xgEdge

Fantasy Premier League

Who to captain

The captain's armband doubles whatever happens, so an average is the wrong number to pick on. Every candidate here carries the shape of his week as well as its middle — how bad it reasonably gets, how good it reasonably gets, and how often it is a haul.

The model's captain

Bukayo Saka

MID · LEE (H) · 9.5m · owned by 14%

4.6

Projected

floor 2 7.5% a haul ceiling 8

Doubled by the captain's armband that is 4.6 points, from a floor of 4.

Gameweek 6 · 3-round window · feed of 01/10 11:39

Bukayo Saka is the model's captain for gameweek 6, at 4.57 projected points. It is a close call — only 0.01 separates him from Erling Haaland, which is inside the model's own error.

His modelled week runs from 2 to 8 points before the captain's armband doubles it, and reaches 10 or more 8% of the time. At safe the board puts Erling Haaland first, owned by 73.8%. The model's own pick is still Bukayo Saka, and he is marked as such in the list below.

His fixture

2 of 5

The model rates his home fixture a 2 out of five.

His ceiling

8.0

His is the highest ceiling on the board at 8.0 points, and he hauls 8% of the time.

His start

92%

He is 92% to start, which is about as close to nailed on as this model ever says.

Best differential

Bobby Thomas is the best differential on the board: owned by 8.5%, projected at 4.42.

See the differentials board →

The top five on this board

The board as your controls leave it — the same five that open the table below.

  1. Ranked 1 Erling Haaland
    FWD vs
    4.6 Pts
    floor 2 ceiling 6
    Haul
    7.8%
    Start
    92%
    Own
    74%

    LIV (A)

  2. Ranked 2 Bruno Fernandes
    MID vs
    4.4 Pts
    floor 2 ceiling 8
    Haul
    7.1%
    Start
    92%
    Own
    39%

    TOT (H)

  3. Ranked 3 Riccardo Calafiori
    DEF vs
    4.1 Pts
    floor 1 ceiling 7
    Haul
    4.1%
    Start
    92%
    Own
    51%

    LEE (H)

  4. Ranked 4 David Raya Martín
    GKP vs
    4.1 Pts
    floor 1 ceiling 7
    Haul
    0.0%
    Start
    92%
    Own
    42%

    LEE (H)

  5. Ranked 5 Bukayo Saka
    MID vs
    4.6 Pts
    floor 2 ceiling 8
    Haul
    7.5%
    Start
    92%
    Own
    14%

    LEE (H)

Enter your squad to narrow this to the players you own →
Showing 40 of 162 players 162 of 487 match your filters Clear filters
FWD vs LIV (A) · 15.6m
2 6
Haul
7.8%
Start
92%
Own
74%
MID vs TOT (H) · 11.9m
2 8
Haul
7.1%
Start
92%
Own
39%
DEF vs LEE (H) · 5.8m
1 7
Haul
4.1%
Start
92%
Own
51%
GKP vs LEE (H) · 6.1m
1 7
Haul
0.0%
Start
92%
Own
42%
5 Bukayo Saka 4.6
MID vs LEE (H) · 9.5m
2 8
Haul
7.5%
Start
92%
Own
14%
6 Bryan Mbeumo 4.4
MID vs TOT (H) · 7.9m
2 8
Haul
7.3%
Start
92%
Own
21%
DEF vs LEE (H) · 8.0m
1 8
Haul
4.6%
Start
92%
Own
23%
8 Bobby Thomas 4.4
DEF vs NEW (H) · 4.0m
1 8
Haul
5.5%
Start
92%
Own
8.5%
MID vs BOU (H) · 7.7m
2 7
Haul
4.4%
Start
92%
Own
41%
10 Pascal Groß 3.8
MID vs SUN (A) · 5.8m
2 7
Haul
3.3%
Start
92%
Own
29%
MID vs LEE (H) · 6.8m
2 7
Haul
3.3%
Start
92%
Own
20%
MID vs MCI (H) · 7.0m
2 7
Haul
4.1%
Start
92%
Own
33%
DEF vs HUL (A) · 6.2m
1 8
Haul
2.2%
Start
92%
Own
17%
MID vs CRY (A) · 8.0m
2 7
Haul
3.9%
Start
92%
Own
16%
15 Thierno Barry 4.0
FWD vs HUL (A) · 5.7m
2 6
Haul
7.3%
Start
92%
Own
8.3%
16 John Egan 4.0
DEF vs EVE (H) · 4.1m
1 8
Haul
3.0%
Start
92%
Own
7.0%
17 Alex Scott 4.0
MID vs CHE (A) · 6.1m
2 7
Haul
2.7%
Start
92%
Own
5.8%
18 Igor Thiago 3.9
FWD vs AVL (A) · 7.8m
2 6
Haul
8.3%
Start
92%
Own
9.1%
19 Yoane Wissa 3.7
FWD vs COV (A) · 6.2m
2 6
Haul
6.2%
Start
92%
Own
19%
20 Enzo Le Fée 4.0
MID vs BHA (H) · 5.7m
2 8
Haul
6.2%
Start
92%
Own
3.4%
MID vs CHE (A) · 6.1m
2 7
Haul
3.6%
Start
92%
Own
6.8%
22 Vitaly Janelt 4.0
MID vs AVL (A) · 5.0m
2 7
Haul
3.3%
Start
92%
Own
2.5%
23 Granit Xhaka 4.0
MID vs BHA (H) · 5.5m
2 7
Haul
3.4%
Start
92%
Own
3.6%
24 Kevin Schade 3.8
MID vs AVL (A) · 6.2m
2 7
Haul
3.3%
Start
92%
Own
12%
GKP vs EVE (H) · 4.7m
1 7
Haul
0.0%
Start
92%
Own
13%
FWD vs IPS (A) · 6.0m
2 6
Haul
6.8%
Start
92%
Own
4.0%
27 Senne Lammens 3.8
GKP vs TOT (H) · 4.9m
1 7
Haul
0.0%
Start
92%
Own
9.9%
28 Lewis Hall 3.6
DEF vs COV (A) · 5.3m
1 7
Haul
2.3%
Start
92%
Own
17%
GKP vs ARS (A) · 5.0m
1 7
Haul
0.0%
Start
92%
Own
5.5%
DEF vs CHE (A) · 5.0m
1 8
Haul
4.5%
Start
92%
Own
0.4%
31 Murillo 3.8
DEF vs CRY (A) · 5.5m
1 8
Haul
3.4%
Start
92%
Own
2.1%
32 Anton Stach 3.7
MID vs ARS (A) · 6.0m
2 7
Haul
1.8%
Start
92%
Own
2.5%
DEF vs NFO (H) · 4.9m
1 8
Haul
2.8%
Start
92%
Own
0.6%
34 Matheus Cunha 3.6
MID vs TOT (H) · 7.9m
2 7
Haul
2.6%
Start
92%
Own
5.8%
35 Jay Dasilva 3.7
DEF vs NEW (H) · 4.0m
1 7
Haul
3.6%
Start
92%
Own
0.7%
36 Emersonn 3.6
FWD vs FUL (H) · 5.5m
1 6
Haul
4.8%
Start
92%
Own
3.5%
37 Oli McBurnie 3.6
FWD vs EVE (H) · 5.5m
2 6
Haul
6.5%
Start
92%
Own
1.9%
GKP vs MCI (H) · 5.5m
1 7
Haul
0.0%
Start
92%
Own
4.1%
MID vs HUL (A) · 6.6m
2 7
Haul
2.8%
Start
92%
Own
6.0%
DEF vs HUL (A) · 5.5m
1 8
Haul
2.3%
Start
92%
Own
2.4%
the week a footballer's usually falls in the projection every bar on one axis of 0 to 12 points

What a floor and a ceiling are here

The floor is the 10th percentile of a modelled week and the ceiling is the 90th — the range a footballer's week usually falls in, not the worst and best he has ever had. The haul column is how often that week reaches 10 points.

Bonus points are projected nowhere on this site, so the ceiling is a ceiling on modelled scoring rather than on Fantasy points. Every other figure here excludes them too.

The spread is composed from the same parts the projection beside it is: its average is that projection, so the two cannot tell a reader different things about the same footballer.

The setting reweights the ranking by ownership and changes no arithmetic of its own. The model's own pick stays on the page at every setting, and the verdict at the top is always his.

Every figure here is what he scores, not what the captain's armband makes of it. Captaining him doubles all of it — the floor as well as the ceiling.

The board is every player the game would let you field, the flagged included — whether a doubtful player is worth picking is a judgement the projection makes with his published chance of playing, not one made by hiding him.

What is measured here, and what is not

  • Measured The projected points

    Replayed against gameweek 2 after it was played: an average miss of 1.617 points a player, against 1.842 for “predict what he has averaged so far”, over 456 players. Recorded Sunday, 6 September 2026 and refreshed by hand, so the date is part of the claim.

  • Not measured The floor, the ceiling and the chance of a haul

    Percentiles of a week composed from the same rates the projection uses. They reconcile with it by construction, which is not the same as having been checked against a round anybody played — nothing on this site has scored a floor or a ceiling.