01Observed
Hook→Flay timing
Shows how quickly Flay follows Q without claiming immediate is always best.
- Measured as
- Median seconds from tracked Q to eligible E.
- Training drill
- Practice immediate and delayed anti-dash Flays.
CHAMPION LAB · EARLY DATA
THRESH LABCreate the angle.
Layer the control.
Prove the value.
Measure control-chain construction, Q/E ordering, follow-up timing, hook activity, and movement without pretending cast volume is accuracy.
WHAT TO WATCH
Control order is measurable before control value is.
The pack identifies Hook→Flay, Flay→Hook, and Box→Flay while withholding contact and displacement claims that need richer replay evidence.
Early dataWe wait for enough players before calling something a benchmark.
Exact actions measured One score per player Compared by current rank Missing evidence stays unavailable
hook conversion
Q-to-Flay timing
- What it measures
- Elapsed time from Q cast to a following E within three seconds. The distribution distinguishes immediate control chains from deliberate delay, but does not isolate landed hooks.
- Why it matters
- This is a decision and style metric, not a simple higher-is-better score. Use its position in the field to find unusual tendencies, then inspect the situations producing them.
MASTER1.90sn=33 · 56 eligible games · benchmark ready
GRANDMASTER1.93sn=15 · 40 eligible games · early sample
CHALLENGER1.90sn=6 · 9 eligible games · early sample
Observed medians run from 1.90s in master to 1.93s in grandmaster.
54 players across 105 eligible replays in 3 ranks.
Curves show observed player distributions. Anything the AI could not see in the replay is unavailable, never scored as zero.THRESH REPLAY LAB
Control sequences, timing, and activity by rank.
530 AI-watched replays from 211 current-rank players. Master is the only spell cohort above the 30-player benchmark gate.
- Your metrics
- 0/12
- Players
- 211
- Player-games
- 530
How Thresh mechanics are measured
Eight sequence and activity metrics calculate. Hook contact, lantern saves, Flay direction, and Box conversion remain withheld.
LEARN THRESH
Turn mechanics into better decisions.
See what matters, when it matters, and one drill to take into your next game.
MECHANICS LIBRARY
The Chain Warden, made measurable.
02Contextual
Flay→Hook setup
Captures creating displacement before attempting Q.
- Measured as
- Eligible E casts followed by Q divided by E opportunities.
- Training drill
- Constrain movement, then aim where the target must go.
03Observed
Control-chain length
Counts contiguous control-spell sequences.
- Measured as
- Mean casts in deterministic Q/E/R chains.
- Training drill
- Review where chains break and whether it was intentional.
04Coverage gated
Lantern value
A cast alone cannot prove a save or correct target.
- Measured as
- Unavailable until ally target, shield, displacement, and threat state validate.
- Training drill
- Review accessibility and ally path, not cast count.
MATCHUP APPLICATION
The same score can mean something different by lane.
Open live Thresh matchup matrix
VS
Leona
ReactiveLane ruleHold Flay to disrupt her committed path.
Best windowAs Zenith Blade connects.
WatchAttack→E · E→Q
VS
Rakan
TimingLane ruleSeparate his entry from escape with directional control.
Best windowDuring his dash commitment.
WatchE timing · chain length
VS
Morgana
PatienceLane ruleDo not make hook volume the goal through Black Shield.
Best windowAfter shield is spent.
WatchQ casts/min · Q→E
GAME PLAN
Use the mechanic for the right play.
Angle before cast
Hook frequency is activity, not accuracy.
See the matching mechanicLayer, do not dump
Preserve control coverage instead of overlapping spells blindly.
See the matching mechanicLantern honesty
Save value stays unavailable until the replay proves danger and outcome.
See the matching mechanicHOW WE KNOW
See exactly what is measured.
We show patterns only after enough real players support them. Early results stay labeled, and missing information stays unavailable.
Measured in replay Coaching guidance Early pattern
Open measurement details
Every result defines what counts, when it applies, and which actions are excluded.
verified hook contact, lantern save value, directional Flay value, Box wall conversion stay unavailable until the replay shows them reliably.