Harris Andrews has the rolling-40 KICK Rating of an average AFL footballer.

Forty-two. Solid but not flashy. Below the median of every other player in a season-best top 100. Brisbane’s defensive backbone for the better part of a decade, multi-time All-Australian, the player every key forward in the comp has to plan around — comes out of the stat sheet at 42.

This is the part of KICK Rating that has driven readers nuts since launch. The formula is box-score-driven. It rewards measurable on-field actions — disposals, marks, tackles, intercepts, one-percenters — and a defender like Andrews who quietly destroys forward-50 entries doesn’t pile up the kind of volume KICK rewards a midfielder for. We’ve explained this on the methodology page and shipped Positional KICK (KICK-M / KICK-D / KICK-F / KICK-R) to address it within the rating’s box-score frame. KICK-D reads Andrews 83 (rolling), 87 (career) — already a much better number than overall.

But there’s another signal entirely we’ve been sitting on. Every match, the two senior coaches vote 5-4-3-2-1 for the AFL Coaches Association’s Champion-Player award (which began in 2003). Coaches see the things the stat sheet misses — defensive structure, decoy running, intercept positioning, the spoils that prevent a contest from happening. We scraped 23 seasons of AFLCA records last year (27,225 individual vote records across 4,137 matches, 2003–2025) and quietly used the data to validate KICK. We didn’t surface it as a public lens. As of today we have.

The Coaches’ eye leaderboard is live. It is not a new KICK. It is not a blended composite. The KICK formula stays unchanged. Coaches’ eye sits alongside as a parallel lens, ranking players by AFLCA votes per game, scaled 0–100 within position — same anchoring philosophy as KICK-P. On every eligible voted-era player profile (80+ career games, 40+ games in the voted era) there’s now a small card with both numbers and the delta.

Below: the 15 active defenders with the biggest “coaches see more” deltas — Coaches’ eye scaled value minus their overall-KICK rolling-40. Top 14 are +40 or more; #15 is +39. None of these names will surprise you. The point is that the surface coaches respect them on now exists publicly, with the data and the formula in the open.

The 15 active defenders the coaches see better than the stats

Active in 2025 or 2026, ranked by the Coaches’ eye − overall-KICK r40 delta.

Player Team Coaches’ eye KICK r40 Δ
Sam Taylor GWS 92 40 +52
Darcy Moore Collingwood 84 33 +51
James Sicily Hawthorn 100 53 +47
Harris Andrews Brisbane 88 42 +46
Tom McDonald Melbourne 85 40 +45
Tom Stewart Geelong 99 54 +45
Tom Barrass Hawthorn 76 31 +45
Jacob Weitering Carlton 85 41 +44
Jake Lever Melbourne 82 38 +44
Steven May Melbourne 91 48 +43
Jeremy McGovern West Coast 92 50 +42
Dane Rampe Sydney 74 33 +41
Noah Balta Richmond 78 38 +40
Alex Pearce Fremantle 71 31 +40
Nick Vlastuin Richmond 85 46 +39

You can rebuild this table yourself: scrape AFLCA’s published vote totals (the Champion-Player award publishes votes after every match), divide votes by games played in the voted era, scale within the defender pool. The full methodology is on the methodology page. That’s the point — the Coaches’ eye lens needs no proprietary data; the AFLCA publishes the inputs, and we just aggregated them.

What this is, and isn’t

It is a parallel ranking by an external signal coaches generate (rather than reverse-engineering it from stats). Same way Positional KICK corrects for “the formula reads defenders low because they touch the ball less,” Coaches’ eye corrects for “the formula misses what coaches actually see on tape.”

It is not an attempt to blend the two numbers. We tested that. A weighted average like 0.85·KICK + 0.15·AFLCA at the headline level actually makes the mid-bias worse, because AFLCA itself is mid-biased — high-vote-getting mids gain more from a blend than moderate-vote-getting defenders. The right move is to keep the lenses separate and let readers triangulate.

It is not a fix for the structural ceiling that closed v1.2, v1.3 and v1.4. KICK’s overall ranking still has to agree with Brownlow and All-Australian (which reward mids 70-80% of the time), so KICK still reads like that too. We can’t fix the box-score bias at the formula level without breaking award alignment. We can surface a different signal alongside, transparently. That’s this.

Where the two lenses agree, where they don’t

Midfielders are where the two lenses converge. The active mid top 8 by Coaches’ eye is Daicos, Bontempelli, Serong, Cripps, Dangerfield, Neale, Rowell, Merrett — essentially the same names as the KICK-M top 8. Coaches and the stat sheet both reward inside-mid grunt, and they agree on the shape of which players have it. The delta column for mids hovers +15 to +25 — meaningful but not transformative; both lenses are reading the same body of work.

Defenders are where they diverge. The active KICK-D top 5 reads McGovern 100, Wilkie 97, May 95, Stewart 95, Sicily 94 (position-fair scaled values); the Coaches’ eye defender top 5 reads Sicily 100, Stewart 99, Sam Taylor 92, McGovern 92, May 91. Same names, mostly, but the ordering shifts — McGovern leads on KICK-D, Sicily on Coaches’ eye. Andrews lands #10 on KICK-D, #6 on active Coaches’ eye. The bigger divergence is players the box score reads as “below average” defenders who coaches keep voting for: Sam Taylor (KICK-D 81, CE 92), Jordan Ridley (KICK-D 93, CE 87 — closer to parity), Mason Redman (KICK-D 86, CE 82).

Forwards sit in between. Heeney leads both lists — comfortably on KICK r40 (71, +9 ahead of Rankine at 62), narrowly on Coaches’ eye (87, with Cameron two points back at 85). The deltas for key forwards (Cameron +28, Greene +26, Rankine +21, Pickett +10) skew positive — coaches respect them more than the box score does — but it’s a smaller correction than for defenders.

Rucks are noisiest. With only 72 voted-era ruckmen in the eligible pool (vs 345 defenders, 302 mids, 142 forwards), the anchor calibration is thinner; tiebreaks are unstable in the back half. Gawn and Grundy stay at the top either way; everyone else moves around. Don’t read too much into Coaches’ eye for rucks below #10.

Historical: who never got the credit on the stat sheet

The all-time list of biggest “coaches see more” deltas is half a roll call of every defender critics have ever said KICK underrates:

(One historical entry worth its own footnote: Scott Thompson the North Melbourne defender would have appeared on every earlier draft of this list at Coaches’ eye 95. He doesn’t — his pid shares a kebab slug with Scott Thompson the Adelaide midfielder, and v1.0 of our compute summed both their AFLCA votes into one bucket. The pid-aware v1.0.1 fix shipped 2026-05-14 untangles them: Scott Thompson NM’s true Coaches’ eye is 77, his Adelaide namesake’s is 73. The cautionary tale is in the methodology page; “elite” historical defender rankings should be cross-checked against the version number that produced them.)

Look at Sam Taylor in the active table at the top of this article. 128 voted-era games, overall KICK r40 40, Coaches’ eye 92, Δ +52 — the biggest active-defender delta on this list. He’s polled near the top of the AFLCA defender pool for the past three years; the stat sheet hasn’t caught it yet. The lens disagreement is not proof he’s elite — coaches can over-vote a fashionable name, KICK can under-weight a real one — but the gap is the kind of signal worth keeping an eye on.

Where to find it on the site

What’s next on the rating side

Coaches’ eye ships as a feature, not a version. KICK itself stays on v1.1, and the next planned version bump is v1.9 Career Greatness — a composite blending peak-40 KICK + career-average KICK + games-played weight + position bonus, aimed at giving the all-time leaderboard a defensible “greatest career” view (rather than the current “greatest peak stretch” view that puts Rowan Marshall at #16). Plan is in V1.9_CAREER_GREATNESS_PLAN.md (internal); ship window opens late May.

We’d rather you argue with the formula than trust it blindly. That includes this one. If a player’s Coaches’ eye reads ten points off where you think it should, tell us — the inputs are public, the formula is on the methodology page, and we still don’t think we’ve answered every question well. We just think we’ve answered one more of them today.

— The Kicker