Gameweek 1 gave the column five winners from ten and still a losing week - the prices didn't cover Hull beating Manchester United or Brighton putting four past Villa, and the ledger opens at -1.93pts. That's the game; it's all on the record.
The shape of the week: City go to Selhurst on Friday night off the back of a routine opening win, Sunday is a four-game pile-up with United trying to wash off the Hull result against Ipswich, and Monday night is the pick of the round - a Villa side that just took the heaviest beating of the opening weekend hosting an Arsenal team that barely broke sweat against Coventry.
Fri 8:00pm
The market makes City 1.65 for this and I think that's a Friday-night price for a Tuesday-night performance. The model has them at 55% - a clear favourite, but meaningfully below what 1.65 implies - and Selhurst under lights is the classic venue for that gap to matter. Palace were poor at Everton, but they were poor in the way opening days are poor, not in the way Villa were poor. Nothing on the match markets offers value at the current prices, so this is a watch-and-nugget game rather than a bet: Mateo Kovacic keeps shooting from range in this team, and 3.40 for 2+ shots is a price the model rates well above its implied chance.
Mateo Kovacic 2+ shots at 3.40 - the model has it at 36%, well above the price's 29%.
Sat 12:30pm
Liverpool drew at Newcastle in a game they twice led, and the early kick-off against a Forest side that lost at home to Leeds is the fixture the schedule-makers hand you to put it right. Not much to argue with here - the model makes Liverpool 58% and the market makes them shorter still, so the win is chalk without value. If you need a Liverpool leg for a builder it's the safest thing on the Saturday slate; as a standalone bet the price does all your winning for you.
Murillo to be carded at 5.00 - carded in two of his last five, and the model rates it above the price.
The bet: Liverpool to win at 1.50 - full match stats
Sat 3:00pm
Everton's 2-0 over Palace was tidier than anyone expected and Bournemouth pushed City closer than the scoreline suggests. The model has both sides scoring at 58% and that's where the interest is - two teams that looked sharper than their opening opponents allowed them to show.
Both teams have scored in the model's read at 58% against a price implying 59% - fine margins, back it small or leave it.
The bet: Both teams to score at 1.70 - full match stats
Sat 3:00pm
Hull's price has shortened everywhere since they turned Manchester United over, and that's exactly when you fade a promoted side. One result, at home, against a United team that spent the second half arguing with itself. The model still makes Coventry 52% here despite their own chastening afternoon at Arsenal - a promoted-club derby where the market has moved on sentiment and the model hasn't.
Coventry at 1.90 is the rare case where the model and the price agree almost exactly - 52% either way.
The bet: Coventry to win at 1.90 - full match stats
Sat 5:30pm
Spurs were dreadful at Brentford and the temptation is to bury them, but 3-0 flattered the hosts and this fixture has produced goals for years regardless of either side's form. Newcastle just traded blows with Liverpool for ninety minutes. The model's highest both-teams-to-score number of the whole round lives here at 62%, and Over 2.5 at 1.67 is the one match price on Saturday where the model actually clears the market. Richarlison scored twice in pre-Brentford August and leads the line while Solanke's fitness is managed - anytime at 2.60 with the model at 41% is the player angle.
Richarlison anytime at 2.60 - the model makes it 41% against the price's 38%.
The bet: Over 2.5 goals at 1.67 - full match stats
Sun 2:00pm
The quiet one of the Sunday slate - the model's lowest goals expectation of the round. Fulham pushed Chelsea all the way in defeat and the model rates them slightly ahead of what 3.10 implies; Sunderland's opener at Ipswich was better than the result. Small stakes territory, but the away price is the right side.
Fulham at 3.10 - the model has 34% against an implied 32%.
The bet: Fulham to win at 3.10 - full match stats
Sun 2:00pm
Brighton's 4-0 over Villa was the statement of the weekend and 3.90 for them here looks big at first glance - but the model doesn't bite, putting them at 25% which is exactly what the price says. Chelsea scored three at Fulham and concede chances in the transition game Brighton live off, so the goals angle is the honest one. Enzo Fernandez is taking more shots than any Chelsea midfielder and the model makes his 2+ shots line at 1.57 a 67% chance against the implied 64%.
Enzo Fernandez 2+ shots at 1.57 - the model's 67% against the price's 64%.
The bet: Both teams to score at 1.62 - full match stats
Sun 2:00pm
The model can't split these two - 37% each - which after Brentford dismantled Spurs and Leeds won at Forest feels about right. Coin-flip matches are goals matches: both teams to score at 60% is the second-highest of the round and the 1.67 price sits level with it.
The model's 60% BTTS here is its second-highest of the round.
The bet: Both teams to score at 1.67 - full match stats
Sun 4:30pm
United at 1.40 five days after losing 2-0 at Hull is the market betting on embarrassment being a coach. The model makes them 69% - a strong favourite, but not a 1.40 one - and Ipswich just beat Sunderland playing forward at every opportunity. The bet here isn't the winner, it's the game: United chasing redemption plus an Ipswich side that scores and concedes makes Over 2.5 at 1.62 the round's clearest goals price, with the model at 64%. Matheus Cunha at 3.25 anytime is the scorer angle if United's response is what Old Trafford demands.
Matheus Cunha anytime at 3.25 - the model has 34% against an implied 31%.
The bet: Over 2.5 goals at 1.62 - full match stats
Mon 8:00pm
Monday night and the model's strongest away call of the round: Arsenal at 60% at Villa Park. That number would have raised eyebrows a fortnight ago; after Villa were taken apart 4-0 at Brighton and Arsenal strolled past Coventry it reads almost conservative. Villa Park under lights against a top side has been Emery's redoubt, and this is the game where he either restores order or the early-season inquest starts in earnest. I think the crowd carries them for an hour and the quality gap decides the last half hour. Arsenal at 1.55 is short but the model says it should be shorter; take it as a single or anchor the weekend acca with it.
Ross Barkley 1+ shot on target at 2.50 - the model's 43% against an implied 40% - if he keeps his midfield spot.
The bet: Arsenal to win at 1.55 - full match stats
One selection per Premier League fixture, settled after every round at 1pt level stakes. Five winners from ten in Gameweek 1, but the prices didn't cover the losers: the season sits at -1.93pts.
Every fixture is run through the Statz projections model, which simulates the match from team strength, form and player-level data to produce win, goals and both-teams-to-score probabilities. The writing adds what the model cannot see: team news, motivation, and whether the market has over-reacted to last weekend.
Every gameweek, published ahead of the weekend once the Friday press conferences have landed and the predicted line-ups have firmed up. Prices quoted are the prices at the time of writing and can move before kickoff.
Yes - one selection per fixture, settled after the round and kept in a public running record at one point level stakes, wins and losses alike. Previous gameweeks stay on the page with their results shown.
Player props are markets on an individual player - shots, shots on target, tackles, fouls, cards. Each fixture note carries one prop the model likes, with the player's recent record behind it, taken from the same board that powers the Statz player props page.