When was the peak? Every top-20 player's best 40-game window

Peak 40 is the best 40-game rolling window in a player’s career — their strongest ~1.75-season stretch, regardless of when it happened. It’s a peak-form metric, not a greatest-career metric. A Brownlow-level 40 in 2018 rates the same as a Brownlow-level 40 in 1998.

We’ve been computing the end-year of every player’s peak window since v1.1. We just hadn’t shown it. A follower asked under Gary Ablett Jr’s #1 rank — “But when was the peak? (Guess would be 2014-15 but would be cool to know)” — and it took us about an hour to realise the answer was already sitting in the ratings JSON, un-rendered. So: a new Peak yrs column on the Peak-form leaderboard, and this article to walk through the top 20 with context.

Ablett’s window: 2011–2013. Bill’s 2014-15 guess was one year late. The 40-game stretch that beat every other career in VFL/AFL history ran across his first three Gold Coast seasons — the Leigh Matthews Trophy year (2011), the Suns captaincy build, and the 2013 Brownlow.

The full top 20, dated

Peak 40 for every player with 150+ games, ranked. Minimum rolling window is exactly 40 games (~1.75 seasons). The Peak yrs column is new today.

#	Player	Peak 40	Peak yrs	Career avg	Games
1	Gary Ablett Jr	93.22	2011-13	70.06	353
2	Patrick Dangerfield	92.64	2016-18	67.97	355
3	Tom Mitchell	89.16	2017-18	69.69	203
4	Dane Swan	87.23	2011-13	71.44	249
5	Marcus Bontempelli	86.07	2022-24	73.20	263
6	Clayton Oliver	84.70	2021-23	71.16	211
7	Patrick Cripps	84.15	2017-19	70.49	233
8	Brodie Grundy	84.12	2018-19	67.34	246
9	Josh Kennedy (Syd)	83.94	2015-17	68.19	285
10	Matthew Boyd	83.67	2011-12	62.83	284
11	Jack Macrae	82.93	2017-19	67.72	262
12	Nat Fyfe	82.88	2014-17	67.02	233
13	Touk Miller	82.65	2021-23	62.46	218
14	Scott Pendlebury	82.04	2011-13	66.53	417
15	Joel Selwood	81.72	2010-12	68.49	351
16	Rowan Marshall	81.46	2023-24	66.66	160
17	Tom Rockliff	81.43	2014-17	67.77	203
18	Dustin Martin	81.17	2016-18	62.51	299
19	Nathan Buckley	80.75	1999-2001	59.48	278
20	Dayne Beams	80.58	2012-14	66.50	176

Live on the leaderboard, with the window shown inline beside every player.

Top 5 in detail — what was happening when

1. Gary Ablett Jr — 2011–2013 (Gold Coast)

The only player in the top 10 whose peak didn’t happen in Melbourne or at a traditional heartland club. Ablett left Geelong for the expansion Suns in October 2010 — a move widely framed at the time as him cashing out on greatness. Instead he immediately produced the highest-rated 40-game stretch in the database. The window contains: the 2011 Leigh Matthews Trophy (AFLPA MVP) and a second-place Brownlow finish behind Dane Swan; the 2012 season where he averaged 30.5 disposals and carried a struggling Gold Coast squad to the edge of finals relevance; the 2013 season where he won his second Brownlow outright (26 votes) and earned All-Australian captaincy. His peak average — 93.22 — is nearly a full Brownlow-medallist season sustained across 40 games.

2. Patrick Dangerfield — 2016–2018 (Geelong)

Mirror image of Ablett: left Adelaide for Geelong ahead of 2016, immediately peaked. 2016 Brownlow (by a record 35 votes, tied then broken in 2023) and AFLCA Champion Player, 33.5 disposal avg, All-Australian. 2017 was All-Australian again plus the Leigh Matthews Trophy, with the Cats losing the prelim to Adelaide. 2018 — still All-Australian, still averaging 27+. The peak 40 window spans the stretch before the ankle-and-hamstring injuries caught up with him from 2020 onward.

3. Tom Mitchell — 2017–2018 (Hawthorn)

The shortest peak on the top 10. Mitchell arrived at Hawthorn from Sydney ahead of 2017 and immediately averaged 37 disposals. In 2018 he won the Brownlow with 28 votes and averaged 35.7 disposals per game — still the highest single-season average in recorded AFL history. That 40-game run from mid-2017 through the 2018 prelim is statistically absurd. A broken leg in the 2019 preseason and his subsequent struggle to regain form means his peak window is tight (2 years) and his career average still ranks 3rd overall despite only 203 games.

4. Dane Swan — 2011–2013 (Collingwood)

Shares the top peak window with Ablett. The 2011 Brownlow (bringing Collingwood’s drought to an end for a man of his era), 2012 All-Australian, 2013 All-Australian again. Swan’s stat profile was pure volume — disposal counts others couldn’t match without the metronomic endurance Collingwood built around him. His peak window bridges the post-2010 premiership core (he won Norm Smith in that grand final replay, just outside this 40-game window) and the decline years from 2014 onward.

5. Marcus Bontempelli — 2022–2024 (Western Bulldogs)

The highest-peaking active player, and the only top-5 entry whose window is still open (his career avg 73.20 is the highest career average on the board). Three AFLCA Coach of the Year awards in the window (2022, 2023, 2024 — a record). All-Australian captain 2022. The window closes in 2024 at the end of a three-season run where the rest of the midfield stratum caught up to him only briefly before he reset the level. A v1.5 or v1.6 recompute mid-season 2027 will likely shift this window forward, not backward.

The 2010s own the list

Peak windows by decade in the top 20:

1990s–2000s: 1 (Buckley)
2010s: 15
2020s: 4 (Bontempelli, Oliver, Miller, Marshall)

That’s not a claim about player quality. It’s a claim about data density. AFL box-score tracking expanded dramatically from 2005 onward — by 2012 the full 17-field KICK formula was feasible against data that simply didn’t exist for Buckley’s era. The pre-2000 caveat we keep on the leaderboard is honest about this: legends of the 1990s — Ablett Sr, Carey, Hird, Lockett — score lower on raw KICK than their era peers knew them to be, because they didn’t get credit for actions (contested marks, rebound-50s, intercepts) that weren’t recorded until later.

Buckley’s 1999-2001 window survives because his disposal counts were so extreme (he averaged 28+ across that stretch) that even the thin input data rendered him top-20. It’s the only pre-2005 entry. By v1.6 — see our decade leaderboards for the era-fair view today — this gap narrows.

The outliers worth flagging

Rowan Marshall (#16, 2023-24) — flagged by a reader last week as looking too high on an “all-time” list. Not a bug. Marshall had a legitimate back-to-back All-Australian calibre 40-game stretch as St Kilda’s primary ruck. Peak 40 isn’t rewarding longevity or career value here — he has neither — just the genuinely elite 1.75-season window he strung together. The test isn’t “is his window real?” (it is) but “should the page title of this leaderboard promise ‘greatest career’ when the metric measures ‘peak stretch’?” We renamed the page to ‘Peak-form leaderboard’ on Thursday to close that expectation gap.

Nat Fyfe (#12, 2014-17) — the only four-season peak window in the top 20. Two Brownlows (2015, 2019 — only the first falls inside this window) spread across a longer form peak that injury interruptions kept fragmented. Peak 40 hunts the best rolling window regardless of continuity, so Fyfe’s 40 games at form come from partial seasons stitched together across four years rather than two consecutive campaigns.

Nathan Buckley (#19, 1999-2001) — three-year window straddling the Y2K boundary. Peak 40 was 80.75 — a genuinely era-elite number on the thinner data available. His career average sits at 59.48, dragged down by the 1993-98 early career where box-score tracking was even sparser. In any pre-1997 recomputation, Buckley rises considerably.

Scott Pendlebury (#14, 2011-13) — shares the 2011-13 window with Ablett and Swan. 417 career games. Peak 40 is only the 14th-best 40-game stretch, but the career average of 66.5 across 20+ seasons is what you’d need a longevity-weighted metric to see properly. Our V1.9 career-greatness composite is the planned answer for “which list is Pendles actually on?”

What this doesn’t mean

Peak 40 is one question with one answer — the best 40-game stretch in a career. It doesn’t tell you:

Who had the greatest career. Longevity and sustained output don’t feature. A composite greatest-career metric is planned for V1.9.
Who was most valuable to their team. Peak 40 is individual output on a per-player rolling window. It’s not Wins Over Without You (WOWY), which is a separate planned metric.
Who was the best at their position. The overall KICK formula is midfielder-biased by design. For position-fair reads see the KICK-M / KICK-D / KICK-F / KICK-R leaderboards.
Who was best adjusted for era. Pre-2000 players are underscored due to data sparsity. See the decade leaderboards for era-relative rankings.

The honest framing for Peak 40 is: whose best stretch was strongest, on the data we have, using a formula that agrees with Brownlow voting? Four lenses, one metric. This one is the peak-form lens.

Ablett 2011–2013 tops it. Dangerfield 2016–2018 second. Tom Mitchell’s 2018 Brownlow third. Everything else is downstream of those three stretches.