FPL Gameweek 3 tips: captain picks and best buys

FPL Gameweek 3 preview: best buys, captain picks and differentials

Fantasy Premier League Gameweek 3 deadline is Friday 4 September at 18:30, and it opens straight away with Ipswich hosting Liverpool at 21:00 the same evening. That early kick-off matters: whatever you do with your Liverpool assets, you have to do it before Friday evening, not over the weekend.

The obvious part of this Gameweek is Manchester City at home to Coventry. Everyone can see it, most managers already own the City players, and simply owning them will not move you up your mini-league. What decides Gameweek 3 is the six or seven squad spots after the template picks — the mid-ownership midfielders, the cheap defenders in good fixtures, and knowing which heavily owned player to move on before his run collapses.

Gameweek 3 fixtures worth targeting

Four fixtures stand out on the model's numbers. Note the one-row-per-team format: each row is a single team's own projection for that match.

Man City - Coventry is the standout. Coventry have conceded 1.27 expected goals per game across their first two Premier League matches and carry a net expected-goal difference of -0.49, the profile of a side that concedes chances without creating them. City's 47% clean sheet chance is the highest in the round.

Brentford are the quiet story. Their opening two league games produced 2.76 expected goals for and 0.95 against — a net expected-goal difference of +1.81, the best number in the division so far. They are barely owned relative to that form, and Sunderland at home is a genuine double-up fixture at both ends of the pitch.

Avoid loading up on Aston Villa attackers despite a friendly-looking trip to Hull. Villa have produced just 0.31 expected goals per game and conceded 2.36 — a net -2.05, the worst in the league. The model still gives them a 52% win probability at Hull, so a Villa defender is defensible; the attack is a leap of faith. Cross-check the team news on the FPL predicted lineups page before committing either way.

Best goalkeepers for Gameweek 3

David Raya (6.0m, ARS) — 4.0 xP, and the best goalkeeper on the six-round view with a decay-weighted score of 15.6. Arsenal have conceded only 0.30 expected goals per game this season, comfortably the meanest defence in the league. A 38% clean sheet chance against Chelsea is not the pick's whole case — Gameweek 4 away at Sunderland projects a 52% clean sheet chance, and Gameweek 6 at home to Leeds another 52%. Raya is a set-and-forget for a month.

Caoimhin Kelleher (5.0m, BRE) — 4.0 xP at a full million cheaper, on a 41% clean sheet chance at home to Sunderland. At 7.2% ownership he is the value route into Brentford's form, and freeing up that million pays for a genuine upgrade elsewhere.

Gianluigi Donnarumma (5.5m, MCI) — 4.0 xP with the round's single highest clean sheet probability at 47%. Worth it if you want the City double-up, though City's Gameweek 6 trip to Anfield takes the shine off the longer view.

Best defenders for Gameweek 3

Defence is where this Gameweek is won, largely because of defensive contribution points. The model's top-rated defenders are not all the ones people talk about.

Nico O'Reilly (6.5m, MCI) is the premium pick of the round at 5.9 xP — the highest of any defender — combining a 44% clean sheet projection with an attacking share City's other centre-backs do not have (0.21 expected goals and 0.18 expected assists in a single match). The catch is the follow-up: 3.2 in Gameweek 4 away at Manchester United and 2.7 in Gameweek 6 at Anfield. Buy him for the fixture, not for the season.

Nathan Collins (5.5m, BRE) is the better multi-week hold. 5.0 xP this week on a 40% clean sheet, 89 projected minutes, and Brentford's league-leading expected-goal difference behind him. At 2.2% ownership he is close to a free differential on a nailed starter.

Michael Kayode (4.6m, BRE) is the best points-per-million pick in the game this week at 1.04 — 4.8 xP from a 4.6m defender in the same clean sheet as Collins. If you need to raise funds, this is the way to do it without losing points.

Hidden gem: Anan Khalaili (5.0m, CRY). Owned by 0.1% of managers, projected for 4.9 points this week, and 16.2 on the weighted six-round score — higher than Gabriel, higher than Virgil van Dijk, higher than any Manchester City defender. The reason is defensive contributions: the model has him at 1.27 expected defensive contribution points and 11.0 expected clearances, blocks, interceptions and tackles per match, the second-highest in the league. His trajectory is 4.9 → 5.2 → 4.3 → 4.1, so he holds his level all the way through Gameweek 6. If you want one pick nobody in your mini-league has, this is it.

Budget alternative: John Egan (4.0m, HUL) — 4.1 xP at the minimum price and a points-per-million of 1.03, driven by the highest projected defensive contribution total in the league (1.81). Hull's fixture against Villa is not pretty, but Villa's attack has produced 0.31 expected goals per game, and at 4.0m he is a pure enabler.

Best midfielders for Gameweek 3

Antoine Semenyo (8.5m, MCI) — 6.5 xP, the second-highest score of any player this week. 89 projected minutes, 0.45 expected goals and 0.28 expected assists against Coventry, and a 46% clean sheet share on top. At 20.7% ownership he is still a differential relative to how good the fixture is. His Gameweek 4 dips to 3.9 at Old Trafford, so this is a two-week window rather than a long-term buy.

Dominik Szoboszlai (7.0m, LIV) — 6.3 xP and Liverpool's designated penalty taker, which is the single biggest reason he stays in at 41.5% ownership. Across his last 64 appearances he has produced 13.98 expected assists alongside 10.43 expected goals, an unusually balanced profile for a midfielder at this price. Gameweek 3-5 total of 17.6 is the third-best in the game.

Bukayo Saka (9.5m, ARS) — 6.0 xP at home to Chelsea, with 81 projected minutes and 0.60 expected goal involvements per 90 across his last 74 appearances. Only 10.5% owned, which is remarkably low for an Arsenal premium in a home fixture the model likes.

Bruno Fernandes (12.0m, MUN) — 5.9 xP now, but the highest weighted six-round score of any midfielder at 23.42, and a rising trajectory: 5.9 → 6.2 → 6.8 → 7.2. He takes United's penalties and has produced 0.74 expected goal involvements per 90 across his last 55 appearances. If you are planning transfers rather than reacting to one week, he is the midfielder to own.

Differential: Phil Foden (7.0m, MCI) — 5.9 xP, 86 projected minutes, and only 5.0% ownership. That is a top-six score at a fraction of the ownership of the players around him, and 2.5m cheaper than Saka.

Budget midfielder: Diego Gomez (5.0m, BHA) — 5.0 xP for 5.0m is a points-per-million of 1.00, one of only three players in the game projected above that mark this week. Brighton host Leeds with a 53% win probability and have generated 2.58 expected goals per game this season. He is 4.5% owned and 88 minutes nailed.

Emiliano Buendia (6.0m, AVL) — 5.1 xP and Villa's penalty taker at 1.1% ownership. The caveat is honest: Villa's underlying attacking numbers are the worst in the league right now, so this is a bet on the fixture at Hull rather than on the form. If he does not start, Diego Gomez is the safer route to the same price bracket.

Best forwards for Gameweek 3

Erling Haaland (15.5m, MCI) — 8.9 xP, 2.4 points clear of the next-best player in the game. 1.19 expected goals in a single match, 89 projected minutes, City's penalties, and 0.87 expected goal involvements per 90 across his last 70 appearances. There is no argument against owning him this week.

Alexander Isak (9.0m, LIV) — 5.8 xP away at Ipswich, whose 2.66 expected goals conceded per game is the worst defensive record in the league. 0.55 expected goals for a single fixture is the second-highest forward number of the round.

Thiago Rodrigues (8.0m, BRE) — 5.4 xP, 88 projected minutes and Brentford's penalty taker. 0.51 expected goals against Sunderland, in the fixture with the second-best clean sheet profile in the round. At 14.5% ownership he is the best-value route into the league's form team.

Hidden gem: Charalampos Kostoulas (5.5m, BHA) — 4.5 xP from a 5.5m forward, 0.38 expected goals against Leeds, and 0.4% ownership. Cheap enough to fund a Haaland-plus-Isak front line.

Gameweek 3 captain picks

Captain Erling Haaland. At 8.9 xP he projects for 17.8 as captain, and the gap to the field is the largest of the season so far. Manchester City at home to Coventry, an 80% win probability, 2.87 expected team goals and penalty duties — there is no case for anything else this week.

If you do not own him, Antoine Semenyo (6.5 xP) is the next-best armband from the same fixture, and Dominik Szoboszlai (6.3 xP) is the best option outside Manchester City, with Liverpool's penalties as the tie-breaker. Bukayo Saka (6.0 xP) is fourth and the highest-ceiling pick if you want to differentiate in the Arsenal-Chelsea fixture.

Look one week ahead too. Haaland drops to 5.4 in Gameweek 4 at Old Trafford, while Cole Palmer (9.6m, CHE) jumps to 7.6 at home to Hull - the joint-highest one-round projection anywhere in the game between now and Gameweek 8. Palmer at 4.0 this week is not a buy right now; he is a Gameweek 4 buy, and planning for that changes which midfielder you sell this week.

Who to sell before Gameweek 3

Riccardo Calafiori (5.6m, ARS) is the clearest sell in the game. He is 43.5% owned — the fourth-most-owned player overall — and his projection falls away: 3.3 → 2.7 → 2.2 → 2.4. Arsenal's defence is excellent, but the model has Gabriel and even Arsenal's midfielders ahead of him for defensive returns, and a third of the league is holding a player whose next four rounds are all below 3.5.

Rayan Cherki (7.7m, MCI) — 25.4% owned and projected for 4.0 in the best fixture of the round, falling to 2.9 in Gameweek 4. With only 48 projected minutes against Coventry he is not a nailed starter, and at 7.7m that is expensive uncertainty when Foden costs 7.0m for 5.9.

Antonin Kinsky (4.5m, TOT) — 20.1% owned, 3.3 xP, and Spurs have conceded 2.36 expected goals per game. Their trip to Nottingham Forest leaves him on a 26% clean sheet chance. Kelleher at 5.0m or Raya at 6.0m are both better uses of the space.

Martin Dubravka (4.0m, TOT) — 18.2% owned and projected for 0.0 in every one of the next six Gameweeks. If he is your bench goalkeeper that is fine; if he is anywhere near your starting XI, that is a dead slot.

Harry Maguire (5.0m, MUN) at 15.1% and Issa Diop (4.0m, IPS) at 15.9% round out the list — 3.7 and 2.3 respectively this week, both trending downwards through Gameweek 6.

Best Free Hit team for Gameweek 3

If you are playing a Free Hit this week, this squad maximises Gameweek 3 expected points alone within the solver's spending ceiling and the three-per-club limit. It costs 99.0m and projects 63.4 points from the starting eleven before the captain multiplier — 72.3 with Haaland captained.

That is a 3-5-2 with the bench filled by minimum-price non-starters. The three-per-club cap binds on Manchester City, which is why Foden and Guehi are absent despite strong individual projections. For the exact optimum against your own budget and current squad, the site solves it live on the optimal teams for FPL page.

Transfer summary for Gameweek 3

The template move this week is out with Calafiori, in with a Brentford or Manchester City defender. If you are taking a hit, make sure it buys you something that survives Gameweek 4 as well — Collins, Khalaili and Fernandes all do; O'Reilly and Semenyo do not. Compare every projection side by side on the FPL player picks page, and check the run of fixtures on the FPL schedule before you commit a transfer you cannot reverse for free. The underlying win and clean sheet numbers behind everything above come from the FPL market probabilities page.

Last week's analysis is in our FPL Gameweek 2 preview, and every round is archived in the FPL preview hub. New to the game? Start with our fantasy football guides.

FAQ

Who should I captain in FPL Gameweek 3?

Erling Haaland. He projects 8.9 points against Coventry at home, 2.4 clear of the next-best player in the game, with an 80% win probability, 2.87 expected team goals and penalty duties. Antoine Semenyo (6.5) and Dominik Szoboszlai (6.3) are the alternatives if you do not own him.

Which budget defenders are best for Gameweek 3?

Michael Kayode at 4.6m is the best value in the game with 4.8 expected points and a points-per-million of 1.04, in Brentford's 41% clean sheet fixture at home to Sunderland. John Egan at 4.0m (4.1 xP) is the cheapest route to defensive contribution points, and Anan Khalaili at 5.0m is the differential at 0.1% ownership.

Should I sell Calafiori before Gameweek 3?

Yes. Riccardo Calafiori is 43.5% owned and projects 3.3, 2.7, 2.2 and 2.4 over the next four Gameweeks. Arsenal's defence is the best in the league, but Gabriel at 8.0m carries the returns, and Calafiori's next month is below the level of a starting FPL defender.

Best differentials for FPL Gameweek 3?

Anan Khalaili (5.0m, Crystal Palace) at 0.1% ownership is the standout, with 4.9 expected points this week and the second-highest projected defensive contribution total in the league. Phil Foden (7.0m, 5.0% owned, 5.9 xP) and Charalampos Kostoulas (5.5m, 0.4% owned, 4.5 xP) are the differentials in midfield and attack.

Is Cole Palmer a good buy for Gameweek 3?

Not this week. Palmer projects only 4.0 away at Arsenal, where Chelsea have a 19% win probability. He jumps to 7.6 in Gameweek 4 at home to Hull, the joint-highest one-round projection anywhere in the game between now and Gameweek 8, so he is a Gameweek 4 buy, not a Gameweek 3 one.

This preview was written by TheFantasyTool AI - verified by Nicolai @ TheFantasyTool