The table says mismatch, the away splits say maybe not
Serbia sit fourth in the standings with zero points from four Nations League games; Germany are third with five. That is the headline tension in Belgrade: a home side that has lost every match in this campaign against a visitor whose own campaign has been built on draws. Nutmegly's AI read makes Germany a heavy favorite, with a 73.0% away-win probability, a 16.2% draw and a 10.8% home win. The lean is clear. The away form is not.
The current table explains the model's confidence. Serbia have no wins, four defeats, three goals scored and eight conceded; at home they have lost both matches, scoring twice and conceding four. Germany have one win, two draws and one defeat, with three goals for and two against; away from home they have drawn both games, 1-1 in the Netherlands and 0-0 in Greece, scoring once and conceding once. Germany's one win in the campaign was the 2-0 home victory over Serbia on 1 October, which is also the single head-to-head record in the data.
What the stored run actually shows
The recent-results list adds context but also warns against overreading. Serbia's five most recent stored results are all defeats: 1-2 away to the Netherlands on 4 October, 0-2 away to Germany on 1 October, 1-2 at home to the Netherlands on 27 September, 1-2 at home to Greece on 24 September and 1-5 away to Mexico on 5 June. The first four correspond to the current campaign's four losses; the Mexico match is older sample. Across the stored eight-game sample, Serbia's points efficiency is 0.38 per game.
Germany's stored run is less dramatic: 0-0 away to Greece on 4 October, 2-0 at home to Serbia on 1 October, 0-1 at home to Greece on 27 September, 1-1 away to the Netherlands on 24 September and 1-1 at home to Paraguay on 29 June. Their stored eight-game points efficiency is 1.50 per game, compared with Serbia's 0.38. In the Nations League sample specifically, Serbia average 0.00 points per match and Germany 1.00. Those numbers support the model's direction, but they do not erase the away-draw pattern.
The case for the model, and the countercase
The case for the AI read is straightforward. Serbia have not taken a point in the campaign, have conceded eight goals in four games and have lost both home matches. Germany have lost once, have conceded twice in four, and have already beaten Serbia 2-0. The model's 73.0% away win is an expression of that gap, not a guarantee. The same model gives over 2.5 goals a 57.4% chance, which is a lean rather than a certainty.
The countercase is stronger than the headline probability suggests. Germany have not won away in this campaign: their two road matches were a 1-1 draw in the Netherlands and a 0-0 draw in Greece. They have scored three goals in four games and one in two away games. A team with that away output can be favored, but a 73.0% win probability is aggressive against a 16.2% draw probability. One supplied matchup signal also notes Serbia averaging 1.00 goal per home game and Germany conceding 1.40 per away game in the broader stored sample, which is a reason to expect Serbia to have some attacking foothold.
The model's own caveats matter. Serbia have no advanced-stat sample in the data and Germany have three, so the attack-quality adjustment is downweighted; historical xG and technical-stat samples are under three matches. An external forecast points in a different direction, and the supplied caveats flag sample-scope differences. There are no injury reports and no confirmed lineups, so personnel impact cannot be assessed either way. That is a gap in the evidence, not proof that both squads are at full strength.
What would change the picture
The match can unfold in two supported ways. If it follows Serbia's campaign pattern — four defeats, eight goals conceded, both home games lost — Germany's favorite status is justified and the away win probability will look reasonable. If it follows Germany's away pattern — two draws, one goal scored, one conceded — the game becomes a low-margin contest in which the draw probability matters more than the headline lean. That is a conditional scenario, not a claim about how either side will set up.
The watchpoint is Germany's away scoring, and Serbia's home scoring. Serbia have scored once in each of their two home matches in this campaign. If that pattern holds, Germany will need two goals to win, a total they have not reached in any away match in this campaign. If Germany score twice on the road, the model's lean has the evidence it needs. Either way, the 40-day gap since both teams' last stored match on 4 October should not be read as a rest advantage; the supplied rest-window signal explicitly warns against that inference.