Position review: Defenders (1 / 2)

[Part 2 focusing on low cost defenders can be found here]

As we discussed in the defender value post, I would guess the majority of managers reading this are aware of the UCL scoring wrinkle around "balls recovered" and how it can be the great equalizer in this game. However, while many of us know the names likely to contribute in the attacking third, we don't necessarily have an instinctive feel for who might do well in this relatively new category, so we'll first dig into some of that data, before getting into specific picks below.

First, a caveat. While the UCL site does show last season's balls recovered numbers, I cannot see a definition nor source for this data. Thiago Silva, for example, is shown to have recovered 69 balls in the Champions League last season but per FBRef he notched 99 "loose balls recovered". Similarly, Leipzig's Klosterman shows 116 at FBRef but only 79 at the UCL Fantasy site. Now the good news is that the two stats appear to correlate well so while we can't really use the FBRef data as a predictor of points, we can use it to see which players do well in this stat on a comparative basis. A second caveat is that FBRef (or any other open site I am aware of) does not offer data on all leagues and so we don't unfortunately have a complete data set, but it's a start. 

While potentially instructive to look at which players did well in the balls recovered category last season in the UCL game, we are risking small samples there with most players having played maybe 6-8 games, so instead let's look at the leaders in FBRef's Recov category within their respective domestic leagues. Of course, it's possible that teams lineup differently in Europe than in the domestic league, but with perhaps the exception of Ligue 1 the quality of opponent week-to-week in these leagues should ensure at least a reasonable level of comparability. We'll refer back to the below visualization as we start to look at some of the defender picks below.

The next factor to think about is, of course, the team's actual defensive value. There are so many good models out there these days I am not going to try and create my own at this stage, and will utilize 538's SPI rating, which I've found to be pretty solid in the past. Rather than look at the abstract concept of which team's have "good" defenses I have used their actual match projections which therefore account for the strength of opposition:

 

I am not really touching on defenders' attacking ability here as (a) we already determined that this is one of the harder pieces of the puzzle to pin down, (b) a lot of that value is already baked into player prices, and (c) we can just simply deal with it on a case-by-case rather than systemic basis in the previews below. 

The selections below are obviously not an all encompassing review of all players but I have tried to reach solid conclusions rather than just suggesting that "all these guys are good".

High end (6.0m+)

I will not guarantee that there is a team I will have two players from this season yet I know I will 100% have three players from one team: Bayern, and it may well end up being two elite defenders: Alphonso Davies (6.0m) and Josh Kimmich (6.5m). Let's start with last year's Champions League defender of the year. As noted in the earlier defender value piece, Kimmich has played in advanced roles this year, takes corners, and has averaged 0.79 xG + xA per game to kick off the Bundesliga season. For context, without penalties, Erling Håland has notched 0.8 per game over the same (albeit brief) period, with Jadon Sancho way back on 0.47. He will of course regress to a degree but even last season's 0.31 mark is extremely impressive, and not a million miles behind attacking midfield options like Sabitzer (0.42) and Perisic (0.45). The one thing he didn't truly excel at last season was recovering balls, but you will note above that in the early games of this season he's actually third among players from the big five leagues. 

If Kimmich isn't a must-own player it's probably because of Canadian teammate Alphonso Davies, who comes in a little cheaper but offers much of the same all around package. His ball recovery numbers are perhaps more reliably high, and his assist potential in open play is comparable. The lack of set pieces plus perhaps some chance of rotation (he didn't start last season as first choice remember) gives Kimmich the edge, but at this stage of the pre-season my intent is to have both.

In the defender value piece we noted that one can forgive a lower balls recovered number only when a player offers a truly exceptional attacking threat. Welcome Achraf Hakimi (6.0m). Familiar to UCL players for his four goal haul last season, Hakimi was surprisingly moved on to Inter in the summer and has hit the ground running at the San Siro in spectacular fashion. Again, small sample caveat, but in his three games to date his is averaging 0.91 xG + xA, sandwiched nicely just between Romelu Lukaku (1.0) and Cristiano Ronaldo (0.85). Again, regression will come, but this is someone who notched 5 goals and 10 assists in 33 Bundesliga games last season and has now joined a team that trails only Bayern and Liverpool in xG/90 to open the new season. The 538 model doesn't love Inter's defense but they are 11th in attacking threat, including two good fixtures to open the campaign putting Hakimi is a good spot to start strong before maybe switching him out before the back-to-back Real Madrid fixtures.

The 538 model doesn't love Liverpool's defense either (and this was before shipping seven to Villa), which coupled with their average balls recovered numbers last season, makes Robertson (6.5m) and Alexander-Arnold (6.5m) start to look a bit too steep. In FPL we have 38 games for them to capitalize on all those crosses and set piece opportunities but in a short schedule against quality opponents I don't see as much value here. Teammate Virgil Van Dijk (6.0m) looks perhaps the pick of the Liverpool bunch with a lower price tag, superior ball recovery numbers and a decent threat from set pieces, but it's still hard to see him in the same class as the names mentioned above right now. The main appeal is that FC Midtjylland game in MD2 but otherwise the opponents look tough here (Atalanta and Ajax round out Group D).

Sergio Ramos (6.5m) has of course enjoyed an incredible goal scoring record over the years (including 11 in the league last season), and seems to be continuing with penalty duties having taken the spot kick against Betis a couple of weeks ago. His non-penalty xG+xA last season was fine (0.18) but nothing to be overly excited about, so the one million premium you are paying over teammates Varane, Carvajal, and Mendy is really tied to those spot kicks. With the way hand ball decisions have been going, one is tempted to back any penalty taker we can find, but it's perhaps worth waiting a couple of weeks to see how renewed guidance in England reverberates across Europe (if at all). Unfortunately Real's best fixture is MD1 and if you want to benefit from penalties, Ramos is maybe someone you need to just pick for the whole group stage and hope the crazy calls continue through this crazy year.

For completeness, despite a likelihood of playing further upfield I don't think we know enough about Marc Roca (6.0m) to consider him over Davies, while Alba (6.0m) and Guerreiro (6.0m) are both fine players, but don't seem to do enough for me to justify the extra cost above teammates such as Pique (5.5m) or Hummels (5.5m). 

This one ended up running longer than planned, so I will split this in two and wrap up the defender review tomorrow.