Gameweek 4 was won on selection, not activity. The 983 managers Bawler tracks at the top of the Overall Fantasy Premier League rankings averaged 69.8 points against a game-wide average of 42 — a gap of 27.8 points — while making just 0.84 transfers each and barely touching the points hit (only 1.7% took one). The squads they had built by the deadline did the heavy lifting. Here is what those squads looked like, and where the armband went.
The captaincy split is the story of the week. Cole Palmer was given the armband by 60.7% of the sample, comfortably clear of João Pedro's 30.3% — even though João Pedro is close to a universal hold among elite managers, owned by 99.6% of them, against Palmer's 79.2%. Ownership and captaincy are different questions: the top sample will happily carry a near-nailed player without trusting him with the extra multiplier. Palmer's ownership gap — the difference between how the elite own him (79.2%) and how the whole player base does (26.8%) — is 52.4 points, one of the widest of the week.
Chelsea's fixture helps explain the pull. They were at home to Hull City, and Palmer went into the deadline in form that made the armband debate close — two goals off the bench in the Carabao Cup in midweek, and more shots in the penalty area than any other midfielder in the league through the first three gameweeks. That context does not show up in an ownership number on its own, which is why it is worth pairing the two.
"Share of scoring" measures how much of the sample's total Gameweek 4 points came from one player, doubled if he had the armband. Palmer was behind 156.9% of it, João Pedro 135.3% — both figures above 100% show how much of the week ran through Chelsea's front line. Morgan Rogers added a further 80.7%, remarkable given only 3.2% of the sample captained him; his points came almost entirely from being owned, not from the armband. Isak's share was 56.6%. Erling Haaland, by contrast, took the armband from just 0.8% of the sample and returned a 28.7% share. His ownership gap of -43.7 shows the top sample already held far less of him than the wider game before a tough away trip to Manchester United, and few backed him with the captaincy either.
Rogers was the gameweek's most bought player by some distance, added by 66.2% of the sample. He arrived at Chelsea in the summer in a British transfer record — £117m from Aston Villa — and the home game against Hull gave the sample an obvious point to bring him in. Palmer (45.8%) and Isak (29%) were the next two most-bought players; Isak's move followed a two-goal, nine-minute burst against Ipswich that took him to 25 points for the season.
David Raya (24.8%) and Arsenal's new defender Ezri Konsa (40.1%, following his own £51m move from Villa) rounded out the week's most popular incomings, alongside two cheaper Premier League defenders, Brighton's Maxim De Cuyper and Newcastle's Lewis Hall.
The selling was concentrated at Manchester United. Bruno Fernandes (47.1%) and Bryan Mbeumo (45.2%) were the two most sold players in the entire sample, with team-mate Harry Maguire (23.1%) also among the week's biggest outgoings. United opened the season with a 2-0 home defeat to Hull City under new head coach Michael Carrick, with a 5-2 win over Ipswich sandwiched in between — form unconvincing enough to push a large share of the elite to move on, even though the ownership gap on Bruno (3.8) shows the wider game had not sold him off nearly as fast.
Cody Gakpo (26.6%) was also cut heavily at Liverpool, with Isak's form taking the nailed forward's shirt. Pascal Groß was bought and sold in almost equal measure — 27.9% in, 26.9% out — a rare case of the sample cancelling itself out on a single player.
Chip activity was unusually heavy for so early in the season. 29.4% of the sample played their Wildcard, which grants free, unlimited transfers for one gameweek, and 23.3% played Triple Captain, which triples rather than doubles one player's points. Both are large numbers for Gameweek 4. The Triple Captain share lines up with the Palmer-João Pedro armband split: with two Chelsea forwards producing gameweek shares above 130%, a lot of the sample's chips will have landed.
None of this is a signal for Gameweek 5. It is a record of what the best-performing end of the game did with the information available at the Gameweek 4 deadline, before a ball was kicked. A transfer made by two-thirds of the elite sample is a consensus, not a certainty — the 27.8-point gap over the game average shows the squads worked this week, not that they will again.