20 August 2013

On Aly Cissokho


Oh wait, that's a different piece.

While every internet denizen was shouting about Spurs ostensibly pipping Liverpool to another attacking target, Liverpool actually finally filled one of their positions of need. Left back Aly Cissokho signed on a one-year loan, reportedly with an option to buy for €5m at the end of the season – the same price that Valencia paid for him one year ago.

As with most transfers from foreign leagues, I'm far more reliant on statistics rather than a firsthand appraisal; I remember Cissokho played well in both meetings against Liverpool in the 2009-10 Champions League, especially the 1-2 loss at Anfield when up against Kelly and Kuyt, but that's about it. So stats it is, all of which are from WhoScored.

Jose Enrique has been a very good defender for Liverpool. His defending isn't much of a concern; it's his ability in attack. All too often, he's a bull in a china shop in the final third, and even more notably, often loses possession during sustained Liverpool attacks, whether through a poor touch or misplaced pass.

That's what Liverpool need to improve on, so crucial to the way Rodgers wants his sides to play. So let's create a "lost possession" metric.

Lost possession = Times dispossessed + Turnovers + Unsuccessful passes + Unsuccessful crosses

Enrique played 2257 minutes in 29 Premiership appearances last season. He was dispossessed 56 times, committed 10 turnovers, made 304 unsuccessful passes, and attempted 61 unsuccessful crosses. By the above, admittedly somewhat facile, metric, Enrique lost possession once every 5.1 minutes.

Cissokho played 1684 minutes in 25 La Liga appearances for Valencia last season. He was dispossessed three times, committed 10 turnovers, made 119 unsuccessful passes, and attempted 67 unsuccessful crosses. Cissokho lost possession once every 8.5 minutes.

In handy table form:

For comparison, Luis Suarez lost possession once every 5.9 minutes: dispossessed 94 times, turning the ball over 79 times, attempting 282 unsuccessful passes, and attempting 42 unsuccessful crosses. Yes, in last season's league matches, Jose Enrique lost possession more often than Luis Suarez. By this metric, more than any other Liverpool player.

I wish I had the XY data to know where these possession losses took place, but it's safe to assume that a fair amount of Enrique's took place in the opposition half. I have to believe that's the main thing Rodgers wants to improve by signing competition at left back.

Enrique completed more key passes and through balls than Cissokho, both in total and per minute, but not a drastic amount more. Enrique created a chance once every 86.8 minutes, Cissokho once every 112.3 minutes in a less creative side. Liverpool created 141 more chances than Valencia did last season, so it's little surprise that Cissokho created 11 fewer than Enrique in 600 fewer minutes.

In attack, Cissokho seems far more limited to the touchline, while Enrique creates chances from all over the left half of the pitch (including a couple from inside Liverpool's half). Which may and may not be a bad thing. Rodgers prefers his players to be multifunctional, to join in attacks from all angles – which Enrique seems far better at – but at the same time, Cissokho staying wide more often than not can provide more space for Coutinho, Sturridge, Suarez, Aspas, Sterling, et al to drift around the final third. Anyone's who's likely to play on the left for Liverpool likes to cut inside from the flank.

Eight of Cissokho's 15 chances created (and both of his assists) came in the last seven matches of the season. 19 of Enrique's 26 chances came in the first half of the season, creating just three in the last seven matches.

Incidentally, only one of Enrique's chances created came in Liverpool's biggest routs – setting up the Suarez injury time goal in the 4-0 win over Fulham. He didn't play in the 5-2 or 5-0 wins against Norwich, and didn't create any chances in the 5-0 win over Swansea, 4-0 win at Wigan, or 6-0 win at Newcastle.

Seven of Cissokho's 15 chances created came from crosses, including both assists, while only two of Enrique's did. In total, Cissokho completed 19 of 86 crosses (22.1%), while Enrique completed seven of 68 crosses (10.3%). Admittedly, Liverpool don't rely on many crosses, not anymore (thankfully), but it's a nice option to have, as Enrique's exceedingly poor at them.

Overall, Cissohko had a better overall pass success rate than Enrique as well: 79.2% to 76.7%. On average, Cissokho completed a pass once every 3.7 minutes, Enrique once every 2.3 minutes, but that's another stat that is partly explained by the two teams' respective playing styles.

It's not a priority for left backs, but Cissokho was more much judicious and much more accurate in his shooting last season. He only took nine shots, hitting the target with 56% of them, scoring two goals. Enrique took 22, hitting the target with 36% of them, also scoring twice. Eight of Enrique's shots came further from goal than any of Cissokho's attempts.

By most defensive statistics, Enrique was better than Cissokho last season. Both averaged the same amount of interceptions per match – 1.2 – but Enrique made 2.5 tackles per match to Cissokho's 1.8, and committed just about half as many fouls. I can't find the relevant statistic on any of the available sites, but it's safe to assume that Enrique shielded the ball into touch more often than Cissokho as well, given that Enrique's one of the best in the league at it.

But one defensive statistic is heavily in Cissokho's favor. And it's one that's very important to Liverpool given the club's weakness in this area. Cissokho won 34 of 47 aerial duels last season, an average of 1.4 per match and a success rate of 72.3%. Enrique won just 13 of 34, an average of 0.4 per match and a success rate of 38.2%, even though Enrique is three inches taller than Cissokho. No Liverpool player had a higher aerial duel success rate than Cissokho last season.

There is a little to be worried about. Cissokho missed a lot of matches last season, making only 25 league appearances last season (seven as a substitute). By the end of the campaign, Andres Guardado seemed first choice left back, with Cissokho making more appearances as a attacking substitute on the left. Valencia's had the opportunity to examine Guardado, fullback/center-back Jeremy Mathieu, and young winger/fullback Juan Bernat, and decided that Cissokho is the expendable one, even though he's younger than both Guardado and Mathieu, after buying the player just one year ago. Yes, Valencia – like all of La Liga aside of Barca and Real – need the money, but that's slightly concerning.

Cissokho's not Liverpool's first choice either. Guilherme Siqueira, Ryan Bertrand, and Lorenzo Melgarejo were all linked before this deal was completed. At least Liverpool's making a move, as a move needed to be a made, but it's not necessarily the move they preferred.

Which is why it's nice that Liverpool are trying before buying. There's no reported clause requiring Liverpool to complete the transfer at the end of the season. And Cissokho will have a fight to dislodge Enrique from the starting XI, proving his worth, playing for a full contract. The addition strengthens Liverpool, in a position that Liverpool needed to strengthen, in facets of the game that Liverpool needed to strengthen. If it comes off, Liverpool will add an excellent player for a low-ish price. If it doesn't, they're out Cissokho's wages and a €1m-or-so loan fee. Small peanuts in the greater scheme of things.

Once again, it appears that Liverpool are acting intelligently in the transfer market, despite the frustrations in trying to sign another attacker. It remains a refreshing change.


Anonymous said...

Nice piece, mate. A very interesting read. Cissokho's an exciting player, and I believe this may prove a good deal for Liverpool.

Hardeep Bhambra said...

Big fan Nate, I have your blog linked too my work outlook so I can have a sneaky read while in the office.

Was really interested with the idea of lose possession, but as Enrique and Cissokho have vastly different playing time , don't you think if you could have a better compairson if you had a weighted average which you could index there performances on, instead of loss possession per minute?

nate said...

I know just enough about statistics to be dangerous, but I thought a per minute average was a fairer way to do it since they played different amounts of time.

Hardeep Bhambra said...

Sorry Nate, didn't mean to cause any offence. Like I said I'm a really big fan and look forward to your blog after each match.

nate said...

Believe me, you didn't cause any offense, and I appreciate the kind words. I just wondered what you had in mind as a better tool. As I said, I know just enough to be dangerous, and thought that using a per minute average (which shows the same discrepancy that a weighted per 90min average would show) did what you were asking for.

Hardeep Bhambra said...

Sorry for the late reply, work/life had taken over!

I have a problem with Min-per-X stat in general. I think it gives an unfair advantage to players that have had less game time.

Too understand a players performance their needs to be a benchmark, with a large enough sample size.

Weighted by playing time, you could use all of the current metrics, but give's you an average for the sample for each metric. This can then be used to benchmark a player to see if he's an above average, average or below average performer.

Hope i'm not just regurgitating what i said before and actually providing a different point of view!