Bangladesh Cricket's Data Revolution: 'All Data is False, But Some Data is Useful'

Sitting in an old press box in Khulna, I first learned how the scoreboard lies. Covering the Wills Cup in 2026, I saw a 47-run innings create a hero who was benched the next week. Nobody said then, 'His strike rate is mediocre, he has issues in the powerplay.' We said, 'His eyes were burning.' Thirty-three years later, I watch that same scoreboard tell an even bigger lie — only now the lie is wrapped in the garb of data analytics. What I am seeing in this season of the Bangladesh Premier League (BPL) is this: teams of analysts, opening data centers on laptops, turning a batsman not into a 'match-fixer' but into a 'match-winner' based on just a single strike-rate metric. The man I am talking about is Khulna's own son — Mehidy Hasan Miraz. Last BPL, his strike rate was 123. This season, the same strike rate is 108. But what nobody is watching is that Miraz is not playing the role of a finisher. He comes into the team in the 12th or 13th over, when spinners are bowling and the boundary is 70 yards. His job is to stay at the crease, rotate strike, and play the big shots in the last six overs. But the metrics say, 'Miraz's form is poor.' This is the danger I want to write about today. I have named this story 'The Arrogance of Data: Crisis of Human Judgement in Bangladesh Cricket's Metric-Dependent Era, Part 1'. Come, from my career experience, let us see how we have moved from the era of 'visual impression' to the era of 'table-dependent analysis', and what we are actually losing in that transition. At the start of 2026, the ninth edition of the BPL is underway. Before the league started last December, a new character joined the management of the franchises — the data analyst. This is truly new in Bangladesh's domestic cricket. Before, there was only the coach, captain, and manager. Now each team has at least one data analyst who has digested every ball, every delivery, every fielding position of the previous BPL seasons. But is this analysis revealing the real truth, or are we again creating a new blind faith in the name of technology? I have been reporting on domestic cricket since 2026. Even then, I had my doubts that in cricket there is no such thing as 'proof', only 'narrative'. But in the era of data analytics, that narrative itself has become a formula. Take young pacer Nahid Rana — this BPL season he has taken 14 wickets at an economy of 8.2. The table says 'excellent', but the day I watched him bowl at Khulna's Sher-e-Bangla Stadium, he was suffering from the biggest problem of fast bowlers — fatigue after the first spell. His second-spell speed dropped from 135 to 128. The data analyst cannot capture this fatigue because his model calculates by the hour, not by the over. But cricket is played over-by-over, ball-by-ball. Data sees only what it sees in front of its eyes. But the coach sitting in the field sees — 'At this moment, there is no flaw in his bowling action, but his chest is rising quickly, he is sweating more.' Can this human perception defeat the data model? I think this question is the real game now. When BPL franchises are hiring data analysts, a latent belief is at work behind it — 'Data is accurate, humans make mistakes.' But if I have sat in the galleries of domestic cricket for 19 years, I can say that Bangladesh cricket's big problem was never 'lack of data', but rather 'not trusting human judgement'. We can write a new epic of T20 cricket overnight if we want, but when data analysts are involved in choosing the hero of that epic, I fear they are turning the hero into just a 'run-machine' and dropping the 'good cricketer'. Let me show some real examples. In the last BPL, young all-rounder Ariful Islam played for Dhaka Capitals. The data analysts said his powerplay strike rate was 135 — so let him open. But he was sent to open and was out for 2 runs off 5 balls. Why? Because the data analysts did not know that Ariful's strength is the off-side, and the opposition bowlers had placed fielders on that off-side. The data model could not capture that field placement, and the opposition bowling unit won the match with that field setup. Who took the blame? Not the data analyst — Ariful. Yet that field placement by the opposition was a simple lesson for a 12-year-old cricket student. In the 2026 ODI World Cup, when Bangladesh's Tigers took the field, the team management carried data sheets from the BPL franchises. The sheet ranked top-order batsmen by 'strike-rate ranking', but there was no column called 'pressure-handling index'. As a result, when batsmen got out under pressure in big matches, our fans said, 'Their technique is poor.' I say, 'Their eyes were not burning.' This 'eyes were not burning' is an important part of human judgement that I have been searching for for 25 years. Data analysts see this 'eye-burning' as an 'outlier'. I say, 'Outliers are cricket's real story.' Take Litton Das. In the last BPL, Litton's batting average was 34, but the strike rate was just 115. For a wicketkeeper-batsman, this strike rate is 'insufficient' in today's T20 era. So the data sheet says. But the person who has seen Litton in the field will say — this 115 strike rate at 34 average means Litton stayed till the 15th over for the team, when there was spin on the pitch, and the team needed 'strike rotation'. The data sheet does not give importance to this 'strike rotation' because it does not capture 'dot-ball pressure' in a metric. I say, these 'invisible metrics' — strike rotation, time at the crease, playing according to the positions of fielders — these are real cricketing skills. They are not captured in tables, but they are visible to the eyes when you are at the ground. I remember in the 2026 BPL final, Mahmudullah Riyad played for Barisal Bulls. In that match, his strike rate was 108, runs were 22. But after the match I said, 'His 22 runs was the match-changing innings.' Why? Because he scored 22 off 12 balls, which had no big shots, only risk-free strike rotation and disrupting the opposition's fielding positions. The match result was a 4-wicket win. The data analyst would see those 22 runs as 'insufficient', but I see them as 'valuable'. This 'seeing' — this is the real difference. Data analysts say, 'We are objective, we are bias-free.' But I say, 'You are objective, and that is why you are making team errors.' Because cricket is not an 'objective' game, it is a 'subjective' experience. This subjectivity is what helps a coach make the right decisions. For example, one of Bangladesh's greatest Test moments — the 160-run win against New Zealand in Mirpur in 2026. On the morning of the fifth day of that match, Bangladesh spinner Taijul Islam was bowling, and the New Zealand batsmen were defending. The data sheet said, 'There is spin on the pitch, Taijul's economy is 3.2, good.' But what coach Russell Domingo said in the dressing room that morning was not in the data sheet — 'Boys, this is the biggest opportunity of your lives, put a smile on the faces of the people of the country.' That human appeal, that emotion, that subjective message — that changed the match. The data analyst would call this dressing-room talk 'noise'. I say, 'This noise is cricket's music.' The current BPL edition is underway. On February 9, 2026, I sat at Khulna's Sher-e-Bangla Stadium and watched the Khulna Tigers vs Chattogram Kings match. Khulna's score was 186, and chasing, Chattogram's young opener Tanzim Hasan was 41 off 23 balls. The net run rate was good, the requirement was 9.2. At this moment, the data analysts' model of Chattogram says, 'Powerplay over, now take risks, increase strike rate.' But sitting in the field, I saw a fielder at deep point and one in the cover region in front of Tanzim. Tanzim played a dot ball, then took a single. I sat and thought, 'This boy understands the match.' In the next over, he took four singles and one boundary. The data model was pressuring him to 'hit big shots', but he was actually managing — rotating strike with singles without taking risks, pushing the pressure of the match onto the opposition. When his teammate hit another boundary in the 15th over, I saw that 'burning' in his eyes — which is not in the data sheet. In the end, Chattogram won the match with 2 balls to spare. After the match, I did not interview that Tanzim — instead, I spoke with him at a tea stall. He said, 'Sir, the data analyst tells us the breakeven rate here is 112 — that is the strike rate. But I know that on a ground where the boundary is 70 yards, rotating the strike is better than increasing the strike rate.' I was impressed by his words and said, 'You have defeated the data analyst; you are a cricketer.' This small incident is the biggest lesson from my twelve years of domestic cricket reporting — data can blind people, but human experience reaches the pinnacle of knowledge. At this moment in the BPL, I am also noticing another thing — the 'Impact Player' concept has given data analysts a new job in every franchise. Impact Player means substituting a player midway through the match. Now data analysts are updating models mid-match and saying, 'Now bring on the spinner' or 'Now bring on the pacer.' But in trying to gain the impact player advantage, teams are actually breaking the important concept of 'batting order'. For example, in the last match, Sylhet Strikers had three spinners in the XI, but in the second innings they brought in an impact player — a batsman. The data analysts said, 'There is spin on the pitch, so drop the pacer.' The result? Sylhet lost by 6 wickets because they sent a 'finisher' in so late that he could not get into the match. Data analysts forget — no matter how much data you study, the pressure inside the crease, the environment of the field, the sound of the gallery — these will never be captured in metrics. In this section, I want to give an important fact that is not in the BPL data analysts' models — the 'fielding position map'. In cricket, each fielding position has a specific zone, and the fielders in those zones actually control the game. For example, if there is a fielder at deep midwicket, the batsman fears playing a shot at midwicket. The data analysts' model does not use this map because it keeps a statistical model of 'player location' but does not give weight to 'fielder influence'. As a result, when they advise increasing the strike rate, the batsman gets out on that exact shot where there was a 'deep cover' fielder in the field. This 'invisible knowledge' of the field cannot be learned by data analysts because it can only be learned by sitting at the ground. In 2026, I went to a primary cricket clinic held in Dhaka, where the former national team captain was present. He was telling the young bowlers, 'Boys, look outside the field, see who is fielding where.' I still remember those words. No matter how modern cricket technology becomes, that 'looking outside the field' can never be recreated with computer visualization. This is the 'human intelligence' vs 'artificial intelligence' battle that cricket is now witnessing. In this battle, the data analysts are actually losing, but they do not feel the pain of losing because they perceive only the 'mathematical model'. I am writing this column in the middle of the ninth edition of the BPL. So far, 24 matches have been played, and I have already seen that teams overly dependent on the data analysts' models are at the bottom of the points table. For example, Dhaka Capitals have won only 2 matches. Yet the data analysts' initial simulation said 'Dhaka Capitals in top-four'. What is the reason for this wrong prediction? Surely, the model does not capture the human element — a batsman's temperament under pressure, a bowler's fatigue, a captain's decisions. The model cannot calculate, 'If Nayeek is bowled in the 17th over, there is an 89% chance...' but the model does not know that this Nayeek will come in and hit 2 balls before being out for 6, because Nayeek's eyes were not burning. The explanation of this 'eyes were not burning' is commonly written off as 'bad form', but it is actually this strange mixture of nerves, skill, experience, and luck that can only be seen by sitting at the ground. I remember during the first COVID-19 lockdown in 2026, when all cricket in the country was stopped, I was sitting in my Dhaka flat and watching old BPL matches on TV. There was a match from 2026 where a young cricketer scored 30 off 14 balls to win the match. The post-match analysts called that innings an 'outlier'. But I had interviewed that young man. He said, 'Sir, that day I was not dependent. There was one at deep fine leg in the field, and I hit the six over that fielder's head.' This confession, this 'fielder-observation' — this is never in data models. That young man does not play in the BPL today, but that innings was the turning point of his career. If the data analysts had discarded that innings as an 'outlier', we would have lost the real story. This is my core point — 'No matter how precise the data, the real subtlety of cricket is the subtlety of the human heart.' The BPL authorities surely hired data analysts thinking they would bring an 'edge' to each team's performance. But I see that this 'edge' is increasing the distance — the distance between human and human. And the result of this distance we are seeing in the field — the wrong player in the wrong team, the wrong change at the wrong time. I am sitting in Khulna, having watched cricket for four decades, and the most important thing to me is — the queen of cricket is a human, not a model. I would request the data analysts, 'Do not hate your model, but do not forget the human.' Because the hero of this show remains the human. Bangladesh's BPL is now entering its youth. However data-driven T20 becomes, the future of this league depends on the 'cricketing curiosity' and 'field education' of the young players. Data sheets can quickly make them 'fan favorites', but cannot make them 'great'. To be great, one must be at the field, learn from the field, and take one's own decisions in the face of danger. I want to tell Tanzim Hasan, Ariful Islam, Miraz — all the young players — 'Read the data sheets, but listen to the voice of your heart.' Because that voice will one day make you the best match-winner of your career. I am writing this column sitting in a tea stall in Khulna, where the evening light was still present. At the next table, a few young cricketers are discussing their matches. One of them says, 'I scored 30 runs today, but my strike rate is 85 in the data sheet, I feel like I failed.' I was listening, and thinking to myself — this 'I feel like I failed' is the biggest ignored area of the data analysts. This feeling changes a player from a 'performer' into a 'performance'. I went to them and said, 'Don't think about your strike rate, think about how you will win the match.' They looked at me and smiled. That smile is enough for me — it is what no data analyst's model can create. Come, let us think about how we can get out of this dangerous dependence. First, franchises must, alongside hiring data analysts, emphasize 'coaching-mentorship', where the coach prioritizes the player's 'mental toughness'. Second, the data analysts' models must be validated through 'human observation'. The model says 'strike rate is bad' — but if the coach sitting in the field sees 'the player is playing great shots, only the fielders' positions are defeating him' — then the model is wrong. Third, players must sit with data analysts and openly discuss their own performances, so they can see the model as a 'companion' instead of an 'accusation'. Above all, data is a machine, a human tool. Humans will control this machine; the machine will not take human decisions — this will be the biggest challenge facing cricket. We cannot avoid this challenge, but we can prepare for it. In conclusion of my column, I say — the BPL has started its data revolution, but for that revolution to succeed, we must understand that the real meaning of 'revolution' is 'change in human consciousness'. Let the data analysts come to the field, but let them listen to the stories of the field. Because the epic of cricket is not written in models — it is written in human hearts, in the burning of their eyes. I am speaking from Khulna — stand by our young people, do not judge them, do not control them either — understand them. Data will then remain just 'numbers', or else it will become 'scripture'. In my 55-year career, I have seen many meta changes, but this meta-change — from human to model — is perhaps the biggest. I am confident that human experience will win this battle too. Because in the end, it is cricketers like Tanzim who win matches, who sit in the field and watch the fielders, who listen to stories at tea stalls, and say to the data sheet, 'You are a guideline, you cannot control me.'