Royal Ascot Recap: What Smartform’s Breeding Data Taught Us
By colin on Sunday, June 21st, 2026Five days, every card read through Smartform’s pedigree data before a horse ran. This is the honest account we promised — where the breeding lens saw clearly this week, where it was right to stay quiet, where it got things wrong, and the one lesson that ties it all together.
The idea was simple, and it wasn’t tipping — and it wasn’t improvised on the morning, either. Behind the week sat a long piece of work. We’d gone back over two decades of Royal Ascot, and a great deal of other top-class racing besides, and run the breeding figures from these same tables across all of it, to learn where the pedigree stats genuinely separate winners from the field and where they tell you nothing. That study is the map we read each card against — and, as much as anything, it’s what told us in advance which races to leave well alone.
So each morning we took the card and read it through the breeding tables — daily_dams_insights, daily_sires_insights and the experimental nicks layer for the cross of sire and damsire — and said, before racing, which races the pedigrees pointed to a winner in and which they couldn’t call at all. Posted in advance, every day, with no hiding from it afterwards. Here’s how it read.
Where the breeding earned its keep
The clearest and most repeatable edge was the sire read in the two-year-old sprints. Three times we pointed past a short-priced favourite to a better-bred sort at a bigger price, and three times it landed: the Coventry, where Great Barrier Reef (No Nay Never) won at 6/1 with the favourite only sixth; the Albany, where Libertango (No Nay Never again) beat an odds-on jolly at 6/1; and on the final day the Norfolk, where Orthodox (Havana Grey) won at 9/2 with the favourite back in ninth. That’s a pattern, not a fluke. At two and three, before the form book has much to go on, the sire line is a real and badly under-used guide — and it tends to be priced as though no one is looking, because mostly no one is.
A caveat ran right through those wins, and it’s an honest one: the sire figure points to the line, not the individual. We said so every time, and every time the winner came out of the line we’d named without always being the one we — or the market — fancied. The Chesham was the clearest example: we said Justify was the sire to be on but couldn’t tell you which of his runners, and after the favourite was withdrawn his other two filled the first two places.
The second edge was reading stamina and trip, where the form book is blind. The King Edward VII was the sharpest call of the week: we doubted a Kingman colt’s stamina at a mile and a half, and he was beaten ten lengths while the two horses we’d named as better bred for the trip fought out the finish in front of him. The Queen Alexandra marathon confirmed the same principle on the last day — Illinois, the deepest-evidenced staying cross on the card, won it, while the sprint-into-mile cross we quietly doubted at the trip trailed in.
The third was the dam, but only where the mare had a real body of runners behind her. The Prince of Wales’s was the standout — Ombudsman confirmed through the dam, with the 15/2 Minnie Hauk flagged for the places behind him — and the Hardwicke on Saturday turned up a 9/1 winner in Giavellotto, with the dam-confirmed favourite second.
Where it was right to say nothing
As much of the week’s value was in the races we left alone as in the ones we read, and that’s the part the prior research mattered for most. These weren’t hunches: across the two decades we studied, the breeding figures simply stop separating winners in handicaps and at the top of the sprinting tree, so we left every one of those races before the meeting even began — and nothing this week argued otherwise. Every elite open sprint went straight into the leave-alone pile — the King Charles III, the Commonwealth Cup, and on Saturday the Queen Elizabeth II Jubilee, where a 25/1 shot won a blanket finish with the favourite only third. That is the textbook proof: at the top of the sprinting tree the fastest horse on the day wins, and pedigree cannot separate a field all bred to the hilt for the job. Every big-field handicap went the same way — the Royal Hunt Cup, the Britannia, the Wokingham, the Golden Gates — where the weights and the market wash the breeding out. Knowing which races the data can’t call is every bit as much the method as knowing which it can.
One race we stood aside from for a subtler reason worth keeping. The Coronation is a fillies’ Classic, and the dams of three-year-old fillies are young mares whose records are dominated by the runners themselves — so the figures look like signal and are really a mirror. We showed the working and left it; the favourite won, which the dam data could neither have predicted nor opposed. Standing aside from a race you genuinely can’t read is a result, not a gap.
Where it got it wrong
The misses clustered, and the cluster is the most useful thing in this whole review. Almost all of them were the nick — the cross of sire and damsire — on a thin sample. The Ascot Stakes, and most of all the Queen’s Vase, our showcase pick of the week, which ran tenth of eleven on a cross that simply had too few runners behind it to mean anything. The nick is the most data-hungry of the three layers and the quickest to mislead when the numbers are short; it earns its place only on real samples, and this week it was the layer that cost us most.
The rest were of two kinds. A couple of soft sire leans that we’d flagged as soft and that duly came up short, the Hampton Court chief among them. And, on the final day, a confirmatory read that backfired: the Jersey, where we endorsed the best-bred favourite and then watched him run sixth while a 20/1 winner came from a sire we’d ranked only mid-field — a clean reminder that simply agreeing with the market leaves you nothing to fall back on when the favourite underperforms. The Gold Cup sat between the two camps: our preferred pair were beaten by the favourite, but we’d said the breeding had no edge among proven older stayers, and that half held — our two were placed, with nothing the market hadn’t already known.
The one lesson, if it comes down to one
Breeding data earns its keep on aptitude and stamina at the top level — the juvenile sprints, the staying trips, the good-class Group races where the field isn’t fully exposed and the market hasn’t fully priced the page. It goes quiet, and should be left quiet, in handicaps and elite sprints, where weights or sheer class wash the edge away. And it should never be trusted on a thin sample, however eye-catching the headline figure: the runs count isn’t a footnote, it’s the first number to read. The Queen Alexandra made that point twice in one race, where the two thinnest figures in the field ran in opposite directions — which is exactly why a low run count is a reason for caution, not a number to back or to oppose. None of this was the week’s discovery: two decades of results pointed to it before the meeting began, and five days of reading the cards in public simply bore it out.
That discipline is the whole thing. The question that turns a pedigree table from a curiosity into an edge isn’t “what does the breeding say?” — it’s “where is the breeding worth listening to?” This week the answer was: in the juvenile sprints, at a trip, and out of a proven mare; not in a handicap, not in a championship sprint, and not on three runs.
All of it is there to run yourself, any day and any card, in Smartform — the same dam, sire and nick tables we read from, the same figures, the same runs counts beside them. We followed it through one particular lens for five days and showed our working, hits and misses alike. It’s a lens worth following race by race, win or lose. Thanks for reading along this week.