Man City vs AEK Athens prediction: the Jarvis supercomputer's verdict
Nhipbong Analytics
Jarvis prediction desk
Man City have just conceded four goals in the Champions League, losing 2–4 to PSG at home. Yet the Jarvis supercomputer model still gives them a 90% chance of beating AEK Athens, with the draw at 7% and AEK at only 3%, and a predicted score of 3–0. That paradox has a fairly simple answer, and it sits on the visitors' side.
- Jarvis gives Man City a 90% chance of winning, the draw 7% and AEK Athens 3%.
- The predicted score is 3–0, also the most likely scoreline at 14%, with 2–0 and 4–0 next at 13% and 12%.
- Man City won four of their last five, scoring 18 but conceding 9, including four against PSG.
- AEK Athens are 20th in the league phase with 3 points from two games, scoring 2 and conceding 3.
- Expected total goals are about 3.8, the chance of three goals or more is 72%, and confidence is medium.
| Outcome | Man City | Draw | AEK Athens |
|---|---|---|---|
| Jarvis probability | 90.4% | 6.8% | 2.8% |
| Metric | Jarvis forecast |
|---|---|
| Expected total goals | 3.76 |
| Predicted score | 3–0 |
| Chance of 3+ goals | 72.5% |
| Confidence | Medium |
Why Man City can lose 2–4 and still be trusted at 90%
Losing 2–4 at home to PSG is the kind of result that makes people doubt a whole team. I doubted too, until I put it next to the opponent in this game. A fall against a strong side is very different from a fall against a weak one.
AEK Athens are 20th in the league-phase table with 3 points from two games, a 1–0 win over LASK Linz then a 1–3 defeat to Shakhtar Donetsk in their latest game, and they have scored only 2 and conceded 3, which is a side that has not yet shown it can take charge of the big nights. That is where the answer sits, on the visitors' side.
I read the 90% as the answer to a very specific question. Do Man City need to be perfect to beat AEK? It seems they do not.
What few people notice: AEK are expected to score less than half a goal
Jarvis expects Man City to score about 3.3 goals and AEK only about 0.5. That gap is far wider than the feeling after the PSG defeat suggests. Feelings get swept along by the latest game, while the model does not.
The chance that both teams score is only 35%. In other words, an AEK goal is already a minority scenario. And if City score enough, a consolation goal would hardly change anything.
I noticed one more thing: Jarvis's top five scorelines, 3–0, 2–0, 4–0, 1–0 and 5–0, are all City wins without conceding. The model leaves almost no room for AEK even in their best scorelines.
The last five games: one side scoring in bunches, the other winning small
Man City beat Norwich 5–0 in the cup, beat Sunderland 5–3, won 3–0 at Liverpool, lost 2–4 to PSG, then beat Ipswich 3–2 in their latest game, scoring 18 goals in all but also conceding 9, with only two clean sheets. A side that scores in bunches but leaks fairly steadily keeps its defence as the real question.
AEK beat LASK Linz 1–0, drew 1–1 with Asteras Tripoli, won 2–1 at Volos, beat OFI Crete 3–0 and then lost 1–3 to Shakhtar Donetsk, scoring 8 and conceding 5: steady small wins. This is a side that knows how to do its job in its own league.
One curious detail: on the same day, 17 September, each side had a night of goals in a cup tie, City beating Norwich 5–0 and AEK beating Panserraikos 8–1. So AEK are not a team that cannot score. It is just that the opponent in this game is on another level.
Why 3–0 tops the table, rather than 4–0 or 5–0
Expected total goals of about 3.8 sounds like an invitation to very big scorelines. But Jarvis's scoreline table gathers its probability on moderate results: 3–0 at 14%, 2–0 at 13% and 4–0 at 12%. I find that fits how big sides usually win, tidily rather than destructively.
A 5–0 has only 8%. A big side often eases off once far ahead. And a three-goal win is still a complete night.
The chance of three goals or more is 72%, but many goals and a heavy scoreline are not the same thing. Many goals means an open game, while a heavy scoreline means one side far ahead, and the two do not always go together.
The rest: the scenarios that could spoil Man City's night
The remainder, a 7% draw and a 3% AEK win, is small but each has a concrete path. Man City just conceded four against PSG, so a bad night at the back would not be strange. Sides that counter-attack well often need only one moment like that.
AEK have beaten LASK Linz 1–0 in this same competition, so this is not a team that does not know how to win in Europe. They just need a night when everything goes right.
That is why I treat 90% as a strong prediction rather than a done deal. Strong is strong, but it is not a promise.
A night for the hosts, but the question is at the back
The game kicks off at 19:00 UTC on 20 October (02:00 on 21 October in Vietnam), at the Etihad Stadium in Manchester, in the third round of the league phase.
If Man City score early, the story could be over by the middle of the first half. If not, I will be watching their defence more than their attack. After four conceded against PSG, every AEK attack is a question.
This is statistical analysis, not betting advice.
Frequently asked questions
Which scoreline is most likely in Man City vs AEK Athens?
3–0 to Man City, at 14%, ahead of 2–0 (13%) and 4–0 (12%). It is also the scoreline Jarvis predicts.
What are Man City's chances of beating AEK Athens?
Man City are at 90% against 3% for AEK Athens, with the draw at 7%. Confidence is medium.
How many goals should we expect in Man City vs AEK Athens?
Quite a few: expected total goals are about 3.8 and the chance of three goals or more is 72%, but the chance that both teams score is only 35%.
What is the recent form of Man City and AEK Athens?
Man City won four of their last five, scoring 18 and conceding 9, with the only defeat a 2–4 loss to PSG in the Champions League. AEK Athens won three, drew one and lost one, the latest defeat being 1–3 against Shakhtar Donetsk.
Where are AEK Athens in the Champions League table?
AEK Athens are 20th in the league phase with 3 points from two games (one win, one defeat), having scored 2 and conceded 3.