Every call, settled in public

A football prediction record is the complete list of every call a publisher has made, each shown with the result that settled it, so that prediction accuracy can be counted instead of claimed. Ours keeps every prediction (soccer picks, for readers in South Africa) with its publication time, market, stated probability and reference price.

Think of it as a bank statement next to a payslip, where the payslip carries the good news and the statement carries every line. Each call starts on the football predictions hub and lands here once the match is over. The figures sit in the tables above and update after every full time whistle, so what follows explains how to read them.

Is This Tips Record Real?

Yes, and the proof sits in a column most readers skip. Every row carries a published at timestamp, shown in WAT, EAT, SAST and GMT, and that time always falls before kick-off. After publication the selection, the probability and the reference price are locked, so nothing is tidied up once the result is known. Readers from Lagos to Kampala have seen enough cropped screenshot wins on social media to be sceptical.

What Does The Headline Prediction Accuracy Figure Include?

The figure at the top is the share of settled predictions that were correct, shown with the number of settled predictions and the date range they cover, because a percentage without a sample size tells you very little. The prediction results are then split by market (1X2 result, over/under goals and both teams to score, the GG/NG of Kenyan and Ugandan slips) and by competition. The Premier League and the Champions League each carry their own line for supporters following African players in European club football. AFCON and World Cup qualifiers have separate records for those following national teams, while Nigeria's NPFL and the South African Premiership each have a local league line, with the smaller local samples labelled as such. That split suits supporters who follow Mohamed Salah or Victor Osimhen from one club to the next.

Do You Show Losing Predictions?

Yes, every losing prediction stays in the table (or log, if you prefer), in the same row format and the same kick-off order as the winners. Nothing is filtered by default, and a loss is marked as plainly as a win. Keeping every miss in the table lets you count the record for yourself. If a bad run looks uncomfortable, then it is doing its job.

How Are Predictions Settled?

A prediction is settled on the score at the end of 90 minutes plus stoppage time. Extra time and penalties do not count, so a cup tie that is level at full time settles as a draw on the 1X2 market, whoever lifts the trophy afterwards. A postponed or abandoned match is marked void, stays visible and is left out of the accuracy count. The same applies to a result awarded by a ruling and not played out on the pitch.

How Is Accuracy Scored, And Against What?

Accuracy is scored with two numbers, and they answer different questions. The hit rate is the share of calls that were right, while the Brier score, introduced by Glenn W. Brier in 1950, is the mean squared gap between the stated probability and what happened, with lower being better. Both are set beside two naive forecasters: a one in three guess and a rule of always picking the home team. That second rule is no pushover, since home sides won 47% of matches in the 2018-19 Premier League, an unusually high figure. Every pick is counted first and only then grouped into confidence bands, so you can see whether the calls given a higher probability really land more often.

What Do The Month By Month Results Tell You?

The monthly and season tables show the record as it happened, bad runs included, and they are the best protection against reading too much into a hot spell. Small samples are noisy, as Leicester City showed by winning the 2015-16 Premier League after being rated 5000 to 1 outsiders, and one outcome cannot judge a probability. Perhaps the most useful habit is to ignore any percentage built on a handful of predictions. A hit rate is a forecasting score and never a return figure, because it says nothing about the price at which each call was made.

How Can You Check The Record Yourself?

The methodology linked beneath the record explains how each probability is produced and how the reference price is chosen. Checking a row is quick: note the fixture and the date, find the official full time score from the league or federation that ran the match, and compare it with the result column. Do that for several rows from a competition you follow, losses included, before deciding how much weight to give the rest.