What Research Says About Football Tipsters, and How to Test One
Out of a total of 17,936 matches in the five major European leagues, betting consistently on the home team resulted in the correct prediction being made 44.0% of the time. Backing the favourites consistently increased this rate to 54.1%, providing readers of football tipster services with two free benchmarks against which to measure track records.
The figures are taken from a Jabula Bets analysis covering the Premier League, Bundesliga, Serie A, La Liga and Ligue 1 from the 2016/17 to 2025/26 seasons. The analysis is based on results and average pre-match odds from football-data.co.uk. In each analysed season, the favourite had a higher success rate than betting on the home team. However, these are prediction success rates only. To assess betting profits, the odds associated with these tips must also be taken into account.
What the Newspaper Tipsters Achieved
In a study published in the International Journal of Forecasting in 2000, David Forrest and Robert Simmons examined three tipsters from English newspapers over a period of 1,694 league matches. The tipsters' success rates were 42.56%, 41.09% and 42.86%, compared to 36.4% for a random prediction strategy. If one had consistently backed home wins in this sample, the success rate would have been 47.5 per cent.
The researchers found that the tipsters did not make full use of the information readily available about the teams' strengths. This information included league positions and current form. Instead, the tipsters relied on their intuition to make clear predictions of a home win, draw or away win. The 2000 study by Forrest and Simmons also found that the three tipsters' collective predictions were more accurate than those of any individual tipster.
The study also found that one of the three participants had apparently used additional, unspecified information relevant to the results successfully. Tipsters can combine publicly available information with contextual knowledge, such as details of injuries to key players.
How Forecasts Compare with Market Prices
In a study published in the Journal of Forecasting in 2009, Spann and Skiera compared the accuracy of prediction markets, betting odds and tipsters over three seasons of the German Bundesliga. They found that prediction markets and betting odds were equally accurate and both performed significantly better than the tipsters.
In their 2019 study, ‘Football Match Prediction with Tree Based Model Classification’, Yoel F. Alfredo and Sani M. Isa tested various models using data from ten Premier League seasons. 'Random Forest' achieved an accuracy rate of 68.55%, outperforming 'Extreme Gradient Boosting' with 67.89% and 'C5.0' with 64.87%. However, despite improvements achieved through fine-tuning, the researchers concluded that their decision-tree-based approach was inadequate for predicting football results.
Price comparisons offer another way of evaluating predictions. Football-Data’s 2016 analysis, which examined 48 tennis tipsters, found that their recommended odds were generally better than the closing odds. However, when the stake remained constant, their bets yielded a return of −0.94 per cent. The analysis focused on tennis, and the collective information advantage was too small to offset the bookmaker’s margin.
If you consistently bet on odds that subsequently decrease, this may indicate information that has not yet reached the market. Football-Data recommends comparing the recommended odds with the closing odds, i.e. the odds at the start of an event.
The Benchmarks Season by Season
According to an analysis of data from football-data.co.uk, the success rate of the market favourite ranged between 52.7% and 57.7%. Meanwhile, the benchmark for the home team ranged from 39.9% to 48.6%. Therefore, the time period chosen by a tipster can influence the comparison, even before the odds and number of tips are taken into account.
Two free benchmarks for any tipster: Europe's top five leagues
| Season | Matches | Always the home team | Always the market favourite |
|---|---|---|---|
| 2016/17 | 1,826 | 48.6% | 57.7% |
| 2017/18 | 1,826 | 45.3% | 55.0% |
| 2018/19 | 1,826 | 44.7% | 53.0% |
| 2019/20 | 1,725 | 44.1% | 53.4% |
| 2020/21 | 1,826 | 39.9% | 53.0% |
| 2021/22 | 1,825 | 42.7% | 52.7% |
| 2022/23 | 1,826 | 45.7% | 54.1% |
| 2023/24 | 1,752 | 43.1% | 55.0% |
| 2024/25 | 1,752 | 42.0% | 53.6% |
| 2025/26 | 1,752 | 44.0% | 53.4% |
| All ten seasons | 17,936 | 44.0% | 54.1% |
Source: The above analysis is based on the results and pre-match average odds from football-data.co.uk.
The favourite benchmark measures the accuracy of following the market-favoured outcome. A tipster who focuses on higher-odds results may win less frequently, but achieve better returns. When making comparisons, it is better to take these odds into account than to assess each service based on its percentage of correct tips.
How to Test a Tipster Fairly
Football-Data’s analysis explains that it can take hundreds or even thousands of tips to distinguish profits from chance. A winning streak alone provides limited guidance, especially since many tipsters have multiple opportunities to achieve one.
1. Start by looking at the total number of bets placed. Take into account any losing streaks, as well as the overall record prior to a successful winning streak. Bets with higher odds are subject to greater fluctuations in winnings and require a larger sample size to determine whether there is a moderate edge.
2. Make a note of the odds for each selection. Alongside the headline win percentage, consider the average odds and the return on investment.
3. Every bet must be published before the start of the match, complete with an unalterable timestamp. An independent record can document losses that might otherwise disappear from an editable spreadsheet. Screenshots of winning tickets do not provide a complete record.
4. Consider returns based on fixed stakes, where the stake remains the same for all tips. Football-Data uses fixed stakes in its tennis analysis. Keep an eye on stakes and returns, and factor in any subscription fees when assessing the cost of using a paid service.
Comparisons with closing prices can provide further insight into the quality of the selection. If prices consistently outperform the final market price, this may indicate good forecasting ability, even before profit figures can be used as a reliable indicator. The recommended price, the actual price achieved, and the closing price should be clearly distinguishable from one another.
Readers already betting with Jabula can use their account history to check previous bets and whether they have been resulted and paid out. The tipster’s complete advance-published record remains a separate part of the assessment.
Keep the Records Behind the Result
The accuracy figures provided are based on different samples and forecasting methods.
A comprehensive betting log includes every selection, the time of publication, the odds, and the stake. Specific football analyses relating to tactics or referees' decisions can also be compared with the predictions. These records allow readers to evaluate the service during both losing streaks and successful periods.
18+ only. Gambling risks losing money. National counselling line: 0800 006 008.