My articles on player value (first, second and third) have highlighted that premium defenders seem to be the most efficient use of funds in FPL, with the it almost always being worth upgrading defenders to the highest scoring, regardless of price.
In this article I will look at whether this still holds true if we look at rotating pairs of cheaper defenders compared to their more expensive counterparts.
Reference Point
Starting with the defenders from the theoretically most efficient team we got:
DEF: Trent Alexander-Arnold | 7.0m | 210
DEF: Virgil Van Dijk | 6.5m | 178
DEF: Matt Doherty | 6.0m | 167
And the season before (2018/19) was:
DEF: Andrew Robertson | 6m | 213
DEF: Virgil van Dijk | 6m | 208
DEF: Aymeric Laporte | 5.5m | 177
DEF: Trent Alexander-Arnold | 5m | 185
And in 2017/18:
DEF: César Azpilicueta | 6.5m | 175
DEF: Marcos Alonso | 7.0m | 165
DEF: Nicolás Otamendi | 5.5m | 156
DEF: Ben Davies | 5.5m | 143
Effectively two rotating £4.5m defenders is equivalent to one £5.0m defender (presuming cheapest bench). If you can get 8 points in a season per £0.5m then an upgrade is usually worth taking (see discussion here) so the aim here is to see if rotating £4.5m can get over 151 – 161 points, at which point they might be worth considering.
Method
I’m going to only look at fairly nailed on defenders (played 60mins in at least two-thirds of their games), and then I’ll take the points per game (PPG) of the top two defenders (not from the same team), average them, then multiply by 38 to get a benchmark for rotating them across the whole season.
This presumes that one of the two will play every game and, to make it a fair fight we will increase the target to 155 – 165 to account for a couple of no-shows for the premium options, presuming a 2-pointer bench fodder coming off the bench in that scenario.
I’ll start with the basic method which gives you a benchmark for arbitrary rotation, then will add some parameters in to simulate rules about which defender you would prioritise.
Results
Random
If you managed to hit upon the two best £4.5m defenders and rotated them randomly, you would get Baldock (3.7ppg) and Dunk (3.6ppg) giving a score of 138.7.
Exclude Man City and Liverpool
The first adjustment is an obvious one – don’t play the defender who is playing Man City or Liverpool. This gives you Baldock (4.0ppg) and Soyuncu (3.9ppg) and a total score of 150.1.
Exclude Top 8
The second adjustment is to only play the defender who is not playing one of the top eight finishers from this season. Although we wouldn’t have known this at the start of the season, excluding the traditional big six plus Leicester and Wolves seems fairly logical.
The top two here come out as Baldock (4.2ppg) and Fernandez (4.1ppg) with a total of 157.7.
Home/Away
The next adjustment is to prioritise the home fixture. The best pair in this setup is Egan (4.5ppg) and Saiss (4.0ppg) giving a total of 161.5 points. This may, however, be slightly misleading as Egan is a full 0.5ppg ahead of anyone else with home form which seems to be a bit of an anomaly, so possibly taking Saiss and Fernandez (4.0ppg) for a total of 152.0 is more realistic.
Combined
If we combine some of the filters we get:
No City/Pool and home: 176.7 points
Not top 8 and home: 220.4 points
In Reality
The results show the most significant boost when not playing cheap defenders against top 8 teams, and if you add in priority for home fixtures the PPG, in theory, goes higher than the top-scoring premium defender. There are a couple of key problems with this.
The first is that finding a rotating pair that lets you always play a home fixture against a non-top-8 team is going to be pretty challenging. Home/away rotating pairs that avoid Man City and Liverpool, however, seems achievable.
The second is that the top two defenders are not going to be the ones that you can actually rotate successfully home and away, even for parts of a season.
Top x Defender Average
It is probably, then, better to look at the average of the top five defenders (again, maximum one from each team) to give a more realistic potential for rotating pairs.
No City/Pool and home: 167.2 points
Not top 8 and home: 193.8 points
What about if we just take the average of the best defender from each team, excluding the teams that would have been obvious to avoid fairly early on (Brighton, Norwich, Southampton, Watford, West Ham)?
No City/Pool and home: 159.6 points
Not top 8 and home: 182.4 points
Conclusion
After doing this exercise, I’m reconsidering rotating £4.5m defenders for one slot over premiums.
It looks fairly reasonable to get 160-180 from rotating pairs by prioritising home and avoiding City/Pool and, ideally, the rest of the big 8.
Please do leave any thoughts on the method, any flaws in the logic or other rules to try to improve PPG performance.
4 years, 1 month ago
Thank you Hedge, this is really good research.
I also prefer rotating 4.5 defenders with usually just 1 premium defender in my squad (if that) but surely this could be even more appetising if we're looking to spend tons on our midfield next season? Nothing like a bit of confirmation bias to make me happy with my plans 😉