0000
Explore · Players
Defender for Chelsea, £4.8m.
DefenderChelsea
Levi Samuels Colwill
Low goal threat, 0.00 xG per 90, 59 of 62 defenders · 450+ minutes in 2026/27.
xPts next
1.4
His projected score in the next gameweek.
xPts window
6.1
5 fixtures in GW8–12.
xMins
28
34% to start, 28% to reach 60 minutes.
Ownership
0.8%
Effective ownership adds captaincy share, which FPL does not publish.
Form
4.0
FPL’s own recent-form average.
Projection · basis: xPts: GW8 H v TOT (FDR 2). fixture adj from fixtures[].team_*_difficulty + teams[].strength_overall_home/away (league mean quality 3.10) - strength_attack_*/strength_defence_* are 0 in preseason. Minutes from history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7. Rates from 1192 weighted minutes of history_past, blended 76% own record / 24% DEF/4.6-5.5m cohort (n=63, fitted on a fixed population of 449 players with a completed season) by share of the points they earn. Scoring from the scoring rules resolved from FPL’s configuration. Unpublished numbers from one shared table of figures FPL has not published. · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
What this projection could see
Projected from his full record: 4 played seasons. The same record every screen reads, so this figure matches the ranked list, the pitch and a comparison.
Ranked among defenders with 450+ minutes on record.
Low goal threat
0.00 xG per 90
59 of 62 defenders · 450+ minutes in 2026/27
The biggest single scoring event, and the part that repeats.
Low chance creation
0.01 xA per 90
59 of 62 defenders · 450+ minutes in 2026/27
The second scoring route, and it survives a missed chance.
The schedule ahead
The schedule the projection above is spread across. Each cell is the opponent, home or away, and how hard the match is on FPL’s one-to-five scale, where one is the easiest.
Set-piece duty
FPL lists him in no set-piece order. That is published, not missing.
Official · basis: bootstrap-static/elements[149].penalties_order, FPL's published penalty order · updated just now · confidence: high
Availability
FPL flags him as fit and has published no note against him.
No chance-of-playing percentage is published, which for a fit player is normal.
Predicted lineups are not modelled here. FPL publishes none, and a guessed eleven could not be checked.
Price
£4.8m
Official · basis: bootstrap-static/elements[149].now_cost, in tenths of a million · updated just now · confidence: high
Since the season opened
-0.2m
Official · basis: bootstrap-static/elements[149].cost_change_start, movement from this season's opening price · updated just now · confidence: high
Owned
0.8%
Official · basis: bootstrap-static/elements[149].selected_by_percent, share of all FPL squads · updated just now · confidence: high
Price history within a season is not published as a series, only the net movement from the opening price.
2025/26
8 pts · 0 goals · 0 assists
Points, by where they come from
P(start) 0.34 x [P(60+|start) 0.84 -> 2pt, else 1pt] + P(cameo) 0.13 x 1pt
Put him in the squad and the rules answer whether he fits, or set him beside others to compare.
Low attacking involvement
0.01 xGI per 90
62 of 62 defenders · 450+ minutes in 2026/27
The single best summary of a team’s attacking play.
Modest defensive work
6.9 defensive actions per 90
39 of 62 defenders · 450+ minutes in 2026/27
Two repeatable points that do not need a clean sheet.
Modest defensive solidity
1.46 xGC per 90
Top 49% of 62 defenders · 450+ minutes in 2026/27
Where defenders and keepers make most of their points.
Low starting security
16% of team games started
39 of 62 defenders · 450+ minutes in 2026/27
Minutes decide an FPL asset more than ability does.
Low bonus pull
9.0 BPS per 90
46 of 62 defenders · 450+ minutes in 2026/27
The BPS table decides bonus points, and this is it.
Low value
2.3 points per £m
59 of 62 defenders · 450+ minutes in 2026/27
What decides whether a player earns his squad slot.
Effective ownership needs captaincy share too, and the public API publishes none. The figure above is ownership alone.
2024/25
109 pts · 2 goals · 1 assists
2023/24
65 pts · 1 goals · 1 assists
2022/23
46 pts · 0 goals · 2 assists
P(60+) 0.28 x P(0 conceded) 0.42 (xGC90 1.33 x fixture 0.78) x 4pt
xG90 0.06 x fixture 1.23 x 0.31 of a match x 6pt
P(start) 0.34 x E[floor(conceded/2)] at lambda 0.86 over 75 mins + P(cameo) 0.13 x E[floor(conceded/2)] at lambda 0.25 over 22 mins x -1pt (xGC90 1.33 x fixture 0.78)
yellow90 0.19, red90 0.002 x 0.31 of a match
bonus90 0.17 x sqrt(fixture) 1.11 x 0.31 of a match
xA90 0.04 x fixture 1.23 x 0.31 of a match x 3pt
P(>=10 defensive actions) 0.07 at 6.9/90 x P(start) 0.34 x 2pt
Projection · basis: xPts: GW8 H v TOT (FDR 2). fixture adj from fixtures[].team_*_difficulty + teams[].strength_overall_home/away (league mean quality 3.10) - strength_attack_*/strength_defence_* are 0 in preseason. Minutes from history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7. Rates from 1192 weighted minutes of history_past, blended 76% own record / 24% DEF/4.6-5.5m cohort (n=63, fitted on a fixed population of 449 players with a completed season) by share of the points they earn. Scoring from the scoring rules resolved from FPL’s configuration. Unpublished numbers from one shared table of figures FPL has not published. · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
The spread around it
Ceiling (90th)
6.0
Projection · basis: ceiling (90th percentile): GW8 H v TOT (FDR 2). fixture adj from fixtures[].team_*_difficulty + teams[].strength_overall_home/away (league mean quality 3.10) - strength_attack_*/strength_defence_* are 0 in preseason. Minutes from history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7. Rates from 1192 weighted minutes of history_past, blended 76% own record / 24% DEF/4.6-5.5m cohort (n=63, fitted on a fixed population of 449 players with a completed season) by share of the points they earn. Scoring from the scoring rules resolved from FPL’s configuration. Unpublished numbers from one shared table of figures FPL has not published. · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
Floor (10th)
0.0
Projection · basis: floor (10th percentile): GW8 H v TOT (FDR 2). fixture adj from fixtures[].team_*_difficulty + teams[].strength_overall_home/away (league mean quality 3.10) - strength_attack_*/strength_defence_* are 0 in preseason. Minutes from history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7. Rates from 1192 weighted minutes of history_past, blended 76% own record / 24% DEF/4.6-5.5m cohort (n=63, fitted on a fixed population of 449 players with a completed season) by share of the points they earn. Scoring from the scoring rules resolved from FPL’s configuration. Unpublished numbers from one shared table of figures FPL has not published. · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
Chance of 10+
1%
Projection · basis: P(10+ points): GW8 H v TOT (FDR 2). fixture adj from fixtures[].team_*_difficulty + teams[].strength_overall_home/away (league mean quality 3.10) - strength_attack_*/strength_defence_* are 0 in preseason. Minutes from history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7. Rates from 1192 weighted minutes of history_past, blended 76% own record / 24% DEF/4.6-5.5m cohort (n=63, fitted on a fixed population of 449 players with a completed season) by share of the points they earn. Scoring from the scoring rules resolved from FPL’s configuration. Unpublished numbers from one shared table of figures FPL has not published. · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
Chance of 2 or fewer
84%
Projection · basis: P(2 or fewer points): GW8 H v TOT (FDR 2). fixture adj from fixtures[].team_*_difficulty + teams[].strength_overall_home/away (league mean quality 3.10) - strength_attack_*/strength_defence_* are 0 in preseason. Minutes from history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7. Rates from 1192 weighted minutes of history_past, blended 76% own record / 24% DEF/4.6-5.5m cohort (n=63, fitted on a fixed population of 449 players with a completed season) by share of the points they earn. Scoring from the scoring rules resolved from FPL’s configuration. Unpublished numbers from one shared table of figures FPL has not published. · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
The minutes model
Starts
34%
Projection · basis: P(start): history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7 · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
Reaches 60 minutes
28%
Projection · basis: P(reaching the full appearance threshold): threshold 60 minutes, in force; history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7 · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
Off the bench
13%
Projection · basis: P(cameo): history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7 · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
Does not appear
53%
Projection · basis: P(unused): history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7 · updated just now · confidence: medium
A projection, not a result. It states the most likely outcome, and football regularly disagrees.
history_past 2025/26 (w0.60, 225min), 2024/25 (w0.25, 3150min), 2023/24 (w0.15, 1794min); status=a (available); age 23.7
What this model does not include
Read before trusting the number
What FPL itself projects for GW8
FPL’s ep_next
4.0
Official · basis: FPL's own expected points for the next gameweek: bootstrap-static/elements[149].ep_next (FPL's own next-gameweek projection) · updated just now · confidence: high
Shown beside xPts, not instead of it. FPL’s figure is quantised to one decimal in preseason, so agreement in order is mild evidence.
What this is built from
Built from FPL’s own published season data for 2026/27, using a fixed population of 449 players with a completed season. Everything else is the stated default.