Frontier language models managing a Fantasy Premier League squad for the whole
season, all reading the same briefing, with every prompt and every decision published.
gameweek 2deadline 2026-08-28T17:30:00Z · Sat 29/8 00:30 WIB· LOCKEDbrief 06bc877ef742…ledger 2036c616fc4a…how this is published
Table after gameweek 2
Manager
Division
GW
Hits
Points
1
Flash
Paid
2
0
164
2
The Crowd
Control
2
0
150
3
Sol
Paid
2
0
147
4
DeepSeek
Paid
2
0
145
5
Kimi
Paid
2
0
141
6
Nemotron Ultra
Free
2
0
138
7
Grok
Paid
2
0
134
8
Opus
Paid
2
0
99
Committed squads — gameweek 2
Every squad below was committed before the deadline and hashed — the deadline has passed, so these are locked. The reasoning and the plan are each contestant's own words, published unedited —
including when they are wrong.
Pick a manager to see the eleven it picked, laid out on a pitch with its
bench underneath. List puts every squad side by side as tables, and the number on each tab is that manager’s season total.
Flash
Paid105
google/gemini-3.7-flash
No transfers — squad unchanged.
3-4-3105 pts
FlashGW2
Verbruggen
0
Calafiori
11
Shaw
2
Aina
0
Saka
11
Ndiaye
4
V
B.Fernandes
23
Mbeumo
11
João Pedro
9
Calvert-Lewin
8
C
Haaland
13
GKP
Dubravka
0
1. DEF
Davis
6
2. DEF
Targett
0
3. MID
Yates
0
fpl.wdnsds.com · gameweek 2 · Flash
DEF
Calafiori
ARS
£5.9
MID
Saka
ARS
£9.6
GKP
Verbruggen
BHA
£4.6
FWD
João Pedro
CHE
£7.8
MID
Ndiaye
MCI
£5.8
DEF
Davis
IPS
£4.0
FWD
Calvert-Lewin
LEE
£6.0
FWD
Haaland (C)
MCI
£15.7
DEF
Shaw
MUN
£4.3
MID
B.Fernandes (V)
MUN
£11.8
MID
Mbeumo
MUN
£7.8
DEF
Aina
NFO
£4.4
MID
Yates
NFO
£4.4
GKP
Dubravka
TOT
£4.0
DEF
Targett
HUL
£3.9
Why this squad
Rolling the free transfer into Gameweek 3 allows greater tactical flexibility with 2 free transfers once clearer trends emerge. Our starting eleven has strong fixtures for GW2, spearheaded by the Manchester United triple-up (Shaw, Fernandes, Mbeumo) at home to newly promoted Ipswich Town, alongside Haaland leading the line and in-form attackers Saka and João Pedro.
What it plans next
Bank the free transfer for GW2 to have 2 FTs available for Gameweek 3. In GW3, evaluate potential upgrades in defense or mid-priced midfield spots depending on early season form and minutes.
The Crowd
Control97
OUT Davis IN Shaw
4-4-297 pts
The CrowdGW2
Raya
6
Gabriel
8
★
Shaw
2
Diop
2
van Ewijk
2
B.Fernandes
23
Szoboszlai
4
Mbeumo
11
Ndiaye
4
C
Haaland
13
V
João Pedro
9
GKP
Dubravka
0
1. MID
Hughes
0
2. DEF
Thomas
8
3. FWD
Kusi-Asare
0
fpl.wdnsds.com · gameweek 2 · The Crowd
FWD
Haaland (C)
MCI
£15.7
FWD
João Pedro (V)
CHE
£7.8
MID
B.Fernandes
MUN
£11.8
MID
Szoboszlai
LIV
£6.8
GKP
Raya
ARS
£6.1
MID
Mbeumo
MUN
£7.8
DEF
Gabriel
ARS
£7.9
GKP
Dubravka
TOT
£4.0
DEF
Diop
IPS
£4.0
MID
Ndiaye
MCI
£5.8
DEF
van Ewijk
COV
£3.9
MID
Hughes
CRY
£4.4
DEF
Thomas
COV
£4.0
FWD
Kusi-Asare
FUL
£4.4
DEF
Shaw
MUN
£4.3
Why this squad
[CONTROL — mechanical, no judgement] 1 free transfer(s) toward the most-owned template, never taking a hit. Captain is the most-owned player in the XI; bench is ordered by ownership.
Sol
Paid103
openai/gpt-5.6-sol
OUT Beto IN Barry
3-4-3103 pts
SolGW2
Lammens
2
Calafiori
11
Thomas
8
Shaw
2
Saka
11
Palmer
7
V
B.Fernandes
23
E.Le Fée
3
★
Barry
2
Calvert-Lewin
8
C
Haaland
13
GKP
Dovin
0
1. DEF
Davis
6
2. DEF
Mitchell
0
3. MID
Hughes
0
fpl.wdnsds.com · gameweek 2 · Sol
DEF
Calafiori
ARS
£5.9
MID
Saka
ARS
£9.6
MID
Palmer
CHE
£9.7
GKP
Dovin
COV
£4.0
DEF
Thomas
COV
£4.0
DEF
Mitchell
CRY
£4.5
MID
Hughes
CRY
£4.4
FWD
Barry
EVE
£5.7
DEF
Davis
IPS
£4.0
FWD
Calvert-Lewin
LEE
£6.0
FWD
Haaland (C)
MCI
£15.7
GKP
Lammens
MUN
£4.9
DEF
Shaw
MUN
£4.3
MID
B.Fernandes (V)
MUN
£11.8
MID
E.Le Fée
SUN
£5.7
Why this squad
I used the free transfer to replace Beto with Barry, who started Everton's opening match, produced 0.52 xG per 90 and has a reasonable fixture at Bournemouth. Lammens, Shaw and Fernandes attack Ipswich at home, while Thomas has the strongest defensive fixture against Hull. Calafiori retains attacking potential despite the difficult trip to Villa. Palmer, Saka and Haaland remain the main premium threats, with Haaland captain because his 0.74 xG in the opening match was the strongest underlying goal threat among them. Davis is first on the bench because he is a secure starter and can cover any outfield absence while preserving a legal formation. No chip is justified this early.
What it plans next
I intend to roll the GW3 free transfer and captain Haaland at home to Coventry. In GW4 I intend to transfer Le Fée to Tavernier, retain the other free transfer, and keep all four chips unused through GW4.
DeepSeek
Paid99
deepseek/deepseek-v4-pro
No transfers — squad unchanged.
3-5-299 pts
DeepSeekGW2
Verbruggen
0
van Ewijk
2
Shaw
2
Aina
0
Saka
11
Kamada
2
B.Fernandes
23
V
Mbeumo
11
Gibbs-White
13
C
Haaland
13
João Pedro
9
GKP
Dovin
0
1. DEF
Mitchell
0
2. DEF
Diop
2
3. FWD
Furo
0
fpl.wdnsds.com · gameweek 2 · DeepSeek
MID
Saka
ARS
£9.6
FWD
Furo
BRE
£4.5
GKP
Verbruggen
BHA
£4.6
FWD
João Pedro
CHE
£7.8
GKP
Dovin
COV
£4.0
DEF
van Ewijk
COV
£3.9
DEF
Mitchell
CRY
£4.5
MID
Kamada
CRY
£5.0
DEF
Diop
IPS
£4.0
FWD
Haaland (C)
MCI
£15.7
DEF
Shaw
MUN
£4.3
MID
B.Fernandes
MUN
£11.8
MID
Mbeumo (V)
MUN
£7.8
DEF
Aina
NFO
£4.4
MID
Gibbs-White
NFO
£8.0
Why this squad
Saving the free transfer. I am starting 3-5-2 because the Man United trio of Shaw, B.Fernandes and Mbeumo have the strongest fixture at home to Ipswich, and loading five midfielders gives a higher floor than using Furo up front. Van Ewijk starts at home to Hull, while Mitchell and Diop drop to the bench because Palace face Man City and Ipswich face Man United away. Haaland keeps the armband, with Mbeumo as vice for the home fixture.
What it plans next
Roll the free transfer into GW3 so I have two free transfers and can respond to team news. Before GW3 I will likely address the Coventry defensive assets ahead of Man City away and look at improving the defence for the fixture swing. I am keeping the wildcard and all chips for later; no chip this week.
Kimi
Paid86
Not reached this gameweek. Its provider could not be
contacted before the deadline, so it keeps the squad it already had. This team will score,
but nobody chose it — it is not a decision and should not be read as one.
moonshotai/kimi-k3
No transfers — squad unchanged.
3-4-386 pts
KimiGW2
Raya
6
Gabriel
8
Guéhi
2
Shaw
2
V
B.Fernandes
23
Ndiaye
4
E.Le Fée
3
Wharton
2
C
Haaland
13
João Pedro
9
Beto
1
GKP
Palmer
0
1. DEF
Van Hecke
1
2. DEF
Diop
2
3. MID
Hughes
0
fpl.wdnsds.com · gameweek 2 · Kimi
GKP
Raya
ARS
£6.1
GKP
Palmer
IPS
£4.0
DEF
Gabriel
ARS
£7.9
DEF
Guéhi
MCI
£6.0
DEF
Van Hecke
TOT
£4.9
DEF
Shaw
MUN
£4.3
DEF
Diop
IPS
£4.0
MID
B.Fernandes (V)
MUN
£11.8
MID
Ndiaye
MCI
£5.8
MID
E.Le Fée
SUN
£5.7
MID
Wharton
CRY
£5.4
MID
Hughes
CRY
£4.4
FWD
Haaland (C)
MCI
£15.7
FWD
João Pedro
CHE
£7.8
FWD
Beto
EVE
£5.4
Why this squad
[AUTOMATIC FALLBACK] Kimi was never reached (HTTP 404 {"error":{"message":"No endpoints found for moonshotai/kimi-k3.","code":404},"user_id":"user_3BxxOAkjYLWzuJAnYm2rte9OnUd"}). Previous squad, lineup and armband held exactly as they were, no transfer, no chip.
Nemotron Ultra
Free71
nvidia/nemotron-3-ultra-550b-a55b:free
OUT Hinshelwood IN E.Le Fée
4-5-171 pts
Nemotron UltraGW2
Raya
6
Gabriel
8
Van Hecke
1
Kerkez
1
O'Reilly
2
Saka
11
C
Palmer
7
V
Wirtz
4
Groß
13
★
E.Le Fée
3
Isak
8
GKP
Steele
0
1. FWD
Watkins
0
2. FWD
Delap
1
3. DEF
Mitchell
0
fpl.wdnsds.com · gameweek 2 · Nemotron Ultra
GKP
Raya
ARS
£6.1
DEF
Gabriel
ARS
£7.9
MID
Saka
ARS
£9.6
FWD
Watkins
AVL
£7.8
GKP
Steele
BHA
£4.0
DEF
Van Hecke
TOT
£4.9
MID
Groß
BHA
£6.1
MID
Palmer (C)
CHE
£9.7
FWD
Delap
NFO
£5.5
DEF
Mitchell
CRY
£4.5
DEF
Kerkez
LIV
£5.3
MID
Wirtz (V)
LIV
£7.2
FWD
Isak
LIV
£9.0
DEF
O'Reilly
MCI
£6.3
MID
E.Le Fée
SUN
£5.7
Why this squad
Used sole free transfer to replace doubtful Hinshelwood (25% chance) with E.Le Fée (Sunderland, 6.0m) who has a favourable home fixture vs Fulham (FDR 2). Formation 4-4-2: Raya in goal despite tough away trip to Villa given Arsenal's defensive solidity; back four of Gabriel, Van Hecke (Spurs home vs Newcastle FDR 2), Kerkez (Liverpool home vs Forest FDR 3) and O'Reilly (Man City away at Palace FDR 3) all have decent to good fixtures; Mitchell benched due to Palace hosting Man City (FDR 4). Midfield pivots on Palmer (Chelsea home vs Brighton FDR 2) and Wirtz (Liverpool home vs Forest FDR 3) with Saka retained for Arsenal's attacking upside; Groß benched due to Brighton's difficult trip to Chelsea (FDR 4). Front two are Isak (Newcastle at Spurs FDR 3) and Watkins (Villa home vs Arsenal FDR 4) with Delap on bench after Forest's daunting away trip to Liverpool. Captain Palmer for Chelsea's home advantage and his GW1 form (13 pts); vice Wirtz for Liverpool's home match.
What it plans next
Hold Wildcard for GW7-8 when fixture swings clarify. Next GW assess Steele's viability and consider upgrading to a Sunderland keeper (Roefs/Kinsky) if fixtures remain kind. Monitor Saka/Arsenal away form; if blank, look to move to Salah or Palmer cover. Target double-gameweek assets in GW12+ window.
Grok
Paid86
x-ai/grok-4.6
OUT Raya, Calafiori, N.Williams, Semenyo, Gibbs-White, Hughes, Calvert-Lewin IN Tzolakis, Kayode, Egan, Palmer, Szoboszlai, Ndiaye, João Pedro
4-4-2Wildcard86 pts
GrokGW2
★
Tzolakis
10
van Ewijk
2
Guéhi
2
Shaw
2
★
Kayode
2
Saka
11
V★
Palmer
7
Mbeumo
11
★
Szoboszlai
4
C
Haaland
13
★
João Pedro
9
GKP
Verbruggen
0
1. MID
★
Ndiaye
4
2. DEF
★
Egan
9
3. FWD
Obi
0
fpl.wdnsds.com · gameweek 2 · Grok
GKP
Verbruggen
BHA
£4.6
GKP
Tzolakis
HUL
£4.7
DEF
Guéhi
MCI
£6.0
DEF
Kayode
BRE
£4.6
DEF
Shaw
MUN
£4.3
DEF
van Ewijk
COV
£3.9
DEF
Egan
HUL
£4.1
MID
Saka
ARS
£9.6
MID
Palmer (V)
CHE
£9.7
MID
Szoboszlai
LIV
£6.8
MID
Mbeumo
MUN
£7.8
MID
Ndiaye
MCI
£5.8
FWD
Haaland (C)
MCI
£15.7
FWD
João Pedro
CHE
£7.8
FWD
Obi
MUN
£4.5
Chip played: Wildcard
Why this squad
Wildcarded in GW2 because the inherited squad was locked into three-club piles and carried two expensive mids (Semenyo, Gibbs-White) with weak underlying numbers plus Forest travelling to Liverpool. The rebuild is a 99.7m spine of Haaland, Saka, Palmer and Mbeumo, with Szoboszlai in for Liverpool’s kind run (Forest, Ipswich, Fulham) and João Pedro for Chelsea’s home match against Brighton. Defence is cheap on purpose: Guéhi for City’s chance volume, Kayode for Brentford’s attack, Shaw against Ipswich, van Ewijk at home to Hull, Egan as a 90-minute Hull defcon piece. Two 4.5 keepers, Tzolakis starting against Coventry. I dropped Raya and the third Arsenal slot to fund Palmer and João Pedro without crossing 100m.
What it plans next
Triple Captain Haaland in GW3 at home to Coventry if he is named to start. No other chip before the GW20 window. The GW3 free transfer is reserved for an injury; if the XI is fit I will roll it. After Chelsea’s trip to Arsenal in GW3 I will review João Pedro — if his minutes drop I move him to a Liverpool or Brentford forward for GW4. Palmer, Saka, Haaland and Mbeumo stay unless injured or the second wildcard is played.
Opus
Paid60
anthropic/claude-opus-5
OUT Rice IN Szoboszlai
4-4-260 pts
OpusGW2
Verbruggen
0
Gabriel
8
Van Hecke
1
Guéhi
2
▼
Senesi
0
V★
Szoboszlai
4
Wilson
1
Semenyo
5
E.Le Fée
3
Thiago
2
C
Haaland
13
GKP
Palmer
0
1. FWD
▲
Calvert-Lewin
8
2. DEF
Diop
2
3. MID
Hughes
0
fpl.wdnsds.com · gameweek 2 · Opus
GKP
Verbruggen
BHA
£4.6
GKP
Palmer
IPS
£4.0
DEF
Gabriel
ARS
£7.9
DEF
Van Hecke
TOT
£4.9
DEF
Diop
IPS
£4.0
DEF
Guéhi
MCI
£6.0
DEF
Senesi
TOT
£5.7
MID
Szoboszlai (V)
LIV
£6.8
MID
Hughes
CRY
£4.4
MID
Wilson
LEE
£6.2
MID
Semenyo
MCI
£8.3
MID
E.Le Fée
SUN
£5.7
FWD
Thiago
BRE
£7.8
FWD
Calvert-Lewin
LEE
£6.0
FWD
Haaland (C)
MCI
£15.7
Why this squad
One free transfer, no money in the bank, so I looked for the single change that raises the ceiling of my starting eleven without costing anything: Rice (7.5m) out, Szoboszlai (7.0m) in. Rice is priced as a premium midfielder but his underlying numbers are those of a holding player — 0.09 xG90, 0.11 xA90 and 8.1 defensive actions per 90, which is below the 12-action threshold midfielders need for a defensive contribution point. He is paying for neither attack nor defcon. Szoboszlai costs half a million less, posted 0.85 xG90 and 0.27 xA90 with a full 90 minutes in GW1, and Liverpool are at home to Nott'm Forest. That is a straight upgrade plus 0.5m banked for future flexibility.
Haaland captains. He is the most expensive asset in the game, took 90 minutes in GW1, and Man City travel to Crystal Palace at a modest difficulty rating — I am not looking for a clever armband in week two. Szoboszlai is vice as the highest-ceiling starter with the friendliest fixture behind him.
The eleven is 4-4-2. I start both Spurs defenders despite their ugly GW1 goals-conceded numbers: Senesi's 16 defensive actions per 90 makes him a live defcon candidate almost regardless of the scoreline, and Spurs are at home in the easiest-rated fixture of my defenders' week. Gabriel and Guehi complete the back line — Guehi's GW1 (a goal, 10 points) came from a genuinely attacking full-back/centre-back role in a Man City side that should dominate at Palace.
Calvert-Lewin is benched ahead of Van Hecke because his profile is thin — 0.13 xG90, one point in GW1 — but he is first sub because he is guaranteed 90 minutes for Leeds at home, which is exactly what you want from the man who covers a non-starter. Diop follows him: another nailed 90 minutes, though Ipswich away at Man Utd is a poor spot. Hughes is third because he has yet to play a minute, and Palmer is last since he is a non-playing backup keeper who can only ever replace Verbruggen anyway.
No chip. Chips are worth using on a fixture or a squad state you have engineered, not on the second week of the season with a squad I have not yet shaped.
What it plans next
Short term I bank my transfer discipline rather than my transfers: I will make at most one move per week and only if it clearly improves the eleven, rolling to two free transfers if nothing does. The two things I actively want to fix are Hughes (4.5m, zero minutes, dead weight on the bench) and Calvert-Lewin (6.0m forward with a 0.13 xG90 who is currently my first substitute). Selling Calvert-Lewin plus the 0.5m in the bank gets me to 6.5m, which is enough for a genuine third forward or a fifth attacking midfielder, and that is my intended GW3 or GW4 move. Hull's cheap defenders (Mendy and Ajayi at 4.0m, both with heavy defensive-action counts) are on my shortlist to replace Diop or Hughes as bench cover that can actually score.
I am not planning a wildcard. If by GW5 I find myself with three or more non-contributing players and no cash to fix them one at a time, I will play it then and rebuild around a defence with real defcon volume rather than reputation. Otherwise it stays in my pocket for the second half of the window.
Haaland keeps the armband every week unless he is flagged; I will not chase the captaincy around the fixture list at this price point. I will hold the Triple Captain for a home Man City fixture against a bottom-half side, not for an away trip, and I will not consider Bench Boost until my bench is four players who all start for their clubs — right now two of them are not, so boosting it would be burning a chip.
Not shown: Nemotron Super (withdrawn by its provider on 2026-08-24), Gemma (withdrawn by its provider on 2026-08-24), GPT-OSS (withdrawn by its provider on 2026-08-24), Human (no squad committed). They hold no squad this gameweek, so there is nothing to display — every
attempt, including the failures, remains in the published ledger.
Withdrawn mid-season
Nemotron Super (nvidia/nemotron-3-super-120b-a12b:free) — stopped being served on
2026-08-24. Never held a squad. Given the identical brief and three corrective attempts in GW1 it returned three different illegal squads (wrong position quotas, then a club-limit breach, then quotas and budget together), taking 32 minutes to do so. It understands the rules — it scored 6/6 on a chip-comprehension check — but could not assemble a legal 15, which is the minimum this competition asks of an entrant.
Gemma (google/gemma-4-31b-it:free) — stopped being served on
2026-08-24. Never once reached. Every request across a week returned HTTP 429 from Google AI Studio’s shared free pool ("temporarily rate-limited upstream"), including three spaced retries in the GW1 run and a further 24 automated attempts over the following day. The model itself was never at fault and is not being judged; it simply cannot be given the brief, and a contestant that cannot receive the brief cannot compete.
GPT-OSS (openai/gpt-oss-20b:free) — stopped being served on
2026-08-24. OpenRouter delisted the :free variant. GET /models no longer returns it, and a request returns HTTP 404: "This model is unavailable for free. The paid version is available now - use this slug instead: openai/gpt-oss-20b". The paid slug is NOT substituted: a paid model cannot stand in for a free-division contestant, and a silent swap would make every gameweek before the change incomparable with every one after it.
Its earlier gameweeks stand in the ledger exactly as they were scored. No replacement
model runs under its name, because a comparison between named models cannot survive one
being substituted for another partway through.
Checked against the real FPL teams
These squads are also entered as ordinary Fantasy Premier League teams, by
hand. That is a transcription, and transcriptions can go wrong — so each gameweek the
entries are read back from the public API and compared field by field. Where they differ,
the ledger is what the contestant decided and the entry is where the mistake is.
3 of 7
match exactly on squad, starting XI and armband: Sol, Nemotron Ultra, Opus.
Flash
armband
decided Haaland (C), B.Fernandes (V)
entered Haaland (C), Saka (V)
DeepSeek
armband
decided Haaland (C), Mbeumo (V)
entered Haaland (C), B.Fernandes (V)
Kimi
bench order
decided Palmer, Van Hecke, Diop, Hughes
entered Palmer, Diop, Hughes, Van Hecke
Grok
squad
decided Verbruggen, Tzolakis, Guéhi, Kayode, Shaw, van Ewijk, Egan, Saka, Palmer, Szoboszlai, Mbeumo, Ndiaye, Haaland, João Pedro, Obi
entered Palmer, Kayode, Guéhi, Egan, Shaw, Saka, Szoboszlai, Ndiaye, Mbeumo, Haaland, João Pedro, Verbruggen, van Ewijk, Tzolis, Obi
xi
decided Tzolakis, van Ewijk, Guéhi, Shaw, Kayode, Saka, Palmer, Mbeumo, Szoboszlai, Haaland, João Pedro
entered Palmer, Kayode, Guéhi, Egan, Shaw, Saka, Szoboszlai, Ndiaye, Mbeumo, Haaland, João Pedro
armband
decided Haaland (C), Palmer (V)
entered Haaland (C), Saka (V)
Points on this page are scored from the ledger — what each contestant
actually chose. A mini-league total that differs is the transcription, and it is left
visible here rather than reconciled away.
Points on this page against the FPL mini-league
The squads are also entered by hand as real FPL teams. This page scores what each
contestant decided; FPL scores what was entered. Where the entry went wrong, the two part
company. FPL's figures below are FPL's own — its points minus its transfer cost, exactly as
its league table adds them up, with its own captain multipliers and automatic substitutions.
Nothing on the FPL side is recalculated here.
Manager
This page
FPL
Difference
Flash
164
164
·
Sol
147
147
·
DeepSeek
145
145
·
Kimi
141
141
·
Nemotron Ultra
138
138
·
Grok
134
128
−6
Opus
99
93
−6
By gameweek, FPL minus this page. A dot means the two agree
to the point.
Manager
GW1
GW2
Flash
·
·
Sol
·
·
DeepSeek
·
·
Kimi
·
·
Nemotron Ultra
·
·
Grok
·
−6
Opus
·
−6
Grok, gameweek 2: −6 — this page 86, FPL 80.
In FPL, a different player was bought and a different starting XI was entered. The decision counted Tzolakis 10, Palmer 7, van Ewijk 2. FPL counted Egan 9, Ndiaye 4 instead.
Opus, gameweek 2: −6 — this page 60, FPL 54.
In FPL, a different substitute came on. The decision counted Calvert-Lewin 8. FPL counted Diop 2 instead.
A cause is printed only when both sides’ player-by-player points add up to their
own totals. The site’s score stands in every case: it is the decision, and a mis-entered
team does not change what the contestant chose.
The rules this runs under
Published in advance, because a protocol written after the results exist is
worth nothing.
One briefing, byte for byte. Every contestant receives an identical brief — the
full player list with prices and published availability, fixtures with difficulty, the
scoring rules, and its own squad state. Nothing else. Its SHA-256 is
06bc877ef742db13… and it is committed before the deadline.
No advice inside it. Players are listed in id order, never ranked. No
recommendations, no highlights, no derived "best" anything. The judgement is the thing
being measured, so the brief must not contain any.
Prior football knowledge counts. Models know different amounts about the Premier
League from training, and that is treated as part of the contest rather than a flaw to
be corrected. Cutoffs are published alongside results.
No tools, no web access. Prior knowledge is fair; looking up results that have
already happened is not.
No sampling parameters. Not "temperature 0" — some providers reject the parameter
outright, so equalising it would succeed for some contestants and fail for others. Every
call uses provider defaults, and reproducibility is not claimed.
The same referee for everyone. An illegal squad is returned with the exact
violation and re-asked, up to 3 attempts. Then the previous squad
is held unchanged — never a fresh pick, which would be a strategy. How often each
contestant needs correcting is published as a result, not hidden as plumbing.
Chips are explained, not advised. From gameweek 2 the briefing states what each
chip does and the gameweeks it is valid in, because those are rules — the same reason the
full scoring table is given. When to play one is judgement and is never suggested. This
was added mid-season and is recorded here rather than slipped in: in gameweek 1 the
briefing named the available chips without explaining them, so every contestant relied on
its own prior knowledge of what they do.
Scores are FPL's own. Points are re-aggregated from the official live endpoint,
never recomputed. Only captaincy, substitutions, chips and transfer hits are applied here.
Checking this yourself
Every claim on this page is meant to be falsifiable by a stranger.
The brief.gw2-common.txt is the exact text every
contestant received. Hash it and compare against the value above.
The decisions.gw2-ledger.json holds each raw response
with the prompt hash and the model that actually served the request. Entries are chained,
so altering an old one breaks every hash after it.
The scores. Points come from the official live endpoint, so a squad and a gameweek
are enough to recompute any total here.