Round of 16 Predictions

My main issue with looking at tournaments is the small sample size that is available for any type of interpretation going into the next round of games. However, I will take the freedom to completely neglect all the intangibles and just look at what SHOULD happen (in my opinion) based on what has happened in the group stage games.

Before we begin, I will introduce the metrics that I will be using to base my predictions.
1) TTS: Time To Shoot, how long a team takes (in seconds) to take a shot when they are in possession of the ball
2) xG: Expected Goals, how likely it is for shot to become a goal
3) xG/S: How many expected goals are added every time the team takes a shot

*Disclaimer: this is as much an experiment for me as it is for you, it has taken me quite a long time to come up with some sort of a metric which I think is appropriate. I may be completely flawed in my conjecture, nonetheless, here’s how it is:

TTSF: Time to Shoot For
TTSA: Time to Shoot Against
xGF: Expected Goals For
xGA: Expected Goals Against
xGF/SF: Expected Goals For per Shot For
xGA/SA: Expected Goals Against per Shot Against

Switzerland vs. Poland

Poland 218.2 237.3 2.5 2 0.074 0.054
Switzerland 234.5 194.9 3.6 2.6 0.090 0.074

The two teams have quietly done well during the group stages, which makes for a game which I am quite looking forward to. Evidently, Poland are a better team when it comes to taking shots quickly and making the most of their possession, whereas Switzerland are giving away too many shots too quickly. This is also reflected in their respective expected goals against, where it is a higher number than that of Poland.
What should also be noticed, however, is that Switzerland are taking better shots than are Poland, because although they take longer to shoot, the quality of their shots are generally better as shown by their higher expected goals for number. Both the teams have a relatively low difference between their xGF/SF and xGA/SA. And that makes sense because majority of the games the teams played in the group stages were very tight and none finished with a margin of greater than one goal.

All that in mind, I predict a game where Poland will come out on top with a narrow margin of a 1-0 win.

Wales vs. Northern Ireland

Wales 190.1 199.4 3.3 3.8 0.087 0.064
Northern Ireland 325.0 175.0 1.1 5.1 0.065 0.084

I think it is fairly clear that Wales are beating Northern Ireland in every stat. With Wales having topped their group and Northern Ireland only coming third, this game has a Wales win written all over it. Wales have not been particularly strong with their 3-man defensive system, but I can’t see Northern Ireland penetrating their back enough to score a goal AND win the game.

Will be expecting a comfortable 2-0 win from Wales.

Croatia vs. Portugal

Croatia 153.7 279.6 5.3 2.7 0.108 0.087
Portugal 141.6 369.7 4.8 2.1 0.069 0.124

Portugal have been largely unlucky, but at the same time, Croatia also had quite an unusual game against Spain and that has made it rather complicated for me to understand what all that data means.
The first standout stat for me is that it takes a long time for a team to get a shot off against Portugal. Again, I have my doubts given the quality of opposition they faced (also reflected by xGA) in the group stages, but we’ll give Portugal’s defence the benefit of the doubt and expect that they will not concede many more shots in the knockout stages. After all the Cristiano Ronaldo shots that we’ve seen, it comes as surprise that Croatia still have a higher xGF number. But Ronaldo has been taking…. odd 40 yard shots almost every second minute so it comes as no surprise that despite shooting so frequently, their statistics are still so lopsided. It maybe of concern that Portugal have such a high xGA/SA, but they only concede a shot every 6 minutes or so which is a very healthy rate. Croatia have got very promising shooting numbers, and just like Portugal, they don’t concede a shot very often. Despite a tough group, Croatia have been very well versed and dominant in how to approach and play all their games.

Portugal have a team that is built solely around Ronaldo, and that imbalance has been shown three times already as they only scraped past into the round of 16 from a group where they really should have won all the games. With Modric & Mandzukic coming back for Croatia, and them coming off the back of a 2-1 win against Spain, I have the feeling that Croatia will win this game 3-1.

France vs. Republic of Ireland

France 176.6 307.1 4 1.9 0.080 0.0079
Republic of Ireland 237.3 305.0 1.7 2.3 0.055 0.0079

Is it just me or does it feel like France still don’t have their best starting xi sorted out? I don’t know, it’s just that I get a very uncertain type of a feeling from France, especially Griezmann who has definitely not been the same player after the Champions League final.
France have performed well though, and they have a very good xGA/SA metric. This coupled with Ireland’s relatively poor xGF/SF number tells me that Ireland will not have plenty of chances to win this game. On the other hand, France have been shooting consistently and at a healthy rate of 0.080 xGF every time they shoot. By the look of it, Ireland and France both have done a good job in preventing shots, which leaves me to believe that this game will depend solely on who can score the first goal, whoever does that will win the game.

Looking at the way things have gone, and how Ireland only beat a weakened Italy side, I will go for a win for the hosts, 1-0 being the final score.

Germany vs. Slovakia

Germany 173.8 387.6 6.4 1 0.103 0.071
Slovakia 287.9 158.5 2.7 4.2 0.104 0.076

Slovakia’s defence has been very good. They’ve blocked all sorts of shots and even scored when they should have. Generally, they look like a team who have defended and known when they needed to get forward. The problem is, when you look at those numbers, it’s a complete mismatch.
It will be expected of Slovakia to play on the counter but even then, they will not get enough of a chance to take a good quality shot. Germany give away a shot roughly every 387.6 seconds when they don’t have the ball and every one of those shots contributes to around 0.071 expected goals against them. Let’s say that Slovakia is able to score on a break like that, even then, Germany’s offensive numbers are so overwhelmingly high that it will be just a matter of time before they can take control and score a few goals themselves.

The World Champions aren’t exactly perfect though, their finishing has been unlucky/poor/lazy but in the knockout stages I believe they will be more serious and clinical and move onto the 1/4 finals. I predict Germany will win 3-0.

Hungary vs. Belgium

Hungary 239.3 182.0 2 2.4 0.057 0.056
Belgium 155.3 220.6 4.1 2.6 0.073 0.076

Hungary have been quite the surprise package of the tournament, scoring 3 really odd goals against Portugal and topping the group has made them fun to watch – but it won’t last for too long. Belgium had a rough start against Italy, but they’ve done very well and have put up impressive numbers to come back from that defeat and assure the bookies that they’re not out of it yet.
Hungary are conceding more shots, but have lower expected goals against… rather weird UNTIL you realize that Ronaldo had been shooting from all over the place (more bad quality shots = lower xGA). Other than that, Belgium seem to be consistently shooting, and shooting from good positions. They are contributing a lot to their xGF/SF but at the back end, they are conceding a worrying amount of expected goals per shot against them. Their saving grace is that they shoot a lot more frequently than they concede a shot, making them favourites coming into the game.

Unless Hungary can pull off yet another unexpected result, Belgium should be winning this game; my prediction is that they’ll win the game 3-1.

Italy vs. Spain

Italy 315.6 244.4 2.4 2.4 0.096 0.071
Spain 202.3 217.8 6.5 1.9 0.127 0.070

I wish this game had happened later in the tournament because two games in, I had the impression that it would be a Spain vs. Italy final. But Spain somehow managed to lose to Croatia and now we’re stuck here with a game which will knockout two teams who don’t deserve to go so early.
Nevertheless, one of them has to go and as it stands, it looks like Spain might edge it. They have better attacking numbers, and they are not leaking chances an unsustainable number of chances at the back. What I will say though, is that in the last group game vs. Ireland, Italy played a significantly weaker team and lost. This hurt their statistics a lot and it does not exactly provide the true picture of what the best Italy side can do (the side we will expect to play on Monday). Spain are putting really impressive numbers on the board, and despite of what I just said, Italy have really only just been good enough till this point.
Italy have scored when they have had the chances (2.4 xGF = good chances), but chances they have been shooting quite a low amount (shot every 315.6 seconds!).

I really don’t know, it’s such a tough game to call EVEN THOUGH EVIDENTLY SPAIN HAVE BETTER NUMBERS, this is just one game where numbers may not be right. Not sure about this prediction, I can be completely wrong, but I’m siding with Spain for a penalties win after the game ends 1-1 in regular time.

England vs. Iceland

England 146.6 378.9 5.1 1.2 0.080 0.067
Iceland 241.6 174.5 2.9 4.2 0.126 0.069

Let’s just say Group F was a total mess. How Austria and Portugal finished last two is still quite astonishing, but fair play to Iceland and Hungary. Iceland, just like Hungary have been quite unpredictable. With help from rather suspect defending they have accumulated a very healthy expected goals number of 2.9. They have though, at the same time given 4.2 expected goals away… thanks once again to Ronaldo’s 39.12 yard efforts! All in all, they have played well and with a bit of luck, deserve to be in the knockout stages.
England on the other hand, have been a little bit like Germany – they have created really good chances, and have posted amazing numbers, but their finishing has been poor which resulted them in finishing second in their group and making the knockout stages tree so lopsided. Somehow, Iceland have one of the highest xGF/SF at 0.126, but they are shooting quite infrequently so they need to get in the opposition box more if they want to make that statistic count.

England will probably win the game; but they will make it hard for themselves and find themselves in a hole of criticism again. Prediction: England wins 2-1.

Conclusively, just a big thank you for reading all of this. It was really fun and frustrating for me to try and figure out how to populate and present all the data.
Let me know your predictions in the comments below!

Follow this blog and connect with me on twitter!

More on TTS: https://deepxg.com/2016/04/05/time-to-shot/
xG data: https://twitter.com/MC_of_A



Leave a Reply