Skrill Final Score: what are your chances of winning? We worked out 2,660 matches
Skrill Final Score is a free game in which you predict the exact scores of six Premier League matches every round. We worked out the chances on 2,660 matches from seven seasons and on 18,433 matches from seven leagues. The best possible pick, the most likely score, lands in 12.9% of matches on average. That gives almost 0.8 hits per round on average and no hits at all in more than 4 rounds out of 10. A perfect 6 out of 6 comes up once in about 293,000 rounds, or once in 7,700 seasons, and 5 out of 6 once in 180 seasons. The game is still worth playing, because it costs nothing. A 38-round season brings about €34 in Knect points on average, and more than 70% of that comes from rounds with one or two hits. "Fan" picks, a 2-1 or 2-0 win for the favourite, earn roughly a third less.
Key takeaways
- 1
1-1 is the most likely score in 60% of matches, because until the favourite has about a 60% chance of winning none of its winning scores beats the draw.
- 2
Always picking 1-1 is worth €29.4 a season, while picking the favourite to win earns only €22.4–24.5.
- 3
The first €10 voucher usually takes 13 rounds to collect, and Knect points only expire at the end of the following calendar year.
- 4
At a bookmaker the same pick with a 12% chance at odds of 6.50 loses 22% of the stake on average to the margin.
Where the data comes from and how we worked it out
The rules of the game, the terms and Knect points are covered in our separate guide to Skrill Final Score. Here we focus on the numbers.
- Matches. 2,660 Premier League matches from the 2019/20–2025/26 seasons for which we have closing odds on the match result and on over/under 2.5 goals. For some of the comparisons we add LaLiga, the Bundesliga, Serie A, Ligue 1, the Eredivisie and the Championship, 18,433 matches in all.
- Chances. We remove the bookmaker's margin from the odds and turn the chances of a win, a draw and the number of goals into the chance of every correct score, using a Poisson model with the Dixon-Coles adjustment. The same model produces the prediction on every match page at SharkBetz.
- Favourite adjustment. The model slightly overrates the favourite's big wins (2-0, 3-0) and underrates its narrow ones (1-0, 2-1). We correct this with multipliers taken from seasons other than the one we test, so no result helps to predict itself.
- Rounds. Skrill picks 6 matches from a matchweek. We recreated that by drawing 6 matches 100 times from each of 234 Premier League matchweeks, 23,400 rounds in all.
- Prizes. We use the table from the terms of the 2026/27 edition (14 August 2026 – 30 May 2027): 1,000 Knect points for 1 hit, 2,000 for 2, 5,000 for 3, 20,000 for 4, 300,000 for 5 and 10 million for 6. 1,000 points are worth €1. We leave the leaderboard prizes out and explain why below.
The chance of a single score: usually 10–14%
Even the best correct score pick misses more often than it lands. In our sample the most likely score had a 12.2% chance on average, and in 8 matches out of 10 between 10.4% and 13.9%. The record was 17.6%, with a very clear favourite. A pick therefore has an 86–90% chance of being wrong.
These chances hold up. In matches where the model gave the pick 12.2% on average, the pick came in 12.9% of the time. That is within the normal spread, and the model even comes out slightly cautious.
The chances of each prize level in one round
Six matches are six independent tries at roughly 12% each. This is how the hits in a round split with the best picks:
| Hits | Chance per round | Roughly | Knect points | What this level adds to a round's value |
|---|---|---|---|---|
| 0 of 6 | 45.8% | almost every other round | 0 | €0 |
| 1 of 6 | 38.2% | two rounds in five | 1,000 | €0.38 |
| 2 of 6 | 13.3% | one round in eight | 2,000 | €0.27 |
| 3 of 6 | 2.47% | 1 in 40 rounds | 5,000 | €0.12 |
| 4 of 6 | 0.26% | 1 in 385 rounds | 20,000 | €0.05 |
| 5 of 6 | 0.015% | 1 in 6,871 rounds | 300,000 | €0.04 |
| 6 of 6 | 0.00034% | 1 in 293,000 rounds | 10,000,000 | €0.03 |
| Total | 0.73 hits per round | €0.90 |
In the actual matches it went slightly better still: 0 hits in 42.2% of rounds, 1 in 40.0%, 2 in 15.5% and 3 in 2.3%, 0.78 hits per round on average. Neither a perfect round nor 5 out of 6 came up in any of the 23,400 rebuilt rounds. That is no surprise with chances of one in several thousand or less.
The last column matters most. 72% of the game's value comes from rounds with one or two hits. The top prize, however much it stands out in the advertising, adds less than 4 euro cents per round to the average.
For comparison: the jackpot of the UK's Lotto, where you pick 6 numbers from 1 to 59 and every draw has two rounds, is about 1 in 22,528,738 per line in a draw. A perfect Final Score round with the best picks is therefore about 77 times more likely, but the prize is incomparably smaller.
A whole season: what you will really collect
A single hit is not the most interesting part; what is left after 38 matchweeks is. So we played 200,000 seasons of 38 rounds each, with the split of hits shown above:
| What | Result |
|---|---|
| Hits in a season | 28–30 on average, between 22 and 36 in 8 seasons out of 10 |
| Knect points in a season | about 34,000 on average, or €34, between 22,000 and 43,000 in 8 seasons out of 10 |
| At least 10,000 points (the smallest voucher, €10) | in practically every season |
| At least 30,000 points | in roughly every other season |
| At least one round with 3 of 6 | in 6 seasons out of 10 |
| At least one round with 4 of 6 | in 4–10 seasons out of 100 |
| A round with 5 of 6 | once in about 180 seasons |
| A perfect 6 of 6 | once in about 7,700 seasons |
You will usually have your first €10 voucher after 13 rounds, and in half of cases between round 10 and round 16. Points only expire at the end of the following calendar year, so nothing is lost during the season.
To be fair: €34 a season is less than €1 a week for a few minutes of typing in scores. But it needs no stake and risks nothing.
Which strategy earns the most
We compared our method with six simple ways of picking on the same Premier League matches. "Hits" is the share of picks that landed in the actual matches; the rest comes from the model for random rounds of 6 matches:
| Strategy | Hits | Round value | Season (38 rounds) | Round without a hit | Chance of 6 of 6 |
|---|---|---|---|---|---|
| Most likely score (with the favourite adjustment) | 12.9% | €0.90 | €34.2 | 45.8% | 1 in 293,000 |
| Most likely score without the adjustment | 12.4% | €0.89 | €33.9 | 46.0% | 1 in 305,000 |
| Always 1-1 | 11.2% | €0.77 | €29.4 | 49.5% | 1 in 546,000 |
| Most likely score, but no draws | 10.9% | €0.75 | €28.5 | 50.4% | 1 in 611,000 |
| Favourite wins 1-0 | 9.9% | €0.65 | €24.5 | 54.6% | 1 in 1.1 million |
| Favourite wins 2-1 | 9.6% | €0.63 | €24.1 | 54.8% | 1 in 1.3 million |
| Favourite wins 2-0 | 9.1% | €0.59 | €22.4 | 56.9% | 1 in 1.9 million |
Three conclusions:
- Picking favourites is the weakest strategy. It earns 28–35% fewer points, and a perfect round is 4–6 times less likely than with the best picks.
- Plain "always 1-1" beats every favourite pick. You are better off picking a 1-1 draw in every match than a win for the favourite.
- Giving up draws costs about 17% of the game's value. That set looks more exciting, but it earns €28.5 instead of €34.2 a season.
Why we pick 1-1 most often
1-1 is the most likely score in 60% of matches. That is not a quirk of the model but of how matches end. We counted it without any model, in 18,433 matches from seven leagues, by the favourite's chance of winning (scores written as the favourite's goals – the opponent's goals):
| Favourite's chance of winning | Matches | Most common scores | All draws |
|---|---|---|---|
| up to 40% | 3,244 | 1-1 15.3%, 1-0 9.6%, 0-1 9.3%, 0-0 8.9% | 30.5% |
| 40–50% | 6,051 | 1-1 13.8%, 1-0 10.3%, 2-1 9.5%, 0-0 7.6% | 28.1% |
| 50–60% | 4,202 | 1-1 12.3%, 1-0 11.2%, 2-1 10.6%, 2-0 9.4% | 25.3% |
| 60–70% | 2,573 | 1-0 11.9%, 2-1 10.8%, 2-0 10.3%, 1-1 9.7% | 22.0% |
| 70–80% | 1,569 | 2-0 11.5%, 1-0 11.0%, 2-1 10.9%, 3-0 9.8% | 16.8% |
| above 80% | 794 | 2-0 13.1%, 2-1 9.6%, 3-0 8.3%, 3-1 8.2% | 10.6% |
A favourite's win is spread over many scores: 1-0, 2-0, 2-1, 3-1, 3-0. A draw almost always ends 1-1, sometimes 0-0 or 2-2. That is why, until the favourite has about a 60% chance, none of its winning scores beats 1-1. Only with a clear favourite do 1-0 and 2-0 become the most common.
In rounds 1–9 of this season 1-1 turned up in our picks even more often, although Skrill fills its rounds mainly with big-six matches, which usually have a clear favourite. Where this paradox comes from we explain in how Skrill picks Final Score matches.
The draw progression system rests on the same pattern: draws come often but unevenly, and they can stay away for long runs.
Is it worth picking differently from other players?
Not for the round prizes. Everyone who gets the same number of scores right receives the same points, and nothing is shared between players. It is different in a tipping contest with a leaderboard, where only the top places win prizes and it pays to stand out.
The matches are independent of each other. A higher chance in one match therefore raises the chance of every prize level at once: of 1 out of 6, of 3 out of 6 and of a perfect round. There is no trade-off here between safer small wins and a better shot at the top prize. The most likely score in every match is the best pick for everything.
The XP leaderboard: a second prize pool
The 2026/27 terms also have a leaderboard. Your results in the game earn you experience points (XP), which reset to zero at the start of each leaderboard period, and at the end of the period Skrill rewards the best players:
| Leaderboard place | Knect points | In euros |
|---|---|---|
| 1st | 300,000 | €300 |
| 2nd | 75,000 | €75 |
| 3rd | 30,000 | €30 |
| 4th–10th | 5,000 | €5 |
| 11th–100th | 2,000 | €2 |
| 101st–500th | 1,500 | €1.50 |
| 501st–1000th | 1,000 | €1 |
| 1001st–2500th | 500 | €0.50 |
When players tie on XP, the pool for that place is split equally. The terms do not say, though, exactly what earns XP or how long a leaderboard period lasts, and we do not know how many people play. That is why we do not count these prizes in the game's value. If XP grow with the number of hits, the most likely scores also earn the most XP. Standing out may make sense at most in the race for the top places, where many players with identical picks would have to share the pool.
What you do need to watch is your own head. After three rounds without a hit it is tempting to "change something", but rounds without a hit happen in almost every other matchweek and mean nothing. More on traps like this in our article on betting psychology.
Our 2026/27 season in numbers
From the first round we have published our set on our Skrill Final Score picks page. We work the picks out from the odds up to the round's deadline, the first match, and then freeze them. This is how the first rounds of the 2026/27 edition went:
| Round | Matches | Our picks | Landed | Expected |
|---|---|---|---|---|
| 1 | 22–24 August | 1-1, 1-1, 1-1, 2-1, 1-1, 1-1 | 1 (Man City – Bournemouth 2-1) | 0.70 |
| 2 | 28–31 August | 1-1, 2-1, 1-1, 1-1, 1-1, 0-1 | 1 (Aston Villa – Arsenal 0-1) | 0.69 |
| 3 | 4–6 September | 1-2, 1-1, 2-0, 1-1, 1-1, 1-1 | 1 (Brentford – Sunderland 1-1) | 0.70 |
| 4 | 12–13 September | 1-1, 1-1, 2-1, 1-1, 0-1, 1-1 | 0 | 0.73 |
| 5 | 19–20 September | 1-1, 1-1, 1-1, 1-1, 2-0, 1-1 | 1 (Fulham – Man United 1-1) | 0.69 |
| Total | 4 of 30 | 3.50 |
4 hits against 3.5 expected is a perfectly ordinary result: with these chances, five rounds bring 4 hits 20% of the time and 3 or fewer 53% of the time. Four rounds with one hit each are 4,000 Knect points, or €4.
The start of the season is interesting. In the first 50 Premier League matches 32% were draws, not the usual 22–27%, yet only 10% of matches ended 1-1. 2-2 came up more often (12%, usually 5–8%) and so did 0-0 (10%, usually about 6%). So there were plenty of draws, just not the ones we pick. In the seven previous seasons 1-1 came up in 11.2% of matches, and that is the level results usually return to.
Before round 6 (10–11 October), the odds of 7 October gave our set a 52.6% chance of at least one hit, 0.70 hits on average and a perfect round at 1 in 407,000. The current set, and how it changes up to the round's deadline, is on our Skrill Final Score picks page.
Why our set has fewer 1-1 picks from round 10
In rounds 1–9 we picked the most likely score of every match, so 1-1 came up in 41 of our 54 picks. From round 10 our Skrill Final Score picks work differently. We pick 1-1 only when the home and away win chances are less than 20 percentage points apart. In the other matches we take the most likely score that is not a draw, and if a clean sheet (1-0, 2-0) is at most 15% less likely than that score, we take the clean sheet. The earlier rounds stay settled as they were.
How this compares with reality shows in the real results of 9,790 Premier League matches from 2000/01 to 2025/26. All we take from the odds here is who the favourite is and by how much. Scores are from the favourite's point of view, so 0-1 is a favourite's defeat.
| Gap between the home and away win chances | Matches | Most common scores |
|---|---|---|
| 0–5 points | 961 | 1-1 13.7%, 0-1 9.8%, 2-1 8.2% |
| 5–10 points | 1,002 | 1-1 13.4%, 0-0 10.0%, 2-1 9.9% |
| 10–15 points | 912 | 0-0 11.0%, 1-1 10.9%, 1-0 10.6% |
| 15–20 points | 857 | 1-1 14.1%, 1-0 10.4%, 0-0 9.5% |
| 20–25 points | 770 | 1-1 11.3%, 1-0 11.2%, 2-0 10.8% |
| 25–30 points | 731 | 1-1 14.5%, 1-0 10.3%, 2-1 8.9% |
| 30–40 points | 1,388 | 1-0 12.8%, 1-1 10.4%, 2-1 10.2% |
| 40–50 points | 1,128 | 1-0 12.7%, 2-0 11.2%, 1-1 10.8% |
| 50–60 points | 864 | 2-0 13.9%, 1-0 12.0%, 2-1 11.5% |
| over 60 points | 1,177 | 2-0 13.6%, 1-0 11.1%, 2-1 10.5% |
1-1 stays on top up to a gap of about 30 points, then 1-0 is the most common score, and from 50 points 2-0. 2-1 is not the most common score in any band. So the 20-point limit is our choice rather than a line in the data: between 20 and 30 points 1-1 still comes up often, and that is where we pay for having fewer draws in the set.
This is not a better forecast but a deliberate choice of a more varied set. Over 2,660 matches from 2019/20 to 2025/26 the old way hit 29.8 times per 228 picks and the new one 26.9 times. That is about 3 hits fewer a season (95% interval: 0.6 to 5.2) and about €6 less in Knect points. In rounds 1–10 of this season 1-1 would be in 27 picks instead of 45. In the 54 matches played that gives 5 hits instead of 7, but the sample is too small to conclude anything. If you want the most hits, enter the most likely score from the match page, even when it is 1-1.
Final Score and a correct score bet at a bookmaker
The same 12% chance looks very different at a bookmaker. The fair odds for a score with a 12% chance are 8.33. Bookmakers pay less than that on correct score, because they build in their margin. At odds of 6.50 a £10 bet returns 0.12 × 6.50 × £10 = £7.80 on average, a loss of £2.20, or 22% of the stake. The same pick in Final Score costs nothing and earns about 15 euro cents on average, a sixth of a round's €0.90.
That is why a correct score bet at a bookmaker is only worth it when the odds are higher than the probability implies. The fair odds calculator shows how to check that, and the market itself is covered in our guide to correct score betting. Why even a good model is not enough on a market with a margin we explain in why a betting model is not enough.
How to get the most out of the game
- Play every round. The game's value is the sum of small hits: every round you skip is almost €1 less on average.
- Pick the most likely score in every match, even if it is the fifth 1-1 in your set. You will find it with percentages on the match page, and from round 10 our Skrill Final Score picks deliberately have fewer 1-1 scores.
- Enter your picks close to the deadline. The odds, and with them the chances of each score, keep changing until the last days before the match, most of all on matchday and after the line-ups are announced.
- Do not count on a perfect round. €10,000 comes once in thousands of seasons. A realistic goal is a voucher every few weeks.
- Redeem your points in time. They expire at the end of the following calendar year.
FAQ
With the most likely score in every match, about 1 in 293,000 rounds, or once in roughly 7,700 seasons of 38 rounds. Picks on favourites (1-0, 2-1, 2-0) have a 1 in 1.1–1.9 million chance. For comparison, the UK Lotto jackpot is about 1 in 22.5 million per line in a draw.
With the prize table of the 2026/27 terms and the best picks, about 34,000 Knect points on average, or €34. In 8 seasons out of 10 it is between 22,000 and 43,000 points. Favourite picks give €22–25 on average. Prizes from the XP leaderboard may come on top.
From round 10 we pick 1-1 only when the home and away win chances are less than 20 percentage points apart, because we want a more varied set. It costs about 3 hits a season, about €6 in Knect points. The most likely score with percentages is still on every match page.
Once in about 6,900 rounds with the best picks, or once in roughly 180 seasons. A round with 4 of 6 comes once in 385 rounds, and 3 of 6 once in 40.
Usually 13, and in half of cases between 10 and 16 rounds. The smallest voucher needs 10,000 Knect points, and every hit is worth at least 1,000 points.
Yes, when the favourite has less than about a 60% chance of winning. In such matches 1-1 is the most common single score: in 18,433 matches from seven leagues it came up in 12–15% of them. The "always 1-1" strategy earned more points than any favourite pick.
Only in one way: pick the score with the highest chance in every match. From margin-free odds that means 12–13% hits per match. No system or gut feeling will raise that noticeably, and fan picks lower the hit rate to 9–10%.
Not for the round prizes: Final Score pays fixed prizes for the number of hits and does not split a pool between players. The matches are independent, so the highest chance in each match raises the chances of every prize level at once, a perfect round included. Standing out may make sense at most in the race for the top places of the XP leaderboard, where players tied on XP share the pool.
Because even the best pick lands in only 12–13% of matches. The chance that none of the six lands is about 46%. Rounds without a hit therefore happen in almost every other matchweek and do not mean the picks are bad.
From bookmakers' odds with the margin removed: we turn the chances of a win, a draw and the number of goals into the chances of every score with a Poisson model. We adjust the favourite's scores with multipliers from seven leagues, because the model alone overrates its big wins. On 2,660 Premier League matches the picks landed in 12.9% of matches against a 12.2% chance according to the model.