Do old horses win races?

Racing loves an old stager the way football loves a one-club man. The twelve-year-old chaser who has been jumping round the same tracks since anyone can remember, ears pricked at fences he knows by heart, draws a cheer no hotshot novice ever earns. And somewhere inside that cheer lives a betting theory. Pascal, the Lab's eternally hopeful punter, states it with feeling: the market is a fashion market, obsessed with youth, with unexposed types and ratings on the rise. The old boys get written off on their birth date, he reckons, and drift out to prices that insult them. But they know their job better than any four-year-old ever could, they jump for fun, they stay all day, and their trainers place them shrewdly where they can still win. Experience, says Pascal, is the last under-priced thing in racing.

It is a theory with warmth on its side, which is exactly why it needs testing rather than trusting. So we tested it at its bluntest: back them all. One flat-stake win bet on every runner aged ten or older in every British race in our database, 8,563 bets in all, settled at the official Starting Price and counted honestly. Note the unit, bets rather than races. A staying handicap chase can field three of the old brigade at once, and each one takes its own stake, because that is what following the rule actually costs.

If the market really does insult the veterans, the prices will be too big for their chances and the bottom line will show a profit, or at least a gentle loss. What the bottom line shows instead is one of the worst results on our entire board, and the reason, when you trace it, turns out to be the cheer itself.

Start with the memories, because everyone has one. The afternoon an old handicapper rolled back the years at a big meeting, the veteran chaser who put in a clear round while the young pretenders fell in a heap, the roar that greeted both. Those stories get retold every winter, and every retelling makes the next veteran look like a fairytale waiting to be collected. The afternoons the old boys trailed home tenth leave no stories at all. Sentiment keeps the wins and bins the losses, which makes it the most selective accountant in betting.

Then the prices do their work. The ten-plus brigade mostly go off at long odds, and long odds promise the two things a punter's heart wants most: a proper payday and vindication. Backing a veteran at a big price is not just a bet, it is a small act of loyalty against a market that has moved on. When it comes off, you were the one who kept the faith, and that feeling is worth more to a belief than any spreadsheet.

There is also a respectable-sounding argument underneath the warmth, and Pascal makes it well. Experience is real. Old chasers do jump more safely, they do know the tracks, their trainers do place them cleverly. None of that is invented, and it reads like value investing: backing the proven article the fashion market forgot.

And the crowd agrees, which is the detail everyone misses. The veterans are the best-loved horses in any race they contest, drawing each-way money for the name and loyalty money from everyone they ever carried home a winner. Affection buys tickets. The old stager is never friendless in the market, and that fact, warm as it is, turns out to be exactly where the bet dies.

Begin with what age honestly means in the formbook. A horse still racing at ten or older is a known quantity in a sport that pays for surprises: usually past its physical peak, thoroughly exposed by years of running, and carrying a handicap mark that describes its ability almost perfectly. Its genuine chance in most races is small, and to be fair to the market, the long prices broadly acknowledge that. If this were the whole story, backing the veterans would lose at about the rate any blind longshot line loses. It loses faster, and the difference is the affection.

Sentiment is not a mood, it is money, and money moves prices. The old stager attracts backing no ordinary outsider gets: each-way loyalty from everyone it ever won for, name-recognition money from once-a-year punters who know exactly one horse on the card, tenners staked for the story by people who would be halfway glad to lose them. Every pound of it shortens the price. So the veteran ends up with the worst combination the board can offer: the small chance of an ageing, exposed longshot, at odds clipped below even that small chance's fair level. You are not paid properly for the risk you are taking, because the crowd's love has already been spent into the price you receive.

On top sits the engine that grinds down every line on this board, the bookmaker's margin. Add up the chances implied by every price in a race and they come to more than 100%, about 12% per race extra on a typical field and climbing towards 30% in big fields. And the margin is not spread evenly: it leans hardest on the longest prices, which is exactly the end of the market where the old stagers live. Small chances, shortened by sentiment, taxed at the market's steepest rate. Three leaks, stacked, and the bottom line shows all three at once.

The rule is blunt on purpose: one flat-stake win bet on every runner aged ten or older, no exceptions. From the 26,839 real British races we hold, every qualifying runner across both codes took a stake, 8,563 bets in all.

The unit matters, so we will spell it out. This experiment counts bets, not races. A single staying chase can put three veterans in the same field, and each one takes its own stake, because that is what the blanket rule costs to follow in real life. Quoting races instead would make the sample look smaller and the outlay kinder than the betting actually was, and this board does not do kinder-than-real.

Settlement is the standard honest set used by every experiment here: official industry Starting Price, non-runners voided, a recorded winner required, and the non-completers, the fallers, unseated riders and pulled-up veterans, counted as the losing bets they are. That last rule matters more here than almost anywhere, because the ten-plus brigade do most of their racing over jumps, where failing to finish is a routine way to lose a stake.

We did not filter for form, class, trainer intent or the right sort of veteran, because the theory under test is the warm blanket one: that experience itself is under-priced, that the market writes horses off on their birth date. A sharper version, veterans in form or dropped to a winning mark, sits on our open-questions list. And as always, SP with no commission and no account restrictions is the kind version of reality, so a real punter following this line would do a little worse than the figure on the next page. The figure does not need the help.

Here is the number, and it is one of the worst we have ever printed. A flat stake on every runner aged ten or older returns -27.54% across 8,563 bets. For every £100 you put through, about £72.46 comes back over the long run. The blanket bet wins 9% of its wagers, fewer than one in ten, and the 95% range is [-34.7,-20.2]: wide, because longshot results are lumpy, and heavy from end to end. Even the kindest reading of this sample is a painful loss.

The comparisons put it in its place. A horse picked completely at random loses -21.6%; the old stagers lose substantially more. The market's own first choice, the favourite, loses -8.6%, a fraction of this. And down at the foot of the board, blind-backing the rank outsider in every race loses -34.6%, which is the company the veterans keep, far closer than anything so well loved has any right to.

The strike rate tells you what the eye already knows: the old boys rarely win. The ROI tells you the part the eye misses: even their long prices are not long enough. Here is the test that proves the sentiment premium is real. If the market merely wrote veterans off, as Pascal claims, their prices would be too big and this line would beat the random one, because over-priced longshots pay better than average. Instead it loses more than random, which means the prices are consistently a notch too short for the results. That gap between the veterans' line and the pin's line is the crowd's affection, made visible and given a cost. Love, at the racecourse, is a lovely thing in the grandstand and an expensive thing on a slip, and now you know roughly what it charges: the difference between -21.6% and -27.5%, on every pound, forever.

So, do old horses win races? Sometimes, and may they never stop. The blanket bet wins 9% of its 8,563 wagers in this experiment, and every one of those wins lifted a grandstand somewhere, because nothing in racing is celebrated like a veteran rolling back the years. Nothing in this experiment says stop cheering. It says stop funding the cheer at the betting window: backed blind, the ten-plus brigade return -27.5%, worse than picking horses at random, one of the heaviest losses on our whole board.

The mechanism is worth carrying away, because it is the purest example of a truth this Lab keeps finding: a price is made of information plus emotion, and only the information part is ever on your side. Age is information, and the market handles it coldly and well. The affection is the crowd's own contribution, and the market simply charges for it. Every tenner staked on a name, a memory or a story shortens the price for the next person who loves the same horse, until the dearest old warriors in training are also the dearest bets in the ring. Sentiment is the one edge the crowd hands over voluntarily.

Keep the good halves of all this. Veterans are racing's soul, and the sport would be poorer without a packed stand roaring one home. If you want a practical rule, it is this: notice when you are betting a story rather than a price, because the story is always the most expensive thing on the card.

The dull rules finish the job, as ever. Past performance is not future returns, SP with no commission flatters even this figure, and no staking plan turns a negative edge positive, it only reshapes the swings. A flat stake you can afford to lose, or a fraction of the Kelly criterion if you genuinely hold an edge. The old stager, bless him, is not that edge. He is the best story in racing, and stories belong in the grandstand, not the ledger.