The table says mismatch; the splits say puzzle
Liverpool sit fourth with 18 points from 10 league games; Manchester United are 18th with 8. Five wins against one. On the surface this looks like a routine Anfield assignment, and Nutmegly's model agrees. But the home and away splits inside those two records make the fixture considerably stranger than the standings imply, and they are the reason this preview is worth reading rather than skimming.
Liverpool's 18 points are built almost entirely on the road: four wins and a draw in five away games, 10 goals scored and 4 conceded. At Anfield in the league the record reads one win, two draws, two defeats, four goals for and six against in five. United, for their part, are without an away win — three draws and two defeats — but they have taken a point from three of five on their travels, scoring four and conceding eight. One side has a home problem, the other has an away ceiling, and the two meet at the same ground.
Two form lines pulling in opposite directions
Liverpool's stored results, newest first: 3-1 at Crystal Palace, 2-3 at Fenerbahçe, 0-1 at home to Arsenal, 2-0 at home to Brighton, 4-3 at home to Villarreal. Three wins and two defeats, with the sample spanning more than one competition. For this fixture the relevant entries are the two home league results in that run — the Arsenal defeat and the Brighton win — because they are the closest thing to evidence about how Liverpool currently perform at Anfield.
United's run reads: 4-4 at home to Aston Villa, 3-2 at home to Roma, 0-2 at Chelsea, 0-3 at home to Bournemouth, 1-4 at Como. One win, one draw, three defeats. Two of those positive results came at home, and the two away entries in the sample are the Chelsea and Como defeats. The defining feature of their league season is not the losses but the draws: five of them in 10 games. That is a side that keeps landing on level terms.
Here is the cleanest number in the file. Both teams have scored 14 league goals. The gap is entirely at the other end: 10 conceded by Liverpool, 19 by United. Identical attacking output, and a nine-goal difference in what happens behind it.
What Nutmegly's model actually says
The AI read leans home: 55.1% Liverpool, 22.9% draw, 22.1% United, with over 2.5 goals at 55.4%. The signals behind that lean are straightforward — 1.88 points per game across Liverpool's stored eight-match sample against 0.75 for United, 1.80 versus 0.60 per league game in the same window, and fourth place against 18th.
It is worth being precise about what 55.1% is and is not. It is a lean, not a majority: the model hands roughly 45% of the outcome distribution to something other than a Liverpool win. And one of its own signals points the other way, flagging the Liverpool home attack against the United away defence as a factor in United's favour. That internal disagreement, rather than the headline probability, is the most interesting thing in the projection.
The strongest case against the lean
The counterevidence is Liverpool's home record itself. Four goals in five home league games is not the output of a team that reliably blows visitors away, and United have drawn half their league matches — a 22.9% draw is a live path, not a formality. If the game is level with 20 minutes left, the away side has already achieved the result shape its season keeps producing.
The second caveat is sample uncertainty. Lineups are not confirmed, and the availability picture reports 15 potential absences across the fixture. Those caveats do not favour either team; they widen the range of plausible outcomes, which is exactly why a 55.1% figure should be read as a direction of travel rather than a verdict. Both sides' most recent stored matches are 14 days old, and a gap that size is not treated as evidence of rest advantage — so do not read freshness into the calendar.
The question to watch
The conditional is this: if the match stays level into the second half, the way United's three away draws suggest it might, does Liverpool have a home gear to change it? Their five home league games produced four goals; the model still expects goals, with over 2.5 at 55.4%, largely because United have conceded 19 in 10. The watchpoint is which of those numbers gives first — Liverpool's modest home scoring or United's porous defence. Whichever breaks tells you whether the model read the fixture or merely the table.