While most seasoned FPL managers are comfortable using stats like xG and shots in the box, there is one powerful stat that is criminally overlooked: percentages.
Take the table below. It shows which players created the most chances last season – in other words, which players are most likely to get you 3 assist points in any given gameweek. Not only did Bruno Fernandes create more chances than any other player last season with 95, but a healthy 21% of these chances were deemed to be ‘big’ chances (these have a much higher likelihood of leading to a goal).
Compare this to Aston Villa’s new signing Emiliano Buendía when he was last in the Premier League with Norwich in the 2019/2020 season. He created 83 chances, which would have placed him a very impressive third for this metric amongst all players last season. At £6.5m compared to Bruno’s £12m price tag, Villa marquee signing appears to offer sensational value. Scratch beneath the surface of the data, however, and the picture is less clear-cut. Only 11% of his 83 chances created were big ones, which is almost half that of Bruno.
There is obviously a big difference in price between Bruno and Buendía, so let’s take a look at Luke Shaw versus Lucas Digne. Both players are £5.5m and some FPL managers may be trying to decide which of these two popular players to put in their gameweek 1 defence. On the face of it, Shaw looks like the far superior pick – he created 72 chances in 30 starts, while Digne only created 44 chances in the same number of starts. But if you look at the percentages, only 10% of the chances Shaw created were those gilt-edged big ones – for Digne it was 25%. So, when the Everton defender creates a chance for his teammates, it is two and a half times more likely to be a big one than Luke Shaw.
Before we end this segment, a special mention must go to Jarrod Bowen. The West Ham midfielder created 29 chances last season (not enough to make the above table) but 11 of them were big chances created, which works out at a whopping 38% (the highest percentage for any player who created 25 or more chances last season). Contrast this with João Moutinho at Wolves who is at the other end of the spectrum – he created 46 chances last season but 0% were big chances created. Southampton’s well owned James Ward-Prowse didn’t do much better: he created 60 chances but only 5% or 3 were deemed to be big chances.
Percentages are also useful when applied to the big chances themselves, which is the actual goal attempt that takes place from the aforementioned big chance created. The table below shows players who had 20 or more big chances last season and what percentage of these chances they missed.
Topping the charts for big chances may seem like a good thing, but if the player fails to score from the vast majority of them, then he may frustrate FPL managers and dodge points on a regular basis. Sadio Mane, Chris Wood and unsurprisingly Timo Werner were especially profligate in front of goal last season. As was Patrick Bamford – he missed over two/thirds (68%) of his big chances and yet he finished as the second highest scoring forward with 194 points. Imagine how many points he would have scored if he had been more clinical and got closer to 50% and missing every other big chance?
Although he didn’t make the table, relegated West Brom striker Mbaye Diagne gets a notable mention. He missed 11 (or 92%) of his 12 big chances. If he returns to the premier league next season and you notice that he is heading to the top of your stats tables for ‘big chances total’, then you’ll do well to remember this 92% figure. Flip this data and Tomas Souček over at West Ham is the polar opposite to Diagne. The Czech defensive midfielder scored 7 of his 8 big chances, missing just 13% of them.
The third table (shown below) displays the players who had the most goal attempts last season with a few interesting additions in the last few rows.
Players like Bruno and James Maddison look like great goalscoring options with 121 and 75 shots respectively last season. However, with Bruno, only 40% of his shots were inside the box and just 2% were inside the six-yard box. For Maddison, it was even worse – 28% of his shots were inside the box and none were in the six-yard box. So here the percentages reveal that, when these two players shoot, they are mostly doing it from range where there is less chance of scoring.
Compare this to Ollie Watkins, Callum Wilson and Jamie Vardy – they, too, have a high volume of goal attempts, but remarkedly at least 90% of their shots are inside the 18-yard box. This is the dictionary definition of a “fox in the box”. Dominic Calvert-Lewin and Michail Antonio must get a very special mention – 25% and 27% of their total shots respectively take place 6 or less yards from goal.
Percentages can also help us pick our Gameweek 1 goalkeeper for the fast-approaching new season. Having a goalkeeper who makes saves from outside of the box is favourable as by their nature these shots are easier to save, and it makes it easier for the goalkeeper to rack up the save points and to not concede. Looking at the final table below, 40% of Dean Henderson’s saves and 38% of Robert Sanchez’s saves are from shots outside the box, whereas for Emiliano Martínez it is 42% and for Illan Meslier it is 38%. Crucially, though, the percentage figures for the Leeds and Aston Villa goalkeepers stay high even after making over 140 saves. This means that their percentage figures for saves made from shots outside the box is more reliable and statistically significant, as it is happening from a far greater number of saves than the Brighton and Manchester United keepers. Leicester’s attractively priced Kasper Schmeichel sits at the bottom of the table at 22% – more than three-quarters of the shots he saves (78%) are from inside the box. This is asking a lot of him, not to let one of these close-range attempts past him and for the clean sheet to be preserved.
Conclusion
We’ve seen how percentages can either reinforce or discredit some commonly used stats like chances created, big chances, goal attempts and goalkeeper saves. On the face of it, a player can be posting some incredible volume stats, but if the stat in question has a close relation – for example, chances created and big chances created – it is always worth seeing if the stat holds up by putting them side-by-side and working out the percentage.
2 years, 9 months ago
Timo Werner 78%, that's my forward 😎
In all seriousness, thanks for writing Virg. Great article focusing on something so commonly overlooked, brilliant example of why I love CAs so much.