State of the Galaxy: What the numbers tell us about this team.

Alright readers, it's time for the first in series of articles I plan to write throughout the year in which I take a step back and take a look at the larger statistical picture and deliver a State of the Galaxy report.

Now this article deals heavily in the concept of expected goals, so if you are unfamiliar with the concept I recommend this article as a good layman's primer. For an even simpler primer, know this: Stats like Goals and Assists are straightforward enough, but they are surprisingly poor when it comes to predicting future outcomes. This is because they are both highly subject to luck.

For this reason, a bunch of eggheads far smarter than me started judging players based on something called Expected Goals (xG for short) which deals in the percentage chance of shot conversion from a given field position and circumstance. Let's say a player takes four shots with an xG value of 0.25 (25%) which adds up to 1 xG. Now here is where luck comes in. It is entirely possible that the player scores twice. If you repeat that game 20 times, however, the player is far more likely to end up with a goal total closer to 20 than 40. For this reason, xG data is good for projecting over long periods of time.

Simple enough, right? Good. Now let's dive into the Galaxy numbers this season.

In the off-season I developed what my doctor calls an "unhealthy fixation" with the question of what the LA Galaxy will look like without Landon Donovan. In this article, I tackled the question from a raw expected goal and expected assist standpoint and determined that Jose Villarreal was the best fit to plug the Donovan offensive gap if he was somehow able to regain his 2013 form. Well, turns out he did and he currently leads the team in offense.   (note: Expected assists are just a measure of how much xG a player directly creates via pass.)

Player	xG + xA
Jose Villarreal	1.72
Robbie Keane	1.45
Stefan Ishizaki	1.38
Gyasi Zardes	1.18
Alan Gordon	0.81
Juninho	0.76
Baggio Husidic	0.76
Omar Gonzalez	0.67
Mika Vyrynen	0.42
Dan Gargan	0.24
Robbie Rogers	0.18
Leonardo	0.15
A.J. DeLaGarza	0.1

Despite this, the Galaxy's offensive numbers are down. Last year we averaged 1.85 expected goals for (xGF) per game. Currently we sit at 1.55.  As you can see below, it still ranks fairly high in relation to the rest of the league, however, we have definitely lost our 2014 claim to being the undisputed top dog. Also, when you consider our relatively weak strength of schedule so far, playing both Chicago and Houston at home, it's entirely possible that our current 1.55 xGF is a tad inflated.

Team	xGF
NE	1.68
CLB	1.59
SKC	1.56
LA	1.55
NYC	1.47
POR	1.4
VAN	1.34
SEA	1.29
NYRB	1.28
CHI	1.27
DCU	1.27
TOR	1.24
SJ	1.22
RSL	1.17
FCD	1.12
PHI	1.07
MTL	0.9
ORL	0.86
COL	0.8
HOU	0.62

The logical explanation for the drop is that it has to do with the absence of Landon Donovan. That was my initial theory until I looked at the data and discovered that this simply wasn't the case.

Below I have broken down the average xG rates for players per 90 minutes and compared them against their 2014 rate, or, in the case of Baggio Husidic and Jose Villarreal, compared them to the production of the players who were playing those positions last year: Donovan and Sarvas.  The blue bars are 2014. The red bars are 2015

As you can see, the drop off from Donovan to Villarreal is slight, at least in terms of expected goals. The same can be said for the drop off from Sarvas to Husidic. Robbie Keane, on the other hand, has seen his xG production almost halved. Of course, Robbie Keane has only played 3 games this year and one would expect that Keane's numbers will start to move closer to last year's averages. In other words, we should not interpret this drop as a reason to panic about Keane's form. We can, however, point to it as the reason why the Galaxy offense has not looked that great for the first 4 games of the season.

The same can be seen when you look at xA numbers, however, the gap between Landon and Jose is a tad more noticeable on this front.

Landon Donovan was a chance creating God, so this is to be expected, however, the Robbie Keane gap is once again the most noticeable. There are a couple of possible explanations for this. Perhaps the biggest take away from my passing analysis of the team last year, was the very specific way in which Robbie Keane creates chances. Mainly, Robbie Keane is absolutely terrible when it comes to facilitating shots from central positions in and around the box.

He is, however, very good at facilitating chances from the wings. His two favorite targets from the wings: Alan Gordon and Landon Donovan. Robbie Keane only created 1 goal for Gyasi Zardes from the wings. This is largely to do with the fact that Gyasi Zardes still has a ton of work to do in terms of his runs. While he has perfected the art of the late far post run to receive a through ball, he has yet to really master getting open for wing players to feed him the ball. Again, this is all supported by the data in my passing analysis linked above.

Now that Landon Donovan is gone, the Robbie Keane wide threat has essentially been limited to the late sub minutes that Gordon is on the field. Alan Gordon may have a reputation as a basher and a bruiser, but he is also very smart in his movements. There is a reason he has made his career off slipping the marks of tired defenders late in games.

Does this mean Gordon should start? No. With each game at left mid, Villarreal has demonstrated a growing level of comfort and awareness when to make those inside runs into the box. As a center forward, he should have no problem getting into scoring positions in these circumstances.

There is also a certain onus on Keane in all of this. In my opinion, Keane looks to be trying to over compensate for the loss of Donovan. He is dropping back deeper into the midfield more often and attempting to create chances around the box in situations he's normally looking to shoot. He gets a pass on the Houston game since they were playing an extra midfielder which affected the Galaxy's shape majorly, however, during the Chicago and Portland games, Keane seemed overly concerned with facilitating (which he is actually statistically poor at from those positions) to the detriment of his world class ability to get into scoring positions or create shooting opportunities for himself.

Until Gerrard gets here, Robbie Keane should put a little more faith in Villarreal, Husidic and Ishizaki and just go back to playing Robbie Keane soccer.

Moving on from Keane, I think a point should be made about Baggio Husidic. Marcelo Sarvas was a great guy who played with a massive amount of heart and this really endeared him to many fans. There seems to be a wide spread belief among fans that part of the reason that the Galaxy haven't looked as good is that we are seeing a drop off from Sarvas to Husidic. From a data perspective, this perception seems completely bogus. While Marcelo Sarvas slightly outpaces Baggio in expected goals (as was also the case last year), Baggio also outpaces Sarvas in expected assists, and by a larger margin. Consider the following chart which I made to compare the Galaxy's passing dynamic to xG output in terms of field position.

As you can see, Husidic is seeing just as much of the ball as Villarreal on the left wing in terms of touches and usage, which means he is involved in the attack. He only averages about 13 passes fewer than Sarvas' per 90 numbers last year. As you can see in the earlier bar charts, Baggio is also outpacing Sarvas in terms of total expected offensive production. In reality, what we are seeing in the fans' reaction to Husidic has more to do with the love of Marcelo Sarvas than it has to do with Baggio's actual performance. Baggio is doing fine.

The final point I'd like to make has to do with the Galaxy's continued expected goal differential (expected goals for - expected goals against) dominance, yet inability to translate this into points per game. The Galaxy were by far the best team in the league last year, however they ended up finishing with fewer points than the Sounders despite the massive expected goal differential gap.

Team	Expected Goal differential per 90	Points per game
2014 LA Galaxy	0.88	1.79
2014 Seattle	0.41	1.88

Why did this happen? The Galaxy dropped a lot of points late in games and that's something that the Sounders simply didn't do. Sure, some of that can be chalked of to a certain randomness in sports. A little luck goes a long way. But it could also very much be a case of poor game management. This is something that I am very much keeping my eye on this season, and, so far, the trend has continued. The Galaxy remain at the top of the league in terms of xGD, however they are still struggling to turn this into points.

Team	xGD
LA	0.84
SEA	0.67
CLB	0.62
SKC	0.5
POR	0.22
NYC	0.2
ORL	0.17
FCD	0.15
NE	0.15
VAN	-0.04
TOR	-0.13
COL	-0.13
RSL	-0.27
MTL	-0.32
CHI	-0.32
PHI	-0.35
HOU	-0.39
DCU	-0.43
NYRB	-0.54
SJ	-0.7

The Galaxy's current points per game total is 1.25. Remember, Seattle's pace last year was 1.88. Now, obviously the sample size for this season is too small and their are a lot of factors to consider, however, it is still continuation of a trend we saw last year so it is definitely something to keep an eye on throughout the season. I know Galaxy fans don't put too much stock in teh supporters shield, but playoff seeding matters. The ability to host MLS Cup matters.

In summation, I think the data shows us that things look far worse than they actually are. Our biggest problem at the moment is that Kean's production is down. Keane happens to be one of the most statistically consistent players in the league, so this is more than likely a problem that will correct itself quite quickly. Keane's production will soon rise and so too will the Galaxy offense. All things considered, for a team that lost Landon Donovan, the state of the Galaxy is actually pretty damn good and on track to get better. Rejoice!