Match spotlight · xGB Core · Mon 31 Aug, 20:00

55%ARS WIN

Aston Villa vs Arsenal

The model's marquee match: 20% home, 55% away. Play this fixture 10,000 times and expect AVL winning 1990, the draw 2490, ARS winning 5520.

⚡ Ollie Watkins 29% to scoreOver 2.5 goals · 54%Full assessment →

AVL WIN20%DRAW25%ARS WIN55%

1–2

Likeliest ARS win

Called 12 of 19finished matches this round — every call frozen before kickoff, graded at full time

Gameweek 2 board — all assessments

Draw watchSUN—FUL29%COV—HUL28%TOT—NEW26%the week's likeliest stalemates

AVL — ARSMon 20:00

H 20%D 25%A 55%

CRY — MCI1–4

H 27%D 24%A 49%

LIV — NOT2–2

H 60%D 22%A 18%

BOU — EVE1–1

H 55%D 24%A 21%

COV — HUL0–1

H 43%D 28%A 30%

TOT — NEW0–2

H 36%D 26%A 38%

SUN — FUL1–0

H 38%D 29%A 33%

CHE — BHA4–3

H 43%D 25%A 32%

LEE — BRE1–1

H 41%D 25%A 33%

MUN — IPS5–2

H 74%D 17%A 10%

Finished — graded

MUN — IPS5–2

✓ calledgave this 74%

SUN — FUL1–0

✓ calledgave this 38%

CHE — BHA4–3

✓ calledgave this 43%

LEE — BRE1–1

✗ missedgave this 25%

TOT — NEW0–2

✓ calledgave this 38%

BOU — EVE1–1

✗ missedgave this 24%

COV — HUL0–1

✗ missedgave this 30%

LIV — NOT2–2

✗ missedgave this 22%

CRY — MCI1–4

✓ calledgave this 49%

FUL — CHE2–3

✓ calledgave this 39%

NEW — LIV2–2

✗ missedgave this 25%

MCI — BOU2–1

✓ calledgave this 58%

BHA — AVL4–0

✓ calledgave this 44%

BRE — TOT3–0

✓ calledgave this 51%

IPS — SUN2–1

✗ missedgave this 26%

NOT — LEE0–1

✓ calledgave this 37%

EVE — CRY2–0

✓ calledgave this 39%

HUL — MUN2–0

✗ missedgave this 16%

ARS — COV3–0

✓ calledgave this 83%

Gameweek 3 — next up

Assessed already, before a ball is kicked. These re-price as this week’s results land.

Draw watchNOT—TOT27%BHA—LEE27%FUL—CRY26%the week's likeliest stalemates

IPS — LIVFri 20:00

H 18%D 18%A 64%

NEW — BOUSat 12:30

H 41%D 25%A 35%

NOT — TOTSat 15:00

H 48%D 27%A 25%

MCI — COVSat 15:00

BHA — LEESat 15:00

H 48%D 27%A 26%

BRE — SUNSat 15:00

H 58%D 23%A 18%

FUL — CRYSat 15:00

H 39%D 26%A 35%

HUL — AVLSat 17:30

H 26%D 25%A 49%

EVE — MUNSun 14:00

H 30%D 26%A 45%

ARS — CHESun 16:30

H 63%D 22%A 16%

Player projections — Gameweek 3


Ollie Watkins AVL	at HUL	0.65	0.18	42%	16%
Erling Haaland MCI	vs COV	0.63	0.08	41%	8%
Thiago BRE	vs SUN	0.60	0.08	40%	7%
Igor Jesus NOT	vs TOT	0.54	0.09	37%	8%
Bryan Mbeumo MUN	at EVE	0.50	0.16	35%	14%
William Osula NEW	vs BOU	0.45	0.10	33%	9%
Morgan Gibbs-White NOT	vs TOT	0.45	0.24	32%	20%
Cody Gakpo LIV	at IPS	0.43	0.48	32%	34%
Kai Havertz ARS	vs CHE	0.41	0.16	30%	14%
Dominik Szoboszlai LIV	at IPS	0.38	0.92	29%	51%
Jack Hinshelwood BHA	vs LEE	0.37	0.13	28%	12%
Bruno Fernandes MUN	at EVE	0.37	0.69	28%	43%
Marcus Tavernier BOU	at NEW	0.31	0.32	25%	25%
Phil Foden MCI	vs COV	0.28	0.61	23%	40%
Antoine Semenyo MCI	vs COV	0.26	0.28	21%	23%
Mathis Cherki MCI	vs COV	0.25	0.49	21%	35%
Dango Ouattara BRE	vs SUN	0.25	0.25	21%	21%
Maxim De Cuyper BHA	vs LEE	0.22	0.35	19%	27%
Nick Woltemade NEW	vs BOU	0.10	0.46	9%	33%
Enzo Le Fée SUN	at BRE	0.08	0.54	8%	37%

Today’s headlines

Manchester United win — The week's most lopsided assessment: MUN vs IPS.

Dominik Szoboszlai projected G+A — The fantasy play of the week (at IPS).

Tottenham Hotspur upset watch — The lower-index side has real upset equity: TOT vs NEW.

Player projections — Gameweek 2


Ollie Watkins AVL	vs ARS	0.38	0.11	29%	10%
Kai Havertz ARS	at AVL	0.35	0.14	27%	13%
Viktor Gyokeres ARS	at AVL	0.32	0.07	25%	6%
Gabriel ARS	at AVL	0.13	0.01	11%	1%
Bukayo Saka ARS	at AVL	0.12	0.25	11%	21%
Ross Barkley AVL	vs ARS	0.12	0.14	11%	13%
Riccardo Calafiori ARS	at AVL	0.10	0.05	9%	5%
Eberechi Eze ARS	at AVL	0.09	0.02	8%	2%
Emiliano Buendía AVL	vs ARS	0.05	0.16	5%	14%
Tyrone Mings AVL	vs ARS	0.05	0.00	5%	0%
John McGinn AVL	vs ARS	0.03	0.21	3%	18%
Ian Maatsen AVL	vs ARS	0.02	0.13	2%	12%
Noni Madueke ARS	at AVL	0.02	0.14	2%	12%
Declan Rice ARS	at AVL	0.01	0.12	1%	11%
Matthew Cash AVL	vs ARS	0.01	0.03	1%	3%
Lucas Digne AVL	vs ARS	0.00	0.04	0%	4%

Projections assume recent minutes patterns — the model doesn’t know lineups. Full method: /model.

What we said last time

Gameweek 1: 7 of 10 called (70%)

ARS — COV3–0

✓ calledgave this 83%

HUL — MAN2–0

✗ missedgave this 16%

IPS — SUN2–1

✗ missedgave this 26%

EVE — CRY2–0

✓ calledgave this 39%

NOT — LEE0–1

✓ calledgave this 37%

BRE — TOT3–0

✓ calledgave this 51%

MAN — BOU2–1

✓ calledgave this 58%

BRI — AST4–0

✓ calledgave this 44%

NEW — LIV2–2

✗ missedgave this 25%

FUL — CHE2–3

✓ calledgave this 39%

Season outlook — 20,000 simulated seasons

Arsenal41%
Man City31%
Brentford8%
Chelsea8%
Man United7%
Liverpool7%

Proportional breakdown — 6 categories totalling 0.8189500000000001.
Category	Value	Share
Arsenal	0.3319	40.5%
Man City	0.2523	30.8%
Brentford	0.0627	7.7%
Chelsea	0.06155	7.5%
Man United	0.0553	6.8%
Liverpool	0.0552	6.7%

Share of simulated seasons each club finishes top. as of 2026-08-31

Bar chart — 6 bars.
Label	Value	Highlighted
Coventry	83%	yes
Hull	54%	yes
Ipswich	51%	yes
Tottenham	24%	no
Fulham	21%	no
Sunderland	16%	no

All three clubs relegated in 2025/26 were in the model's top four candidates before a ball was kicked. as of 2026-08-31

Model scoreboard — judged against the market

RPS 0.1982 · 54% correct

RPS 0.1983 · 54% correct

RPS 0.1989 · 54% correct

algorithm of the week

RPS 0.2005 · 54% correct

The market's closing prices score 0.1949 on the same matches — the one benchmark still ahead of us, and the running record is published either way.

Data through 2026-08-31 · every prediction on the record at /model.