As so many things have in my Fantasy Premier League career, this good idea started with a Twitter conversation with my partner in FPL crime, Jeremy Spitzberg. The discussion was about the value of what has become the standard forward looking projection about the upcoming schedule facing players you might be considering moving in or out of your Salary Cap or Draft fantasy team. The format is very generic assigning weights to where teams are in the table and whether the match is slated to be home or away. Certainly, there is information to be gained from this sort of analysis – if a player you’re considering is playing City away then that’s probably not too good for his prospects. Unfortunately, that’s about where the value of such analysis begins and ends in my book.
The hypothesis that the tables we’ve all been going off of is that match-ups are equally valuable across teams (i.e., all bottom half teams behave essentially the same) and by position (i.e., all teams concede fantasy points similarly across positions). Written out that way, it doesn’t make much sense does it? Are all relegation strugglers conceding points at a similar rate? Do all mid-table clubs concede points the same way? Do all clubs have the same variance between home and road performances?
I understand the simplicity of the notion that the worse a team is, the more likely an opponent is to have a good match against them. It also makes for a nice, pretty chart. It is also a lot of work to compile the data to examine whether the statements above are reasonable or not. Fortunately, I have been able to snag myself some data so that I can start examining our match-up assumptions a little more carefully. As it turns out, this is also about a reasonable time in the season to start making some conclusions that might not be attributable merely to small sample sizes.
This is obviously a lot of data to go through, so I’ll start off with some of my initial observations and, as time progresses, I’ll figure out better ways to harness this into specific fantasy recommendations from week-to-week.
[[ad:athena]]
No Teams Are Created Equal
Whether you’re looking at the top of the table or the bottom, the notion that all teams are created equally based on where they sit in the table is just wrong. Would you like a few examples?
Brighton and Huddersfield Town sit a few spots apart in the table (12th for Albion and 15th for the Terriers after 13 weeks) but facing off against them is a significantly different proposition thus far this season. Whereas the Terriers have conceded 635 fantasy points this season (3rd bottom in the Premier League Fantasy Points Against table), Albion have conceded a much more respectable 515 (12th in my mythical table but significantly closer to Manchester United in 11th than West Ham in 13th). Over the 12 match weeks I have data for, that’s a difference of exactly 10 more fantasy points conceded per match by Huddersfield Town than Brighton.
Certainly, you don’t know where those extra 10 points will come from when you’re going onto the waiver wire or transfer market. If you asked me, though (and you have because you’re here reading), I’d rather pick up someone from the waiver wire playing against a side likely to concede 53 fantasy points or 43 in a given match. Looked at in those terms, that’s a 20% difference in total fantasy points conceded in a given week.
You could be wondering if I’ve just picked the one outlier in the data and written a few paragraphs just to win a months-old argument with Jeremy. While I am in no way above doing something like that, that’s not the case here. Facing off with Manchester City yields an average of just under 23 fantasy points/match while moving down the table only two spots to where Spurs reside the per/match rate rises a whopping 35% to just a shade under 31 points/match. You’d still clearly rather have someone playing a midtable side but if you’re deciding on players in your line-up going against City or Spurs, the decision isn’t nearly as close as the chart tells you it is.
Positions Aren’t Created Equal Either
Even when we have clubs that are similarly situated in the table have yielded very similar aggregate point totals, it doesn’t mean that they distribute those points evenly to their opponents. Some clubs struggle to score and their opponents pick up clean sheet points where as others ship goals like crazy and it is the opposing forwards and midfielders who prosper. Examples? Of course:
Two of our surprise teams in the early going have been Wolves and Watford. As of the end of Week 13, they sit 11th (Wolves) and 9th (Watford) in the real-life table. They are similarly situated in the Fantasy Points Against table at 7th (Watford) and 10th (Wolves) having conceded 433 and 460 points respectively for a difference of under 3 points/match between them. Does that mean you should evaluate players playing against them similarly? I’m sure I wouldn’t have written the opening sentences of this paragraph if you should.
Quite simply, Wolves do not give up points to opposing forwards where Watford ships them by the ton. Wolves are actually the second stingiest club in the Premier League as far as points from forwards go with a mere 50 conceded through 12 weeks. That’s less than Manchester City who are, in aggregate, only giving up a shade over the two points/match that teams would get for fielding a full line-up for 90 minutes.
While Wolves are shutting down opposing forwards at an elite rate, fellow mid-table surprise package Watford who, remember, are actually conceding fewer points/match than Wolves overall, are very close to Fulham’s neighborhood when it comes to conceding points to opposing forwards. The Cottagers are currently “leading” the Premier League in total fantasy points allowed, fantasy points allowed to opposing forwards, and fantasy points allowed to opposing defenders while 2nd to the “top” in points allowed to opposing midfielders. Fulham are bad. Watford are nearly as bad ranking as the fourth most generous club for opposing forwards to face off against. They have conceded more than twice as many points to opposing forwards over 12 matches than Wolves have. That’s a real difference.
Oh, if you’re wondering where the counterbalance comes to get Wolves and Watford back on par as far as aggregate points conceded then the difference is in midfield where the Hornets are relatively stingy (3rd fewest points conceded to opposing midfielders through 12 matches) while Wolves come in where you’d expect as middle of the road in points conceded to midfielders.
Bringing this back to making actual fantasy decisions, if you’re looking at a Waiver Wire claim in Draft and you’re looking to add a 5th midfielder for Week 14 and you have the choice between Callum Paterson against Wolves or Demarai Gray against Watford (both reasonably likely to be available in 8-team leagues) then you should be focused on Paterson because Wolves are more inclined to give up points to midfielders. If you’re making the same choice between forwards in those same match-ups, you’d be much more inclined to take a risk on a forward facing off against Watford (granted, Jamie Vardy isn’t going to be available but if it looks like Okazaki is going to get the start again, you’d have to like his chances.
Location, Location, Location (it matters too)
So, not all teams similarly situated in the table are created equal and not all teams concede at the same rate at all positions. As it turns out, teams also differ significantly in how their generosity to opposing players varies based on the location of the match. The presumption that goes into our favorite forward-looking chart is that clubs play better at home and, therefore, fantasy players on their travels are less attractive across the board. Would you be shocked to learn that the numbers don’t support this assumption?
There are certainly teams – Brighton (16.66 more fantasy points conceded per road match than home match), Everton (14.17), Cardiff City (12.91), and Manchester City (9.54) – that fit the narrative of being more generous on the road than at home. There are also some clubs – Crystal Palace (15.67), Spurs (13.63), Chelsea (7.23), and Manchester United (4.91) – who stand the narrative on the road and actually concede more fantasy points at home than they do on the road. There are other still – Arsenal, Leicester City, West Ham, and Wolves – whose propensity to concede fantasy points is almost exactly the same on the road and at home.
If you’re looking for some overall wisdom as far as home/road splits, you might be surprised to learn that overall this season, Premier League teams have conceded a mere 3 fantasy points more per match on their travels as compared to at home.
So What?
So, where is all of this fun with pivot tables and formulas taking us? The top level conclusion is that there is far more nuance to be had when it comes to playing at making short term changes to your squad, especially when it comes to the draft version of the game where your choices are limited. Are you going to sit/sell Eden Hazard or Raheem Sterling because they’re playing against an opponent who is particularly stingy against midfielders? Of course not. What you are likely to do when scanning the waiver wire for a one or two week starter, is look a little more deeply into what sort of opportunity your waiver wire choices are facing.
Week 14 Waiver Wire Advice
Hopefully, I’ve made a reasonable case that there is more to a waiver wire decision than a top line analysis of the quality and location of an opponent. With that in mind, here are some waiver wire thoughts ahead of the deadline for Week 14 based on the more detailed look at who tends to concede fantasy points and to whom:
Huddersfield Town Midfielders – Brighton are an interesting side as a fantasy opponent, at home their “points against” looks similar to Arsenal and Chelsea while on their travels they look like more like West Ham or Newcastle – not abject but certainly not good. This will certainly be a case of a movable object facing a resistible force as the Terriers have been among the least prolific teams in the Premier League at scoring fantasy points (2nd bottom in “Points For”) so it’s no guarantee that the Terriers will profit from Brighton coming to town but Schindler, Mooy, and Lossl – all three presentable fantasy performers thus far this season – are far more valuable in Week 14 than they usually are.
Brighton Defenders – If you’re looking for a hedge for your investment in someone like Aaron Mooy, it turns out you can pick up some insurance cheap by going with the numbers that tell us that Huddersfield are among the most generous hosts in the Premier League for opposing defenders. Only Fulham and Southampton have allowed more points to defenders overall than the Terriers and they are actually the worst in the division at giving up points to opposing defenders on home soil. If Duffy and Dunk are both taken then it might be the weekend to pick up Bruno.
Wolves Midfielders over Forwards – Yes, Cardiff City are near the bottom of the table in the real world and their generosity when it comes to conceding fantasy points is similar. What won’t be as obvious is that the Bluebirds have been far more likely to concede fantasy points to midfielders than they have to forwards when playing at home. If you’re choosing between Wolves players to start then perhaps Neves (and the hope of a penalty) or Moutinho over Jiminez.
A Burnley Bounce? – The Clarets have been abject thus far this season, especially compared to their fantasy-defender-friendly recent past that saw guys like Michael Keane, Ben Mee, and James Tarkowski not to mention a parade of goalkeepers be fantasy-relevant over the past few seasons. It has all fallen apart this season but a trip to Selhurst Park, where Crystal Palace are conceding the most points/match to opposing goalkeepers and the second most to opposing defenders, could be a tonic for readily available Burnley defenders and Joe Hart.
Callum Wilson – Manchester City have been amazing across all fronts and there is almost zero value in playing a goalkeeper or defender from Bournemouth for the Cherries’ trip to the Ethiad in Week 14. What is surprising is that City look almost the same as West Ham on a points/match basis when it comes to giving up fantasy points at home to opposing forwards. I’m not guaranteeing Wilson a goal but starting him isn’t the fools errand that starting Begovic would be.
Avoid Newcastle vs WHU – This is more of a trend item than an aggregate “points conceded” item but after both clubs started off the season shipping fantasy points to their opponents for fun both of these clubs have limited opposing output significantly over the past few weeks. Playing “momentum” in mid-table clubs is always a dangerous thing but both clubs are currently looking more like Top Four clubs than bottom half teams when it comes to conceding fantasy points to their opponents.
Newcastle’s Forwards Have Value – Much has been made of Newcastle’s selling of Alexsander Mitrovic and their subsequent lack of a talisman. As it turns out, they have managed to cobble together a rather representative output from their forwards despite the lack of a clearly defined star attacker. After Arsenal and Bournemouth, Newcastle are tied for 3rd in total fantasy points from their forwards. It requires knowing who is actually going to start but if Rondon ends up getting a run of starts there is statistical evidence that it is likely to go better than the popular narrative would lead you to believe.
How Abject is Fulham’s Defense? – Just how bad have the Cottagers been at the back? West Ham have the second fewest aggregate points from their defenders at 103 over the first 12 matches of the season. Fulham have 40 points in aggregate from their defenders. That’s just obscene. At the same time, Fulham have conceded a whopping 702 fantasy points thus far with the next highest total being Southampton’s 637. That’s a over 5 points/match different between 19th and 20th place. Given how good Fulham have been to fantasy managers thus far, it is worth paying enormous attention to Claudio Ranieri’s progress. If we, as managers, lose such an obvious source of points, we will have to start thinking a lot harder about where to go for fringe players each week.
Rest assured that there will be much more to come as the data continues to come in and I have more time to work with it. To tide you over and allow you to draw some of your own conclusions, here is how the Points Against tables work out overall, vs Forwards, vs Midfielders, vs Defenders, and vs Goalkeepers after 12 weeks.
Points Against Table (Total/F/M/D/G)
- MCI 274 53 129 63 29
- LIV 317 41 167 79 30
- CHE 341 53 155 97 36
- TOT 371 66 148 124 33
- ARS 391 60 189 112 30
- EVE 419 88 183 111 37
- WAT 433 105 154 138 36
- LEI 434 71 187 138 38
- BOU 441 94 182 124 41
- WOL 460 50 213 151 46
- MUN 484 83 240 114 47
- BHA 515 67 237 171 40
- WHU 563 80 233 200 50
- NEW 578 58 228 230 62
- CRY 616 83 225 239 69
- BUR 622 110 248 217 47
- CAR 632 82 287 208 55
- HUD 635 117 221 238 59
- SOU 637 68 250 248 71
- FUL 702 117 264 268 53
Are there trends that equate Points Against with overall standing in the real-life table? Absolutely. Is the relationship as similar as the better the team, the fewer fantasy points conceded? Certainly not. When applied by position, is there even more nuance? Absolutely. Look for more in the weeks to come.
