3,102,447 squads already exist.
The season starts on Friday 21 August - Arsenal host newly promoted Coventry to open it - and over three million managers have already made their calls. Before a single ball is kicked, the crowd has spoken. This issue is about what it said, and where it might be wrong.
What the crowd decided
Haaland sits in 73.6% of squads at £15.5m. That's not a pick anymore, that's infrastructure. João Pedro is next among forwards at 55.8% (£7.5m), and in midfield it's a two-horse template: Bruno Fernandes (48.2%, £12.0m) and Szoboszlai (43.6%, £7.0m).
The pattern I find interesting: the crowd's conviction collapses fast below the obvious names. Third-most-owned forward? Calvert-Lewin at 25.2%. Third goalkeeper? Lammens at 19.7%. One or two consensus picks per position, then a shrug. Three million squads, and most of them are the same squad with different budget haircuts.
What the schedule says
My fixture model - built on team strength, not the official FDR - hands Arsenal the kindest opening six (2.0 difficulty average), with Newcastle (2.17) and a Chelsea/Palace/Liverpool trio (2.33) close behind.
The short straw goes to Bournemouth: 3.33, the hardest opening run in the league, with Spurs (3.0) and Fulham (2.83) next. Among the promoted sides, only Hull make the ugly end of the table (2.83) - Coventry and Ipswich land mid-pack, a welcome you can at least breathe in.
One honest footnote: until roughly GW4 this model runs on pre-season priors. Once real xG data lands, it starts thinking for itself - and the first disagreements between my numbers and the official difficulty ratings should be worth the ticket alone.
The moves that are actually new
Plenty of "new" faces aren't new at all - the market has had half a season to watch anyone who moved in January. Strip those out and the genuine summer arrivals tell two very different stories.
The £75m latecomer. Bruno Guimarães completed his move to Arsenal on Friday - the champions' most expensive piece, walking straight into the league's kindest opening run. The market's response so far: 9.4% → 10.3% ownership since the announcement. No stampede yet - but most squads get rebuilt in the final days before the deadline, and at £7.0m he's the cheapest ticket into a title-winning midfield. Rice, for context, sits at 21.8%.
The early consensus. Rogers joined Chelsea on 21 July and is already the third-most-owned midfielder in the game - 26.7% at £7.5m. Three weeks, no minutes, a third of the player pool on board: the fastest consensus of the summer, built entirely on last season's memory and a blue shirt. Elsewhere, Spurs quietly signed three of the ten most-owned newcomers - Dubravka (22.6%), Senesi and Van Hecke - a rebuilt spine the market is pricing one budget piece at a time.
What I'll be watching
Three things between now and GW4: whether the template's thin conviction below the premium picks starts moving, which of the hard-run teams (Bournemouth, I'm looking at your fixtures, not your squad) gets punished by schedule rather than quality - and the moment my model wakes up and starts arguing with the FDR.
New here? The Throughball lands twice a week: a Tuesday deep-dive on one thing the numbers noticed, and a brief the day before every deadline. All analysis, no tips - I show you the pitch, you make the calls.
Three million squads are guessing. From Friday, the ball starts grading them.