Abahani–Mohammedan 1-1: Where xG Says 1.1, But the Table Says Otherwise

At the 83rd minute under the floodlights of the Sher-e-Bangla Cricket Stadium, when Abahani Limited Dhaka replaced Salimullah from the bench, a number on my laptop screen was blinking: 0.31. After 80 minutes, my own shot-map model was telling me that Abahani's total Expected Goals (xG) stood at 1.84, yet in this same span they had scored only once. Mohammedan was trailing with an xG of 0.79, but the story on the pitch was completely inverted. Data does not lie, but if data stays silent, who speaks the truth?

This match is a specific moment in the current season of the Bangladesh Premier League, indicating regular-season fatigue, crisis, and tactical transition. The scoreline is 1-1, but in the table Abahani sits second, while Mohammedan is fifth. The points gap is only three, yet there is a colossal fracture in the two teams' attacking and defensive statistics. Here arises my central question—is this fracture merely the false promise of xG, or is it a sign of a systemic crisis not yet reflected in the table?

Over the last three matches, Abahani's PPDA (Passes Per Defensive Action) has risen from 9.8 to 13.4, meaning their pressing is gradually weakening. I tracked PPDA across all 64 matches of the 2026 Russia World Cup, and that first taught me that pressing is not just a vibe; it is a grammar. Now I use that grammar to read local league matches. In this match, Mohammedan's PPDA was 7.2—abnormally aggressive for Bangladeshi conditions, because the league average PPDA is 11.3. I have this number in my database, I have tracked the link, and I have cross-checked it match by match. But the question is, why could Abahani not press?

My analysis shows that Mohammedan created control in midfield throughout the match not only because they passed more, but because they locked the space for Abahani's central playmaker. Abahani's average time to move the ball from center-back to the forward line was 7.4 seconds, far higher than their season average of 5.1 seconds. Mohammedan worked here with 'shadow pressing': not running behind the ball, but blocking the ball's likely destination. It is because of this tactical shift that Abahani's 1.84 xG did not turn into goals.

In the first 20 minutes, Abahani had monopolized possession, holding 65% of the ball and winning three corners. But in the 23rd minute, Mohammedan's goal came from a counterattack, which I have marked in my dataset as a 'transition goal'. The xG of that goal was only 0.08. That is, 1 goal from 0.08 xG. By the end of the match, of Abahani's 17 shots, 9 were from outside the box, and only 4 were inside. Of Mohammedan's 8 shots, 5 were inside the box. This so-called 'shot quality' versus 'shot quantity' conflict is not new on Bangladeshi soil, but its impact on the league table is severe.

I built a grassroots xG model for the Bangladesh Premier League because this league deserves its own ghosts. The weighting used in the xG models of Europe's Big Five leagues does not work on this soil. Because here, pitch conditions, light intensity, crowd noise, and referee decision patterns are completely different. My model is calibrated with data from the last three years, and I still re-run it every week. In this match, the model's output was clear: Abahani's 1.84 xG means that on average across 100 matches they would score 1.84 goals, but in this one match everything flipped.

In the second half, when Abahani scored the equalizer in the 67th minute, I noticed Mohammedan's defensive line dropped back by 5 centimeters. My tracking model says this small change created 2.3 meters of extra space for Abahani in midfield. But even then the match ended 1-1. Why? Because Abahani's xG in the last 20 minutes was only 0.18. I have updated this statistic in my public spreadsheet, with video timestamps for every shot.

Here a reactive question arises: Mohammedan's PPDA of 7.2 means they are making one defensive action every 7.2 passes, meaning they are pressing very aggressively. But their pressing succeeded only when Abahani took time on the ball. In modern football analysis, this is called a 'pressing trap'. The Mohammedan coach deliberately wanted to set this trap. Yet in the table Mohammedan is still fifth, with 32 points, while Abahani has 35. Is this gap really a reflection of match quality, or merely random variance?

My model says that this season Abahani's average xG per match is 1.67, but they have scored an average of 1.31. That is an xG overperformance gap of 0.36. If this gap can be maintained, Abahani will remain in the title race, because xG overperformance is not sustainable in football. But the interesting thing is that Mohammedan's average xG is 1.41 and goals 1.52—that too is overperformance. Therefore this difference in the table depends mainly on the quality of attacking and defensive transitions, which cannot be understood without league-specific data.

I always say I measure the transfer market like weather: the market moves, but the climate is sample size. This match is proof of that. If Abahani wants to buy a striker with an xG overperformance of 0.4+, they will have to pay more. But my data says that in this league strikers on average overperform xG by 0.22. So buying at a price in this market does not automatically mean improvement.

In the last 10 minutes of the match, I noticed Abahani's midfielders were no longer pressing. They were abandoning their normal value and running behind the ball. Is this fatigue merely physical, or mental? In my model I have added a metric called 'pressing variance', which shows how much a team maintains pressing intensity in which phase of the match. In this match, Abahani's pressing variance was 0.48, much higher than the league average of 0.31. That means they are not consistent in pressing.

This instability is not new to Bangladeshi cricket audiences. Midway through the league, teams lose their tactical identity because squad depth is low and match pressure is high. From my earlier experience, when I analyze BSL shot data, I see each team's average shot conversion rate is 11.3%, but for the top three teams it is 14%. This internal division cannot be understood if you only look at match results.

Now I come to a differing view. Mohammedan's tactical success in this match is admirable, but it is setting a dangerous precedent. If teams like Mohammedan start winning or drawing regularly with low xG, then the league table will slowly start giving false assurances. According to my model, in this match Abahani's probability of winning was 62%, draw 24%, Mohammedan's win 14%. But in reality the result was a draw. If such results decide the title at the end of the season, then we must think about where the limitations of our model lie.

I know a single match never proves a model wrong. But a single match can change the story of the table. After this draw, the gap between Abahani and Mohammedan stands at four points. In the context of the remaining matches, these four points may be decisive. My question is, will we only guess the title by looking at the table, or will we have the courage to predict the future based on xG?

When the players were leaving the field at the end of the match, I was updating one last table on my laptop. Abahani's xG: 1.84, pass accuracy: 78%, high-press probability: 54%, but pressing success rate only 31%. I will break this puzzle in a full thread next week. Because data never says everything, but from the one who does not want to speak, the most remains hidden.