How to Keep Horse Racing Betting Records

Keeping accurate horse racing betting records is one of the simplest ways to understand how you actually bet.

Memory is unreliable.

A memorable 12/1 winner can make a month feel successful. A frustrating losing run can make a sound approach feel disastrous. Neither impression tells you much without the numbers.

A proper betting record tells you:

• how much you have staked
• how much you have won or lost
• your return on investment
• your strike rate
• the prices you normally take
• whether you beat the closing market
• which race types suit your approach
• where your losses occur
• how severe your drawdowns become
• whether your perceived betting edge appears in the results

More importantly, good records separate what you thought before the race from what happened afterwards.

That distinction matters.

A good bet can lose.

A bad bet can win.

If you record only winners, losers and account balances, you miss much of the information required to judge your betting properly.

This guide explains how to build a horse racing betting record, which information to track, the calculations that matter and how to analyse your results without being misled by short-term variance.

Why Keep Horse Racing Betting Records?

The obvious reason is to know whether you are winning or losing.

But that is only the beginning.

Imagine you have placed 300 bets and made a £250 profit.

That sounds encouraging.

Now suppose you discover:

• £700 profit came from one 50/1 winner
• your other 299 bets lost £450
• your bets below 10/1 consistently lost money
• you regularly accepted shorter odds than were available elsewhere
• your selections generally drifted before the off

The original £250 figure suddenly looks very different.

Now consider another bettor who lost £150 over 300 bets.

Their records show:

• prices taken consistently beat the closing market
• their maximum drawdown remained within expectations
• their strongest results came from one clearly defined race type
• the underlying selection process remained consistent
• the sample included an unusually poor run of results

That bettor should not automatically conclude that their strategy works, but their records provide much more useful evidence than the headline loss.

This is closely connected to betting variance in horse racing. Short-term results can differ substantially from the underlying quality of your decisions.

Without records, it is difficult to tell the difference.

What Should a Horse Racing Betting Record Include?

Your spreadsheet can be as simple or sophisticated as you need.

At minimum, record:

Field	What to Record
Date	Date of the race
Race	Course and race time
Selection	Horse backed
Bet type	Win, each-way, place etc.
Stake	Total amount risked
Odds taken	Price when bet placed
Result	Win, lose, placed etc.
Return	Total amount returned
Profit/loss	Return minus stake
Bookmaker	Where the bet was placed

That creates a basic financial record.

For serious analysis, add more.

The Betting Record I Recommend

A more useful horse racing betting spreadsheet could contain:

Date

Racecourse

Race time

Race type

Distance

Class

Going

Field size

Horse

Bet type

Stake

Bookmaker

Odds taken

Your estimated probability

Your fair odds

Minimum acceptable odds

Starting Price or chosen closing price

Result

Return

Profit/loss

Running betting bank

Reason for bet

Strategy/category

Post-race notes

You do not necessarily need every column from day one.

The objective is to collect information you will genuinely analyse.

Recording 50 variables that you never use simply creates unnecessary work.

Record the Bet Before the Race

This is one of the most important rules.

Do not reconstruct your reasoning afterwards.

Before the race, record something such as:

Selection: Horse A

Fair odds: 4/1

Available odds: 6/1

Minimum acceptable price: 5/1

Reason: likely uncontested lead and suitable ground

Then leave it alone.

Afterwards, you can evaluate what actually happened.

Why does this matter?

Because hindsight changes how we remember decisions.

A horse wins easily and suddenly all the positive factors seem obvious.

A horse finishes last and you start remembering concerns that barely influenced you before the race.

Recording your reasoning beforehand creates an honest record of the decision.

If you need a more structured framework for reaching that decision, start with How to Analyse a Horse Race Like a Professional.

Record Your Fair Price

One of the most useful additions to a betting record is your own price.

Suppose you assess a horse as having a:

20% chance of winning.

Fair decimal odds are:

1 ÷ 0.20 = 5.00

Equivalent fractional odds:

4/1.

You then find:

6/1.

That gives you a clear reason for considering the bet.

You believe the horse should be:

4/1

but the market allows you to back it at:

6/1.

Recording only:

Horse A, £10 at 6/1

loses that information.

Recording:

Fair price 4/1, bet price 6/1

allows you to assess your pricing process later.

Our How to Price a Horse Race guide explains how to create your own probabilities and fair odds.

Record the Price You Actually Took

Never replace the price you actually received with a better price you noticed later.

If you backed a horse at:

5/1

record:

5/1.

Do not record 6/1 because another bookmaker offered it.

Your betting record needs to represent what actually happened.

This also exposes an important part of betting performance:

price execution.

You might be good at finding horses but poor at obtaining the best available odds.

Suppose over 500 bets you repeatedly take:

4/1 instead of 9/2

6/1 instead of 7/1

10/1 instead of 12/1.

Those differences can accumulate.

This is why comparing bookmakers matters once you have decided a bet represents value. Our horse racing bookmaker comparison guide explains the factors worth comparing beyond the headline odds.

For readers deciding which regulated operators best suit their racing betting, our main Best Horse Racing Betting Sites rankings compare the leading options.

Record Which Bookmaker You Used

The bookmaker column can reveal more than you might expect.

Over time you can analyse:

• average price obtained
• frequency of Best Odds Guaranteed adjustments
• extra-place usage
• withdrawal experience
• which accounts you use most
• where you most frequently find your preferred prices

Do not interpret this as meaning the bookmaker causing the most winners is somehow lucky.

The race result is independent of which account you used.

Instead, assess whether your choice of bookmaker affects:

price + terms + execution.

If you regularly bet on mobile, it can also be useful to compare how easily you can check prices and place bets through different products. British Racecourses maintains separate rankings of the Best Horse Racing Betting Apps for this purpose.

Record Best Odds Guaranteed Correctly

Best Odds Guaranteed can complicate record-keeping.

Suppose you take:

6/1.

The horse wins at:

8/1 SP.

Your eligible bookmaker settles the bet at 8/1 under BOG.

What should you record?

Keep separate columns for:

Price taken: 6/1

Settlement price: 8/1

This preserves the original betting decision while ensuring your financial return remains accurate.

Do not change the price-taken column to 8/1.

You did not make the original decision at 8/1.

Our Best Odds Guaranteed guide explains how the concession works and why eligibility and terms matter.

Record a Closing Price

The next useful column is:

Closing price.

You need a consistent definition.

Depending on your method, you might use:

• Starting Price
• Betfair Starting Price
• a chosen exchange price close to the off
• another consistently recorded market benchmark

Consistency matters more than switching retrospectively to whichever benchmark makes your results look strongest.

Why record it?

Because it allows you to measure Closing Line Value in Horse Racing.

Suppose:

Your fair price: 4/1

Price taken: 6/1

Closing price: 7/2

Result: Lost

Financially:

you lost.

But your price moved from:

6/1 → 7/2.

That provides information about the market’s later assessment.

It does not prove your original probability was correct.

It does provide another way to evaluate your betting beyond individual race results.

Keep Results and Decision Quality Separate

Consider two bets.

Bet A

Your fair price:

3/1

Available price:

5/1

Result:

Lost

Bet B

Your fair price:

5/1

Available price:

2/1

Result:

Won

Which was the better decision?

According to your own pre-race probabilities:

Bet A.

Bet B produced the better financial outcome, but you accepted a price substantially shorter than your own estimate of fair value.

This distinction is fundamental to Expected Value in Horse Racing Betting.

Your records should therefore let you analyse both:

financial results

and:

decision quality.

How to Calculate Horse Racing Betting Profit and Loss

The basic calculation is:

Profit/Loss = Return – Stake

Suppose you place:

£10 at 5/1.

If it wins:

Total return = £60

Profit = £50.

If it loses:

Return = £0

Profit/loss = -£10.

For a series of bets:

Total Profit/Loss = Total Returns – Total Stakes

Example:

Total stakes:

£2,000

Total returns:

£2,180

Profit:

£180.

Simple profit matters, but it does not tell you how efficiently the money was staked.

For that, calculate ROI.

How to Calculate Betting ROI

Return on investment is one of the most useful basic betting metrics.

Use:

ROI = Profit ÷ Total Stakes × 100

Suppose:

Total stakes = £2,000

Profit = £180

Calculation:

180 ÷ 2,000 × 100

ROI:

+9%.

If instead you lost £180:

ROI:

-9%.

ROI allows better comparison between periods and strategies using different amounts of turnover.

A £500 profit from £2,000 staked is very different from a £500 profit from £50,000 staked.

Turnover Matters

Always record total stakes.

Consider two bettors.

Bettor A

Staked:

£1,000

Profit:

£100

ROI:

+10%.

Bettor B

Staked:

£10,000

Profit:

£100

ROI:

+1%.

Both made £100.

Their performance was not identical.

Turnover provides context.

It also helps you understand how much money is being repeatedly exposed to risk.

Calculate Your Strike Rate

Strike rate measures the percentage of bets that win.

Formula:

Winners ÷ Bets × 100

Suppose you place:

200 win bets

and:

40 win.

Strike rate:

40 ÷ 200 × 100

=

20%.

Strike rate becomes much more informative when compared with your average odds.

A bettor backing short-priced favourites should normally win more frequently than someone targeting horses at 10/1 and above.

A low strike rate is not automatically bad.

A high strike rate is not automatically good.

Price determines whether the winners compensate adequately for the losers.

Average Odds Give Strike Rate Context

Imagine:

Bettor A strike rate:

40%.

Bettor B strike rate:

15%.

Which is better?

You cannot know.

If Bettor A generally backs horses at:

6/4

and Bettor B generally backs horses at:

8/1

their required strike rates are completely different.

Record average odds and, better still, analyse results by price band.

For example:

Odds Range	Bets	Strike Rate	Profit/Loss	ROI
Odds-on
Evens to 3/1
Above 3/1 to 8/1
Above 8/1 to 20/1
Above 20/1

This can expose weaknesses that overall results hide.

You may discover that your analysis is strong around:

3/1 to 8/1

but that you consistently overestimate outsiders.

That would be particularly relevant when considering Favourite-Longshot Bias in Horse Racing.

Track Your Betting Bank

Your betting record should also include a running bank.

Example:

Starting bank:

£1,000

Bet 1 loses £10:

£990

Bet 2 wins £40:

£1,030

Bet 3 loses £10:

£1,020.

This lets you see the path your results take rather than only the final figure.

That path matters.

Two strategies can both finish the year:

+£500

while experiencing completely different levels of volatility.

One may have suffered a:

£150 maximum drawdown.

Another:

£1,500.

Those are very different risk profiles.

Read our complete guide to Horse Racing Bankroll Management before deciding how much money to allocate to betting or how stakes should relate to the size of a bank.

Calculate Maximum Drawdown

Maximum drawdown measures the largest decline from a previous betting-bank peak to a subsequent low.

Suppose your bank develops like this:

£1,000

£1,100

£1,250

£1,180

£1,050

£900

£980

The peak was:

£1,250.

The subsequent low was:

£900.

Maximum drawdown:

£350.

Percentage drawdown from the peak:

350 ÷ 1,250 × 100

=

28%.

That figure tells you something that final profit alone cannot.

A profitable strategy can still expose you to uncomfortable or financially damaging losing periods.

Record Your Longest Losing Run

A betting record also lets you calculate consecutive losses.

Suppose your results are:

L – L – L – W – L – L – L – L – L – W.

Your longest losing sequence is:

five.

This matters because losing runs are inevitable when betting on uncertain events.

The typical length depends heavily on your strike rate.

Someone backing horses with a 50% genuine winning probability has a very different losing-run profile from someone betting at average odds of 12/1.

Do not increase stakes simply because you believe a winner is “due”.

Each race needs to be assessed on its own merits.

Categorise Every Bet

One of the biggest improvements you can make is adding a:

Strategy

or:

Reason

column.

Possible categories include:

• value bet
• pace angle
• draw angle
• sectional angle
• form bet
• handicap angle
• favourite
• outsider
• each-way
• model selection
• market-based bet

Use categories that match how you genuinely bet.

Then analyse them separately.

You might discover:

Strategy	Bets	Profit/Loss	ROI	CLV
Pace	250
Draw	180
Sectionals	140
General form	300
Model	500

That is far more useful than knowing only your overall profit.

Analyse Results by Race Type

Horse racing is not one homogeneous betting market.

Your results may vary substantially between:

• Flat
• National Hunt
• All-Weather
• handicaps
• non-handicaps
• maidens
• novices
• sellers
• claimers
• two-year-old races
• staying races
• sprints

Suppose your overall record shows:

+4% ROI.

But deeper analysis reveals:

Flat handicaps:

+11%.

National Hunt:

-9%.

That raises an important question.

Do you genuinely possess the same level of understanding in both areas?

Your betting record can identify where your apparent edge is concentrated.

Analyse by Racecourse

Course configuration can materially affect races.

Different tracks present different:

• turns
• gradients
• straight lengths
• draw effects
• pace characteristics
• surfaces

Record the racecourse and analyse results by venue once you have enough data.

Do not overreact to tiny samples.

A record of:

three wins from four bets at Chester

does not prove you have discovered a permanent Chester edge.

But hundreds of consistently categorised bets may reveal areas worth investigating.

If draw forms part of your analysis, our Horse Racing Draw resource explains why stall position can matter differently according to course, distance and field.

Analyse Your Pace Bets Separately

If you regularly use pace to find selections, tag those bets.

Then ask:

• Did the expected pace develop?
• Did the horse obtain its anticipated position?
• Were your pace predictions accurate?
• Did front-runners perform as expected?
• Did your strongest pace opinions outperform weaker ones?

Our Horse Racing Pace Maps Explained guide shows how to estimate likely race shape before the off.

This is a good example of why post-race notes matter.

A pace-based selection losing because the expected lone leader was unexpectedly challenged is different from repeatedly predicting the race shape incorrectly.

Analyse Sectional-Based Bets

The same principle applies if you use sectional times.

Tag:

Sectional

as a strategy category.

Then track whether horses identified through sectional analysis:

• produce positive ROI
• beat the closing price
• perform better under particular conditions
• are repeatedly overbet next time

The Horse Racing Sectional Times Explained guide explains why finishing position alone can hide important information about how a horse actually ran.

Analyse Results by Going

Add a going column if ground forms an important part of your analysis.

Possible categories include:

• Firm
• Good to Firm
• Good
• Good to Soft
• Soft
• Heavy

For All-Weather racing, record the surface separately.

Again, the objective is not to mine the data until you find a profitable historical pattern.

It is to test sensible hypotheses.

If you believe you analyse testing ground particularly well, your betting record gives you a way to examine whether the evidence supports that belief.

Analyse Results by Distance

You might also split your bets into broad distance categories:

Sprints

Middle distances

Staying races

Your strengths may not be uniform.

Someone with a strong understanding of early pace and draw might perform differently in five-furlong handicaps from three-mile chases.

Your records should help identify those differences.

Analyse by Field Size

Field size affects:

• race shape
• traffic problems
• draw dynamics
• each-way terms
• pace pressure
• probability distribution

Possible groups might be:

2-7 runners

8-11

12-15

16+.

Choose categories relevant to your strategy rather than creating arbitrary groups purely because one produces the best historical ROI.

Analyse by Odds Range

This is one of the most valuable breakdowns.

Suppose your record shows:

Odds	Bets	ROI
Under 2/1	120	-6%
2/1-5/1	260	+8%
Above 5/1-10/1	180	+12%
Above 10/1	140	-18%

That gives you something to investigate.

Perhaps:

• you overestimate outsiders
• market margins affect certain price ranges differently
• one large winner is distorting a band
• your fair-price calculations become less accurate at low probabilities

Do not immediately stop betting one category because of a small negative sample.

Investigate first.

Analyse by Bookmaker

Your bookmaker records can answer questions such as:

• Where do I obtain my best prices most often?
• How frequently do BOG settlements improve my returns?
• Which accounts do I barely use?
• Do I repeatedly accept a weaker price because one app is more convenient?

This is where record-keeping connects directly with bookmaker selection.

British Racecourses’ Best Horse Racing Betting Sites comparison assesses operators across racing coverage, odds, promotions, mobile experience, payments and support rather than treating every betting account as interchangeable.

The objective is not to accumulate accounts unnecessarily.

It is to understand whether the accounts you use serve a genuine purpose.

Analyse Your Mobile Betting Habits

Mobile betting can also create behavioural patterns worth tracking.

You might discover that bets placed:

at home after proper analysis

perform differently from:

impulsive bets placed shortly before races.

If that happens, the lesson is not necessarily that mobile betting itself is bad.

The issue may be the decision process.

You can add a simple column:

Planned / Unplanned

or:

Pre-analysed / Impulsive

That can reveal whether convenience is affecting selection standards.

Readers who primarily bet through phones can compare product usability through our Horse Racing Betting Apps rankings, but app quality should never replace disciplined race and price analysis.

Record Why You Did Not Bet

This is an advanced but extremely useful idea.

Suppose your system identifies a horse, but the available odds are too short.

Record:

No bet – price below threshold.

Why?

Because later the horse might win.

Without a record, you may think:

“I knew I should have backed it.”

But if your fair price was:

4/1

and the market offered:

2/1

your original decision may have been completely consistent with your rules.

Recording rejected bets helps stop results from rewriting your decision-making process.

It can also help when backtesting a horse racing betting system, because you need to distinguish genuine qualifying bets from selections that never met the required price.

Keep a Separate Record for System Bets

If you are testing a defined betting system or building your own horse racing betting model, give it a clear identifier.

For example:

Model V1

Model V1.1

Model V2.

Do not mix results from changing systems and then evaluate them as though every bet came from the same rules.

Suppose:

Model V1 = 300 bets

You change three variables.

Model V2 = 200 bets.

Your complete 500-bet record remains useful financially.

But for model evaluation, those are two separate samples.

Version control prevents you from accidentally comparing different strategies as one.

Historical Backtesting and Live Betting Records Are Different

A backtest asks:

What would have happened if these rules had been applied historically?

A live betting record asks:

What actually happened when I applied them in real time?

The second can expose problems that historical testing misses.

These include:

• unavailable advertised prices
• late non-runners
• execution delays
• changing ground
• practical bookmaker restrictions
• human errors
• missed bets
• emotional decisions
• rule deviations

A strategy showing attractive historical performance still needs prospective testing.

Historical data is evidence.

It is not a guarantee of future results.

Do Not Delete Losing Bets

This sounds obvious, but selective record-keeping destroys the value of the dataset.

Do not exclude a bet because:

• you placed it in a rush
• the horse was unlucky
• the jockey made a mistake
• you regret placing it
• you broke your normal rules
• you forgot to record it initially

Record it.

If the bet broke your rules, tag it:

Rule violation.

Then analyse rule-compliant and non-compliant bets separately.

That itself can reveal something valuable.

Record Mistakes Separately

Not every loss is a mistake.

And not every winner is a good decision.

Create a column such as:

Process error? Yes/No

Examples of genuine process errors might include:

• wrong stake
• wrong horse
• failed to compare available odds
• ignored minimum price
• misread race conditions
• used outdated going
• broke strategy rules
• chased a previous loss

That creates an important distinction:

bet lost

versus:

process failed.

The two are not the same.

Track Rule Compliance

If you use a defined strategy, add:

Rules followed?

Yes/No.

After a few hundred bets, compare:

Rule-compliant bets

with:

Non-compliant bets.

If your disciplined selections outperform your impulsive deviations, the records provide strong behavioural feedback.

If both groups lose badly, you may have a more fundamental strategy problem.

Beware of One Huge Winner

Suppose you place:

500 bets.

Overall profit:

£1,000.

One 100/1 winner generated:

£2,000 profit.

Without that horse:

you would have lost £1,000.

Does that mean the 100/1 winner should simply be removed?

No.

If it was a genuine qualifying bet, removing it would distort the record.

But you should analyse:

results with and without extreme outcomes.

This shows how dependent your headline profitability is on rare events.

That is especially important for high-odds strategies.

Do Not Cherry-Pick Your Analysis

Once you have a large spreadsheet, it becomes easy to find apparently profitable combinations.

You might discover:

Bay horses

aged four

drawn 3-7

at left-handed tracks

between 7f and 1m

on good-to-soft ground

produced:

+27% ROI.

Does that combination have a convincing racing or market rationale?

Or did you keep filtering until something looked profitable?

The more categories you search, the greater the chance of finding random historical patterns.

This is one of the major dangers discussed in our betting-system backtesting work.

Start with a hypothesis.

Then test it.

Do not start with the result and invent the explanation afterwards.

Sample Size Matters

Imagine you analyse:

10 bets

and make:

+40% ROI.

That tells you very little.

Now imagine:

1,000 bets

placed under consistent rules.

The larger sample provides much more evidence.

There is no universal number of bets that proves a strategy works.

Required sample size depends on:

• average odds
• strike rate
• size of expected edge
• variability of results
• consistency of the method

Higher-odds approaches generally produce greater volatility because winners occur less frequently.

Be particularly cautious about strong conclusions from small samples.

Review Records at Fixed Intervals

Do not completely reassess your strategy after every race.

That encourages result-driven decision-making.

Instead, establish review periods.

For example:

every 100 bets

every 250 bets

monthly

quarterly.

The correct frequency depends on how often you bet.

At each review, calculate:

• bets
• stakes
• returns
• profit/loss
• ROI
• strike rate
• average odds
• maximum drawdown
• longest losing run
• average CLV
• performance by strategy
• performance by odds
• performance by race type

Then investigate meaningful changes.

A 100-Bet Review Example

Suppose your latest 100 bets show:

Total bets: 100

Total stakes: £1,000

Returns: £1,080

Profit: £80

ROI: +8%

Winners: 21

Strike rate: 21%

Average odds taken: 5.2

Longest losing run: 12

Maximum drawdown: £160

Average closing price: shorter than price taken

That is much more informative than:

“I made £80.”

Now segment the sample.

Perhaps:

Pace bets:

+15% ROI.

General form bets:

-8%.

Perhaps bets below:

5/1

performed strongly.

Outsiders:

poorly.

Perhaps your prices consistently beat the closing market.

Those findings create questions for further investigation.

They do not automatically justify changing the strategy.

A Useful Monthly Betting Review

At the end of each month, ask:

Financial Performance

What did I stake?

What did I return?

What was my profit/loss?

What was my ROI?

Selection Performance

What was my strike rate?

What were my average odds?

Which strategy categories performed best?

Price Performance

How often did I beat the closing price?

Did I regularly obtain the best available bookmaker price?

Did BOG affect returns materially?

Risk

What was my maximum drawdown?

What was my longest losing run?

Were stakes consistent with my betting bank?

Discipline

How many bets broke my rules?

How many were impulsive?

Did I chase losses?

Analysis

Which assumptions were supported?

Which require more data?

Which appear wrong?

That creates a structured review rather than an emotional reaction to the month’s final balance.

Should You Use Excel or Google Sheets?

For most bettors, a spreadsheet is enough.

Excel and Google Sheets can both handle:

• bet records
• formulas
• filters
• pivot tables
• charts
• running bank calculations
• ROI calculations
• category analysis

You do not need sophisticated betting software simply to keep accurate records.

A spreadsheet has another advantage:

you decide exactly what is recorded.

Start simple.

Add complexity only when it answers a useful question.

Example Horse Racing Betting Spreadsheet

A practical starting structure might look like this:

Date	Horse	Course	Strategy	Stake	Fair Odds	Odds Taken	Closing Odds	Result	P/L
1 Aug	Horse A	Course A	Pace	£10	4/1	6/1	7/2	Lost	-£10
2 Aug	Horse B	Course B	Value	£10	2/1	5/2	7/4	Won	+£25
3 Aug	Horse C	Course C	Draw	£10	8/1	10/1	12/1	Lost	-£10
4 Aug	Horse D	Course D	Sectional	£10	5/1	7/1	9/2	Lost	-£10

Four bets tell you almost nothing about long-term performance.

But maintain the same record across:

100

500

1,000

or more bets

and you begin building a useful dataset.

Advanced Metrics Worth Tracking

Once your basic record is reliable, you can consider additional measures.

Average CLV

Are the prices you take generally bigger or smaller than the closing market?

Expected Value

What theoretical edge did you estimate when placing each bet?

Expected Versus Actual Winners

If your bets collectively implied 100 expected winners, how many actually won?

Calibration

When you assign horses approximately 20% probability, do they win around 20% of the time over a sufficiently large sample?

Maximum Drawdown

How severe was the largest decline from peak bank?

Profit Factor

How much gross profit did winning bets generate relative to gross losses?

These measures can add depth, but only if your underlying data is accurate.

Complicated calculations cannot rescue poor records.

Probability Calibration Is Particularly Valuable

Suppose you create your own prices.

You can group selections by estimated probability.

For example:

Your Probability	Bets	Expected Win Rate	Actual Win Rate
5-9.9%
10-14.9%
15-19.9%
20-24.9%
25-29.9%
30-39.9%
40%+

If horses you consistently price at around:

20%

win only:

11%

across a substantial sample, something deserves investigation.

Your probabilities may be too confident.

If they win approximately:

20%

that is more encouraging.

Calibration does not guarantee profitable betting because the available market price still matters.

But it tells you whether your probability estimates resemble realised outcomes.

Profit Is Not the Only Feedback

This is worth repeating.

Suppose after 200 bets you have:

-5% ROI.

Should you abandon your method?

Not necessarily.

Should you assume variance explains everything?

Also no.

Investigate:

• sample size
• average odds
• CLV
• calibration
• fair-price accuracy
• rule compliance
• changes in market conditions
• whether the original analytical rationale still makes sense

Likewise, a:

+20% ROI

does not automatically prove you have discovered a durable edge.

Ask whether:

• one winner dominates results
• the sample is small
• your prices were realistically obtainable
• the edge persists out of sample
• the closing market supports or contradicts your assessments

Good record-keeping encourages scepticism in both directions.

Records Can Improve Your Race Analysis

Your spreadsheet is not only an accounting tool.

It becomes feedback on your racing judgement.

Suppose you believe you are particularly good at identifying horses that ran better than their finishing position suggests.

Tag those bets.

Over time, ask:

Do these horses outperform market expectations next time?

Do I obtain value prices?

Does the market already recognise the same hidden performances?

This connects record-keeping directly to race review rather than treating betting results as a separate subject.

The better your categories reflect your actual analytical process, the more useful the feedback becomes.

Records Can Expose Behavioural Mistakes

Some of the most valuable findings may have nothing to do with handicapping.

You might discover:

Saturday bets:

-12% ROI

Weekday bets:

+5%.

Why?

Perhaps Saturdays contain more competitive races.

Perhaps you place more recreational bets.

Perhaps you increase stakes.

Perhaps nothing meaningful is happening and the sample is noisy.

Your records allow you to investigate.

Other behavioural categories could include:

• planned bet
• late bet
• festival bet
• tip followed
• own analysis
• chasing previous loss
• bet outside normal strategy

The objective is not to find excuses.

It is to make your actual behaviour measurable.

Do Not Use Records to Chase Losses

A detailed spreadsheet can become counterproductive if you obsess over recovering a previous peak.

Suppose your betting bank falls from:

£1,500

to:

£1,250.

Do not think:

“I need £250 back.”

The next race has no obligation to restore your previous balance.

Increasing stakes simply to recover losses adds risk.

Use your record to control behaviour, not justify chasing.

Should You Increase Stakes When Your Record Is Profitable?

Not automatically.

A profitable historical record does not guarantee future profit.

Before increasing stakes, consider:

• sample size
• drawdown
• average odds
• estimated edge
• betting-bank size
• probability uncertainty
• whether results are repeatable

If you use probability-based staking, our Kelly Criterion for Horse Racing Betting guide explains how estimated edge can be connected to stake size.

Full Kelly can be aggressive when probability estimates are uncertain, so staking decisions still require caution.

What Does a Good Betting Record Look Like?

A good record is:

Complete

Every relevant bet is included.

Accurate

Actual stakes, prices and returns are recorded.

Contemporaneous

Important information is entered at the time of the decision rather than reconstructed later.

Consistent

Categories and closing-price benchmarks do not constantly change.

Detailed enough

It contains the information required for useful analysis.

Simple enough

You can realistically maintain it.

Honest

Losing bets, mistakes and rule violations remain in the dataset.

The best spreadsheet is not necessarily the one with the most columns.

It is the one you maintain correctly.

Horse Racing Betting Record Checklist

Before placing a bet:

☐ Record the selection

☐ Record your reasoning

☐ Record your fair odds where applicable

☐ Record your minimum acceptable price

☐ Compare available prices

When placing the bet:

☐ Record the actual bookmaker

☐ Record the actual stake

☐ Record the actual price taken

☐ Record the bet type

After the race:

☐ Record the result

☐ Record the return

☐ Calculate profit/loss

☐ Update the betting bank

☐ Add the chosen closing-price benchmark

☐ Record relevant post-race observations

During reviews:

☐ Calculate ROI

☐ Calculate strike rate

☐ Review average odds

☐ Measure drawdown

☐ Check losing runs

☐ Analyse CLV

☐ Segment by strategy

☐ Segment by odds

☐ Segment by race type

☐ Check rule compliance

☐ Look for extreme winners distorting results

☐ Avoid drawing conclusions from tiny samples

Frequently Asked Questions

What is a horse racing betting record?

A horse racing betting record is a structured log of your bets. It normally includes selections, stakes, odds, results, returns and profit/loss. More advanced records can also include fair odds, closing prices, race conditions, strategy categories and post-race analysis.

Why should I keep betting records?

Records show whether you are actually winning or losing and help identify where your results come from. They can also reveal weaknesses in pricing, strategy selection, staking and betting behaviour.

What should I record for every horse racing bet?

At minimum, record the date, race, horse, bet type, stake, odds, bookmaker, result, return and profit/loss. For deeper analysis, also record your fair odds, closing price, strategy, race conditions and reason for betting.

Should I record losing bets?

Yes. Excluding losing bets makes the record unreliable. If a bet broke your normal rules, record it and tag it as a rule violation rather than deleting it.

Should I record bets that I nearly placed?

You can maintain a separate record of qualifying selections that were rejected because the price was too short. This can be particularly useful when evaluating a value-based strategy.

What is ROI in horse racing betting?

ROI measures profit relative to total stakes.

The formula is:

Profit ÷ Total Stakes × 100.

If you stake £5,000 and make £250 profit, your ROI is +5%.

What is a good betting ROI?

There is no universal figure. ROI must be considered alongside sample size, average odds, variance and whether the results are repeatable. A spectacular ROI across 20 bets tells you much less than a robust record across a large relevant sample.

How many bets do I need before analysing my results?

You can review records immediately, but strong conclusions generally require much larger samples. The number needed depends on your average odds, strike rate, strategy and expected edge.

Should I track Starting Price?

It can be useful, particularly if you want a consistent market benchmark. Some bettors instead use Betfair Starting Price or another closing-market measure. Choose a relevant benchmark and apply it consistently.

What is closing line value?

Closing line value compares the price you took with a later market price near the start of the event. Consistently securing bigger prices than the closing market can provide useful information about price selection, although it does not guarantee profit.

Should I keep separate records for Flat and jumps racing?

You can keep one master spreadsheet and categorise bets by code. This makes it easy to analyse Flat, National Hunt and All-Weather results separately while preserving a complete financial record.

Should I record which bookmaker I use?

Yes. This lets you analyse where you obtain prices and terms and keeps the financial record accurate when using multiple accounts.

Can I use a betting app’s history instead of a spreadsheet?

Account history can help, but it usually lacks your pre-race reasoning, fair odds, strategy categories and cross-bookmaker information. A separate master record gives you greater control over the analysis.

Should I record free bets?

Yes, but distinguish promotional bets from ordinary cash stakes so they do not distort analysis. Record the actual economics of the promotion rather than pretending a free bet involved the same cash risk as a normal wager.

Can betting records prove I have an edge?

No single record or metric proves a permanent edge. A large consistent dataset containing profitability, price performance, calibration and out-of-sample results can provide evidence, but betting markets change and future results remain uncertain.

What is the biggest betting record mistake?

Incomplete or retrospective record-keeping. A spreadsheet becomes much less useful when bettors forget losing bets, alter categories after seeing results or reconstruct their original reasoning with hindsight.

Summary

Horse racing betting records turn impressions into evidence.

At the simplest level, record:

selection + stake + odds + result + return + profit/loss.

For more useful analysis, add:

fair odds + closing price + bookmaker + race type + strategy + betting bank + reasoning.

Then review the record systematically.

Measure:

profit/loss

ROI

strike rate

average odds

maximum drawdown

longest losing run

closing-line performance.

Break the results down by:

strategy

odds

race type

racecourse

distance

going

and other categories that have a genuine analytical rationale.

Do not search endlessly for profitable historical combinations.

Do not delete inconvenient losses.

Do not judge a strategy by one big winner.

Do not assume a losing month proves your approach has failed.

And do not assume a winning month proves it works.

The purpose of a betting record is not to make past performance look impressive.

It is to create an honest dataset that helps answer a much more important question:

Are my betting decisions actually as good as I think they are?

That record then feeds naturally into Expected Value, Closing Line Value, Betting Variance, Bankroll Management, Betting Models and proper system backtesting.

The sequence becomes:

Analyse → Price → Compare → Bet selectively → Record → Review → Test → Improve.

That process cannot remove uncertainty from horse racing.

It can make your decisions more measurable, consistent and accountable.

18+. Gambling involves financial risk. Only bet with money you can afford to lose.