Welcome back to the Stats Corner for Gameweek 19! If you’ve ever read one of my articles, you will know that I love statistics. Statistics do not lie. The people interpreting them may do so (intentionally or accidentally), but the numbers themselves do not. We’re at a point now where trying to prevent the infiltration of data in football is about as successful as Marco Silva trying to prevent the arrival of his P45. So let’s embrace it. Statistics have helped Liverpool return to relevance and Manchester City assert their dominance, but more importantly, they could also help you win your fantasy football league.
Every few weeks – between the Underlying Numbers and EPL Points Against articles – I will bring you Stats Corner. In Stats Corner I will provide the latest breakdown of Points per 90 (PP90), Lost Points, and Expected Goals: Bookends Removed; three statistics that can help you in your quest for fantasy glory…and which can (probably) only be found here. Oh, and I’ll also pop in a bonus statistic just for good measure.
Gameweek 19 Stats Corner
PP90
PP90 stands for ‘Points per 90’ and is simply the FP/G of a player if they were to play the full 90 minutes. By and large, it is probably a better reflection of a player’s fantasy value than FP/G, which does not take into account how much of a game a player has played, only that they played. PP90 better appreciates the value of those players who do not always start. Knowing this can then give us an advantage on those occasions when they do make the starting eleven.
City players dominate the PP90 leaderboard – as they have all season – but it is the name of Michail Antonio that should scream out to fantasy managers. A 10.79 FP/G that rises to 15.95 as a PP90 may just be keeping him under the radar for now, but that won’t last much longer. He’s started the last three, scoring 38.5 points, and has looked good in the process too.
On the right-hand side, there are a few interesting options. I said in my last article that I wasn’t going to fall into the trap of promoting Christian Benteke just yet…but all that has changed. I don’t care if he can’t score; when you win 25 headers in two games it doesn’t matter. It’s a similar story for Andy Carroll, who is just 16% owned; though if you do pick him up, check to see if he continues to start – with his injury history you’d expect him to be rested at least once over the next few games. Lucas Moura was poor against Chelsea, but he may see his minutes increase given that the competition in the Spurs attack is set to diminish. If so, he could be one to target with a trade – a current FP/G of just 7.31 might make doing so pretty easy. Finally, you’d expect Reece James to start a couple of games over the festive period, and if his PP90 (10.93) is anything to go by, he could prove a very nice present for one manager.
Missing out on the PP90 leaderboard (because they haven’t yet played the 400-minute criteria) are monsieur’s Origi, Hudson-Odoi, Gudmundsson, Albrighton, and Foden. All have PP90’s north of 12, with CHO’s a huge 18.42 and Foden’s a massive 20.43. If ever these players are to be given a full 90 minutes this season, the next few weeks are it – so set your alarms for 2 pm and make sure you’re the first to check those lineups!
Lost Points
One of the many ways in which Fantrax EPL is the best form of fantasy football is in the types of scoring system that can be implemented, most notable of which is the Togga scoring system. Togga scoring allows for negative points – or Lost Points – and includes dispossessions, cards, and own goals (note, they do not include goals conceded). Cards and own goals are relatively rare occurrences, and therefore knowing which players have FP/G’s that are skewed by such things could give you a leg up on the competition.
A few big names entering the leaderboard in Richarlison, Son, and Grealish. Richarlison has been cultivating his inner-Holebas in recent weeks, with three yellow cards, whilst Villa’s main-man is (understandably) having to take matter into his own hands…but that’s coming at the cost of increased dispossessions. I’d prefer not to mention the Tottenham forward after he single-handedly cost me victory this weekend with his sending off 30 minutes from time. That’s his second of the season. He’s now been sent off three times in his last 17 league games for Spurs after not seeing red (and only seeing yellow four times!) in his previous 129. Oh, and of course, Wilfried Zaha still leads the way.
Fantasy Options: Philip Billing is owned in just 47% of leagues, but his 6.85 FP/G should increase as the season goes on as it is currently being brought down by an unusually high amount of yellow cards (even for him).
Expected Goals: Bookends Removed
Expected goals is “a statistical measure of the quality of chances created and conceded” (Understat.com). These expected numbers are based on averages, so we should expect the teams that have above-average players to consistently outperform them. But for everyone else, we should see their performances (and therefore their league position) more closely align with the expected numbers as the season goes on. Obviously fantasy points and league performance go hand-in-hand (see the PP90 above for evidence of that!), so knowing this information can help us stay one step ahead in finding that breakout player (or player who has already peaked!).
Many football sources use expected goals tables as an indicator of which teams have the best/worst defenses and attacks. The problem with this is that every team in the Premier League has the capability to play excellently or atrociously on the odd occasion (the old “every dog has his day”/”it never rains but it pours” idioms comes to mind). Grouping such occasions into our sample of data distorts the final picture being presented. One way around this is to simply remove the bookends: the best and worst performances of each team during the season. For instance, Manchester City has an average xg of 3.13, but if you remove their best (6.63 vs Watford) and worst (1.48 vs Liverpool), then it becomes “just” 2.95. This is, in my opinion, gives you a better estimate of what to expect from an attacking perspective next time City play. Now, with 18 games of the season gone, we will be removing the TWO best and TWO worst performances; leaving the XG and XGA tables to be based on the middle 14 games.
The alarm bells should be going off if you’re a fan of Aston Villa or West Ham. Not only do they – rightly – sit bottom of the XGA table (see next table), they both see their new numbers drop precariously in the XG table above, too. Simply put, everyone knows they can’t defend, but actually, they’re much less potent in attack than it appears too! Anyone looking to target Wesley, Trezeguet, Snodgrass, Yarmolenko, etc. should be aware of that.
Elsewhere, Steve Bruce may have found a genius tactic that can win games but fail miserably to pass the stats-test…or Newcastle may be massively overperforming, with a drop back down to earth due. Whilst I think they’ve definitely improved since November, I’m still backing it to be the latter. Conclusion: hold off jumping back on the Almiron hype train for now. Leicester is renowned for defying the expected stats, but that is likely largely down to Jamie Vardy, who is one of the most clinical forwards in Premier League history. Whether the likes of Maddison, Tielemans, and Barnes can maintain their overachieving numbers is less certain. Oh, and look at the position of Southampton in all of this! This might not just be a purple patch for Danny Ings…
Fun fact: both Arsenal and Everton have had their two worst attacking performances of the season in their last two games. Top work by caretakers Ljungberg and Ferguson…anyone would think the two were defenders back in their day!
“Normal” XGA tables will likely continue to underappreciate the defenses of Watford and Southampton all season (because of those hammerings against Man City and Leicester, respectively). These tables won’t, and it may tempt you into bringing in a Kabasele or Vestergaard if the matchup is right. Leicester’s defense is already receiving all the plaudits, but even then, they may be undervalued. Acquiring a fantasy asset there might come at too great a cost though. As mentioned last time, it is at Spurs where you might be able to find some cheap talent. Yes, Chelsea took them apart, but Chelsea is a good team – especially in attack. With games against Brighton and Norwich coming up, their defenders could be worth streaming.
Fun fact: Southampton have an XGA of just 0.86 over their last five games. To demonstrate just how exceptional this is, league leaders Liverpool have the best defense in the league, with an XGA of 0.97 this season.
Bonus Corner: WhoScored Rating vs FP/G
We seem to have a million stats in football these days. Whether it’s the basic ones such as goals, shots, and tackles, or the more complex ones like expected threat and ball progression, there are numbers for pretty much everything. What would be nice, though, is for someone to package all these different things up and give us a simple score – out of 10 – indicating how a player performed in a game. As it happens, a number of sites do just that, and WhoScored is one of the most prominent. How accurate their models are in reflecting the performance of a player is clearly up for debate, but what’s more important to us Fantrax Fantasy managers is whether these numbers can give us some insight and help in managing our teams to glory.
The figure below shows the correlation between WhoScored rating and FP/G (Togga scoring) for the 198 players that have played in at least 50% of their club’s league games this season.
As we see, there is a clear, strong relationship between WhoScored rating and FP/G. The better a player performs (according to WhoScored!), the more points they earn. This, obviously, is a good thing, and it goes to show how the intelligent Togga scoring system (and other similar ones afforded by the customizable points feature of Fantrax) nicely reflects what’s going on out there on the pitch.
But what, if anything, does this mean for you and your fantasy team? To help explore the options, let’s take a closer look at some of the overvalued (FP/G far exceeds what would be expected, given the WhoScored rating) and undervalued (FP/G considerably lower than would be expected, given the WhoScored rating) players.
What immediately stands out is that eight of the nine most undervalued players come from a top-three team (and Traore’s Wolves sit an impressive 6th place too), suggesting that WhoScored ratings favor performances if they are part of a win/draw. So whilst using WhoScored as a gauge of player form (and therefore fantasy value) is a very good idea, just be aware that the form of the team will likely be contributing somewhat to their average rating. Oh, and also note, KDB’s absurd 20.82 FP/G is actually lower – by about 1.5 points – than what you’d expect from someone with a 7.99 rating. (Quick Quiz: Since WhoScored started keeping performance records in 2009/10, how many players have finished the season with a rating higher than KDB’s current 7.99? A free lifetime subscription for these articles awaits anyone who guesses correctly; answers at the bottom of this page).
The undervalued list is more ambiguous. There is not a pattern immediately obvious (at least to me anyway!), which suggests that either, a) these players are doing something on the pitch that happens to accumulate fantasy points but is not necessarily “good” (at least according to WhoScored), or b) the fantasy scores for these players randomly happen to be greater than what they should be getting…which suggests that we could see some regression in “luck” in the second half of the season. If it’s the latter, then this list essentially becomes an “avoid” or a “trade-out” list…particularly interesting for the last guy there.
“The idea that I [should] trust my eyes more than the stats, I don’t buy that because I’ve seen magicians pull rabbits out of hats and I know that the rabbit’s not in there.” (Billy Beane; Moneyball)
Predominant sources used include www.understat.com (source for Expected Goals: Bookends Removed), www.premierleague.com, www.sofascore.com, and www.whoscored.com.
Answers:
Three. Luis Suarez had a rating of 8.43 in 2013/14, Didier Drogba had a rating of 8.32 in 2009/10, and Cesc Fabregas had a rating of 8.25, also in 2009.10.
Follow me on Twitter for all the latest: @the_innergeek
Fantrax was one of the fastest-growing fantasy sites of 2018 and we’re not slowing down any time soon! With multi-team trades, designated commissioner/league managers, and drag/drop easy click methods, Fantrax is sure to excite the serious fantasy sports fan – sign up now for a free year at Fantrax.com.