Sunday's Premier League slate delivers four fixtures on September 20, and the Both Teams to Score market deserves a careful look across each of them. If you are new to this market, here is the baseline: Both Teams to Score means you are betting that each side scores at least one goal. A 1-0 result loses. A 1-1 result wins. The final scoreline does not matter beyond that threshold. That is all there is to it. Now the numbers.

One preliminary note on the historical BTTS data available for this slate. All seven clubs involved show a 0% BTTS hit rate through four matches each -- an extreme figure that almost certainly reflects early-season variance and very small sample sizes rather than genuine league-wide defensive dominance. Expected goals and expected goals against are more reliable inputs at this stage in the season. When a handful of results conflict with the underlying xG model, the model wins. That is the framework here.

Liverpool vs AFC Bournemouth -- BTTS: Yes

Liverpool travel to Vitality Stadium carrying emotional weight that does not show up in any model. Andoni Iraola, the manager who built AFC Bournemouth into a Premier League fixture, now sits in the Liverpool dugout and returns to face the club he shaped. He stated his intention plainly: "From a personal side, I'm really looking forward to see them, players that I love and staff I enjoyed working with. I wish them all the best, but we need to go there and beat them and be as ruthless as possible." The sentiment is registered. The model does not price it.

Both defensive profiles meet the threshold for a BTTS Yes recommendation. Bournemouth's xGA sits at 1.43 per game. Liverpool's sits at 1.50 per game. Both figures exceed the 1.2 per game marker the model uses as its primary BTTS Yes signal. Bournemouth's xG output of 1.79 per game reflects genuine attacking quality at home. Their ability to threaten on the counter through Tavernier and Kluivert gives them a realistic path to scoring even against a recalled Liverpool defensive spine featuring Van Dijk and Alisson. Liverpool's press at PPDA 7.1 -- the most aggressive figure in this dataset -- creates transitions in both directions. A high press opens space behind the line, and Bournemouth are built to exploit exactly that.

The complicating factor is schedule. Bournemouth's Europa League win at Real Sociedad on Thursday creates a three-day turnaround before this fixture. Marco Rose is expected to rotate the backline. Rose acknowledged the inherited squad quality directly: "He left the club in top shape and that's what I knew. That's what I got when I came here -- a top team with top staff. Everyone is professional and keen to work hard to be successful." The depth is real. Whether that depth holds against a full-strength Liverpool side with their first-choice creative midfield restored after cup rotation is the live question. The score projection of 1.5 for Liverpool and 1.1 for Bournemouth lands in BTTS Yes territory. The xGA profiles on both sides support the pick. Yes at -200. Medium confidence.

Crystal Palace vs Leeds United -- BTTS: Yes

Crystal Palace arrive at Elland Road in their most compromised defensive shape of the season. Disasi is serving a three-match ban. Riad is out for the season with an ACL injury. Goalkeeper Henderson remains unavailable. With two starting centre-backs absent, Palace's xGA of 2.23 per game -- already the second-highest figure in this dataset -- becomes even less flattering as a defensive baseline. Leeds United have the attacking profile to take full advantage.

Daniel Farke did not mince words when asked about Leeds' early-season results: "It's not a coincidence. It's quality. The table is currently not important. Eight points are important." That confidence has statistical grounding. Leeds are unbeaten through seven home fixtures across all competitions. Their striker has scored in five consecutive home Premier League appearances. They are operating as the most reliable home side on this matchday card. Palace's xGA hands them an open invitation to score.

The Palace side of this equation is equally supported by the data. Palace's away xG of 2.30 per game is among the highest figures in the entire dataset. A depleted, fatigued Crystal Palace side is still generating chances at significant volume on the road. The blended score projection of 1.5 Leeds and 1.1 Palace is a BTTS Yes result. Palace concede at a rate that says Leeds score. Palace attack at a rate that says they score too. Yes at -135. Medium confidence.

Sunderland vs Manchester City -- BTTS: No

The model says no here, so I say no. The edge is not there for BTTS Yes, and I am not going to pretend otherwise.

Manchester City carry a 1.02 xGA per game -- the lowest defensive expected goals conceded figure in this dataset by a meaningful margin. Their press at PPDA 8.2 combines elite defensive structure with a 2.09 xG output per game. They are the strongest defensive unit on this slate and one of the strongest in the league. Sunderland have to score against that structure while arriving in the worst possible situational shape.

The situational case against BTTS Yes is unusually clear. Sunderland have scored three goals in four league matches -- 0.75 per game. They arrive for their fourth match in 12 days. Left-back Reinildo is suspended. Centre-back Alderete is a major doubt after suffering a muscle injury in their Europa League win on Wednesday. Dayann Methalie could make his Premier League debut at left-back, facing City's attack. The fixture calendar has compounded every available disadvantage for Sunderland before kickoff. City's xGA at 1.02 per game means this defence is expected to concede roughly one goal against an average opponent. Sunderland are not functioning as an average opponent right now -- they are depleted, fatigued, and scoring at a rate that puts even reaching the BTTS threshold in doubt. The market prices BTTS No at -106, which reflects a near coin-flip. That pricing does not account adequately for Sunderland's actual situational position. The lean is No. Medium confidence.

Manchester United vs Fulham -- BTTS: Yes

Manchester United travel to Craven Cottage with one of the wider process-to-results gaps in the league. They sit second in the Premier League for xG output at 2.12 per game. They lead the league in touches in the opposition box. They own four points from four matches. The model shows an elite attacking side. The table shows a frustrating return. Against a Fulham defence conceding 2.08 xGA per game -- and 2.47 xGA at home -- United will score. The model is confident on that side of the ledger.

Fulham sit at the bottom of the table with one point from four matches. Their xGA at 2.08 per game is among the highest in this dataset. They are conceding at a rate that makes clean sheets nearly impossible against competent opposition. The relevant counterpoint is their attacking output. Fulham generate 1.56 xG per game overall, but their 11.0 deep completions per game at home is the highest figure in the entire dataset. This is a genuinely dangerous attacking team operating behind a defence that cannot hold. United's xGA profile combined with Fulham's home attacking volume says this fixture ends with both sides scoring.

Sesko does not travel with United. That is a notable absence -- an established goalscorer removed from the squad. However, Fernandes arrives with three goals and one assist on the season, taking set pieces against a defence that has conceded seven league goals. Mbeumo has recorded 14 shots and two goals. The attacking output is sufficient against this backline even without Sesko. The score projection of 1.5-1.1 or similar sits squarely in BTTS Yes territory. Yes at -182. The odds are short, but both xGA profiles support this result without ambiguity. Medium confidence.