How We Log and Settle Every Prediction We Publish

The tipping industry has a trust problem, and it was earned honestly. Scroll any social feed on a Saturday and you will find accounts claiming strike rates that would embarrass professional syndicates — with no archive, no timestamps, and a remarkable tendency to delete November. Against that background, the only honest answer a predictions site can give is not "trust us" but "check us". This article lays out exactly how FikrBank logs and settles every published pick, so you can verify the record instead of taking our word for it.

Why a public record matters more than a promise

Anyone can be right about a football match. The coin-flip nature of single games means a stopped clock nails a prediction several times a week. What cannot be faked cheaply is a long, timestamped sequence: hundreds of picks, each published before kick-off, each settled after the final whistle, wins and losses alike. A record like that lets you answer the questions that actually matter — how often the reasoning holds, how the picks behave in bad months, whether the strike rate at real prices could survive the bookmaker's margin.

That is why we treat the archive as the product, not the picks themselves. A pick is an opinion; a settled, unedited history of picks is evidence.

The logging rules we follow

A track record is only worth auditing if the rules are fixed in advance. Ours are simple enough to fit in one list, and we hold ourselves to every line of it.

• Timestamped reasoning. Each pick goes up with its full written reasoning before kick-off. Nothing is edited after publication except typo fixes marked as such.
• One pick, one grade. Every prediction is settled as win, lose or push against the stated market and price taken at publication. No "close enough" grades.
• Nothing deleted. Losing weeks stay in the archive with the same prominence as winning ones. A record that hides its bad months is marketing, not tracking.
• Flat staking for grading. Picks are graded at one unit each, so the record measures judgement, not stake-sized bravado after a good run.
• Price honesty. We record the price available at publication time, not the best price that existed for twelve minutes overnight.

How a pick gets settled

Settlement happens after the final whistle and follows the market the pick was made in. The grading table below shows the only three outcomes a pick can have on this site — there is no fourth category for excuses.

Void or abandoned matches follow the bookmaker's own settlement rules, and the pick is marked void rather than quietly removed — the entry stays visible so the archive stays complete.

How to read our record without fooling yourself

Three cautions before you draw conclusions from anyone's archive, ours included. First, sample size: fifty picks prove almost nothing, because variance can fake skill for months. Second, price quality: a 60% strike rate at short odds can lose money while a 42% strike rate at bigger prices makes it — always read the record against the prices taken, not just the win column. Third, survivorship: a site that launched three times and kept only the successful archive is showing you its best coin flips. Ask how long the same unbroken record has been running.

We would rather you apply that scepticism to us than skip it. A prediction you have checked is the only kind worth following — and the day our record stops standing up to it, you should stop reading. That is the whole point of publishing it.

One more distinction the archive makes visible: strike rate and yield are different animals. A bettor hitting 60 per cent of short-priced picks can still bleed units, while a 42 per cent strike rate at bigger prices quietly profits. Because every settled pick carries its price taken and its flat one-unit stake, the archive lets you compute both numbers yourself — wins divided by picks for the strike rate, units returned minus units staked for the yield. When you evaluate any tipster, demand the second number over a long sample. Strike rate is for headlines; yield is for bankrolls.