How to Analyse a Football Match Before Betting: 12-Point Checklist

Analysing a football match is not reading everything written about it. It is running the same twelve checks in the same order and stopping the moment one of them fails. Most fixtures die on the first four points, and that is the intended result: the checklist exists to remove matches, not to justify them. Whatever survives gets a probability written as a number, and that number gets compared with the price on the coupon after the bookmaker's mark-up has been taken out.

This page is the entry point for the analysis section of the blog. It sets the order of work; the pages linked along the way go deeper into single steps, and the terminology used here is defined in the betting glossary.

In short.

• Twelve checks, run in a fixed order. The first four are filters and remove roughly two thirds of a weekend card before any thinking starts.
• On the card worked below, 24 fixtures become 7 after the filters, 2 after the numbers and 1 after the price.
• You do not need a model to state a probability. Start from your league's own outcome split and adjust it in whole percentage points, with every adjustment written down.
• Compare that figure with the coupon only after removing the margin: at 1.85 / 3.60 / 4.20 the book totals 105.6%, and the home side's true market price is 51.2%, not 54.1%.
• The worked card ends with one ₦500 bet at +8.0% expected value and one fixture passed at −4.2%, and the pass is the more useful of the two decisions.

What twenty minutes of analysis should produce

The output of an analysis is not an opinion about who wins. It is three percentages that add up to 100, plus a note of what would change them. Anything that does not end in a number cannot be compared with a price, and anything that cannot be compared with a price cannot be a bet.

That framing also settles how much research is enough: enough to fill the ledger in the twelve points, and no more. The match you watched last night gets the same checks as the match you have never seen.

The twelve points and what each one removes

The right-hand column is the part people skip. A check that cannot rule anything out is not a check, it is a habit.

#	Point	What you look at	What it rules out
1	Competition and stakes	Is this a normal league fixture that both sides want to win?	Friendlies, dead rubbers, cup ties in a congested week, sides already promoted or relegated
2	Sample	Eight or more matches for each side at this venue in this season	Opening weeks, newly promoted clubs, divisions you have no data for
3	Team news	Are the line-ups confirmed or safely predictable an hour before kick-off?	Fixtures where a first-choice goalkeeper or striker is a question mark
4	Price sanity	The book percentage of the market you intend to bet	Padded markets and thin divisions where the mark-up eats any edge
5	Venue-split form	Last 6–10 home matches for the hosts, last 6–10 away matches for the visitors	Conclusions drawn from a blended season figure that describes neither venue
6	Goal profile	Goals and expected goals for and against, per venue	Sides whose results are running far ahead of their chances created
7	Style match-up	Tempo, possession, whether one side needs the other to come forward	Totals bets on fixtures where both sides are content to sit
8	Schedule and travel	Rest days, a third match in eight days, distance travelled	Fixtures where one squad is being rotated and you cannot tell how far
9	Motivation and context	Title race, relegation, European places, a cup replay four days later	Matches priced on the table that will be played on the calendar
10	Referee and discipline	Cards and fouls per match for this official, derby history	Card bets placed without knowing who is refereeing
11	Weather and pitch	Rain, wind, heat, pitch condition, artificial surface	High totals in conditions that flatten passing football
12	The price and its movement	Where the line opened, where it is now, what moved it	Bets taken into a move you cannot explain

The first four points do most of the work

Run the filters over a full weekend card and the arithmetic is blunt. Take 24 fixtures across the divisions you follow.

• Point 1 removes 4: two cup ties with rotation likely, one dead rubber between mid-table sides, one fixture involving a club already relegated. 20 left.
• Point 2 removes 6: two promoted clubs without a venue sample, and four fixtures in a division whose season is five matches old. 14 left.
• Point 3 removes 4: three with an unresolved injury to a first-choice forward, one where the manager has said the goalkeeper is a fitness test. 10 left.
• Point 4 removes 3: three-way markets priced with a book above 107%, which is a mark-up wider than any edge you are likely to find. 7 left.

Seven fixtures now get points five to twelve. Two of them will end with your figure meaningfully different from the market's, and one of those two will clear the betting threshold. That is a normal weekend: 24 matches, one bet.

Important. The filters are not a ranking. A fixture failing point 3 is not a weaker bet than one that passes — it is not a bet at all, however attractive the price looks. The moment you start weighing a failed filter against a good price, the checklist has stopped working.

Points five to nine: building the picture

These five points produce the raw material for your number. Each one is a question with a numeric answer, not an impression.

Venue-split form. Take the hosts' last eight home matches and the visitors' last eight away matches, and never average them together. A club taking 2.00 points a game at home and 0.75 away is two different teams, and the blended 1.38 describes neither of them.

Goal profile. Write four numbers per side: goals scored, goals conceded, expected goals for, expected goals against, all at the relevant venue. The gap between goals and expected goals is the interesting part — a side four goals ahead of its chances created has been finishing well, being lucky, or both, and the market has usually priced the results rather than the chances.

Style match-up. The practical question is who is forced to attack. A home favourite against a defensive visitor will have the ball for an hour and may still not create much.

Schedule. Count rest days for both squads and look one fixture ahead as well as one behind. A third match in eight days is where rotation and late fatigue live, and it matters more to the side with a thinner bench.

Motivation. Ask what each side needs from this match specifically. A team needing a point and a team needing three points is a different fixture from two teams needing three, and it shows up in the draw price more than anywhere else.

Points ten to twelve: context, weather and the price

The referee matters for one family of markets and barely at all for the rest. If you are pricing cards or fouls, the official's average is half of the estimate; if you are pricing the result, it is close to noise. Wind and heavy rain suppress passing accuracy and favour the side that wanted a scrappy match anyway, which makes weather a totals input rather than a result input.

Point twelve is the discipline point. Note where the price opened and where it is now, and only then decide. A market that has moved four or five points against your view since the morning is telling you that money — often better informed money — disagrees. That is not a reason to abandon a well-built number, but it is a reason to check which of the eleven earlier points you answered with stale information.

Putting a number on it without a model

You do not need a spreadsheet to state a probability. You need a baseline and a ledger of adjustments, each written in whole percentage points, each with a reason you would be willing to defend afterwards.

The baseline comes from the division itself. Suppose the league's completed season contained 380 matches: 167 home wins, 99 draws, 114 away wins. That is 43.9%, 26.1% and 30.0% — round it to 44 / 26 / 30 and use it for every fixture in that competition. League tables and season records for the competitions we cover sit on the tournaments pages.

Then adjust. Two rules keep the ledger honest: every adjustment must move points from one outcome to another so the three figures still total 100, and the sum of all adjustments is capped at 15 percentage points. The cap is what stops a checklist from becoming a way of talking yourself into a fixture.

Finding from the checklist	Home	Draw	Away
League baseline	44	26	30
Point 5: hosts 2.00 points per home match, visitors 0.75 away	+6	0	−6
Point 3: home first-choice centre-back suspended	−3	+1	+2
Point 8: visitors playing a third match in eight days	+3	+1	−4
Point 9: visitors safe in mid-table, hosts chasing a European place	+2	+2	−4
Your figure	52	30	18

Total movement is 14 points, inside the cap. The fair prices follow directly: 1 ÷ 0.52 = 1.92 on the hosts, 1 ÷ 0.30 = 3.33 on the draw, 1 ÷ 0.18 = 5.56 on the visitors. Write all three down before opening the coupon.

Taking the mark-up out of the coupon

A bookmaker's three prices describe more than 100% of the outcomes, and the excess is the margin. Comparing your 52% against a raw 54.1% is comparing a probability with a price, which is not the same unit.

Example. The coupon reads 1.85 / 3.60 / 4.20. Turn each into a percentage: 1 ÷ 1.85 = 54.05%, 1 ÷ 3.60 = 27.78%, 1 ÷ 4.20 = 23.81%. They sum to 105.64%, so the book carries 5.64% of mark-up. Divide each by that total: the market's own view is 51.17% / 26.29% / 22.54%.

Now the comparison is fair. Your 52 against the market's 51.2 on the home side is a gap of 0.8 points — nothing. Your 18 against the market's 22.5 on the visitors runs the wrong way, so that side is out. Your 30 against the market's 26.3 on the draw is a gap of 3.7 points, and that is the only line worth pricing. The conversion itself, including two-way markets, is covered in turning odds into probability.

Two fixtures, one bet and one pass

Seven fixtures survived the filters. Two produced a gap worth examining; here is what happened to both, with a ₦500 stake as the unit.

Fixture A — the bet. The ledger above gave 52 / 30 / 18. The draw is priced at 3.60 and your figure says 30%. Expected value is 0.30 × 3.60 − 1 = +8.0%, or ₦40 on a ₦500 stake. It clears both thresholds — at least 3 percentage points of gap against the cleaned market price, and at least 3% of expected value — so the bet goes on at ₦500 to return ₦1,800 if it lands.

Fixture B — the pass. A stronger home side, and the ledger comes out at 61 / 22 / 17. The coupon reads 1.57 / 4.20 / 5.50, which converts to 63.69% / 23.81% / 18.18% and a book of 105.69%; cleaned, the market says 60.27% / 22.53% / 17.20%. Your 61 against 60.3 is a gap of 0.7 points, and the expected value on the home price is 0.61 × 1.57 − 1 = −4.2%. Every check passed, the analysis was sound, and the answer is still no.

Fixture B is the point of the whole exercise. A checklist that only ever produces bets is a filter with the sieve removed. Over a season the passes are where the money is saved, which is why they belong in your log alongside the bets, with the price you refused written next to them.

Where to run it. The fixture list with line-ups, form and the full set of prices sits on the match pages, and the fixtures our analysts have already worked through are collected in today's predictions. Use them as raw material for the checklist, not as a substitute for it.

The mistakes that survive a checklist

• Opening the coupon first. A price of 1.60 makes any side look like a winner. Build the ledger before you look, or the ledger will land wherever the price already is.
• Reordering the points. The order is the method. Doing the interesting work on a fixture that fails point 2 wastes twenty minutes and creates a sunk cost you will want to justify with a bet.
• Adjusting past the cap. Fifteen points is already a strong claim about a single match. A ledger showing a 30-point swing usually means one finding has been counted three times in different words.
• Blending home and away. The single most common way to produce a number that is confidently wrong in both directions.
• Analysing until the answer changes. If two hours of reading turned a pass into a bet, what changed was your patience, not the fixture.

The rest of this section

This page is the map; the two below are the individual steps in more detail.

• Expected goals explained — what the metric measures at point 6, and the six situations where it will mislead you.
• Which football stats actually predict results — the stable metrics, the noisy ones, and why the window at point 5 is six to ten matches rather than three or thirty.

FAQ

How long should analysing one match actually take?

About twenty minutes once you know where the numbers are, and far less for most fixtures because the first four checks end the process early. If a match is taking an hour, you are usually re-reading the same inputs hoping they will say something different rather than adding new information.

Can I skip points I do not have data for?

You can skip a point only if it cannot affect the market you are betting. A missing referee record does not matter for a totals bet; a missing venue split does. If the gap sits in points one to four, the fixture is a pass, not a gamble on incomplete information.

What if my number and the cleaned market price are almost identical?

Then the market has priced it and there is no bet. That is the outcome for most fixtures and it is a sign the method is calibrated, not a failure. Set a minimum gap — three percentage points is a workable threshold — and treat anything below it as a pass.

Does this checklist work for markets other than the result?

Yes, with different weights. Totals lean on points 5 to 8, cards on points 9 to 11, and the result on points 5, 6 and 9. The first four filters and the final price check stay the same whatever you are betting, because they are about whether a fixture is analysable at all.

This article is for information only and is not an inducement to gamble. Betting involves the risk of losing money — never stake more than you can afford to lose. 18+. If gambling stops being entertainment, read our responsible gambling guide and seek help.