The Table and the Split Point in Different Directions
Romania and Poland meet in Bucharest on Nov. 14 with a table that seems to point one way and venue splits that point another. Romania are fourth in the League B table with zero points from four matches, three goals scored and 13 conceded. Poland are third with four points, seven scored and four conceded. The obvious read is that Poland should be favored. The less obvious read is that Poland's four points have all come at home: away from home they have played twice, lost twice, scored once and conceded four.
The match matters because Romania's points column is empty after four games. A home fixture at the Arena Națională is the most direct chance to change that, though the data offers no confirmed lineups or injury reports to assess how they might do it. Poland, meanwhile, can add to a total built entirely at home. The split is the story: Poland's overall record is better, but their away record is not. That is the tension Nutmegly's model has to resolve.
Form: The Edge Is Overall, Not Away
Romania's stored recent results are bleak. Their five listed matches are all defeats: 0-1 at home to Sweden on Oct. 5, 0-6 away to Poland on Oct. 2, 2-4 at home to Bosnia & Herzegovina on Sept. 28, 1-2 away to Sweden on Sept. 25 and 0-1 away to Türkiye on March 26. Across the wider eight-match stored sample, Romania have averaged 0.38 points per match. The dates matter: this is a stored sample running from March to October, not a complete current-season record.
Poland's listed results are mixed: a 0-1 away defeat to Bosnia & Herzegovina on Oct. 5, a 6-0 home win over Romania on Oct. 2, a 1-3 away defeat to Sweden on Sept. 28, a 0-0 home draw with Bosnia & Herzegovina on Sept. 25 and a 2-3 away defeat to Sweden on March 31. Their wider eight-match sample yields 1.12 points per match. In the Nations League sample specifically, Romania have averaged 0.00 points per match and Poland 1.00. That same-competition gap is the strongest numerical case for the away lean.
The head-to-head sample is one match: Poland's 6-0 home win over Romania on Oct. 2. It is a striking result, but a single observation cannot establish a pattern in Bucharest. It does, however, give the model a recent, high-scoring data point in Poland's favor. The caveat is that the same result came at home, where Poland have taken all four of their points; their two away matches produced zero.
The Model's Read: A Clear Lean, Not a Guarantee
Nutmegly's AI read leans away. The model gives Poland a 51.2% win probability, Romania 23.5% and the draw 25.3%. It puts over 2.5 goals at 44.9%, so the under side of that line is the more likely outcome. The draw is marginally more likely than a Romania win, a useful reminder that the model's non-Poland outcomes are close together. This is a leading outcome, not a guarantee, and the probabilities are estimates rather than observed results.
The case for the model is built on overall efficiency. Poland's 1.12 points per match over the stored eight-match sample is nearly three times Romania's 0.38. In the Nations League sample, Poland's 1.00 points per match compares with Romania's 0.00. Romania have lost all four of their current table matches and have conceded 13 goals. Those are the numbers that make a Poland lean defensible even before venue splits are considered.
But the away evidence cuts the other way. Poland have zero points, one goal and four conceded in two away matches. Romania have averaged 0.75 goals per home match, while Poland have conceded 2.00 per away match. That matchup does not support the idea of a comfortable away win; it suggests Romania can find a goal. The model's counterpoints acknowledge thin historical xG and technical-stat samples, and an external forecast that points in a different direction. Those are reasons to treat 51.2% as a lean, not a verdict.
Lineups are not confirmed in the data, and there are no injury reports. That is a gap, not evidence that either squad is fully available. The absence of confirmed XIs also means the model's read was generated without knowing the final selection. Historical attacking-quality data is limited to two advanced-stat matches for Romania and three for Poland, so any correction for chance quality carries a lower weight. The away lean is therefore best read as a statement about overall results, not about a settled tactical picture.
What Would Change the Picture
A plausible way this becomes a different match is if Romania's home attack performs closer to Poland's away concession rate than to its own home scoring average. Romania have averaged 0.75 goals per home game; Poland have conceded 2.00 per away game. If the hosts score, the 51.2% away probability becomes a live test rather than a comfortable script. If Poland's away scoring remains at one goal in two matches, their win case will depend on keeping Romania out, a task their away record has not managed.
The watchpoint is specific: Romania's home attack against Poland's away defense. The model says Poland are more likely to win, but the venue split says Poland's away form is the weakest part of their case. If Romania score at home, the match could flip toward the 25.3% draw or the 23.5% home win. If Poland score first, their overall efficiency may take over. That is the conditional script to follow, not a second prediction.