How we grade confidence on every tip

Every tip on GameTipster carries a confidence grade from 1 to 5. After the odds, it is the most important number on the card, and it is also the easiest one to misread. This is exactly what a grade means, how we arrive at it, and how it turns into a stake you can check against our results.

What the five grades mean

A grade measures one thing: the gap between the probability we give an outcome and the probability implied by the price. We call that gap the edge, and we measure it in percentage points. The bigger the gap, the higher the grade.

• 5 · Maximum · edge 10%+

  Rare. The market has missed something or badly mispriced a match. If we were handing these out every day, the problem would be our numbers, not the bookmakers'.

• 4 · Strong · edge 6-10%

  A clear disagreement with the price, backed by a reason we can write down in a sentence or two.

• 3 · Solid · edge 3-6%

  The working grade. A real but modest edge. Most of our singles should sit here or one step below.

• 2 · Lean · edge 1-3%

  We think the price is slightly too big, but the margin for error is thin. Small stakes only.

• 1 · Speculative · edge under 1%

  Marginal. Published because we like the angle and want it on the record, staked at a level that cannot hurt you.

A grade is a claim about the price, not about the team. We are never confident in a club. We are confident that a price is too big.

That distinction matters. A grade 5 on an outsider can still lose far more often than it wins, and a grade 1 on a heavy favourite can still win most of the time. The grade tells you how wrong we think the market is, not how likely the bet is to land.

How a grade is decided

Four rules, applied to every tip, whatever the sport.

1. We price it ourselves before we look at the market

We write down our own probability first. Form, injuries, schedule, motivation, style match-ups: all of it goes into a number before we open a single odds screen. Look at the price first and it anchors you. After thirty years of doing this, I still catch myself nudging my number towards the market when I have seen it. So we don't look.

2. Edge is our probability minus the implied probability

Implied probability is simply 1 divided by the decimal odds. Say we make a home win 55% and the best available price is 2.10. The implied probability is 1 ÷ 2.10 = 47.6%. The edge is 55 − 47.6 = 7.4 points. That is a grade 4.

Second example: we make an away win 40% and the price is 2.62. The implied probability is 1 ÷ 2.62 = 38.2%, so the edge is 1.8 points. That is a grade 2.

We use the price you can actually get, which includes the bookmaker's margin. So the edge has to beat the margin, not just a theoretical fair price. If you want the background on implied probability and margin, read our guide to decimal and fractional odds.

3. Team news can move a grade one step

Line-ups and late fitness news arrive after we have done our pricing. When they do, we re-check the number. News can move a grade by one step, up or down, and no further. If a piece of news changes the picture by more than that, the honest conclusion is usually that our original view was fragile, and we would rather not publish the tip at all than bend the grade to fit.

4. Grades are locked at publish

Once a tip is live, the grade, the odds and the stake are fixed. If the price shortens after we publish, we don't upgrade it to look clever. If it drifts, we don't quietly downgrade it. Every tip is settled against the grade, odds and stake we printed.

Turning a grade into a stake

We stake in units, and 1 unit is 1% of your betting bank. The grade decides the number of units.

On a £500 bank, 1 unit is £5. A grade 4 is 2 units, so £10. A grade 1 is 0.25 units, so £1.25. A grade 5, the biggest stake we ever recommend on a single, is £15.

Some people find that timid. Here is why it isn't. The Kelly criterion, a standard formula for sizing a bet when you have an edge, says that at a true 55% chance and odds of 2.10 you should stake about 14% of your bank. We stake 2%. The reason is that 55% is our estimate, not a fact. If we are a few points too optimistic, a 14% stake does real damage, while a 2% stake barely registers. Small stakes are how you survive your own mistakes.

How to check us

Every tip is settled publicly on the Results page, with the stake and odds exactly as published. Wins, losses, voids, all of it. Nothing is removed after the event.

Tracking started on 15 September 2026, when the site relaunched. That means there is no long record to point to yet, and we are not going to pretend otherwise. Results appear on /results as tips settle, and the record will build in plain sight.

When there is enough of it, here is how to judge the grading:

• Look at results by grade. Over a few hundred bets, grades 4 and 5 should return more per unit staked than grades 1 and 2. If they don't, the grading isn't doing its job.
• Ignore short runs. Twenty or thirty bets prove nothing either way. A good grading system can have a bad month, and a bad one can have a lucky month.
• Compare our price with the price at the off. If the prices we publish are regularly bigger than the prices available when the event starts, that is a decent sign we are finding value before the market does.

Accumulators work differently. They are staked at a flat 0.25 units regardless of the legs, and you can see them on the accumulators page. The grades above apply to singles.