Roma-Inter, the big one: who to start in fantasy

The pick of matchday five follows a stated criterion, not a personal taste: among all the fixtures of the round we take the one with the lowest sum of the two teams' league positions. Roma first and Inter second add up to 3, and nothing else comes close: the runner-up, Frosinone-Como, is on 10. Kick-off is Saturday 19 September at 18:00 Italian time (16:00 UTC) at the Stadio Olimpico.

Both arrive with a perfect record: 12 points each, four wins out of four. The way there differs. Roma score 3.00 goals per game and concede 0.30, with three clean sheets in four; Inter score more, 3.30, but concede 1.50 per game and have kept a clean sheet only once. The team xG tells the same story: 2.78 produced and 0.62 conceded per game by Roma (over 3 of the 4 games with statistics), 2.61 and 0.87 by Inter (over 3 of the 4 games with statistics). Inter run the game — 66.0% possession against 59.5%, 25.0 shots per game against 21.2, 91.0% pass accuracy against 86.3% — but shots on target are almost level: 8.0 against 7.8.

The probable line-ups are built on the last three matches: they are probable, the official ones come out an hour before kick-off and we do not have them. Gasperini towards a 3-4-2-1: Svilar; Mancini, Hermoso and Balerdi (62%, competing with Wesley at 60%); Lulli (88%, with Koné at 61%) and Molina (64%) on the flanks, Cristante and Pisilli (62%, Rodrigo Mora at 37%) in midfield, Dybala and Soulé (67%) behind Malen (88%). Chivu sticks to the 3-5-2: Josep Martínez; Akanji, Bastoni and Dimarco (61%, with Bisseck at 42%); Diouf (78%, competing with Sučić at 54%), Barella, Zieliński (79%), Mkhitaryan (61%) and Carlos Augusto (62%); up front Esposito (88%) with Thuram (64%) slightly ahead of Lautaro Martínez (60%).

Who to start. In our listone the role letters are Italian: P goalkeeper, D defender, C midfielder, A forward. Malen (A, price 33) is Roma's hottest name: an 11.50 fantasy average, 70% likelihood of starting and 2.12 goals plus assists per 90 minutes in this Serie A — a figure over few minutes, but above our 180-minute threshold — after 0.97 in 2025-26. Behind him Dybala and Soulé, both on 8.38, with 1.14 and 1.43 goals+assists per 90: Dybala, though, starts 90% of the time, Soulé 50%. At Inter the best return is Thuram's (9.17), but his per-90 figures this season are too small a sample and he starts 50% of the time: safer are Esposito (8.12, starting 70% of the time, 0.93 goals+assists per 90) and the midfielders, Diouf at 1.06 and Zieliński at 0.69 per 90, with Barella on a 7.62 average. Among the defenders, Hermoso (7.50) and Wesley (7.00) on one side, Bisseck (8.00, 0.48 per 90) on the other.

The risks. Roma concede almost nothing — one goal in four games, Svilar on a 6.00 average with 10 saves in 366 minutes — while Josep Martínez sits on 4.38 with 6 goals conceded in 360 minutes and a single clean sheet: in fantasy he is the goalkeeper to avoid. On cards, the 2025-26 numbers warn about Hermoso (0.39 per 90), Carlos Augusto (0.31), Mancini and Wesley (0.26), Barella (0.25) and Bastoni (0.24): in a game this heavy, defenders and holding midfielders are easy penalty points.

Penalties. The data say who has taken them, not who will take them: in 2025-26 Malen scored 3 and missed 1, Pellegrini scored 2 out of 2, while Dybala missed one without scoring any. At Inter the first taker was Çalhanoğlu (4 scored, 1 missed), but he is out, and the only other player with penalties taken is Zieliński (1 out of 1).

The absentees and what I would do. Roma report only Bah (knee); Inter lose Çalhanoğlu with a muscle problem, plus Stones and Spence, none with a stated return date — our data do not include suspensions, so the absence of a report is no guarantee that everyone else is available. In fantasy I would start Malen, Dybala and Svilar without a second thought on one side, Esposito and Barella on the other; I would leave out Josep Martínez and decide at the last moment between Thuram and Lautaro Martínez, with that one still wide open. Among defenders, go for Roma's: a side conceding 0.62 xG per game is a better clean-sheet bet than one conceding 0.87.