Last updated: 21 June 2026
Some backers look at the horse first. Plenty of others look at the people around it. They scan a race, find the trainer they trust and the jockey they rate, and start there. It is a reasonable instinct. The trainer gets the horse fit, picks the race it can win, and decides when it is wound up to go. The jockey has to get it into the race, settle it, and time the run. Between them they shape what happens in a way no single bit of form figures can capture, and that is exactly why connections are such a useful way to cut a crowded card down to a handful of names worth a proper look.
The trouble is that the moment you start weighing trainers and jockeys, you have tripled the amount of information in front of you. Is this the right trainer for this horse? Are they trying today, or giving it a run? Does the jockey suit it, and do they go well at this track? You can drown in those questions on a busy Saturday. A connections shortlist is the answer to that overload, but only if it is built the right way.
How do you shortlist horses by their connections?
Start with the trainer and jockey, not the horse. Pull out the runners whose trainer and jockey are both performing above a sensible strike-rate floor, ideally under today’s course, distance or going, and treat that as your shortlist. Then check the thing the strike rate hides: are those winners beating their prices, or just short favourites doing what favourites do? Connections narrow the card quickly. They do not decide the bet on their own.
Why connections are a shortlisting tool, not a verdict
The point of leaning on connections is speed. On a day with seven meetings and eighty races you cannot study every runner properly, and you should not try. What you can do is sweep the card for the horses whose trainer and jockey are both in good order, mark those as worth investigating, and let the rest go. That is shortlisting, and it is a genuinely powerful way to spend your limited attention where it has the best chance of paying off.
What it is not is a verdict. A strong yard and a top rider get you onto the page, nothing more. The horse still has to suit the conditions, handle the going, and be competitive at the class. Treat a connections shortlist as the list of runners that have earned a closer read, then do the closer read. The fastest way to lose money with this approach is to mistake the shortlist for the selection and back a name because the silks are familiar.
The strike-rate trap
Here is where most connections-based betting quietly leaks. A strike rate, the bare percentage of runners that win, feels like the obvious number to filter on. Race Advisor’s analysis of more than 1.4 million UK runners between 2013 and 2026 shows why that instinct is half right and half dangerous. Runners whose trainer has been winning at fifteen percent or better over the previous fortnight win 15.1 percent of the time, against 9.9 percent for runners from colder yards. The strike rate clearly moves. But a strike rate only means something once you set it against expectation, and expectation lives in the odds. Line those same in-form trainers up against their Betfair starting prices and the actual-versus-expected figure comes out at 1.04, exactly the same as the colder yards on 1.04. The extra winners were already in the price. Twenty-five percent from a stream of odds-on shots is not an edge, it is the market being roughly right.
The same applies to riders. The leading jockeys are on the best-backed horses almost by definition, so their strike rates are high and backing them blind tends to grind a slow loss, because the price has already swallowed the advantage. A modest strike rate is not automatically a warning either, because a trainer sending out bigger-priced horses that win their fair share can be far more profitable to follow than a fashionable yard whose runners are always too short. So the bare percentage is a starting filter, useful for sweeping the card, but it is the wrong number to bet on. The number that matters is whether those runners win more often than their prices said they would.
Building a connections shortlist
The practical method is a combined filter. Take the runners whose trainer and whose jockey are both clearing a strike-rate floor you have set in advance. Jockeys behave just like trainers here: riders winning at fifteen percent or better over the last fortnight win 14.7 percent of their races against 10.1 percent for the rest, and again the actual-versus-expected figure barely shifts off the field. Where you put that floor is your call, and it should flex with the type of racing, but somewhere around a sixth to a fifth of runners winning is a reasonable starting line for both. Requiring both halves of the partnership to clear it does a lot of work on its own. Across the same sample of more than 1.4 million runners, those with both trainer and jockey in form win 18.0 percent of the time, nearly double the 9.3 percent managed when neither is firing, because a strong trainer paired with a cold jockey, or the reverse, is a much weaker signal than the two pulling together.
Then tighten it with conditions. A blanket strike rate is less telling than the same figure read under today’s circumstances, so favour connections who also go well at this course, over this distance, or on this going. A trainer and jockey who clear your floor in general and clear it again at the specific track are a much firmer shortlist than one built on overall numbers alone. The code matters as much as the conditions. That combined in-form signal is priced almost perfectly on the Flat, where both-firing runners return an actual-versus-expected figure of around 1.00, but over jumps it pays. In chases, runners with both connections in form have won 23.1 percent of the time across more than twenty-three thousand runners, at an actual-versus-expected figure of 1.15, comfortably clear of the level where a factor is genuinely beating the market. Recency helps too: in-form trainers and the jockey strike rates of the last couple of weeks tell you who is in good order now, which matters more than a season-long average that may be carried by a spell three months ago. None of this is a single magic threshold; it is a set of filters you stack until the list is short enough to study properly.
From shortlist to selection
Once the list is down to a few names, the connections work is done and the racing work begins. This is the point to build a racecard that puts the trainer and jockey win and place strike rates, and ideally their profit at the course, alongside the horse’s own claims. Race Advisor’s own most-suited-trainer flag is a useful sharpener here: when a runner’s trainer is the one the data marks as best suited to the race, those runners have won 21.3 percent of the time, the highest strike rate of any single connections filter, though at an actual-versus-expected figure of 1.05 it stays a shortlisting signal more than a standalone bet. A trainer showing a level-stakes profit at a track over a real run of runners is more interesting than one with a flashy strike rate and nothing to show for it, because the profit is the clue that the market has been underrating them there. Read that against a course record for both halves of the partnership and you can see quickly whether the conditions today are the ones that produced the edge.
From there it is ordinary form study. Does the horse want this trip and this ground? Is it well in at the weights, fit, and competitive at the class? Sometimes the answer is that all your shortlisted runners are worth a bet; sometimes it is that none of them are quite right today, and the honest move is to pass. Both are fine outcomes. And connections should never be your only sieve. Run it alongside other ways of reading the form, because a runner that survives two or three independent shortlists is a far stronger proposition than one that only ever looked good through a single lens.
Let the data keep you honest
Every judgement above is a hypothesis, and the only way to know whether a connection is a real edge or a comforting habit is to measure it. The measure that does this honestly is the actual versus expected ratio, the ratio of a connection’s actual wins to the wins their runners’ Betfair starting prices implied they should have managed. Above the value threshold and the trainer or jockey is paying their way; at or below it, you are backing a memory of past winners rather than a live edge, however good the headline strike rate looks. It is also the cleanest way to tell an in-form yard that is genuinely overpriced from one that is winning exactly as often as the market already expects.
Race Advisor’s data covers more than ten years of UK racing and is built to answer exactly this kind of question, putting win-rate and actual-versus-expected figures on trainer and jockey records by recent form, course, race type and going, so a connections shortlist can be judged on evidence instead of reputation. The headline from over 1.4 million runners is the lesson in one line: in-form connections nearly double the win rate, from 9.3 percent when neither is firing to 18.0 percent when both are, yet the actual-versus-expected figure holds at about 1.04 across almost every split. The strike rate is real, the value mostly is not, and the one place it clearly tips over is jumps, chases above all. That is the principle worth betting on: connections are a fast way to shortlist, a strike rate alone is not a reason to bet, and the runners worth your money are the ones whose people beat the market, not just the field. Pull up today’s racecards, build your connections shortlist, and let our free Bet Tracking keep an honest record of whether it holds up.
