Twenty-Five in an Over: Why Test Cricket's Most Expensive Over Is the One Nobody Bids For

I didn't want to begin with the 25-run over. I wanted to begin with the man who had no business batting in it.

December 2026, Ellis Park, Johannesburg. On the night of 24 December, a New Zealand passenger train was swept off a bridge at Tangiwai by a lahar; more than a hundred people died. Bob Blair's fiancée, Nerissa Love, was on that train. Four days later Blair walked out to bat at No. 11, against Hugh Tayfield, an off-spinner whose entire reputation was built on denying runs.

The first four balls went six, six, dot, six off Bert Sutcliffe's bat; the fifth brought a single. Blair hit the sixth for six. Twenty-five off the over. It was the first time in Test history that a bowler had conceded 25 in an over. The number that matters, though, is not 25. It is 47.

Because from that December in 2026 until the turn of the century — roughly 47 years — no bowler in men's Tests conceded 26 in an over. Wisden's quiz record is plain about it: the 26 barrier was not breached until the 21st century. Every Test of the 20th century, every bowler, every hitter, and the wall held.

Tayfield needs to be understood separately. A South African off-spinner famous for parsimony, a man whose overs were where batters went to breathe. It was his over that produced the first 25. It was also an eight-ball over — South Africa used them then — and the 25 was complete inside the first six balls; Blair did not score off the last two.

Remember Sutcliffe's part in it too. He batted with a bandaged head after taking a blow. So the two men central to the over were one half-fit batter and one grieving one. Test cricket has rarely produced a messier, more human set of conditions.

About 27 years later, at Port of Spain in 2026/81, Ian Botham bowled to Andy Roberts. Four, six, two, six, six off the first five balls — 24. A leg-bye off the sixth. Twenty-five off the over, the record equalled, but Botham's figures recorded 24.

That one-run gap is the most important piece of information here. The record book says 25; the bowler's figures say 24. What we are measuring is not a bowler's merit but an event, in which byes, leg-byes and no-balls all land in the same bucket. One run in a data table; a turning point in a history.

Watching those two overs again and again, one pattern becomes obvious. In Tests, the bowlers who concede the most expensive overs are not bad bowlers. One was the most parsimonious spinner of his age; the other is one of the greatest all-rounders the game has had. That is no coincidence. A bad bowler gets taken off before he reaches 26. It is the good bowler who is kept attacking, with mid-off up and a slip in place. This over is not a weapon. It is a receipt — proof of what a captain bought and what he paid for it.

The eight-ball arithmetic deserves attention, because it is where every comparison collapses. Sutcliffe and Blair's 25 came off six balls, 4.17 per ball. Roberts's 25 also came off six balls, exactly the same. The only difference is that the first over carried two extra deliveries on which nothing happened. On a per-ball basis the two overs are identical, yet the record book sits them side by side. Anyone who builds a leaderboard on runs per over is measuring a rulebook, not cricket. A proper cross-era table needs runs per ball, the match run rate, the bowler's role and the field setting — four inputs, not one.

So what does the 47-year silence say? It says 20th-century Test batting had a limited appetite for risk. No. 10 and No. 11 were taught to survive, not to swing. Captains protected strike bowlers with sweepers back. Attacking a spinner carried an uncertain reward: an old ball came off the bat more easily, but wickets were expensive.

The wall broke in the 21st century for three reasons, none of which is that bowling got worse. Volume first — more Tests and more overs bowled since 2026; rare events stay rare when you roll the dice less often. Vocabulary second — the slog-sweep and the reverse scoop migrated from T20 into the bag of a No. 9, whose job used to be seeing off the ball. Reward third — a six is broadcast inventory now; nobody clips a forward defensive.

Having spent the last few years building my own over-by-over Test datasets, mostly at night in a Levenshulme flat with scorecards open on two screens, one thing keeps showing up. Analytical models work on the mean — expected runs per over, economy, win probability. Records live in the tail. The mean tells you who bowls well; the tail tells you which over changed the temperature of a dressing room. A 26-run over inside a 450-run innings is under one per cent of the match. It is also the six balls in which a captain's plan dies, the field scatters and the next batter starts believing.

Here is the part the numbers miss entirely: a 26-run over rarely changes the result of a Test. What it changes is who plays the next Test and who doesn't. These overs are not outcomes; they are career verdicts. The most expensive over in the record book is usually the one with the least bearing on the scoreboard and the most bearing on the balance of power inside a squad.

Now apply the logic of a transfer window. Agents in this window sell death-over specialists; auction prices are set by who bowls the 19th over. Nobody prices a Test over. The reason is simple: in Tests the price of an over is set not by the bowler's skill set but by field placement and match situation. The man who defends the 19th over in a T20 can go for 26 in a Test, because there is no sweeper and his captain is running an experiment. The Test over is the last unpriced over in cricket, and that is exactly where the extremes live. Price something and coaching templates arrive; templates arrive, players know the answer before the question, and the room for the extreme shrinks.

And Blair's six? That was not technique. To a man whose life had been taken apart four days earlier, being careful is not an instruction. Risk appetite here is a state of mind, not a method. The over that entered the record book was bowled to someone with nothing left to lose. Time does not wash away a thread; it makes the ink run deeper.

What is the strongest case against me? This: the 47-year silence may not be an event at all, but an artifact. Fewer Tests were played in the first half of the 20th century, ball-by-ball records were not kept, uncovered pitches favoured bowlers, and a No. 9 in that era had a completely different reason to hold a bat. So was the 26 barrier breaking about a change in risk culture, or simply more matches, more overs and more cameras?

The second objection cuts deeper. Modern analytical teams do build over-level models — expected runs, win probability, field-setting simulations. My claim that nobody prices a Test over is overstated. The reverse may also be true: attacking fields in the 20th century were often forced, because a captain without a fifth bowler had to keep mid-off up. What I am reading as a captain's courage may simply have been a shortage of resources.

And one test should be run on my own argument. If the rate of 26-plus overs per 100,000 overs is flat across eras, my thesis collapses. In that case I will have dressed T20-era nostalgia in the clothes of a record book — and admitting that is the most honest part of this piece.

Still, I'll put a marker down, with a date. Before December 2030, men's Tests will produce at least two more 26-plus overs, and one of them will come in Asia, off a frontline spinner, with the field up, to a batter between No. 8 and No. 11. When it happens, will anyone ask for the bowler's name — or will they finally ask for the captain's?