How it works

Every published match carries four things:

• The pick — what we think happens, on a named market.
• The odds we recorded — the price available when we published. The same call at a worse price is a different bet, so we log it.
• A share of your bank — how strongly we hold it, as a percentage you can act on directly.
• The reasoning — written out, so you can disagree with it.

If we do not have a strong read on a match, we publish nothing. The point is not to fill a page.

Every pick carries a share of your bank. Not a "unit", not a rating out of five — a percentage, because that is the only form of the number that works the same for everybody.

The percentage is always taken from what the bank is worth right now, so stakes rise as you win and fall as you lose. That is the whole plan, and it is the one thing that keeps a losing run survivable.

Put your number in on your account page and every pick shows the amount in $ instead. You can change it, top it up, take money out or stop tracking whenever you like.

The share of settled picks that won. 62 correct out of 100 settled is 62% correct.

On its own it proves nothing. Back nothing but heavy favourites and you can be right 75% of the time and still lose money, because the winners pay so little. That is why the headline figure on this site is what happened to a bank, not how often we were right.

We take every settled pick in order, stake the percentage we published, and see where $100 ends up.

After a hundred picks that lands somewhere, and wherever it lands is the number at the top of the track record. It compounds, because a percentage of a bigger bank is a bigger stake.

What is normal: a real edge in this market looks like a bank drifting upward over months with ugly weeks in it. Anyone showing a straight line is showing you a drawing.

The deepest hole the bank sat in along the way, measured from its highest point.

A record can end the year up and still have put you through that. This is the number that decides whether you can actually follow a service, and it is the one most of them do not print. If a −30% dip would make you stop, plan for it now rather than discovering it in week six.

Nothing on this site is typed in by hand. Settling a pick recalculates everything at once:

1. We publish a pick with a market, a share of bank and the odds available then.
2. The match is played and the score goes in.
3. The pick is marked won, lost or void.
4. Every figure on the site is recomputed from every settled pick.

Because the figures are derived rather than stored, they cannot drift out of step with the results underneath them, and a pick cannot quietly disappear from the record.

Check the sample size first — a hundred settled picks tells you something, twenty tells you almost nothing. Then look at the odds column: a record built on 1.20 favourites is a different animal from one built on 2.50 outsiders. Then look at the worst dip, and ask yourself honestly whether you would have stayed.

Then the important bit. A track record describes what already happened. It is evidence that a process has worked so far, not a forecast. Variance here is large enough that good and bad months both happen without the underlying quality changing at all. Never treat past performance as a rate of return you are going to receive.