All seven headline fixtures from Champions League matchday 2 are done, so as usual let’s go back and check the pre-match predictions. The seven games produced 38 goals, an average of 5.4 per game — a proper night of high scores, where even a three-goal lead wasn’t safe.
7 predictions: 5 hits, 1 partial, 1 miss
(For the pre-match predictions, the scores and the result of each, see the results table above.)
The five hits: the direction was right, and in some games the substance was closer than the scoreline suggests
Arsenal vs Lille (3-0). The first half was actually very tight; the gap only opened in the second half, when Arsenal scored three goals in seven minutes between the 64th and 71st. The numbers back it up: Arsenal’s xG was about 1.85 against Lille’s 0.4, shots were 16 to 4, and Lille didn’t register a single shot on target all game — in line with my pre-match line that Lille’s midfield couldn’t be compared with Arsenal’s.
Atlético Madrid vs Manchester United (2-0). Atlético scored in the first minute, United’s Yoro was sent off in the 74th, and Atlético added a second in the 79th to seal it. Before the game I wrote that United’s form was awful, so I backed the in-form Atlético, and that was right. But the numbers say the gap was smaller than the scoreline (xG about 1.4 to 1.0, shots 11 to 9), and United even hit the crossbar with one effort. I certainly didn’t predict conceding inside the first minute or the red card, but the general direction was right, and United’s finishing was poor.
Galatasaray vs Barcelona (1-4). Barcelona were briefly behind in the first half, levelled before the break and then scored three more after it to kill the game. I wrote that a Süper Lig side couldn’t withstand 90 minutes of sustained attack, and the numbers support that: 22 Barcelona shots, 12 on target. What’s interesting is that I expected Galatasaray to park the bus, and I was curious how Barcelona would break it down — instead Galatasaray scored first and played much more proactively than I imagined. I underestimated them there.
LASK vs Liverpool (3-6). I wrote that even Liverpool’s bench would show a class gap, and the numbers were even more extreme: roughly 26 Liverpool shots, xG about 3.3, and 19 shots in the first half alone. The pre-match note that Isak might be rested also held (though that was always the likely outcome) — Isak, Gakpo and Szoboszlai all stayed home, and Koumas led the line. The surprise was LASK scoring three with an xG of about 2.0; Liverpool’s defence was top-heavy, but conceding three is more than I expected and a signal worth tracking. The “look at the goals” fallback also landed (9 total goals).
Manchester City vs PSG (2-4). I wrote that, even with City at home, I preferred PSG personally, and PSG duly went into Champions League killer mode to lead 3-0 in the first half (Doué, Dembélé and Vitinha each scoring), with Doué, who I’d singled out, opening the scoring — his form really is good. But on the numbers, City were actually the better side (xG about 2.6 to 1.8, shots 27 to 15), and Haaland’s two second-half goals made it a contest again. Still, PSG are two-time winners for a reason: their big-game ability, finishing and creativity were all on display. According to reports, it was the first time since January 2012 that City had trailed by three goals at half-time at home. The numbers didn’t back this pick, yet it still landed — a reminder that in football you can’t rely on data alone, and that some things that can’t yet be quantified still need human interpretation, which is why human predictions still have a place.
The partial: the Bayern direction missed, but “look at the goals instead” landed
Viking vs Bayern (3-3). This was the biggest twist of the round. Before the game I wrote that I couldn’t see how Bayern lose, that 1.08 on the win had no value, and that if you had to bet, the over was better. Bayern didn’t win, but six goals meant the goals fallback landed, so it’s a partial.
The game itself was wild: Viking led 2-0 in the first half through a corner and a lovely counter-attack; Bayern came back in the second half through a Kane penalty and a fine header from Kane (his second), before Luis Díaz put them ahead. The turning point came in the 73rd minute, when Olise was sent off for a second yellow card, and with Bayern down to ten men, Viking won the ball and scored a 90th-minute equaliser. I’d specifically mentioned that Olise was in red-hot form; I didn’t expect him to steal the spotlight like this. The numbers were in line with the pre-match read: Bayern had 66% possession and an xG of about 3.5 against Viking’s 0.6, completely one-sided on paper — but Viking’s set-piece and counter-attack plan was well executed, and with an unpredictable red card, Bayern, with a no-vig win probability of about 87% from the pre-match 1.08 / 11 / 23 line, ran into the remaining 13%. But I preferred the over from the start, so call it a partial.
The miss: Roma led by three, and the last seven minutes wiped it out
Roma vs Real Madrid (3-4). This was my only miss. I backed Roma because Real Madrid had many injuries, Vinícius was in a slump, and I thought Roma’s discipline would trouble them. For the first 83 minutes the game did follow my read: Roma led 3-1, and although Mbappé briefly levelled it at 1-1 thanks to a defensive error and a fortunate deflection (xG about 0.03), Vinícius was still in a slump, contributed nothing in attack and was substituted in the 59th minute, with Real Madrid completely outplayed.
Then the last seven minutes turned everything around: Real Madrid bombarded the Roma box with corners and scored three times to complete a 4-3 comeback. The final xG was about 1.8 for Real Madrid and 1.4 for Roma, so the chances created in those last seven minutes pushed them ahead both on the numbers and on the scoreboard within 90 minutes. Before the game I said Real Madrid were in bad shape and their main players had fit problems; having watched the game, I haven’t changed my mind — they are in bad shape and the fit problems are real. But I have to admit I underestimated two things: first, Real Madrid’s bench depth — even with the main players struggling, their young players (Espí, Endrick) could still change the game late on when both teams tired; second, their corner and set-piece execution was better than I thought. Winning multiple corners in a row in the last seven minutes was no accident; it looks like specific work on the training ground. This was a blind spot my pre-match analysis didn’t cover at all.
Summary: when the gap is clear, the over is often worth more than the win — and bench depth decides games
Two things were most useful to me this round:
First, when the gap is clear, the over is worth more than the win. Bayern at 1.08 and Liverpool at 1.20 — I said before both games that the win had no value and the over was the better bet, and those two games produced 6 and 9 goals. Six of the seven games went over 2.5 goals. It shows that in lopsided games, whether the strong side dominates or the weak side tries to cause an upset, the game tends to head toward the over. So when the win price isn’t attractive, the over is definitely worth considering.
Second, bench depth matters. Liverpool used plenty of substitutes with Isak, Gakpo and Szoboszlai out, and Real Madrid turned the game around with their bench. Both show how much it matters: a bench can change a game late on, or keep a team competitive when the starters are injured. Bench players are often less famous, and with so many teams, even long-time fans can’t always name the important ones on each side — but if you want to predict more accurately, paying attention to bench players can definitely improve your hit rate.
But as always: betting carries risk, so please bet responsibly.