There is a particular sound a racecourse makes when a handicap goes off with five runners instead of twelve. Not silence, quite — but a thinning, as though somebody had turned the volume down on the crowd along with the field. You hear it best on a wet Tuesday somewhere unglamorous: no Sky cameras, no fanfare, just a handful of horses circling under a sky the colour of wet slate, and a betting ring so quiet you can hear the tic-tac men doing the crossword. Twenty years ago that was the exception, tucked away at the unfashionable end of the card. Now, in parts of the jumps programme built to carry the sport's future — its novice chasers, its next King George horse — it is fast becoming the house style.

New figures from RaceMetrics, drawn from the Proform Racing database and setting British handicaps in 2008 against 2025, show precisely how we arrived here. They do not flatter the story British racing prefers to tell about itself.

A pyramid turned upside down

At the bottom of the Flat scale, races carded per hundred horses at that level have more than doubled since 2008 — from 16.1 to 39.7 in the lowest band, and from 15.5 to 32.2 over jumps. ("Horses at that level" means horses whose average rating that year rounds up to the band's ceiling: the horses for whom that band is home.) More opportunity for the cheapest horses sounds, on paper, like exactly what a struggling training industry ordered. But racing did not stop at the bottom. It kept building upward, adding volume too at the very top of the jumps scale, in the 0-120/130 and 0-135+ bands, where the number of races rose by two-thirds or more for a population that grew by only 12 and 27 per cent — so races per horse there rose by a third to a half. The one place the programme actually shrank was the solid, unglamorous middle of the Flat — the 0-75 to 0-85 bands, the only ones with fewer races than in 2008 (down 4 per cent) despite 8 per cent more horses to fill them, so that races per horse fell 11 per cent — precisely where a sensible programme would want depth.

More at the bottom. More at the top. Less in the middle that used to hold the whole thing together. It is not a pyramid any more. It is an hourglass, and hourglasses run out.

The arithmetic of emptiness

Here is where the story turns from curious to damning. A second dataset tracks handicaps run with fewer than eight runners — the threshold at which each-way punters lose their third place, and the product itself gets quietly worse. In 2008, that was a minor irritant almost everywhere: 24.4 per cent of jumps handicaps in the 0-120 to 0-130 bands, 19.5 per cent in the 0-105 to 0-115 bands below them, 12.1 per cent at the very top of the jumps scale, and single figures in the cheapest Flat handicaps. By 2025, the 0-120 to 0-130 bands — among the very ones racing chose to expand — were filling under eight runners 61.6 per cent of the time. The bands below them: 58 per cent. The top of the jumps scale, the 0-135+ band: 48.6 per cent, four times its 2008 figure. Even the bottom rung of jumps racing, supposedly awash with new opportunity, now sees more than a third of its handicaps go short.

Racing built more races precisely where it had the fewest horses capable of filling them, and is now surprised that the fields it built are empty. It is rather like doubling the number of motorway lanes on the stretch with the least traffic, and wondering why the new lanes stand vacant while the old ones jam.

The regulator's own confession

Credit where it is genuinely due: unlike much of the sport's messaging, the BHA has not tried to dress this up. Its 2027 fixture list, published in early September, holds race volume almost flat — 1,440 fixtures, down just 18 — despite years of tactical tinkering aimed at improving field sizes. Racing's director of racing, Richard Wayman, was candid that the board chose to prioritise prize money and income over cutting the programme, accepting that fields would stay "under pressure." The National Trainers Federation, in its response, was blunter still, warning that consumers should expect "fewer each-way betting opportunities than at any point in recent memory." The Professional Jockeys Association added its own warning about the declining broodmare band and foal crop sitting underneath all of this. Nobody at board level is now pretending the trend in these charts is about to reverse. They are simply choosing to live with it, for reasons that have everything to do with income and very little to do with the horse.

Tomorrow's card offers a live illustration. Kelso stages six jumps races on 16 September with fewer than thirty runners across the whole afternoon — barely five runners a race. It is not an outlier; it is what "under pressure" now looks like in practice, on an ordinary Wednesday, at an ordinary track.

Who actually wants the extra race?

Nobody sets out to card a five-runner handicap. But somebody benefits from the volume regardless of whether it fills: racecourses paid on fixtures and broadcast hours, media-rights arrangements that need content whatever the field size, a calendar built to service the machine rather than the horse population feeding it. The bettor loses a place. The trainer loses a knowledgeable opponent to test a horse against. The sport loses the thing field size actually measures — competitive depth — while the fixture count itself keeps up appearances. Nobody is hiding this. It is simply nobody's job, within the current structure, to stop it.

What racing owes the horse

The instruction now writes itself. Publish field-size targets by rating band, not just aggregate averages that hide exactly where the collapse is worst. Bring forward the levy-funded breeding and retention incentives the PJA says are still awaiting approval. And the Horserace Betting Levy Board, under Alan Delmonte, should finally publish turnover by fixture rather than as one industry-wide number — because until we can see whether all this extra racing is actually generating betting turnover or merely diluting it across thinner fields, "boosting opportunity" remains a slogan, not a strategy.

Racing has always asked a great deal of the horse. It might, for once, ask a little less of the fixture list — and a little more honesty of itself.

Data supplied by RaceMetrics, Cutting Comment's data partner, drawn from the Proform Racing database: every British handicap, 2008 and 2025.

Readers can get 20% off Proform's RaceMetrics Pro for three months with code CUTTINGCOMMENT.