How to Pick a Winner in Horse Racing: My Form-Study Checklist

Home › Form Guide › How to Pick a Winner

By The Guv'nor · Last updated October 2026

To pick a winner in horse racing, start with the race, then put every runner through the same checks: going, trip, class and handicap mark, then draw, pace and course, then trainer form and kit changes. Price the horses that survive yourself, and only bet when the odds are bigger than your price. Nobody finds every winner.

A worked example: the 2026 Ebor, read line by line

Learning how to pick winners in horse racing is easier with a real race in front of you, so here's one first. It's a past race and I'm walking through it with the result in hand, so treat anything that looks clever as hindsight.

The Sky Bet Ebor Handicap was run at York on Saturday 22 August 2026. It's a Class 2 heritage handicap for four-year-olds and up over 1m 5f 188y (about 2,787 metres, a shade under 1m 6f), and the ground was Good to Soft, Good in places. Twenty-two ran for a first prize of £300,000.

Daiquiri Bay (10/1, Alan King, ridden by Rossa Ryan) beat Hopewell Rock (13/2) by a head. Opportunity (14/1) was a neck away third and Ascending (28/1) was two lengths further back in fourth. The 7/2 favourite, Love Me Tender, finished tenth.

Here's how the first four and the market leaders looked on the morning, from public form only. The last column is from the in-running comments, so that part is hindsight.

Nobody ticks every box. The winner had the trip and course form but a question over softer ground, the favourite was a guess on almost everything, and the next two in the market had solid claims and ran poorly. That's normal. The checklist doesn't hand you the winner; it tells you which horses you can put a sensible price on.

Start with the race, not the horse

Before you open a form line, work out what sort of race it is. In a handicap the official ratings have already tried to give every horse the same chance, so you're looking for the one the handicapper has got wrong. If that idea is new to you, read how handicaps, ratings and weights work first.

Next, look at what usually wins it. The Ebor goes back to the 1840s (York's own history says it had been introduced by 1846); here are the last six winners, compiled from Sporting Life's results.

Six races is a tiny sample, so read only broad shapes into it. Every winner was rated between 97 and 108 and carried between 9st 3lb and 9st 8lb. Three of the six were four-year-olds, two were favourites, and none went off bigger than 10/1. Four of the six were run on Good to Firm (Good in places), so 2026's Good to Soft was the softest ground in the set.

Handicaps are also where favourites struggle: my own count found handicap favourites won 29.8% of the time (N=8,333, Betfair SP, 1 Oct 2025 to 30 Sep 2026). More on how often favourites win.

The checklist: going, trip, class and mark

These four checks do most of the work on any card. If a card's figures are new to you, my guide on how to read horse racing form goes through a racecard line by line.

Going

Check the official going on the morning, then each horse's runs on similar ground. Fourteen of the 22 Ebor runners had last run on good to firm (my count from Sporting Life's form), so Good to Soft was a question for plenty of them, the winner included. The descriptions are in horse racing going explained.

Trip

Has the horse won or run well at the distance? Opportunity had never gone beyond 1m 4f and still finished third, which is the point: an untried trip isn't a no, it's a reason to want a bigger price.

Class

Is the horse going up or down in grade? Group-race form can read well in a handicap. Opportunity had been second in the Group 3 Glorious Stakes 21 days earlier and was the highest-rated horse in the field on 110, but Emit had Group placings in France too, so class alone won't separate them.

Mark

Compare today's official rating with the marks the horse has won or run close off. Daiquiri Bay had won off 96 and 100 and was now on 105, so the handicapper had caught up a bit. Ascending was back on 98, the mark he'd finished second off in the race a year earlier. More on spotting a horse the handicapper has underrated is in well handicapped horses.

Draw, pace and course

Draw. The first four in 2026 came from stalls 1, 3, 6 and 9, so it's tempting to say "low is best". But the six winners in the table above came from stalls 1 to 24. Six races can't show a draw bias, which is why my draw bias guide works by course and trip, and members get the full Draw Biases In Flat Racing special.

Pace. Ask who is going to lead and whether they'll go too quick. In the Ebor, Burdett Road (80/1) made the running until two furlongs out and faded to seventh. The four others ridden close to the pace early finished 12th (Emit), 19th, 21st and 22nd, and the first four all came from mid-division or further back. That's one race and pure hindsight, but you can sketch a pace map beforehand from recent running comments.

Course. Some horses simply run better at certain tracks. Daiquiri Bay and Ascending both had placed form over course and distance, but Klassleader had won at York in May and finished eighth. Course form is a plus, not a guarantee.

Trainer form and equipment changes

Check whether the yard is in form from its last couple of weeks of runners. Sporting Life's racecard flagged three Ebor runners as "hot trainer" (two for Willie Mullins, one for Simon and Ed Crisford). They finished 10th, 19th and 20th. That's three horses in one race, so it proves nothing, except that a flag on a racecard isn't a bet. My leading trainers guide has the season-long numbers.

Equipment changes are worth a look too. Two Ebor runners wore cheekpieces for the first time, Poetic Sound (40/1) and Furthur (33/1), and they finished 17th and 18th. Again, a sample of two. Whether it helps is covered in first-time blinkers, cheekpieces and wind ops.

Last, check the gap since the last run and any change of stable. Love Me Tender had been off 299 days and was making his debut for a new yard. That's two unknowns in one horse at 7/2.

Turning your contenders into prices

Every price is a probability. For fractional odds a/b, the implied chance is b ÷ (a + b), so 4/1 is 1 ÷ 5 = 20%.

Those seven alone add up to 97.3%. All 22 runners come to 143.5%, matching the on-course book percentage Sporting Life published. Everything over 100% is the bookmakers' margin.

To price your own shortlist, give each contender a percentage chance, make sure the whole field adds up to 100%, and turn each into fair odds with 100 ÷ your percentage.

You don't need to be sure a horse wins; you need to rate its chance higher than the price does. The maths is in value betting in horse racing, and staking is in my horse racing betting strategy rules.

How I pick a winner from a full card

Look, I was a Quantitative Researcher at a Tier 1 investment bank before all this, so I'm essentially a statistics/maths geek, and my version of this checklist is mostly prices and probabilities. It's worked well enough that most of your favourite bookies won't take a bet from me.

I don't try to find a winner in every race. My newsletter is a short list for your betting of the day. It usually has 'banker' bets, which is what I call the shorter-priced ones I rate as solid, plus some outsiders. The outsiders come from a new but proven system that picks out horses that have won or come close off a higher weight or mark than they're running off today. It's the same idea as the mark check above. When I've got stable information or a 'smart money' move, that goes in too. Some days nothing makes the list and there are no tips at all. That's a stinky day for the ponies.

I bet on every horse I tip and I stake in points: 1pt might be £5 on a £100 bank. I never chase, and I never go all in because I have a feeling about one. If I am honest, an each-way outsider that only places often makes us more than an odds-on shot winning. Losers are a thing, and understanding that is the key to being a winner longer term. If you're unsure about a bet, skip it. If you'd rather have my shortlist done for you, that's what my tips service is for.

Picking winners: questions people ask

Is there a way to pick a winner every time?

No. Anyone who says they can is a fraudster. Even the 7/2 favourite in the Ebor was given only a 22% chance by its own price. The aim is to bet at prices bigger than a horse's real chance and to come out in profit over months, not to win every bet.

How do I pick a horse to bet on if I'm new to racing?

Pick one race. Run each horse through going, trip, class and mark, cross off the ones with two or more question marks, and price what's left. Stake small, in points, and keep a record.

Should I just back the favourite?

Favourites win about a third of races, but the price already reflects that. In my count of 12,991 UK and Irish races (1 Oct 2025 to 30 Sep 2026), the Betfair SP favourite won 35.0%, and backing every one of them still lost 1.6p in the pound before commission.

More in this guide

• How to read horse racing form: a racecard explained
• Horse racing going explained
• Horse racing handicaps explained
• First-time blinkers, cheekpieces and wind ops
• Top horse racing trainers by the numbers
• Horse racing abbreviations A-Z