Friday 23 October 2026 Premier League 2026/27 · Matchweek 8 This week's calls

Inkwell & Algorithm

Data-driven Premier League predictions, with the reasons. New every Thursday.

Match of the week · Sun 14:00

Nothing to split Liverpool and Brighton

We call Liverpool win, 2–1. If Liverpool's attack were only average, this would swing to an away win (41%).

Liverpool 2–1 Brighton
Our call Home win Confidence: too close to call
Liverpool 37%Draw 28%Brighton 36%

Why we're calling it

  1. 1Liverpool's attackLiverpool create and score 22% more than an average side. Worth +7 points to our call.
  2. 2Brighton's defenceBrighton concede 6% more than an average side. Worth +2 points to our call.
  3. 3Brighton are leaking chances away from homeExpected goals (xG) measures the quality of chances. Over their last 5 home games Liverpool created 1.5 and allowed 1.5 a game. Over their last 5 away games Brighton created 2.1 and allowed 1.9.

Read the full breakdown: form, xG, team news and head to head

The rest of the round

9 more matches · kick-off times UK
Fri 20:00Confidence: too close to call
Ipswich v Nott'm Forest 1–2
Away win
Home 32% · Draw 27% · Away 41%

Nothing to split Ipswich and Nott'm Forest.

Sat 12:30Confidence: solid
Aston Villa v Man City 1–2
Away win
Home 26% · Draw 26% · Away 47%

Man City's attack should be too much for Aston Villa.

Sat 15:00Confidence: strong
Arsenal v Everton 1–0
Home win
Home 61% · Draw 26% · Away 13%

Arsenal's attack should be too much for Everton.

Sat 15:00Confidence: too close to call
Coventry v Fulham 0–1
Away win
Home 33% · Draw 30% · Away 36%

Nothing to split Coventry and Fulham.

Sat 17:30Confidence: strong
Chelsea v Tottenham 2–1
Home win
Home 57% · Draw 24% · Away 19%

Chelsea's attack should be too much for Tottenham.

Sun 14:00Confidence: too close to call
Crystal Palace v Newcastle 2–1
Home win
Home 40% · Draw 26% · Away 34%

Nothing to split Crystal Palace and Newcastle.

Sun 14:00Confidence: too close to call
Hull v Brentford 0–1
Away win
Home 31% · Draw 30% · Away 39%

Nothing to split Hull and Brentford.

Sun 14:00Confidence: too close to call
Man United v Bournemouth 2–1
Home win
Home 44% · Draw 27% · Away 30%

Nothing to split Man United and Bournemouth.

Sun 16:30Confidence: too close to call
Sunderland v Leeds 0–1
Away win
Home 28% · Draw 30% · Away 42%

Nothing to split Sunderland and Leeds.

The round in charts

Hover a mark for the numbers
Attack against defence
Expected goals created and allowed per game this season. Top right is where you want to be.
Better defence (xG allowed)
1.0Better attack (xG created)2.4
Our record, week by week
Correct results each week this season, against simply backing the bookmakers' favourite.
UsBookies' favourite
5
7
3
4
5
5
6
21 Aug28 Aug04 Sep12 Sep18 Sep10 Oct17 Oct

We've beaten the favourite in 2 weeks out of 7, and fallen short in 2.

Last week's verdict

6 of 10

Games from 17 October, with 2 exact scores.

MatchOur callResultVerdict
Brentford v LiverpoolHome win 2–11–2Wrong
Everton v ChelseaHome win 2–10–2Wrong
Fulham v HullHome win 1–01–0Exact score
Man City v IpswichHome win 3–03–2Right result
Newcastle v Aston VillaHome win 2–10–1Wrong
Bournemouth v SunderlandHome win 2–12–0Right result
Brighton v Crystal PalaceHome win 2–15–0Right result
Leeds v Man UnitedHome win 2–11–1Wrong
Nott'm Forest v ArsenalAway win 0–10–1Exact score
Tottenham v CoventryHome win 1–03–1Right result
Owning the misses

What we got wrong

Verdict: Our mistake

Newcastle v Aston Villa: we said Newcastle to win at 43%. It finished 0–1.

The chances went the way of the result too (xG: Newcastle 0.9, Aston Villa 2.1). We misread this one, and the ratings will move to reflect it.

Each week we pick our most confident wrong call and check it against the chances both sides created. Unlucky means the chances backed us; our mistake means they didn't.

The table

Results up to 19 October
#ClubPGDPts
1Man City7+1221
2Arsenal7+618
3Brighton7+1816
4Chelsea7+413
5Liverpool7+112
6Brentford7+510
7Leeds7+310
8Everton7+110
9Hull7+19
10Ipswich7-49
11Newcastle7-48
12Aston Villa7-48
13Man United7+07
14Crystal Palace7-87
15Bournemouth7-46
16Tottenham7-46
17Coventry7-86
18Fulham7-35
19Nott'm Forest7-45
20Sunderland7-84

The table they deserved

The same season ranked by the points each side's chances were worth, from expected goals (xG). A plus means results have run ahead of the chances. This describes the season so far. It is not a forecast of where anyone will finish.

#ClubPPtsDeserved+/−Real place
1Man City72113.8+7.21
2Brighton71613.4+2.63
3Arsenal71813.0+5.02
4Bournemouth7612.2-6.215
5Nott'm Forest7512.1-7.119
6Liverpool71210.5+1.55
7Everton71010.4-0.48
8Brentford71010.2-0.26
9Man United7710.0-3.013
10Fulham759.8-4.818
11Leeds7109.5+0.57
12Chelsea7139.4+3.64
13Tottenham768.9-2.916
14Sunderland748.5-4.520
15Aston Villa788.2-0.211
16Coventry767.4-1.417
17Ipswich797.2+1.810
18Newcastle787.1+0.912
19Crystal Palace776.8+0.214
20Hull796.6+2.49