We show every pick.
Wins and losses.
Every prediction is timestamped before kick-off and locked. No deletions, no edits, no retroactive changes. If a pick was wrong, it stays on the record permanently. That is what accountability looks like.
Most tipsters delete their losses. We keep every single one.
30-day rolling results
How our picks land
Every settled prediction in the last 30 days, scored independently. No cherry-picking the window.
15,719 picks scored
Exact predicted outcome landed. Counted as a full win.
Right direction, marginally off the threshold.
Wrong call. Counted fully against us, no exceptions.
The numbers don't lie.
We just make sure of it.
130,724
Predictions tracked
Every single pick logged before kick-off. Not one entered after the fact.
12,253
Settled matches
Fully verified results. Every win and every loss counted equally.
1,100+
Leagues covered
Premier League to Liga MX to J1 League. If it is on the schedule, we have a prediction.
70%+
Over / Under accuracy
Our strongest market, verified across tens of thousands of settled picks.
Accuracy by market
Five markets. Every result public. We do not hide the ones we find harder.
Over / Under
Our flagship market. We analyse expected goals, pace, defensive shape, and head-to-head scoring patterns before every pick. Backed by tens of thousands of settled picks across every major league in Europe, South America, and Asia. This is not a good month — this is the all-time average.
70%
all-time accuracy
Double Chance
Home/Draw or Away/Draw combined. A high-confidence market where our form models consistently find value.
Accuracy65%Both Teams to Score
We model attacking threat and defensive vulnerability from both sides. Consistent performer across all major leagues.
Accuracy62%Match Result (1X2)
The hardest market to call. We use Elo ratings, Dixon-Coles, and AI reasoning to find edges in deeper leagues.
Accuracy60%Draw No Bet
Backing a winner with your stake returned on a draw. Reduces risk while keeping strong upside on clear favourites.
Accuracy61%
Over / Under — month by month
Every month shown. Including the slower ones. Consistency over cherry-picked snapshots.
Apr 2001
70.6%
49,418 picks
Mar 2001
71.9%
2,481 picks
Jun 2001
68.6%
691 picks
May 2001
69.6%
33,422 picks
How we count
The rules are simple and they never change. Every number on this page is calculated the same way, every time.
Locked before kick-off
Every prediction is timestamped the moment it is generated — before the match starts. There is no window to enter a pick after the result is known. The database timestamp is immutable. If you had the app open at 14:58 on a 15:00 kick-off, you saw the same pick we are counting now.
Binary scoring, no partial credit
A pick is correct only when the exact outcome lands. Over 2.5 goals requires three or more goals. BTTS requires both sides to score. 1X2 requires the precise result. No rounding up, no close-enough. A result of 2-1 does not make an Under 2.5 pick a partial win. Win or lose, nothing in between.
All markets shown, including the hard ones
Every market we predict is reported publicly — Over/Under, 1X2, Both Teams to Score, Double Chance, and Draw No Bet. We do not surface only the markets where we look good. If a market has a lower hit rate, it is on this page. That is what a public record means.
Same rules, every market, every day
The accuracy formula does not change based on how a market is performing. We do not recalculate when results look bad, apply different windows per market, or exclude leagues from the count. The same logic runs for every pick, in every league, from day one.
Global coverage, not cherry-picked leagues
We generate predictions across 1,100+ leagues worldwide — not just the Premier League or La Liga where data is richest. Our models work harder in lower-division football where less public information is available. Every league, every tier, counted the same way.
Results settle automatically, not manually
Match results are pulled directly from the API after the final whistle and scored against our predictions without human involvement. There is no manual review step where a correct result could be entered and a miss discarded. The pipeline runs the same way for every match, every day.