You open the race card and one rating says the horse is the fastest in the field. You feel good about it. Then your eye drifts to the next column and a second rating has the same horse near the bottom. Which one do you believe?
Many bettors quietly pick the rating that agrees with the bet they already wanted to make. There is a better way, and it does not involve deciding which rating is correct. It involves accepting that none of them is, and using them together. That is comprehensive analysis, and it is a skill you can build on a couple of horses a day.
What is comprehensive horse racing analysis? It means judging a horse from every angle at once rather than trusting a single number. Each rating, whether it measures speed, form, class or connections (the trainer and jockey), is an estimate that sees one slice of the horse. You combine five to ten of them, add what the market price is telling you, and read them together as one 360 degree picture before deciding the horse is worth backing.
By the end of this guide you will be able to take any runner, build a one-page 360 picture of it in about ten minutes, test that picture against the market price, and keep a record that shows whether your method is earning its place. No method can promise a profit. This one gives you a clear, checkable way to compare a horse with its price.
In this guide
- What is a horse racing rating, really?
- Why do two ratings disagree about the same horse?
- Which rating is the right one?
- How do you build the 360 picture, step by step?
- A worked example, from ratings to a verdict
- Why is the market your final rating?
- How do you know the picture is worth backing?
- What mistakes make a 360 picture misleading?
- Frequently asked questions
- Do this today: your checklist
What is a horse racing rating, really?
Open any newspaper, racing site or race card and you will see numbers. The most familiar are the Official Rating from the BHA, the Racing Post Rating and Top Speed. Everyone can see those, so the market has already priced them in. That is why ratings like ours exist: to measure the things the obvious numbers miss.
There is no perfect rating, and every rating is an estimate of a horse’s expected performance. That is true even for ratings worked out after the race. Take a speed figure. We know exactly how long the horse took, but raw time means little on its own.
A finishing time is shaped by the course, the distance, the going (the state of the ground), the weight carried, the size of the field and the way the race was run. The same effort gives very different times on a fast track in a strongly run race and on soft ground in a crawl. So a speed rating adjusts the raw time for all of that, to estimate how fast the horse really went in a way you can compare across races. It is genuinely useful, and it is still a smart, conditioned estimate of one aspect of the horse.
The four families of rating
Almost every rating you will meet belongs to one of four families. Knowing which family a number comes from tells you what it can and cannot see.
| Family | What it measures | What it cannot see |
|---|---|---|
| Speed | How fast the horse ran, adjusted for conditions | Whether it will handle today’s ground, trip or pace |
| Form | Recent results and how they were achieved | Whether those runs came against weak or strong rivals |
| Class | The level the horse has competed at and where it is rated | Fitness, mood or a trainer’s plans for the day |
| Connections | Trainer and jockey patterns and current form | The horse’s own ability |
Notice that each family has a blind spot that another family covers. That is the whole reason to use more than one.
Why do two ratings disagree about the same horse?
Once you accept that a rating is an estimate, disagreements stop being confusing and start being informative. Two ratings disagree because they are estimating different things from different positions.
Picture a horse in the paddock, and four people who may each look from only one side. The person at the front sees a strong, well-muscled animal and expects a big run. The person behind spots something not quite right in a hind leg. The person on one side watches it walk sweetly with a glossy coat. The person on the other side notices an old surgical scar. Four honest observers, different verdicts, and none of them is wrong. They simply cannot see what the others can.
Ratings work the same way. A speed rating only looks at speed. A recent-form rating weighs the last few runs. A collateral-form rating compares the horse with rivals it has already met. A class rating asks whether it has competed at this level. Each one reports honestly on its slice. When they disagree, they are handing you different pieces of the same horse.
Three reasons for disagreement, and what to do about each
- Different slices. Speed and class are measuring different things. Write down what each one is seeing and move on.
- Different time windows. One rating may use the last run, another the last five. A horse that has recently improved will look better on the short window and worse on the long one. Note which way the trend is pointing.
- Different conditions. A rating built mostly on good ground will say little about a soft-ground race. Check whether today’s going, distance and course match the conditions behind the number.
In every case the disagreement is a question to answer, not a problem to hide.
Which rating is the right one?
None of them, and that is good news. It stops you hunting for the one true number and points you towards building the full picture.
Think of a good doctor. They do not glance at one symptom and call it. They take a temperature, listen to the chest, check the history and run the bloods, then decide. One reading can mislead, and together they tell the story. A racehorse deserves the same treatment. A horse might be the fastest in the race on raw speed, but only on good ground, and today is good to soft. It might have beaten half the field before, which lifts its chance, but it might also make mistakes at its jumps. None of those facts cancels the others. Stacked together, they show whether the bet is worth making.
That is why combining several ratings gives you a fuller assessment than leaning on any single one. One rating is one angle. Six ratings chosen to measure different things come close to the whole horse. A fuller picture helps you judge how likely a horse is to run well. Whether the price is worth taking is a separate question, which the last sections cover.
How do you build the 360 picture, step by step?
The routine is deliberately simple, because a method you will use beats a clever one you will not. As you look at each rating, write one plain sentence about what it tells you. One line per rating.
With six ratings you write six sentences. “Fastest in the race on adjusted time, but only on good ground.” “Form is improving, won two of its last three.” “Dropping out of handicap company into a weaker race.” “Trainer is in red-hot form this fortnight.” “Has beaten three of today’s rivals already.” “Stays the trip well but can be slow away.” Read them back together and you are no longer looking at six numbers that argue. You are looking at one rounded portrait, with strengths and question marks side by side.
Step 1: Choose your races and your runners
Do not try to analyse a whole card. Pick one or two races where you already have a reason to look, then shortlist the two or three runners that interest you most. If you need a quick way to narrow a field first, the guide to finding the most competitive runners in a race shows how to do it in about five minutes, and the four-step method for finding winners explains where this fits in a full routine.
Step 2: Cover the four categories
Cover the four big categories so you see the horse from every side:
- at least one speed rating
- at least one form rating (and reading a horse’s form properly is a skill in itself)
- at least one class measure, such as the Official Rating or a class-movement read
- at least one connections angle covering the trainer and jockey
Five to ten ratings across those four categories is plenty. Fewer leaves blind spots, and many more adds noise that says the same thing twice.
Step 3: Write one sentence per rating
Keep each sentence plain and specific. Say what the rating shows, and add the condition that qualifies it. “Fast” is not enough. “Fast on good ground over this trip” is a sentence you can use.
Step 4: Mark each sentence green, amber or red
Green means it supports the horse today. Amber means it is neutral or depends on a condition you cannot yet confirm. Red means it is a genuine concern. This takes seconds and turns a paragraph of text into something you can scan.
Step 5: Look for independent agreement
When two independent ratings that measure different things point at the same horse, pay attention. It is like two camera angles lining up on the same picture. Race Advisor’s PR rating (which weighs recent form in the context of class and distance) and VDW rating (which compares class and ability) are built to be an independent check on each other. When both rank a horse top of the race, the agreement is a stronger sign that it is likely to run well than either rating alone.
Ratings that share the same inputs do not count as independent. Two speed figures built from the same times are one opinion said twice.
Step 6: Write the verdict in one line
Finish with a single sentence that starts “I like this horse because” or “I am passing because”. If you cannot complete the sentence in plain words, you do not yet understand the horse well enough to back it. The rule that stops this turning into paralysis is covered in the piece on why less information often wins.
A worked example, from ratings to a verdict
The numbers and horse below are made up to show the method. They are illustrative, not real results, and they are not a tip.
| Angle | One-line sentence | Colour |
|---|---|---|
| Speed | Second fastest on adjusted time, and both fast runs came on soft ground | Green |
| Form | Beaten a short head last time after a clear run of improvement | Green |
| Class | Stepping up two grades, never tested at this level | Amber |
| Connections | Trainer quiet this fortnight, jockey booking is new | Amber |
| Collateral form | Has finished ahead of two of today’s rivals before | Green |
| Course and trip | First run over this distance | Red |
Read down the column. Three greens, two ambers and one red. The picture says this is a capable horse with two open questions: the step up in class and the new trip. A fair verdict might be: “Interesting, but only if the price pays me for the trip risk.”
That last clause matters. The picture has told you how the horse looks. It has not yet told you whether the odds are fair. That is the job of the next two chapters.
Why is the market your final rating?
There is one more angle, and it is the one most form students forget to write a sentence for. The live market price is itself a rating: the crowd’s composite estimate, with everyone’s speed, form, class and connections reads boiled down into the odds.
So make the odds your last sentence, every time. If your picture is strong and the market agrees by making the horse favourite, your read is sound, though there is probably little extra value because everyone can see it. If your picture is strong and the price is generous, treat the gap as a prompt to investigate: has the market spotted a problem you missed, or is it undervaluing something you have checked? If your picture is full of warning signs but the horse sits near the head of the market, that is a horse to take on rather than back. The market is the benchmark that shows whether your analysis matches what the market price is telling you or has found something it missed.
A simple decision rule for the market line
- Picture strong, price short: your read is sound, but expect little extra value. Usually pass or keep the stake small.
- Picture strong, price generous: investigate first. Find out what the market might know before you act.
- Picture weak, price short: a horse to take on, or simply leave alone.
- Picture weak, price long: nothing to do. Move to the next race.
If you cannot explain why the market and your picture disagree, that is a reason to wait, not a reason to bet.
How do you know the picture is worth backing?
A well-rounded picture of a horse is not the same thing as a bet with value. If the picture is so obvious that the market has priced the horse to the floor, a lovely analysis has little left to earn.
The test that settles it is actual versus expected, the A/E ratio. You take a group of selections built a particular way and compare how often they actually win with how often their prices said they should, at Betfair SP (the exchange’s closing price). A figure of 1.00 means the group is market-fair. Above 1.00 means the method is beating the prices. Below 1.00 tells you the method needs a fresh look.
Race Advisor’s database can answer these questions. In an earlier analysis of more than 1.5 million UK runners with a Betfair starting price, between July 2013 and May 2026, the top horse on our PR rating won 32.3% of its races, easing to 4.6% for those ranked fifth or worse. There is a clear gradient. The A/E at every rank sat between 0.99 and 1.01, so the market already reflects what a single rating can see.
Agreement is where it gets more interesting. When PR and VDW independently ranked the same horse top of the race, across roughly 69,000 of those runners, the strike rate climbed to 35.6%, comfortably clear of either rating on its own. The A/E stayed at 0.99, because the crowd can see the same agreement you can. The fuller picture finds more winners, and the price moves with it. The process is still worth owning, because it helps you understand why you like a horse and lets you check the price against your own reasons, which is the foundation for any edge you later find. So build the picture from every side, write your sentences, fold in the market, and use actual versus expected to see whether you have an edge or a good story.
How to run your own A/E check
You do not need a database to start. You need a record and some patience.
- Define your selection rule in writing before you look at results, for example “green on at least four of six angles”.
- Log every selection that meets the rule, with the price you could actually get.
- Wait for a sample large enough to mean something. A handful of bets proves nothing either way.
- Compare how often the group won with how often the prices implied. Work out the implied chance from each price, add them up, and set that against the actual number of winners.
- Change nothing mid-sample. If you adjust the rule halfway, you have started a new test.
An illustrative, not real, example: if your rule produced 100 selections and the prices implied about 25 winners in total, but 25 actually won, the A/E is 1.00 and the method is market-fair. Above 25 would be encouraging, below would be a prompt to rethink. Treat any single result with caution, because luck moves small samples a long way.
What mistakes make a 360 picture misleading?
The method is simple, and so are the ways it goes wrong. Build these checks into the routine.
- Counting the same information twice. Two ratings built from the same inputs are one opinion. Choose ratings that measure different things.
- Adding ratings until you like the answer. Decide your set of five to ten before you look at the runner, then stick to it.
- Skipping the market line. A beautiful picture with no price check is only half an analysis.
- Forgetting today’s conditions. Check going, distance, course and likely pace against what each rating was built on.
- Over-reading a small sample. A great week or a poor week tells you very little about the method.
- Never writing it down. A picture held in your head drifts towards what you wanted to see. Words on a page keep you honest.
If you find yourself drowning in columns, go back to the idea that why less information often wins. Five to ten well-chosen ratings beat thirty that overlap.
Frequently asked questions
How many ratings should I use?
Five to ten, spread across speed, form, class and connections. Fewer leaves blind spots and more mostly repeats itself.
What if the ratings all agree?
Agreement is a good sign for how likely the horse is to run well, but the market can see it too. Check the price before deciding the bet has value.
What if the ratings are split?
Write down why. A split usually points to a condition, such as ground, trip or class, that you can check. If you cannot resolve it, pass and move on.
Do I need to analyse every race?
No. Two or three horses a day is enough to build the skill. As it gets faster, the guide to how to analyse a race quickly gives you a short routine for busy days.
Where do I start with form and the racecard?
Start with how to read a racecard, then move on to reading a horse’s form. Both feed straight into the picture you build here.
Does a complete picture guarantee a profit?
No. It helps you understand a horse and compare it with its price. Profit depends on whether your selections beat the odds over a large sample, which is what your own record will show.
Do this today: your checklist
- Open today’s race cards and choose one or two races.
- Shortlist two or three runners. Do not try to cover the whole card.
- List five to ten ratings across speed, form, class and connections.
- Write one plain sentence per rating, and mark each green, amber or red.
- Check whether any two independent ratings agree on the same horse.
- Add the market price as your final sentence and apply the decision rule.
- Finish with one line: “I like this horse because” or “I am passing because”.
- Record every selection in the free Bet Tracking tool, with the price you took.
- Review your record after a meaningful sample, not after a handful of bets.
Your own record, built over time, shows you how your fuller picture performs against the market. That is the real test, and the only one that matters.
