Original data studies from the Tips.GG Data Team. We track predictions, odds and match data across football and esports, and publish what the numbers show — including when they contradict us.
Every study on this page is built on data we collect or calculate ourselves, with the method stated in the article. Where a study replaces an earlier edition, the newer one is listed first in its section.
Using our data
Journalists and researchers are welcome to use our findings, charts and datasets. We ask for one thing in return: attribution to Tips.GG with a link to the original study.
If you need the underlying data, a chart in a different format, a methodology clarification or a comment from our team, write to [email protected]. We answer data requests from media at no cost.
Prediction accuracy and upsets
We monitor published football predictions from hundreds of tipsters and prediction services, then measure what actually happened. These studies look at where the consensus failed and which competitions punish favourites most often.
| Study | What we found | Published |
|---|---|---|
| One 35.00-odds result defined the month | Across 6,120 matches tipped by 592 different tipsters, Nove Sady beat Sigma Olomouc at odds of 35.00 with no mainstream tipster backing them. Cup competitions produced upset rates roughly double those of the Premier League and Bundesliga. | October 2025 |
| The 20 most improbable results of November | Ranked from more than 73,000 tracked predictions, showing which underdog wins the market and the crowd both missed. | December 2025 |
| 78,000 predictions in one month: which top leagues are least predictable | Top-league upsets ranked by winning odds, and leagues ranked by average winning odds to show where favourites are least safe. | February 2026 |
Discipline and referees
Our Strictness Index measures disciplinary points per match rather than raw card totals, because raw totals reward teams that survive longer in a tournament rather than teams that break the rules more often.
| Study | What we found | Published |
|---|---|---|
| Ghana, not Argentina, play the dirtiest World Cup football game for game | The expanded 2026 tournament set a raw record of 266 yellow cards across 104 matches, yet recorded the lowest cards-per-match rate of the century at 2.97. Argentina hold the most total disciplinary points; Ghana lead on a per-match basis at 2.58. | July 2026 |
| The Premier League’s strictest referees | Attwell and Taylor topped the 2025/26 rankings. The study tracks yellow cards, second yellows, red cards and penalties awarded for every referee in the competition. | April 2026 |
Clubs, squads and players
Studies on where talent comes from, what it costs, how much of it clubs actually use, and what players physically produce on the pitch.
| Study | What we found | Published |
|---|---|---|
| Who sustains the most demanding workload per 90 minutes in the Champions League | We weighted the running output of 72 players by speed zone rather than counting total distance, then compared players against others in the same position. Intensity turns out to depend far more on role than on total kilometres covered. | September 2026 |
| The Mechanics of Football Corners: Why Bookmakers May Analyse No Better Than You — Yet Still Win — and Why Yours Matters | A single bet presents the bettor and the bookmaker with completely different tasks. The bettor is trying to predict the number of corners. The bookmaker is deciding how much it will pay if the bettor turns out to be right. For most bookmakers, that decision is shaped primarily by three factors: competitors’ lines, betting activity and margin. | August 2026 |
| Twenty academies built Brazil’s 2026 World Cup squad | The 26 players in the squad came through 20 different academies with a combined market value of €1.1 billion. Flamengo developed three of them; Santos produced one, and he accounts for 79 of the 150 international goals the squad has scored between them. | June 2026 |
| The eleven players whose European finals ended in defeat | Every footballer in history who reached four or more European Cup or Champions League finals and won fewer than half of them lost their final appearance. All eleven of them. Between them they form a complete starting eleven by position. | May 2026 |
| Nottingham Forest fill their ground better than any other Premier League club | The ten best-attended Premier League grounds of the 2025/26 season, measured as a share of capacity rather than raw attendance, so smaller stadiums are not penalised. | January 2026 |
| The most valuable under-21 players in the Premier League | A ranking of the league’s most expensive young talent, with the market value method used to build it set out in the article. | November 2025 |
Esports
Tips.GG began in esports, and we apply the same prediction-tracking methods to Counter-Strike, League of Legends and Dota 2 that we use in football.
| Study | What we found | Published |
|---|---|---|
| StarLadder Budapest Major 2025: the upsets and the consensus flops | Vitality took a third Major and ZywOo a record third MVP award. The study breaks down the tournament’s biggest upsets, the matches where tipster consensus was most wrong, team volatility and the final tipster leaderboard. | December 2025 |
How we work with data
Three sources sit behind these studies. The first is our own tracking of published predictions, which runs continuously and covers hundreds of tipsters and prediction services. The second is match and player data from our data providers. The third is public reference data, including market values and historical competition records.
Where a study uses a calculation of our own, we name it and explain it, so anyone can reproduce the result or disagree with it. Where a study has limitations, we say so in the article rather than leaving readers to find them.
Media enquiries
For data requests, methodology questions, interviews or comment from the Tips.GG Data Team, contact [email protected].