A Table-Topping Start Against a Model That Disagrees
Two rounds into the Bundesliga season, Saturday's meeting at the WWK Arena throws up a genuine oddity. FC Augsburg arrive top of the standings with six points from two games, seven goals scored and one conceded. Bayer Leverkusen sit seventh on three points, with six scored and three against. And yet the read from Nutmegly's AI model leans away from the home side, handing Leverkusen a 50.9% chance of winning against 25.5% for Augsburg and 23.6% for a draw.
That gap between the standings and the model is the central storyline of the fixture, and part of it has a simple explanation. The prediction was generated in the early hours of 6 September, before Augsburg's 4-1 win at Eintracht Frankfurt that afternoon. Its inputs therefore reflect a snapshot in which the two clubs looked much closer than the current table suggests. Read the probabilities as a considered estimate built on that snapshot, not as a verdict on the table as it stands now.
Two Convincing Wins, One Awkward Away Day
Augsburg's two most recent recorded results are both victories: 3-0 at home to FC Schalke 04 on 30 August and 4-1 away at Eintracht Frankfurt on 6 September. Those two games account for the whole of their current-season record, and the goals line up exactly with their table row. Sitting behind that is a previous-season sample from May: a 0-4 defeat at Union Berlin, a 3-1 home win over Borussia Mönchengladbach and a 3-1 win at Werder Bremen. Useful context, but it belongs to the stored record rather than to this campaign.
Leverkusen's form line runs the other way at the back of it. They opened with a 4-0 home win over Union Berlin on 5 September, but the match before that was a 2-3 defeat away at SV Elversberg on 29 August — the away loss that appears in their table row, three goals conceded in a single away outing. Further back in the sample sit a 1-1 home draw with Hamburger SV, a 1-3 defeat at VfB Stuttgart and a 4-1 home win over RB Leipzig, all from the previous season. Both clubs, then, arrive on the back of a convincing home win, but only one carries a recent away defeat.
Where the Home-Side Case Bites Hardest
The supporting numbers lean heavily toward Augsburg. Across their last eight stored matches they have taken 1.88 points per game, against 1.62 for Leverkusen. Narrow the sample to Bundesliga fixtures and the gap widens to 2.00 points per game for Augsburg against 1.40 for Leverkusen. The venue split points the same way: Augsburg average 2.20 goals per home match in the sample, while Leverkusen concede 2.00 per away match. Four of the six signals recorded alongside the prediction favour the home side.
Two caveats keep that from being decisive. Those averages are drawn from stored samples that span more than one season, so they are a lean rather than a season record. And the model is not forecasting a tight, low-scoring contest: it puts the probability of more than 2.5 goals at 63.2%. Goals are baked into its expectation. The disagreement is purely about who finishes with more of them.
The Away Lean, and What Would Break It
Why does the model still side with the visitors? The counterpoints are explicit. An external forecast runs in a different direction from the model, which suggests the two are not reading the same evidence the same way. More importantly, lineups are not confirmed for either side; that is logged as a missing input, so the picture could still move before kick-off. The injury feed reports no absentees, but an empty report is not the same as a guaranteed fully fit squad.
There is also a rest-window signal that favours the home side, though the data itself cautions that gaps in the record prove neither extra rest nor that a team has not played. Add the smaller samples: five competition matches per side, five same-venue matches each, and only two recorded head-to-head meetings. None of that settles the argument. What it does mean is that the away lean rests more on how the model weighs its evidence than on any single overwhelming number.
The Watchpoint
The question to hold on to is whether Leverkusen's away record is a pattern or a single bad afternoon. Augsburg have scored in both of their recorded matches this season; if that continues, the model's away lean will be under real pressure early, especially with the model already expecting goals. If instead Leverkusen reproduce the scoring form that produced four goals in two of their stored home matches, and Augsburg's early-season output stalls, the 50.9% will look well judged. Watch the first half hour, not the final scoreline, to see which version of this fixture shows up.