Tap a team to set them as the winner for a match, or tap "vs" for a draw. Tap again to remove a result. Knockouts will populate as the group qualifiers are decided.
Hover undecided knockout spots to view the distribution of possible qualifiers.
Groups
Knockouts 🏆
Last 16
Quarter-finals
Semi-finals
Final
Estimated Odds
The probabilities are generated by simulating the tournament a ~million times. The Betfair odds are refreshed every minute. Results will be added as matches complete.
The Delta column shows the difference between the estimated probability and the implied Betfair probability.
| Last 16 | Quarter-Final | Semi-Final | Final | Winner | Betfair | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Team | Odds | Prob. | Odds | Prob. | Odds | Prob. | Odds | Prob. | Odds | Prob. | Delta | Odds | Prob. |
|
Spain |
1.0 |
100.0% |
1.0 |
100.0% |
1.0 |
100.0% |
1.0 |
100.0% |
1.0 |
100.0% |
|
|
|
How it works
Suppose a tournament is taking place, and suppose a match in that tournament is completed. There’s a winner and a loser, or maybe it’s a draw.
There are two distinct impacts that the result has on the tournament as a whole:
- The quantitative impact: the team who lost might be eliminated immediately, or their chances of qualifying from the groups will be reduced. The team who won are through to the next round, or more likely to win the group. There might be a knock-on effect on the chances of teams who they are likely to play in the next round. A draw might qualify or eliminate both sides. Overall, this impact is fairly easy to estimate/quantify, because we know the structure of the tournament and we know roughly how good each team is beforehand.
- The qualitative impact: we might learn things about the relative ability of the teams involved. A star player might be injured or an unknown youngster might break through. We may learn information that causes us to update our priors.
This model provides estimates for the quantitative impact, as matches are played. The internal ratings for each team were calibrated based on the pre-tournament odds. Implicitly, this gives us a way to estimate the qualitative impacts, by comparing the estimated odds above to the real odds displayed on betting exchanges. In the past it has disagreed with the exchange on the likelihood of long-odds qualifiers to make it out of groups, although that seems to be less true this year. There are often mispricings on Betfair between two markets (e.g. to win vs to qualify from group), and this can be a useful way to spot these.
It would be incorrect to bet based only on these estimates, as you would be implicitly assigning zero value to qualitative impacts. However, having a way to break down the changes in odds between quantitative and qualitative aspects could be a useful input into a more holistic strategy.
Note that this model may struggle during the final stage of group games, as teams will often know exactly what they need to do to qualify, and this can skew the numbers.
There are a vast number of simplifications above, both in the model and in the explanation. Do not use this for anything serious.