clock menu more-arrow no yes

Filed under:

Mathematical Musings: Optimizing Your Official BPL Line-Ups by Oppositional Targeting

New, comments

We all have a core group of match-up proof players who we start regardless of the opponent. However, that leaves us with several tough choices every week regarding who to play in the starting line-up, which players to put on the bench and who to captain. These ranking charts utilizing Official BPL scoring comprising the first half of the season will help you make those difficult decisions from here on out in the second half.

An old adage: Never get involved in a land war in Asia. A modern one: Bench your attacking players when visiting White Hart Lane.
An old adage: Never get involved in a land war in Asia. A modern one: Bench your attacking players when visiting White Hart Lane.
Matthew Lewis/Getty Images

There is a multitude of ways you can choose your starting line-ups and captains. You could pick names out of a hat, but then again, you probably wouldn't be visiting and interacting with our handy dandy NMA blog site if you wanted to rely purely on chance. You can patronize your local fortune teller, but of course that might leave your wallet significantly lighter as you attempt to find the basement of the Alamo. Or, in what is probably a better bet, you could actually take a gander at something a bit more concrete. Say, for instance, data!

A good, albeit simple way to start is by looking at the standings for which teams allow the most goals. After all, Official BPL scoring relies heavily on goals scored (and resultantly, assists made) for strikers and midfielders, whereas minimizing goals allowed (primarily via clean sheets) is a huge part of scoring for defenders and keepers. But just how effective is that, and is there a better way?

Actual Premier League Scoring Ranks vs. Official BPL Fantasy Scoring Ranks

Strikers/Midfielders

The following chart looks at how actual EPL team scoring stacks up to Official BPL fantasy scoring (for Strikers/Midfielders only for the latter) through Game Week 19, which covers the first half of the season in which every team played each other team once.

Actual EPL Rank Team Most Goals Allowed/Gm Official BPL Fantasy Rank Team Most Pts Allowed/Player/Gm
1 Sunderland 2.00 1 Sunderland 4.47
2(T) Aston Villa 1.79 2 Aston Villa 4.30
2(T) Bournemouth 1.79 3 Newcastle 4.10
2(T) Newcastle 1.79 4 Bournemouth 4.00
5 Norwich 1.68 5 Norwich 3.93
6 Chelsea 1.53 6 Swansea 3.75
7 Everton 1.47 7 West Brom 3.63
8 Leicester City 1.32 8 Everton 3.51
9(T) Swansea 1.26 9 Southampton 3.50
9(T) West Brom 1.26 10 Chelsea 3.49
11(T) Southampton 1.21 11(T) Man. Utd. 3.20
11(T) West Ham 1.21 11(T) Man. City 3.20
13 Liverpool 1.16 13 West Ham 3.12
14(T) Man. City 1.05 14 Leicester City 3.10
14(T) Watford 1.05 15 Stoke 3.06
16 Stoke 1.00 16 Watford 3.04
17 Arsenal 0.95 17 Liverpool 2.91
18(T) Crystal Palace 0.84 18 Crystal Palace 2.85
18(T) Man. Utd. 0.84 19 Arsenal 2.77
20 Tottenham 0.79 20 Tottenham 2.66

Note: I chose to compile the Official BPL fantasy scoring numbers from starters only. For one thing, this puts per player per game averages on a relatively equal playing field, so to speak. Some teams regularly use more substitutes than others, which would otherwise skew averages. For another, most of the players that we are interested in for our fantasy teams will be expected to start the vast majority of games, and even if an important player unexpectedly comes off the bench rather than starting, he will usually make little impact and will not score well in fantasy for that game. In fact, if we have a player in our starting line-up who does not actually start, we would probably prefer that player to remain out of the game so that another player from our fantasy bench who actually started can slot in and score points for us.

The correlation is somewhat strong, but far from perfect. Several teams are very close, particularly at the top and the bottom, but there are some notable disparities. Leicester City gives up the 8th most goals in real life but allows only the 14th highest average to midfielders and strikers in fantasy, while Manchester United goes the other way, allowing the second fewest goals (tied) but giving up the 11th most points (again, tied) in fantasy, while there is also a noticeable gap for Liverpool (#13 vs. #17) and Chelsea (#6 vs #10).

How about on the other end? The following chart looks at how actual EPL team goals allowed ranks stack up to those for Official BPL fantasy scoring (for Goalkeepers/Defenders only in the latter) through Game Week 19.

Goalkeepers/Defenders
Actual EPL Rank Team Fewest Goals/Gm Official BPL Fantasy Rank Team Most Pts Allowed/Player/Gm
1 Aston Villa 0.79 1 Aston Villa 4.02
2 Swansea 0.84 2 Newcastle 3.87
3 West Brom 0.95 3(T) Swansea 3.81
4(T) Newcastle 1.00 3(T) Sunderland 3.81
4(T) Sunderland 1.00 5 Stoke 3.69
6 Stoke 1.05 6 Crystal Palace 3.68
7(T) Bournemouth 1.16 7 Liverpool 3.63
7(T) Liverpool 1.16 8 Bournemouth 3.44
7(T) Man. Utd. 1.16 9 West Brom 3.42
7(T) Norwich 1.16 10 Norwich 3.40
11(T) Chelsea 1.21 11 Man. Utd. 3.34
11(T) Crystal Palace 1.21 12 Southampton 3.33
13 Watford 1.26 13 Watford 3.04
14 Southampton 1.37 14 Chelsea 2.93
15 West Ham 1.47 15 Arsenal 2.76
16(T) Arsenal 1.74 16 Everton 2.72
16(T) Tottenham 1.74 17 West Ham 2.64
18 Everton 1.84 18 Man. City 2.58
19(T) Leicester City 1.95 19 Tottenham 2.45
19(T) Man. City 1.95 20 Leicester City 2.05

The situation is similar on the defensive end, where there is a decent correlation overall but there are several gaps. West Brom has scored the third fewest goals but has allowed just the 9th most fantasy points to defenders and keepers, while Crystal Palace comes out the other way, scoring the 11th fewest goals (tied) but allowing the 6th most points. Meanwhile there is also a gap for Manchester United (#7T vs. #11).

There may be several reasons for the disparities. To start with, there is obviously more to fantasy scoring than merely goals and assists, as yellow and red cards, own goals, penalty kicks missed/saved, clean sheets, saves, bonus points, etc. also come into play. Another thing, there are variances between teams regarding the percentage of assisted goals compared to unassisted goals (such as penalty kicks or broken plays) scored or allowed, as well as own goals which not only have no assist credited, but no goalscorer on the offensive end. Additionally, some teams may have personnel and formations which influence opposing teams to play different formations than they usually do, utilizing fewer or more strikers, midfielders or defenders, and thus introducing inferior players who are not part of their favored first 11.

As well, certain teams may play in such a way as to impact playing time for opposing squads. If a team tends to play with a hectic, energetic style or get out to early, comfortable leads, this may influence the other team to make substitutions earlier than otherwise, diminishing the scoring potential for its starters. There is a small clean sheet bonus for midfielders, rewarding not just offense but defense. Similarly, going the other way, defenders score goals and get assists, too! Finally, there is not a pure correlation between goals scored (or allowed) and clean sheets earned (or suffered). As well, of course, many goals are scored by substitutes (again, who you are generally not interested in having in your fantasy squads anyway) or allowed after your player may have left the field.

On top of that, naturally we can expect that there will be variance in teams' balance between striker and midfield scoring on offense, as well as between keeper and defender scoring on the other end. So, if you are looking for an accurate predictor of fantasy points, you will need to look beyond the simple goals scored/allowed table and go directly to the source: fantasy points themselves. I know, big shocker! Additionally, it should be helpful to do so with a mind to home and away splits since not only do teams generally perform much more poorly away than at home, there is a measure of variance among teams in their difference between home and away performance.

Positional Targeting of Opposition with Home and Away Splits

Strikers
Rank Team Pts Allowed/Striker/Gm (at Home) Rank Team Pts Allowed/Striker/Gm (Away)
1 Southampton 5.27 1 Newcastle 6.67
2 Newcastle 5.17 2 Sunderland 6.00
3(T) Bournemouth 4.57 3 Aston Villa 5.62
3(T) Everton 4.57 4 Norwich 5.59
5 West Brom 4.54 5(T) Bournemouth 5.00
6 Chelsea 4.50 5(T) West Ham 5.00
7 Man. City 4.25 7 Man. Utd. 4.75
8 West Ham 4.09 8 Swansea 4.29
9(T) Aston Villa 4.00 9 Liverpool 4.00
9(T) Liverpool 4.00 10 Crystal Palace 3.69
11 Norwich 3.92 11 Watford 3.64
12 Stoke 3.67 12 Chelsea 3.56
13 Swansea 3.14 13 Arsenal 3.46
14 Leicester City 3.09 14 Leicester City 3.13
15 Sunderland 3.00 15 Man. City 3.00
16 Man. Utd. 2.83 16 Tottenham 2.77
17 Arsenal 2.77 17 Everton 2.58
18 Crystal Palace 2.41 18 Stoke 2.47
19 Watford 2.25 19 West Brom 2.45
20 Tottenham 2.20 20 Southampton 2.31

Not surprisingly, you should not be shy about starting your strikers when playing Newcastle or Bournemouth, whether home or away. More interestingly, though, there are some unequal home/away splits with many other teams such as Sunderland, which is much tougher to crack at home; meanwhile, Southampton and Everton are easier for strikers to smash when traveling.

Midfielders
Rank Team Pts Allowed/Midfielder/Gm (at Home) Rank Team Pts Allowed/Midfielder/Gm(Away)
1 Aston Villa 4.05 1 Sunderland 5.28
2 West Brom 3.87 2 Newcastle 4.56
3 Everton 3.86 3(T) Aston Villa 4.18
4 Swansea 3.83 3(T) Bournemouth 4.18
5 Sunderland 3.33 5 Norwich 4.16
6 Bournemouth 3.23 6 Southampton 3.83
7 Leicester City 3.18 7 Chelsea 3.80
8 Watford 3.10 8 Swansea 3.61
9 Stoke 3.00 9 Man. City 3.54
10 Norwich 2.98 10 Man. Utd. 3.53
11 Southampton 2.93 11 West Brom 3.34
12 Chelsea 2.92 12 Stoke 3.16
13 Newcastle 2.82 13 Crystal Palace 3.06
14 Man. City 2.65 14(T) Everton 2.98
15 Arsenal 2.63 14(T) Leicester City 2.98
16 Crystal Palace 2.49 16 Tottenham 2.95
17 Man. Utd. 2.48 17 West Ham 2.91
18 West Ham 2.45 18 Liverpool 2.90
19 Liverpool 2.43 19 Watford 2.86
20 Tottenham 2.42 20 Arsenal 2.71

Aston Villa, Bournemouth and Sunderland are fodder for midfielders whether home or away, while you may want to rest your playmakers when going up against Arsenal or West Ham.  Interestingly, Manchester City and Manchester United, teams which do not give up many goals, are both in the top 10 in terms of allowing fantasy points to midfielders on the road.

Defenders
Rank Team Pts Allowed/Defender/Gm (at Home) Rank Team Pts Allowed/Defender/Gm (Away)
1 Sunderland 4.39 1 Man. City 4.11
2 Aston Villa 4.30 2 Newcastle 4.06
3 Swansea 3.78 3 Norwich 4.00
4 Crystal Palace 3.75 4 Stoke 3.98
5 Newcastle 3.46 5 Chelsea 3.91
6 Stoke 3.35 6 Liverpool 3.79
7 Man. Utd. 3.34 7 Swansea 3.75
8 Bournemouth 3.33 8 West Brom 3.69
9 Watford 3.18 9 Aston Villa 3.64
10 Liverpool 3.14 10 Crystal Palace 3.48
11 Southampton 3.08 11 Bournemouth 3.28
12 West Brom 3.05 12 Sunderland 3.26
13 Norwich 2.86 13 Southampton 3.23
14 Everton 2.45 14 Man. Utd. 2.95
15 Tottenham 2.40 15 Everton 2.94
16 West Ham 2.32 16 Watford 2.84
17 Arsenal 2.29 17 Arsenal 2.83
18 Leicester City 1.90 18 West Ham 2.73
19 Chelsea 1.80 19 Leicester City 2.09
20 Man. City 0.80 20 Tottenham 1.80

You are probably shocked seeing Manchester City atop the list of giving up fantasy points to defenders when away from the Etihad. Then again, seeing that City has been shut out at Manchester United, Aston Villa, Stoke and Leicester City, while also allowing goals to Tottenham defenders Toby Alderweireld and Eric Dier at White Hart Lane, and that notion becomes less fanciful. After all, managers tend to adopt a conservative, defensive-minded approach when Manchester City comes to town, something that does not happen when entertaining Newcastle or Norwich. City is truly a Dr. Jeckyll and Mr. Hyde team, giving up the fewest points to defenders at home. The lesson, feel free to start your defenders at home against the Citizens, but leave them on the bench when traveling to the blue side of Manchester.

As for Leicester City, it scored in each of the first 17 game weeks, and frequently so at that, making things tough on its opponents' defenders. Of course, the well has dried up since then, and it will be interesting to see if can hold its position on the chart going forward.

Keepers
Rank Team Pts Allowed/Keeper/Gm (at Home) Rank Team Pts Allowed/Keeper/Gm (Away)
1(T) Bournemouth 4.78 1 Man. City 5.11
1(T) Crystal Palace 4.78 2 Newcastle 5.00
3 Tottenham 4.40 3 Liverpool 4.70
4 Man. Utd. 4.11 4 Southampton 4.56
5(T) Aston Villa 4.00 5 Chelsea 4.33
5(T) Sunderland 4.00 6 West Brom 4.22
7 Swansea 3.90 7 Swansea 4.11
8 Newcastle 3.80 8 Stoke 4.00
9 Watford 3.78 9(T) Aston Villa 3.60
10 Arsenal 3.67 9(T) Man. Utd. 3.60
11 Southampton 3.50 9(T) Norwich 3.60
12 Liverpool 3.33 9(T) Sunderland 3.60
13 Stoke 3.33 13(T) Bournemouth 3.30
14 West Brom 3.10 13(T) Crystal Palace 3.30
15 Chelsea 2.90 13(T) West Ham 3.30
16 Norwich 2.89 16 Everton 3.11
17 West Ham 2.78 17 Tottenham 3.00
18 Everton 2.70 18(T) Arsenal 2.80
19 Leicester City 2.20 18(T) Watford 2.80
20 Man. City 1.90 20 Leicester City 2.44

You may have noticed that Chelsea has had a horrible season so far. Even so, it has managed to stymie defenders and goalkeepers at home, thanks to avoiding clean sheets at Stamford Bridge for the most part, though away has been an entirely different story. As for what stands out on the road table, avoid Watford when it travels away from Vicarage Road and generally ends up playing in a more open style.

Choosing a Captain

While the decisions regarding the rotational starting positions is difficult, the most excruciating choice each week is probably the choice of captain. It's a huge difference between doubling a great performance around 15 points and a two point dud. Since bonus points are awarded to the game's best players and provide an very nice boon with doubling, a good guide for captaincy choice would be through looking at teams that have given up the most bonus points by position. There are not enough data to justify using home and away splits or individual averages, so the table combines all games and positional totals. Generally we tend to choose strikers or midfielders, and rightly so since they tend to score the most points, so let's start there.

Strikers/Midfielders
Rank Team Bonus Pts Allowed to Strikers Rank Team Bonus Pts Allowed to Midfielders
1(T) Aston Villa 21 1 Aston Villa 40
1(T) Norwich 21 2(T) Sunderland 36
3 Sunderland 17 2(T) West Brom 36
4(T) Bournemouth 16 4 Bournemouth 33
4(T) Southampton 16 5 Norwich 28
6 West Ham 16 6 Chelsea 26
7 Man. Utd. 14 7 Everton 24
8(T) Chelsea 13 8 Swansea 23
8(T) Newcastle 13 9 Newcastle 22
8(T) Swansea 13 10 Southampton 20
11(T) Arsenal 10 11 Watford 19
11(T) Crystal Palace 10 12 Stoke 18
11(T) Man. City 10 13(T) Crystal Palace 16
11(T) West Brom 10 13(T) Leicester City 16
15(T) Everton 8 13(T) Man. Utd. 16
15(T) Liverpool 8 13(T) West Ham 16
15(T) Watford 8 17(T) Man. City 15
18 Leicester City 6 17(T) Tottenham 15
19 Stoke 2 19 Liverpool 13
20 Tottenham 0 20 Arsenal 8

It's mostly the usual suspects up there, but there are some interesting disparities. For instance, if you are targeting West Brom, do so with a midfielder, not a striker, whereas the opposite holds true for Southampton.

If you are feeling bold and want to captain a defender or keeper, here's the opposition you should have an eye on.

Defenders/Keepers
Rank Team Bonus Pts Allowed to Defenders Rank Team Bonus Pts Allowed to Keepers
1(T) Aston Villa 33 1 Southampton 9
1(T) Sunderland 33 2 Newcastle 7
3 Crystal Palace 29 3(T) Liverpool 6
4(T) Newcastle 27 3(T) Tottenham 6
4(T) Stoke 27 5 Bournemouth 5
6 Swansea 25 6(T) Chelsea 4
7 West Brom 23 6(T) Crystal Palace 4
8(T) Norwich 22 6(T) Man. City 4
8(T) West Ham 22 6(T) Man. Utd. 4
10(T) Liverpool 20 10(T) Arsenal 3
10(T) Southampton 20 10(T) Stoke 3
12(T) Everton 18 10(T) Sunderland 3
12(T) Man. Utd. 18 10(T) Watford 3
14 Watford 16 10(T) West Ham 3
15(T) Bournemouth 15 15(T) Aston Villa 2
15(T) Chelsea 15 15(T) Leicester City 2
17 Arsenal 14 15(T) Swansea 2
18 Leicester City 13 18 Norwich 1
19(T) Man. City 10 19(T) Everton 0
19(T) Tottenham 10 19(T) West Brom 0

A key takeaway here is that likely relegation fodder Aston Villa and Sunderland are worthy opponent choices for targeting when selecting your captain for strikers, midfielders and defenders, but not for keepers who clearly are not kept busy enough making saves in their own nets. Meanwhile, Newcastle offers an appetizing captain target when it comes to your defenders and keepers, but is merely middling when it comes to strikers or midfielders.

What do you think? Is this going to impact how you select your starting line-ups and bench, as well as your captain? Is there anything in the charts that surprised you? Join in the discussion!