FC (GK-stat reliability & sample size) · Science for Sport (DSI, ACWR, 505) · VALD/Hawkin (asymmetry, NordBord, ForceFrame) · Timmins 2016 & Opar/Bourne (Nordic) · Sole 2018 (RSI-mod). Full URL list in the project research docs.

Add keeper

Adds to the active squad for this session (demo).
Cancel+ Add