Saturday, December 17, 2016

Quick Preview of Arsenal

If your Twitter feed is anything like mine (and it probably is, since you're reading this), you likely have heard all about Arsenal riding their luck so far this season. They have outperformed their xG numbers and posted very high conversion rates, meaning their current goal difference is probably inflated above what it "should" be. They currently average a goal difference of 1.25 per game, but if they and their opponents converted Shots on Target at a league average rate, their average goal difference would be just .41. That's still okay, but it's worse than all the rest of the top six teams. In comparing this season to last, possession has stayed pretty much constant, but the shots they have generated per that time of possession has fallen a bit, and their finishing has gone way up (45% G/SOT versus 29% last season).

So what's causing the drop-off in chance creation, and the increase in finishing? One major change that could explain part of it is the playing of Alexis Sanchez as a central striker this season. As this article shows, Sanchez has been a big reason Arsenal have scored more than their xG number would expect. But while the change to striker might have been good for him personally (particularly since he's in contract negotiations with Arsenal at the moment), I'm less sure it's good for Arsenal. The following table shows Arsenal's usage rates from last season:

Player Successful Passes Total Passes Unsuccessful Passes Key Passes Assists Goals Shots Take-Ons Take-On Success % Unsuccessful Take-Ons Minutes Possessions Used Usage Rate Positive Outcomes Negative Outcomes % Positive
Sanchez 1139 1430 291 58 4 13 107 101 61 65 2436 525 18.53% 169 356 32.22%
Giroud 571 839 268 30 6 16 101 11 44 14 2430 419 14.84% 137 282 32.70%
Ozil 1936 2230 294 125 19 6 47 46 57 35 3036 520 14.73% 191 329 36.75%
Ramsey 1809 2106 297 33 4 5 69 39 60 26 2623 429 14.08% 106 323 24.71%
Cazorla 1081 1198 117 34 3 0 24 37 80 9 1290 187 12.49% 61 126 32.58%
Oxlade
Chamberlain
360 437 77 19 0 1 23 36 71 15 935 134 12.31% 42 92 31.41%
Iwobi 314 368 54 10 2 2 15 16 67 8 639 89 11.97% 27 62 30.38%
Welbeck 114 149 35 5 2 4 20 14 52 13 581 75 11.10% 27 48 36.04%
Joel Campbell 265 337 72 11 2 3 19 17 63 10 893 114 10.99% 32 82 28.07%
Walcott 246 309 63 15 2 5 45 25 51 24 1382 149 9.28% 62 87 41.61%
Monreal 1551 1817 266 25 3 0 12 16 67 8 3252 314 8.31% 40 274 12.74%
Bellerin 1423 1664 241 17 5 1 13 63 66 32 3240 308 8.19% 35 273 11.35%
Flamini 536 619 83 3 0 0 7 2 50 2 1094 95 7.47% 10 85 10.53%
Elneny 677 728 51 2 1 0 13 4 80 1 792 68 7.39% 16 52 23.53%
Koscielny 1440 1654 214 8 0 4 19 10 91 1 2845 242 7.32% 27 215 11.16%
Coquelin 980 1099 119 6 0 0 8 30 83 6 1656 139 7.23% 14 125 10.06%
Mertesacker 1086 1218 132 2 0 0 8 3 100 0 2047 142 5.97% 10 132 7.04%
Paulista 972 1080 108 1 0 1 11 3 75 1 1764 121 5.90% 12 109 9.92%
Cech 302 491 189 1 0 0 0 2 100 0 3060 190 5.34% 1 189 0.53%

Last season, Arsenal had a pretty fearsome attacking triumvirate of Sanchez, Giroud, and Ozil. Sanchez actually posted the highest usage rate in the league, but all three were quite high and used the ball efficiently. This season, their usage rates now look like this:

Name Successful Passes Total Passes Unsuccessful Passes Key Passes Assists Goals Shots Take-Ons Take-On Success % Unsuccessful Take-Ons Minutes Possessions Used Usage Rate Positive Outcomes Negative Outcomes % Positive
Sanchez 476 650 174 38 5 12 48 38 63 22 1345 287 17.57% 91 196 31.71%
Oxlade Chamberlain 242 307 65 10 4 2 20 27 68 13 700 112 13.17% 34 78 30.36%
Iwobi 343 400 57 8 3 1 16 10 48 11 606 95 12.91% 27 68 28.42%
Ozil 762 879 117 35 2 5 25 15 65 8 1246 187 12.36% 62 125 33.16%
Walcott 285 375 90 9 2 7 44 28 56 22 1251 167 10.99% 55 112 32.93%
Mustafi 578 697 119 2 1 0 9 7 78 2 1014 133 10.80% 12 121 9.02%
Cazorla 497 545 48 8 2 2 8 9 60 6 618 72 9.59% 18 54 25.00%
Xhaka 640 720 80 8 0 1 10 7 88 1 869 99 9.38% 18 81 18.18%
Bellerin 524 618 94 9 2 0 8 21 72 8 1146 121 8.69% 19 102 15.70%
Monreal 615 736 121 5 1 0 4 6 60 4 1333 135 8.34% 10 125 7.41%
Coquelin 543 618 75 7 0 0 7 10 91 1 998 90 7.42% 14 76 15.56%
Elneny 371 401 30 2 1 0 6 3 100 0 453 39 7.09% 9 30 23.08%
Koscielny 642 741 99 0 0 2 6 1 50 1 1350 106 6.46% 6 100 5.66%
Cech 130 218 88 0 0 0 0 2 100 0 1440 88 5.03% 0 88 0.00%

Note that Giroud isn't even on here; he didn't meet the 360 minute threshold I'm using as Sanchez has completely taken over his role. Sanchez's numbers are quite similar to last season, but Ozil is using a lot fewer possessions than last year. You can also see the percentage of possessions used with a positive result is down for just about everybody on the team, partly explaining why S/TOP is down for them.

The switch of Giroud for Sanchez probably has a part to play in Ozil's usage rate, the drop in shot volume and the uptick in finishing. Giroud is the prototypical target man who needs to be fed the ball. Sanchez, as the Ringer article makes clear, receives the ball outside of the danger zone and brings it in himself. As a result, Ozil has created a lot fewer chances: only 5 of Sanchez's 11 open-play goals have been assisted, and none have been from Ozil. Last season by comparison, Ozil assisted Giroud 6 times before the New Year. Moreover, because Sanchez was able to generate such high shot volume from the wing, that has to be replaced by some combination of Walcott and Oxlade-Chamberlain. Walcott is averaging about 3.1 shots per game and the Ox 2.6, both lower than the nearly 4 Sanchez was averaging last year from the position. That loss of shots and Ozil's reduced role are probably combining to reduce the number of shots Arsenal produce overall.

Arsenal's improved finishing can also be attributed to the switch. This is not because Giroud is an especially bad finisher, but because Sanchez has been out of his mind so far. However, one thing I have noticed while reviewing Sanchez's goals is how much he has benefited from things outside his control. For example, three of his first six Premier League goals came from rebounds or deflections from other shots. While finding space in the box is certainly a skill, the fact that the ball made it to him is almost entirely luck. Only 5 of his 11 open-play goals have been assisted, as in addition to the rebounds he has been able to profit from defensive errors (notably Gary Cahill in the Chelsea game).  Partly as a result, a lower percentage of Arsenal's shots are being created by a pass this season, 75% down from 81% last season. The fact that Arsenal are converting their SOT at such a high rate when not creating the chances deliberately makes me think in the long run this is unsustainable.

Even though I think it's doubtful the change will prove useful to Arsenal over the long run, I do think it may do the trick against City. As Pep has been pointing out recently, we have struggled with second balls and our defense does make errors, and Sanchez's pace, trickery, and power make him the perfect candidate to exploit those flaws. His tendency to drop into midfield will also be a problem, since I don't see any of Fernando, Toure, or Aleix Garcia being able to stop him. Honestly, I think the best defense is ball retention here, so I would go with a lineup geared to maintain possession. That would probably mean Nolito up front and Toure/Fernando as a defensive midfield combo, as much as it sickens me to type those words. We'll see what Pep has planned in order to beat a Top 5 opponent for the first time this year.