CASE
STUDIES
Examples that show step by step how to read our percentages. The aim is not to boast; it is to explain openly what a probability means — including the cases where we are wrong.
The cases below are illustrative scenarios that teach how to read the model's output — they are not presented as the real result of any particular match. For real, verifiable performance see the Track record and Past results pages (Turkish).
When our calculation thinks differently from the odds
The home side is mid-table, facing an away side in good form. The betting odds see the visitors as clear favourites.
Our calculation takes into account the chances the home side has created in recent matches and the chances the opponent concedes away from home, and gives the home side about 8 points more than the odds do.
This gap is not a sign of profit; it is two different views placed side by side. Which one is right in a single match proves nothing; what matters is how close our percentages come to reality over hundreds of matches.
▸ Lesson: Thinking differently from the odds does not mean being right. A single match can go either way.
When the 68% we gave doesn't happen
We gave a 68% probability of 3 or more goals in a match. It finished 1-0.
We expected both teams to score a lot; but in football low-probability outcomes happen too. A 68% probability means that in about 1 of 3 similar matches the opposite happens — if the percentages are right, that is exactly what is expected.
Reading this match on its own as “they were wrong” is misleading. The right question: in all the matches where we said 68%, did 3+ goals really happen in about 68% of them? The consistency chart in the Track record shows this.
▸ Lesson: A single match is not evidence. Transparency means showing the predictions that missed and sharing accuracy openly.
When two cautious teams meet
Two teams near the bottom of the table that both score little. The odds are balanced; our calculation expects a clearly low-scoring match.
When both teams are expected to score little, most of the probability gathers in scorelines such as 0-0, 1-0 and 0-1. “At most 2 goals” and “at least one team doesn't score” come from the same scoreline probabilities, so they agree with each other.
That is the benefit of deriving all 13 probabilities from the same calculation: it won't say “yes” on one and contradict itself on another.
▸ Lesson: Consistency is our strong side; but low-scoring matches are open to surprises too — a percentage is not certainty.
This is not betting advice.
Past performance is no guarantee of the future. Numerical probabilities offer a mathematical point of view; the decision is entirely the user's.