Can the LA Galaxy be a top team without Landon Donovan?

Landon Donovan was a legend. A superstar. A once-in-a-generation player. Over the last 10 years, LA Galaxy fans have had the honor and privilege of watching him suit up for their local club and put in inspiring performances week in and week out. But those days are over. Landon Donovan is now happily retired and the LA Galaxy are venturing into the strange uncharted waters of the post-Donovan era.

It's hard to imagine what a Landon Donovan-less Galaxy looks like, but that's exactly where we are. So what are the Galaxy losing in Landon Donovan? Is it asking too much for them to reload so quickly after losing someone so integral to team identity? What will a Donovan-less Galaxy lack, and how should that affect who is chosen to play left mid for the 21 games prior to the arrival of Steven Gerrard?

These are the big questions hanging over the Galaxy in this post-Donovan era and seemingly everyone has their opinions on them. In times like these -- times of uncertainty and rampant subjective postulating -- I always like to retreat into one of two places: My spreadsheets and my tactics books. Often times, a certain clarity can be found in their warm embrace that is simply lacking from your run-of-the-mill eyeball analysis.

But the questions surrounding the Galaxy are questions which can not be satisfactorily answered by data analytics or tactical theory alone. However, each school sheds its own unique light on the problem, and, when combined, the faint glimmer of  an answer can be seen to many of these questions.

What we are losing?

Landon Donovan accounted for one-third of the Galaxy's offense last year. He accounted for a third of the total number of chances the Galaxy created. Last year he accounted for a third of the Galaxy's expected goals numbers, via a shot or a pass leading to one. His expected assist numbers to Robbie Keane alone accounted for 6% of the team's total expected goals numbers. His expected assists to Gyasi Zardes made up 5%. The following chart is mind blowing.

But let's drop percentages and talk raw numbers. Landon Donovan was either the shooter or passer behind 19.86 expected goals last year. The only midfielder to top that last year was Pedro Morales, and that's basically because he took 38 more shots which pumps up the xG totals quite a bit. Landon made 110 key passes last year off of which the average shot leverage was 0.11, or stated another way, had an 11% chance of being scored on average. He finished the season with 19 assists and 10 goals.

Using math to state the obvious, Landon Donovan was very good at soccer. The Galaxy stand to lose a huge portion of their offense now that Landon has retired.

So the Galaxy's main chance creator is gone along with the 17% of our offense created by his passes and 15% created by his shots. How in the world can the Galaxy recover from such a loss?

Left Mid Options

Right now, the big question for the Galaxy is who is going to play left mid until Steven Gerrard arrives and the lineups get reshuffled? Judging by preseason minutes, it appears to be a race between Bradford Jamieson IV, Jose Villarreal, and Ignacio (Nacho) Maganto. Each player has their strengths, weaknesses, a certain degree unknown, and, most intriguingly, developmental ceilings. This is by no means an easy choice. It's a choice, however, that Bruce Arena will have to make, and one that will have to be made based on preseason performance.

Bradford Jamieson IV is a US U20 prospect who projects as a forward, but due to size reasons, is getting his minutes on the wings as his body grows. This is not uncommon around the world. The first thing that stands out to you about Bradford is his pace and agility. He's technically gifted one v one, but not for reasons of trickery, step-overs, etc, but rather for his ability to make quick lateral cuts which open up shooting windows for him. This skill, combined with his nose for goal and wicked shot, make him incredibly dangerous around the goal. He had 6 goals for Los Dos last year and was a finalist for the USL Pro's rookie of the year award.

By far the most intriguing thing about him is he is only 18 years old and has only one real season of professional soccer under his belt, and that was at the USL level. Who knows how much he can develop if given similar minutes at an even higher level? Although player development is very hard to project, it's hard to look at Bradford and not see a ceiling that is incredibly high. If Bradford can get significant minutes in the 21 games the Galaxy will play without Gerrard (unless there is a work stoppage), it's very possible we see him develop to a point that he is significantly affecting the teams expected goals output in a positive manner.  In terms of replacing Donovan's 9.24 xG from last year, it's entirely possible that Bradford Jamieson could fill a significant chunk of that void.

The drawback to Bradford's game is one that is understandable considering his age and positional versatility. Having a winger that is goal dangerous is an incredible asset; however, wingers must also be involved in the build-up play. This is especially true in Galaxy soccer where possession and width are stylistic hallmarks. This is one area where I feel that Bradford needs to improve. He needs to better read the game to a point where he is getting in positions to receive the right pass sooner, and when he receives it, needs to be reading the game fast enough to know what to do with it.

Of the preseason games I have seen him in, in addition to my observations of him for Los Dos, this isn't happening for young Bradford yet, although it's important to stress his age. Gyasi Zardes had similar struggles adapting to the wing, and although I would argue that he still isn't a great winger, his improvement at that position over such a short period of time has been undeniable, and, in my opinion, Bradford has a better soccer brain than Gyasi so a similar or greater improvement is not farfetched. While that improvement is occurring, however, his inclusion would have major stylistic ramifications and from a tactical perspective, certain adjustments would have to be made to maintain the integrity of Tiki-Taco soccer.

One such adjustment is to recognize that more of the chance creation load is going to be running through Stefan Ishizaki on the right. Whether Keane can be as effective drifting into pockets of space in-between the lines on the right, rather than the left, is debatable, but  in the interest of putting our best players in close proximity to each other for combination, this is an adjustment Keane will have to make.

More importantly, however, the Galaxy should shift Robbie Rogers to right back, which they have been experimenting with in preseason; although they claim the reasons aren't tactical. Part of Donovan's renaissance at left mid last year was the fact that he had Robbie Rogers overlapping him, opening up channels for Donovan to run through. Similarly, he also had Keane drifting left. The same dynamic can be mirrored by moving Rogers to right back; however, this will require Todd Dunivant to return to his previous form.

Ishizaki is going to be asked to do more next season, (at least in the first 21 games) and moving Robbie Rogers to right back is a move that will make that easier for him. From what I've seen in preseason, the two already have a great understanding of each other's movements and have a pretty cool little dynamic where sometimes Ishizaki drifts wide and Rogers cuts inside, and sometimes Rogers makes the outside overlap and Ishizaki drifts inside. The following chart shows the movements I'm talking about as well as the adjustment of Keane's movements to the right in order to retain the blessed Tiki-Taco triangle which opens the holes for Zardes to run through allowing him to continue to be the team's secondary scorer. It's the same old Tiki-Taco, only now operating on the right side.

The second left mid option for the Galaxy is Ignacio Maganto. Now, based on personal observations in preseason, which, granted, is a small sample size, I think Ignacio Maganto has the potential to have a breakout rookie season in terms of chance creation if given the proper minutes. His vision is superb and he's supremely creative. He also fits our possession style nicely. He works well under pressure and knows how to circulate the ball. Obviously more has to be seen from him in preseason, and quite honestly, the fact that he received such a limited number of minutes against Hammarby is worrying.

Maybe Bruce is just rotating players, however, it also could be a sign that things are happening behind the scenes in training that are making him fall behind in the depth chart. Whatever is the case, if Nacho does find himself as the Galaxy's starting left mid on opening day, very few tactical adjustments will be required to make such a lineup run smoothly. Like Donovan, he'll keep the ball moving in the build-up and provide chance creation in the final third. In terms of filling the Donovan numbers gap, I can easily see Nacho filling a good portion of that 17% of the offense which Donovan generates with his passes which we are now missing.

The final left mid choice, and the one which is by far the easiest to analyze, is Jose Villarreal. Villarreal is a known quantity among Galaxy fans, and we actually have pretty good data from 2013 on which to judge him. In 2013, Jose played 1,109 minutes and totaled 4.34 expected goals and 2.2 expected assists in that time. If you take those rates and project them over the amount of minutes Donovan played last year, that comes out to 10.6 expected goals and 5.4 expected assists.That would put his projected contribution to 16 goals which is not that far off Donovan's 19.86 mark from last year.

Eureka! Has analytics found our solution? Possibly, but there are some major caveats to consider. While 1,109 minutes is not an insignificant sample size, it's also not ideal. Villarreal is also a left footed player, which makes it harder for him to be goal dangerous if he makes inverted runs from the left. His 2013 numbers were primarily as a withdrawn striker, where shots are easier to come by. There is also the matter of possible regression over his Cruz Azul period, not to mention Bruce's recent comments about his professionalism in terms of fitness.

What exactly is going on there is anyone's guess, but it's certainly not good to hear. In the game against Hammarby, Jose had a good number of minutes on the left, and I think he accounted for himself well. His passes were nothing special in terms of opening up the defense, but they were timely and kept the ball moving. Aside from that, however, more observation is needed.

Right now we know that Jose Villarreal can put up impressive numbers as a withdrawn forward, and his link-up play in those areas is pretty darn good aswell. As an inverted winger, his link-up game should adapt well, which means it's not unreasonable to project his expected assist numbers to translate well to left mid. His expected goals numbers, however, are highly in question. Again, him being left footed is an issue. In 2013, only 3 of his 37 shots came off of his right foot, and as a left mid, the majority of his chances are going to be shots best taken with the right.

Tactically this could be minimized to some extent if Keane drifts more towards Ishizaki's side rather than to the left as he currently does. This would allow Villarreal to step into the vacated space, where, as a natural center forward, he should feel most comfortable. In this scenario, I think Rogers should remain at left back in order to maintain some manner of left sided width. I have drawn up a loose movement chart to demonstrate.

All of this may seem kind of complex, however, the basic tenants of Tiki-Taco as outlined so well by Matthew Doyle in this article, remain true. In said article, he states the following:

Their preferred method of operation is to stack two players along the left flank, about 30 yards out, and two attackers along the 18, one in front of the other. When the low cross comes in, the first guy dummies it and rolls into space, while the second guy makes a simple wall pass into the roll:

He uses the following sequence to illustrate this point.

The two players on the left were usually Robbie Rogers and either Robbie Keane or Landon Donovan, although Sarvas would occasionally jump into the mix. The two at the top of the box would then be some combination of  Keane, Donovan, Sarvas, and Zardes depending on the situation and  which player is left with Robbie Rogers.

In this version, however, the attack now runs down the right hand side, and Sarvas' runs to the top of the box are now supplemented by the left mid making inverted runs when Keane drifts wide. This not only puts Villarreal into positions where he can be goal dangerous, but it also leaves two midfielders back to provide ample defensive cover on counter attacks. It's a minor tweak, but the same principle of two pairs of two still applies.

If Jose Villarreal can regain his 2013 form, and if the proper adjustments to the system are made to highlight his strengths, then he may just be the best candidate of the three. But those are some mighty big ifs, and, ultimately, Jose will have to prove himself worthy of the position in preseason, just like the other two candidates.

There are a great deal of unknowns with each choice, and it's probably fair to say that the Galaxy offense will lose some expected goals, regardless of who is chosen. Picking the right player, then,  is really a choice of who can best minimize the loss.  But this brings up an interesting question.

How much of a loss can the Galaxy take and still remain a top MLS team?

In the case of this question, analytics may hold the key. If you consider the Galaxy's expected goal differential numbers from last year, you'll see that the answer to this question is a lot more than you would expect, given the right circumstances.

Last year the Galaxy's beastly offense and defense combined to create goal differential totals that were absolutely obscene. Here are their expected goal differential per game numbers from last year compared against Seattle's as well as both the average expected goal differential numbers of the Supporters Shield winners from 2012-2014 and the team's which finished the season with the most points in the west over that same period.

	Expected Goal differential per 90
2014 LA Galaxy	0.88
2014 Seattle	0.41
2012-2014 Avg. SS	0.28
2012-2014 Avg. West	0.34

Aside from being just plain impressive to look at, these numbers are relevant to the Donovan discussion because it demonstrates that, if the Galaxy's defense can maintain similar numbers to last year, the Galaxy can lose a lot of expected goals and still be a top team in the conference and league. Just how many? Well, if we're discussing left mid, it's possible that Steven Gerrard may bump out whoever ends up filling that role from the lineup, so the following graph takes the per game numbers from the table above and demonstrates the expected goal differential numbers, or xGD, over a projected 21 game period and then over the full season. This gives you a sense of just how large the gap is, and from there, finding an exact number is just a matter of simple subtraction.

If we use the average of the Supporters Shield winners from 2012-2014, the Galaxy can lose 12.53 xG over the 21 games without Gerrard, and 20.29 xG over the entire season and still be in the running for the Supporters Shield. Keep in mind that Donovan accounted for 19.86 of the Galaxy's xG by means of shot or pass. From a numbers standpoint, the Galaxy could replace Donovan with a player that contributes 0 xG by way of shot or pass and would still have a 0.43 xG lead on the Supporters Shield average over the course of a full season.

If you use the xGD of the teams that have finished the regular season at the top of the western conference as your benchmark, the bar is raised a tad higher, however, the final deficit is marginal. If you apply the same neutral left mid scenario as before, the Galaxy only fall under the Western mark by 1.39xG-- an output that any of our three left mid candidates should be able to easily surpass.

Finally, if Seattle has a similar season to last year, which is entirely possible, the deficit would be 3.88 xG, which is also a marginal number for any Galaxy winger to contribute. Now, if you assume that Gerrard will significantly boost team numbers and thus only consider the first 21 games, the deficits are: +0.27 for Supporters shield, -0.85 for the West, and -2.39 for Seattle. All in all, these are very reachable numbers for our left mid candidates.

Of course, a good expected goal differential doesn't guarantee results. I think that's nicely evidenced by how the Galaxy somehow managed to lose the Supporters Shield last year when their xGD was double that of Seattle. I'm merely suggesting that, with the right goal distribution, the Galaxy can lose a huge chunk of Donovan's contribution and still be a top team. For further evidence of this, look no further than the 2013 and 2011 Supporters Shield winners. New York won the shield with an xGD of 0.10 and in 2011, the Galaxy managed to win the shield with an xGD of -0.05. This is both a combination of luck and having the type of team that simply knows how to win a game 1-0, like the 2011 Galaxy did. The 2014 Galaxy, on the other hand, sloppily dropped points left and right and still managed to win the cup and almost win the shield. The Galaxy simply will not have that luxury this year.

Last year's team was so good that they were able to compensate for losing 4 points on missed penalties, 4-6 points on Leonardo bungling a play, and 2 points on a Dan Gargan throw in to a keeper taking a knee. If this year's team drops 10-12 points like that, don't expect them to be able to find a way out of that hole.

Conclusions

You know, I started this article telling you how great Landon Donovan was and then shifted gears and started telling you how unnecessary Landon Donovan is to the Galaxy's success. Let me attempt to rectify this dichotomy with the following few paragraphs. The numbers tell us that Landon Donovan was practically a demigod. Those numbers could not have been more essential to rectifying a team that dropped 10-12 points they simply had no business dropping. Had they been as clinical as the 2011 team, they would have won the shield as handedly as they beat out the rest of the league in terms of goal differential. Their play was so brilliant, they were able to compensate for a level of incompetence that would have buried any other team.

From a tactical standpoint, Tiki-Taco was built on two pillars, Donovan and Keane, however, there is a supporting arch between those two pillars that is often forgotten-Robbie Rogers. Now that one pillar is gone, the temple of Tiki-Taco will almost certainly need remodeling. In order for Ishizaki to take on more of the structural load on the opposite end of the temple, it would behoove the architect to move the archway to that side, else try to build a new pillar on the left side out of nachos. Villarreal also works but there was no pun to be had there.

In the end, the Galaxy can still play Tiki-Taco if the pieces are put in the right place. If an xG oriented player like Bradford Jamieson IV is placed out there, then the attack will need to shift heavily to the right as build up play on the left will happen far less. Robbie Rogers has shown he is more than capable of being a dangerous right back, and Ishizaki has looked more dangerous for it on those occasions. If Villarreal can regain form or if Nacho proves to be able to make the impact I think he can, then Rogers needn't necessarily move for the reasons I stated earliar.

From a numbers standpoint, with the right luck and the right level of clinical game management, the Galaxy don't really need all that much production from any left mid in order to be one of the league's top teams. Essentially, anything over 3.88 xG from that position is just an added bonus; especially if Ishizaki contributes more, as most people expect. And then there is the matter of Gerrard. If we look at our left mid as simply someone buying time until he arrives, then the Galaxy need only find about 2.21 extra xG from someone or a combination of players, in order to have an xGD over those 21 games on par with last season's Supporters Shield winner.

Sure, this year's team will not be the Galaxy strike force they once were, (at least until Gerrard arrives), but if they can clean up the sloppy mistakes, that's almost a moot point. If they batten down the hatches and play a solid season from start to finish, there is no reason why they can't be a force to be reckoned with in 2015, no matter who plays on the left. That's how good this team is.