The Magnificent Seven Captaincy Theory

Intrigued by a recent Hot Topic from Net Gains outlining a captain strategy, I decided to take a closer look at the best way to assess candidates for the armband.

The majority of the data was close to hand anyway, it was just a matter of teasing it out (and giving a few tweaks to minimise my workload).

The Magnificent Seven Captaincy Theory

As a recap, the theory was to find the attacking player that does the best across seven equally weighted categories:

• Points per minute (PPM)
• PPM @ Fixture
• PPM @ Fixture for last three games (i.e. if next game is at home, PPM from last three home games)
• % involvement in team goals (incl. assists)
• Team goals scored @ relevant fixture for last three games
• Opposition goals conceded @ relevant fixture for last three games
• Count of unblanked games (i.e. games where the player got Goals/Assists/Bonus points)

Some considerations/tweaks:

1. We are only considering attacking players so defenders are out.
2. I much prefer to multiply PPM by 90 to get more normal looking numbers. Sergio Aguero for instance gets 0.066 PPM, which can be multiplied up to give 5.95 PP90M.
3. When considering PPM or PP90M, I only count ‘GAB’ points – those from Goals/Assists/Bonus. I find including minutes/cards/own goals to detract from the overall picture.
4. To make this interesting for people other than myself, I want to expand beyond just looking at my own eight attackers, and instead look at all the possible captains.

NET GAINS METHOD

The challenge is in this last point. If we stick purely with the rules as set out by Net Gains, then the winner of each category is as follows:

• PP90M = Heung-Min Son (8.77)
• PP90M @ Venue = Nacer Chadli (23.1)
• PP90M @ VenueX3 = Nacer Chadli (23.1)
• % Involvement = Jermain Defoe (44%)
• Team goals scored @ VenueX3 = Liverpool/City players (3)
• Opposition goals conceded @ VenueX3 = Chelsea players – facing Leicester (3.33)
• Unblanked games = Diego Costa (6)

Therefore Net Gains’ system picks out Chadli and Costa as the winner of two categories each. Chadli has eight GAB points from the last two weeks, Costa has 10, so Costa is the pick. In joint third place are Son, Defoe and all the Liverpool/City attackers.

AN ALTERNATIVE METHOD

An alternative method is to be less clear-cut than Net Gains had suggested (winner of each category is the only player to score a point). To get around this, for each category I have divided the players score by the maximum score in that category. Therefore the ‘winner’ of each category gets one point, and everyone else gets somewhere between 0 and 1 depending on how close they are to the maximum score.

For example, Heung-Min Son got the highest PP90M score of 8.77, so he gets one point. Nacer Chadli is in second place, with 7.06 PP90M. 7.06 divided by 8.77 equals 0.81, so Chadli gets 0.81 points for the PP90M category.

So having done this, we can give all candidates a score between 0 and seven. If a player was top in every category he would get the full seven points.

Gameweek 8’s Top Picks

The top 10 scorers with this method are as follows:

1. Nacer Chadli – 4.60
2. Heung-Min Son – 4.59
3. Diego Costa – 4.34
4. Philippe Coutinho – 4.01
5. Raheem Sterling – 3.87
6. Adam Lallana – 3.81
7. Kevin De Bruyne – 3.80
8. James Milner – 3.71
9. Alexis Sanchez – 3.41
10. Jermain Defoe – 3.40

So once again, Chadli is listed as top scorer. At this point I feel obliged to pour an ocean of salt onto the exercise – Nacer Chadli’s next game is at home (to Spurs), and he has only played one game at home so far, so essentially is being expected to continue to get two goals and two assists every home match, which obviously will not happen. This is the curse of the small sample size unfortunately. There are ways to mitigate this, such as not being as focussed on the fixture venue, though even then Chadli’s total PP90M is second in the rankings, so his strong performances in limited minutes really skews the outcome in his favour. Certainly he may be worth a look, and at just 2.8% ownership, he could be a great differential. Though his fixtures do look horrendous.

An additional pile of salt comes from the fact that the last three home games is the same (at this stage of the season) as the total home games. So a good result in that category is essentially being doubled, when I personally would not be inclined to consider returns from the specific home/away venue to be as crucial a factor. As the season goes on, this will be minimised.

What about the big names that did not make the top 10? The three most owned players all missed the cut, as did a few other high ownership players that you may have expected (or hoped) to feature:

• Zlatan Ibrahimovic – 2.61 (28th overall)
• Sergio Aguero – 3.01 (21st overall)
• Etienne Capoue – 3.36 (11th overall)
• Romelu Lukaku – 3.07 (18th overall)
• Eden Hazard – 3.34 (12th overall)
• Michail Antonio – 3.02 (20th overall)
• Sadio Mane – 3.20 (15th overall)
• Dimitri Payet – 3.16 (16th overall)
• Roberto Firmino – 3.29 (14th overall)

CONCLUSION

Nacer Chadli has great PPM, and ridiculous PPM at home. He’s got a good involvement in his team’s goals, and his team scores a respectable amount (1.66 per game for the last 3). So it’s understandable that he’s tipped strongly. But I don’t believe that all these categories should be considered equal, and I won’t captain him myself (not just because he’s not in my team). Heung-Min Son‘s stellar away record should be considered despite facing a Pulis team. Diego Costa has gone up in my estimation, with returns in six matches and facing a Leicester team with a poor defensive record away from home – a factor I was previously unaware of. So all three players highlighted by the Net Gains theory are well worth a closer look – I do think there may be some merit in this, though I’m sure more improvements can be made.

Further Statistics

For a closer look at the statistics click here.