Beware the 'defcon' MIRAGE...!

 

I have a problem with the new 'defensive points', introduced into the Fantasy Premier League game for the first time last year. I went into some detail on this when the change was first announced; but here's a brief recap - and an expansion.

Why chasing 'defcon points' is probably misguided

1)  They're opaque.  

FPL has still provided no definitions or illustrations of how the various eligible actions are identified (and distinguished from each other!). And there are so many different game actions that might fall under one or more of the eligible categories that it is simply impossible to know from watching a game whether a player of ours has reached the necessary threshold of 'defensive contributions' to earn us extra points or not. We have to take it entirely on trust from Opta. [It is often as arbitrary and mystifying as the operation of the widely reviled 'Bonus Points System' in the game.]

2)  They're unreliable.  

Opta is extremely fallible: they do frequently miscount things. And with these 'defensive points' a miscount of one can make the difference between your guy getting 2 extra points (perhaps unjustly) or getting 0 points (perhaps unjustly).

3) They're unduly sensitive to fine margins of difference in performance (and margins of error in the counting!), and hence very unpredictable. 

Because the thresholds for earning the points are set so high, and being deemed to have come up even one contribution short of that mark means you get no additional points (while you cannot earn any further points beyond the basic 2 points, even if you exceed the threshold by 2x or 3x!!), it is impossible to predict whether a player is likely to attain the points in any particular game; much of the time, the best 'defcon' performers are hitting a number around their 'season average' - which is slightly below the required threshold to earn the points.

4)  The extreme sensitivity of the eligibilty threshold referred to above also makes it difficult to infer with any confidence how players' performance on these points will hold up from one year to the next.

It is impossible to rely on how much consistency players will achieve from one season to another in earning 'defensive points': a player might produce a similarly high level of consistent defensive solidity in his performances, and achieve an almost identical total 'defcon' tally for this season as he did for last, yet come up with a vastly lower return of 'defensive points' - simply because he had unluckily come up just short of the threshold in several matches this time around.

5)  They tend not to be earned in combination with points for anything else. 

Players usually only earn 'defensive points' in games where they're being put under a lot of pressure, where their team is generally having the worst of things. In games like that, they're unlikely to be keeping clean sheets, or featuring well in the Bonus Points - or coming up with any attacking contributions. 'Defensive points' can be a very nice consolation in a game in which your player's team take a bit of a battering; but a good 'defcon' number is rarely allied with a really good weekly points return. (And so, since you're usually rotating out your weakest 1 or 2 defenders, and sometimes your weakest midfielder - you probably often won't be playing your 'defcon monster' picks anyway: because the games that give them the best prospect of earning the 'defensive points' are games in which you don't fancy them to pick up more than 4 points, and the rest of your squad really should be able to do better than that.)

6)  'Defensive contributions' don't register very strongly under the BPS.

You get massively more credit under the 'Bonus Points System' for overall game involvement: a defensive player who gets on the ball a lot and registers a high percentage for pass completion will almost always out-rank one who just has a big day on the 'defensive contributions'. The 'defcons' may occasionally have enough of an impact to raise a player into the bonus points (though, usually, just a single one), in a sterile, low-scoing game where nobody achieves much of a BPS score; but that's only going to amount to a handful of extra points a season, at best. 

7)  The impact of the 'defensive points' last season was massively amplified by other - freakish! - factors.

Last season was - if you allow for the 'defensive points' (probably worth 100-150 points for most people) and the new additional chips (perhaps worth an extra 10-50 points together) - one of the lowest-scoring in the history of the game; and the lowest-scoring ever for midfielders: almost all of the usual highest points-producers - Palmer, Saka, Mbeumo, Cunha, Maddison, Wirtz, Gakpo, Maddison, Mitoma, etc., etc. - suffered seasons blighted by patchy, disppointing form and/or repeated or long-term injuries. Thus, the fact that so many central defensive midfielders finished fairly near the top of the midfield rankings last year, largely as a result of their 'defensive points', is unlikely to be repeated. In fact, only Rice and and Anderson crept into the Top Five midfielders (and Casemiro and and Garner were just in the Top Ten), so - that's still not heavy representation in the upper rankings. And in a more 'normal' season, you'd expect that at least 10 or 12 of the more prolific goalscorers and creators in the position category would comfortably outscore all the defensive midfelders. (Moreover, all of those top CDM performers last year had their ranking buoyed by an unusual number of attacking contributions as well; that's also unlikely to be repeated this year. And a key part of their appeal was their overall consistency, a steady drip-drip-drip of points through most of the season. This means that they were almost always being outperformed over shorter runs of games by several midfielders on a short-lived streak of good attacking form, and thus never really deserved a place in the squad - even though their end-season totals were quite impressive.)

So,.... I'm inclined to think that the allure of the 'defensive points' is a dangerous distraction.

The online forums at the moment are full of people asking who are the best 'defcon' players?? 

As I said above, we can't necessarily have very high confidence that any of last season's best 'defcon' players will get the same return on these points this year, even if their level of performance is identical. (And there are other issues too: Van Dijk and Tarkowski are getting old, Senesi and Lacroix and Anderson have changed clubs, Garner is injured going into the start of the season, and... Crystal Palace and their previously impressive defence might fall apart without Oliver Glasner.)

And there are a few players who return pretty well on the 'defensive contributions' but offer little prospect of regular points for anything else. Some of the most consistent producers of these points are from weaker teams, who are regularly taking a battering; you generally want to stay away from teams that are going to be foundering down in the bottom third of the table.

Is is best to target defensive players from teams who concede fairly few goals. Teams who concede few goals will also pick up a decent number of clean sheets. No-one this season is going to get anywhere near Arsenal's remarkable haul of 19 clean sheets last time (not even Arsenal!!); but even 7 or 8 or 9 over the season is worth having. And for defenders, it's crucial to escape the slow attrition of shedding points as more goals are conceded; you have to try to avoid players who often take a pasting (this is why gambling on defenders from the promoted clubs, just because they're so cheap, very rarely pays off.)

And because they're 'in games' a lot of the time, not constantly being overwhelmed and only fighting rearguard actions, these sorts of teams have better prospects of picking up an above-average number of attacking contributions too (Sunderland's defenders did particularly well on this last season, for example). The best defenders at teams like this will probably be posting good 'defcon' numbers as well - but they'll be picking up plenty of points for other things too. And those other points are likely to be more reliable and more predictable.  So, that's what your main focus should be on.  'Defensive points' are nice to have, but they're a collateral benefit of being an all-around 'good defensive prospect' in FPL - not the main deciding factor for the selection. 'Defcon' potential can break a tie between two players who appear to you to be inseparable on every other metric; but if you choose a player for their 'defcon' record alone, there are almost certainly going to be alternative picks who might fare slightly less well on the 'defcon' numbers but are likely to do much better overall.