Set Pieces in Daily Fantasy Football

Instead of previewing the UCL slates this week, I decided to write an article about set pieces in daily fantasy football. With each cross being worth 0.75 DraftKings points at the time of writing, it’s important to know who takes set pieces when building lineups. The noise volume in football stadiums also increases significantly whenever fans see their team win a corner, with many fans viewing it as a great opportunity to score regardless of who they’re facing.

A couple of things happened last weekend that prompted me to look into this in a bit more detail. Firstly, MOTD2 did a feature that suggested Manchester City were bad at defending set pieces (around 20 minutes in for those interested), and that narrative has gained some traction elsewhere in the media as well. As a City fan, I’m not willing to accept Phil Neville criticising my team without looking at the evidence myself!

Secondly, West Brom’s Sam Field was anywhere between 10-40% owned in GPPs last weekend, which seems a bit nuts even when you factor in his $3400 salary. I deliberately didn’t mention Field in my GW15 article for several reasons. Whilst Field scored in GW14, I’m not ready to accept him as being a guy with that sort of upside on a regular basis. We don’t have much data on Field, but he’s only averaging 0.08 xG90 so far in his short career, and West Brom just aren’t a good attacking team to begin with. Field took a couple of corners in GW14, but we had no idea whether that would continue into GW15, especially with Pardew coming in as manager. I was concerned that Pardew would substitute Field early or switch up the set piece takers for his first game in charge, and it turned out that Field didn’t register a single cross against Crystal Palace. Field’s cheap price tag kept him from being a horrible play in cash games, but it’s amazing to me that nobody seemed at all concerned that he might not be a great play even at that salary. I’ll dive into this, and Man City’s defence, more throughout the article.

Which teams are winning the most corners?

For this article I decided to focus on corners specifically, as this is the area where Man City were accused of being vulnerable on MOTD2. It’s also easier to predict corners than free-kicks for DFS purposes, and teams sometimes use different players to take corners than free-kicks as well. Corners are also always taken from the same place on the field (duh), so we can more accurately compare the performance of different teams to each other.

To start with, let’s take a look at which teams are winning lots of corners, and vice versa. The dashed lines show the median on all plots in this article.

We can see that Man City are winning the most corners per game, and conceding the fewest. This is what you’d expect from a team that dominates possession and spends a lot of time in the opponent’s final third. In fairness to Neville he did mention that teams would still struggle to score against City from corners because they just won’t have many opportunities, and he’s right.

The rest of the consensus “big 6” teams are also averaging at least 5.9 corners per game, although it’s perhaps surprising that Southampton and particularly Crystal Palace are keeping pace with them. Leicester, Stoke and Burnley rank 17th-19th in possession, so it makes sense that they are conceding the most corners per game to opposing teams (defensive style can also influence this, of course). West Brom are taking fewer than 4 corners per game, so the early signs aren’t looking good for Sam Field here even if you assume will take every single one.

Which teams are scoring and conceding the most from corners?

Next, let’s take a look at which teams are scoring and conceding goals from corners. We’re about 40% of the way through the season, but we’re still dealing with a fairly small sample size when it comes to goals, and some teams are still yet to find the back of the net from a corner this season. I’ve therefore adjusted the data in the following plot towards the mean to attempt to account for the sample size issues.

We can see that Man City haven’t been elite at preventing their opponents from scoring from corners (unlike Chelsea), but there’s hardly cause for concern here. City have one of the smallest starting XIs in the league (if not the smallest), so the fact that they are only conceding goals from corners at an average rate suggests that they probably have a decent defensive setup. This excellent article by @From_The_Wing goes into detail on exactly how City defend set pieces (link). The best way to exploit City’s system is at the near post, especially if Fernandinho is on the bench (as he was vs West Ham in the last game). However, this still hasn’t translated to a lot of goals for opposing teams yet since Pep took over. Maybe teams haven’t caught on yet, but it’s also much harder to score when you’re running away from the goal at a tight angle. City generally do a good job of preventing teams from attacking the middle of the goal, where they can do the most damage.

There are a few other teams that stand out on this graph. Klopp has often been criticised due to Liverpool’s inability to defend set pieces well, and we can see that they’ve been consistently bad at it since the start of last season (at least). Fortunately for Liverpool, they’re only conceding 3.3 corners per game. Watford are arguably the worst team in this time frame defensively despite having plenty of tall CBs in the squad, so it makes sense to target them in DFS if you’re looking for set piece goals and assists. Stoke and Leicester have been nearly as bad this season, which doesn’t bode well when you’re also conceding plenty of corners each game. Stoke’s drop off from last season is particularly alarming; it could just be variance, but there are plenty of other reasons why their performance could have dipped (e.g. a change in personnel, or perhaps a new system). The opposite can be said for Bournemouth, who are now conceding the fewest goals per corner after being below average last season. More detailed analysis on these teams (including video) is needed to know for sure why changes have occurred.

On the attacking end, West Ham have been good at scoring from corners both this season and last, so it’s perhaps unsurprising that they were able to cause City some problems last weekend. Newcastle have also been impressive so far this season, and fortunately for Watford, Leicester and Stoke the goals have been going in at the other end of the pitch too. West Brom were excellent last season from set pieces but have taken a big step back in this campaign, which is likely a major contributing factor to their current table position just above the relegation zone. Maybe they’ve just been unlucky, but with Pardew in charge who knows whether they will have the same focus on scheming set pieces going forward. Another negative for Field there, then.

From the next chart, we can get a better idea of which teams to target in DFS when looking for set piece goals.

The top right of the graph is where you want to be, and that’s exactly where the best teams in the league are. Perhaps unsurprisingly, it’s the worst teams in the league that are in the bottom left portion of the graph. No real edge there, then, assuming you don’t rely on the current league table to judge team strength.

What about crosses from corners?

Ah, the short corner – every DFS player’s favourite. Unfortunately, the better teams in the league tend to favour short corners a lot more than the worst teams, particularly Manchester City and Chelsea. This means that although these two teams are winning lots of corners per game, they aren’t actually getting more crossing opportunities than some of the worst teams in the league. Crystal Palace and Southampton stand out here, with set piece takers from these teams being favourable targets in DFS. Palace are actually leading the league currently with 4.5 DK points per game from corners on average, which is a nice boost to Cabaye’s floor when he’s on the field. Everton, on the other hand, are last with 2.55 DK points per game, which is perhaps one of the reasons why Gylfi Sigurdsson has been a lot less reliable in DFS this season.

So what about West Brom? Well, they’re near the bottom of the pile, with just 2.75 DK points per game from crossed corners. That’s still not an awful boost for a $3400 player like Field if you make the very questionable assumption that he will take 100% of the corners, but he will still need to generate quite a few points from open play to be worth using, especially in GPPs. I certainly think it’s a stretch to say that he has at least a 40% chance of being in the winning GPP lineup at high stakes even with only 29 entrants, and if you’re being results oriented he didn’t hit cash game value based on my calculations (different to the 2x salary multiplier most people use).

I actually played Swansea’s Ki.S.Y in my cash game team in GW13 (don’t ask). Whilst I wouldn’t recommend any of you do that, he was cheaper than Field in GW15, had a similar matchup and was around 0.25% owned in the large GPPs (Field was anywhere between 12-25% in similar contests). Ki outscored Field in GW15, and there is absolutely no good reason why he should have been 50-100x lower owned.


So what are the key takeaways here?

  • Manchester City aren’t bad defensively from corners, as evidenced by the numbers in this article and the excellent piece by @From_The_Wing. They’ve conceded a couple of good opportunities in recent games, but right now I wouldn’t be concerned about their ability to prevent set piece goals over the course of the full season.
  • Set pieces alone are not enough reason to justify using a player. On average teams are only earning about 3.25 DK points from crossed corners (excluding assists). The matchup will be a key factor here, and we want our set piece takers to generate most of their points from open play in as many ways as possible (e.g. shots, crosses, goal/assist upside etc). DMs/CMs aren’t necessarily good options even at a cheap price tag just because they have started taking corners.
  • Crystal Palace, Southampton and Newcastle set piece takers stand out as potentially being a bit undervalued currently, and if West Ham can somehow right the ship we can potentially add them to the list also.
  • West Brom, Everton and Chelsea are probably overvalued currently in terms of set pieces. These teams aren’t currently crossing enough from set pieces to give them a safe floor, and their goal upside has dropped off significantly so far this season after excellent results in 16/17. Of the three, Chelsea are the most likely to turn things around based on what we saw from them last season – the other two have new managers and significantly different starting XIs to the previous campaign.
  • Stoke, Leicester and Watford appear to be good teams to target in GPPs when looking for an assist from a set piece taker. Liverpool aren’t good at defending corners either, but they simply don’t concede enough of them to make it worth going out of our way to pick on them.
  • Arsenal probably have the best combination of opportunity and upside from a good team, and yet Ozil is typically a bit overlooked compared to set piece takers from the other “big 6” teams.
  • Football fans probably shouldn’t be that excited when their team wins a corner. On average, teams are only scoring about 3.5% of the time per corner since the start of the 16/17 season.

That’s it for now! Thanks for reading, and please share this article if you enjoyed it.

All data by Opta.

Leave a Reply

Your email address will not be published. Required fields are marked *