Why Do MLS Teams Suck at Drafting Goalkeepers?

By Bill Reno (@letsallsoccer)

Last year I interviewed John McCarthy, who at the time I was sure would be an MLS SuperDraft pick. McCarthy had just graduated from LaSalle University, and while his school didn't make that year's NCAA tournament, it was obvious he was good enough to play professionally. Still, the SuperDraft came and went, and McCarthy went untouched in the four rounds.

Andre Blake was the heralded newcomer, but the other three selections---along with other MLS combine invitees---were largely unknown. McCarthy responded to the setback by signing with the Rochester Rhinos, where he unseated an MLS-loaned player on his way to being named both USL Goalkeeper and Rookie of the Year. So how did MLS miss this one?

Every January, MLS teams draft collegiate players in the aptly named MLS SuperDraft. But for being in a country that is renowned for producing elite goalkeepers, MLS has a miserable time of identifying the talent. Consistently good collegiate goalkeepers take an unnecessarily long road to get to MLS while clubs select goalkeepers that never make an appearance. Here is a list of every drafted goalkeeper since 2006.

Names highlighted blue were invited to MLS combine
GS - Games started, orange numbers have at least five starts a year since draft
LG.2, LG.4 - If GK continually stayed in MLS two, four years later
TM.2, TM.4 - If GK continually stayed with team two, four years later
Y1, Y2, Y3, Y4 - Status of GK in first four years

The list looks at the first four years of each drafted goalkeeper. In the nine years covered, 60 goalkeepers were taken in either the SuperDraft or the following Supplemental Draft. Of those 60, only 31 (52%) finished the first year with the team that drafted them. After four years, two thirds of the goalkeepers weren't even in the league anymore. And with Billy Knutsen and Luis Soffner's contracts not being extended, it's looking like only four of the ten goalkeepers drafted in the past two years will still be with their original team.

What's even more bizarre, goalkeepers drafted after the 40th pick have a higher collective number of starts than those drafted before it.

Drafted	Appearances	# GKs	# GKs (1+ GS)
1 - 40	408	20	12
41+	501	40	12

Even though the number of post-40 draft picks double those drafted earlier, when you remove the goalkeepers that never started, both groups have twelve goalkeepers. Half of the goalkeepers selected in the first two rounds didn't get more than ten starts. Not only are goalkeepers being poorly selected in the draft, but it doesn't matter all that much when they were drafted or in what round. Late or early, they have surprisingly low chances of ever starting in MLS. The inefficiencies of the SuperDraft continue when you survey the current goalkeeper pool.

Orange names are starters
Blue numbers were in the first 40 draft picks

Sixty goalkeepers were on an MLS roster this past year. Excluding the 18 goalkeepers that were not able to be selected in a collegiate draft (including Marcus Hahnemann, who graduated two years before MLS started), 16 current MLS goalkeepers went completely undrafted and entered the league later, including six current starters. Another three starters couldn't, or wouldn't, agree to a contract with an MLS side.

What if more than a quarter of all NFL quarterbacks were originally undrafted and represented 30 percent of last weekend's starters? That's ten Kurt Warners! It's not that there aren't enough rounds in the draft for goalkeepers to be drafted, it's that the SuperDraft is incredibly ineffective in scouting MLS talent. And the talent is definitely there. Fourteen of the 19 starting MLS goalkeepers came from NCAA, and 75 percent of all goalkeepers in MLS played college ball, including Canadians. Even on the international scene, NCAA has served the USMNT as well.

USMNT Caps	Player	USMNT Years	College
104	Tim Howard	2002–2014	Did not attend college
102	Kasey Keller	1990–2007	University of Portland
100	Tony Meola	1988–2002	University of Virginia
82	Brad Friedel	1992–2004	UCLA
28	Brad Guzan	2006–2014	South Carolina
16	Nick Rimando	2002–2014	UCLA
15	Mark Dodd	1988–1998	Duke
9	Marcus Hahnemann	1994–2011	Seattle Pacific University
8	Juergen Sommer	1994–1998	Indiana University
8	Zach Thornton	1994–2001	Loyola University Maryland
7	Troy Perkins	2009–2010	South Florida
5	Kevin Hartman	1999–2006	UCLA
4	Sean Johnson	2011–2013	University of Central Florida
3	Jonny Walker	2004	Louisville
2	Joe Cannon	2003–2005	Santa Clara University
2	Bill Hamid	2012–2014	Did not attend college
2	Matt Reis	2006–2007	UCLA
1	Jon Busch	2005	Charlotte
1	Tom Presthus	1999	Southern Methodist University
1	Zach Wells	2006	UCLA
1	David Yelldell	2010	Did not attend college
1	Luis Robles	2009	University of Portland

Drafted players regularly don't work out in any league. There are only so many spots on rosters so not everyone is going to make it. But good goalkeepers are consistently coming out of the NCAA---ones good enough to play for the national team---yet MLS still hasn't figured out who they are.

There are too many goalkeepers getting invited to the combine, but then not drafted. There are too many goalkeepers that go undrafted and yet eventually do make it. MLS clubs are making bizarre trades for eventual starting goalkeepers. Teams are overpaying aging goalkeepers. Teams overstock on goalkeepers they can't unload. The league's entire approach to goalkeeping is mind-boggling, and few are getting it right. With the expansion of MLS-USL affiliations, goalkeepers are getting more secured playing time, but it doesn't matter much if MLS continues to pass on top goalkeepers and mishandle the ones they currently have.

Plus-minus stats in MLS

Most analytically inclined sports fans are aware of plus-minus metrics and what they mean. For those new to the concept, Plus-minus metrics combine the individual and the team. For soccer, it represents the goal differential for a player's team only while he was playing. I was going to provide a plus-minus chart for players in MLS going back to 2011, but as I'm going to explain, it's mostly useless. Even when using expected goal differential (xGD) instead of actual goal differential to create xPlusMinus, not much can be gleaned from the metric.

Here is a chart of the top 25 MLS players in 2014 by xPlusMinus. It is followed by a brief discussion on its current lack of meaningfulness, and then some foreshadowing as to what adjustments will be made in the future to create a worthwhile plus-minus metric. These metrics are on a per-96-minute scale.

Player	Team	Starts	Appearances	Minutes	xPlusMinus	PlusMinus
Todd Dunivant	LA	5	7	390	1.58	1.72
Alan Gordon	LA	5	14	566	1.53	1.70
Omar Gonzalez	LA	22	22	2014	1.12	1.10
Robbie Rogers	LA	15	19	1466	1.05	0.98
Marcelo Sarvas	LA	25	28	2338	1.04	1.07
Gyasi Zardes	LA	26	32	2541	1.02	1.21
Landon Donovan	LA	30	31	2870	0.95	1.07
Jose Leonardo Ribeiro da Silva	LA	20	24	1949	0.95	1.28
Dan Gargan	LA	27	29	2580	0.94	1.19
Chance Myers	SKC	7	7	609	0.87	1.26
Vitor Gomes Pereira Junior	LA	31	34	2906	0.86	1.06
Baggio Husidic	LA	26	34	2298	0.84	1.17
Robbie Keane	LA	28	29	2696	0.78	0.85
Jaime Penedo	LA	29	29	2768	0.77	0.90
A.J. DeLaGarza	LA	28	29	2642	0.76	0.80
Oriol Rosell Argerich	SKC	7	7	641	0.76	0.45
Helbert (Fred) da Silva	PHI	3	11	389	0.72	0.99
Clint Dempsey	SEA	23	26	2268	0.70	0.42
Zach Scott	SEA	16	16	1478	0.69	0.71
Rob Friend	LA	4	10	408	0.67	-0.24
Stefan Ishizaki	LA	22	30	1994	0.65	0.91
Obafemi Martins	SEA	29	31	2811	0.58	0.55
DeAndre Yedlin	SEA	25	25	2362	0.57	0.33
John Berner	COL	5	5	479	0.55	0.20
Chad Marshall	SEA	31	31	2967	0.50	0.49

Here's an example of how to interpret the chart. During Omar Gonzalez's 2,014 minutes on the field, the LA Galaxy recorded an xGD of +1.12. Since the Galaxy as a team recorded a league-best xGD of 0.88, one could come to the conclusion that Omar Gonzalez is one of the best players on the best team. That particular conclusion probably isn't too far from the truth, but what about the rest of the table?

Control for a player's team

Basically, this statistic reiterates that the LA Galaxy were the best team in the league. 16 of the top 25 players are from LA, while five of the remaining nine play for the Seattle Sounders. There are fewer lineup combinations in soccer than in basketball and hockey because of the restrictions on substitutions, and this leads to a mostly redundant plus-minus metric. However, by looking at this table, we can start to figure out how to make it better, and the first thing we need to do is control for a player's teammates.

Control for strength of opposition

If we focus just on Galaxy players, we still see some weird things. Todd Dunivant and Alan Gordon are not LA's best players, so why did they end up on top? It's very possible that when these two were on the field, it also independently happened to be when LA's other players were playing exceptionally well. Neither player has a large sample size of minutes with LA, and that can lead to additional random variance in any metric. 

But there is a more concrete factor that likely influenced their plus-minus metrics: an individual's strength of schedule. Weighted by minutes, Dunivant and Gordon played against teams with expected goal differentials of -0.23 and -0.15, respectively. Compare that to Landon Donovan and Robbie Keane's -0.05. Those discrepancies don't make up the whole difference in xPlusMinus, but show that even players on the same team may face different opponents. 

Control for home/away and gamestate

Two other key components to adjust for are whether or not a player's team is at home, and the gamestate in which he typically plays. Robbie Rogers is probably not more valuable than Donovan or Keane, but he did play 54% of his minutes at home and 69% of his minutes while tied or losing. That compares to Donovan's 50% and 63% figures and Keane's 49% and 63%. I know that was a smattering of strange numbers, but the point is that Rogers was given both a home-field advantage and a gamestate advantage relative to many of his teammates. We must also adjust for these factors.

We hope to create a refined plus-minus statistic such as the one ESPN uses for NBA players that controls for specific lineups at any given time. For now, I have put up some of the aforementioned plus-minus metrics in our Plus-minus tab in the upper right. Or just click this link. Proceed with caution.

Shots: Confusion in correlations

By Matthias Kullowatz (@MattyAnselmo)

Much of the research I do for this site revolves around predictive analysis. I like to know which individual and team skills can be measured with stable metrics, metrics that hold true month after month. However, it's still worthwhile doing what I call explanatory analysis. Explanatory analysis involves finding the variables that explain an outcome which has already happened, even if these variables may fluctuate randomly in the future.

I have shown before that shot quality and quantity correlate well to future outcomes. But with that in mind, it is somewhat confusing that the same shot information doesn't correlate so well to the outcomes of the very same games from which the data were gathered. Here are some interesting facts about shots. 

Over the past four seasons in Major League Soccer, home teams averaged more shots in games they lost than in games they won (14.5 to 14.2). Conversely, away teams averaged more shots in wins than in losses (11.9 to 11.1). When the data are combined, the correlation between shot differential and goal differential within a match is virtually zero (CI: 0.02, 0.13). Superficially, this information seems more confusing that helpful.

This finding has led some to reason that shots are a less important metric when it comes to team evaluation. The fact that shot information is predictive is enough to convince most people (including me) that it is useful information to have. But how can it be that something predictive is not also explanatory? How can shots help to predict future outcomes, and yet not be able to explain the outcomes of those games in which they occurred?

You've probably already spotted the subtle differences between explaining and predicting, but let me take a shot (I promise that was an accident). Within a match, correlations are confusing due to all kinds of confounding variables. The answer to this question would clear some things up: "Who was winning the game when all these shots were happening?" Let's explore.

Typically, home teams outshoot away teams 14.3 to 11.3 per 96 minutes of play, and 14.1 to 10.7 in even gamestates. But when they're winning, home lose the shots battle 13.0 to 12.2, likely more content to sit on their leads. When they're losing and desperate for points,  home teams outshoot the visitors by a huge margin, 17.2 to 9.3. So I would argue that the goal differential (gamestate) influences the shot differential as much as the shot differential influences the goal differential.

It's no wonder that in-game correlations between shots and goals are non-existent. Early on in games, the team that gets more shots tends to take the lead. But once they have the lead, those teams tend to  ease up on shots. Thus whenever a team "holds on" to win a game, it very likely had a shot advantage at some point, and then relinquished that shot advantage in attempting to preserve the lead. Without taking into account the gamestates, a superficial analysis would suggest that shots do not correlate to wins. 

I have done nothing with shot quality here, but that wasn't really the point. The point was to show that in-game correlations have to be treated with a lot of care if you want to come to any conclusions about causation. But for the curious, the in-game correlation between Expected Goal Differential 3.0 and final goal differential was 0.37 (0.32, 0.42). Though gamestates are still an issue, shot quality is able to account for the fact that the losing team will be taking lower quality shots, and we get something sort of intuitive.

Do goals stimulate goals?

By Matthias Kullowatz (@MattyAnselmo)

I've heard it said before that a soccer team is most vulnerable after a goal has been scored. My coaches often said this, anyway. Perhaps it was just to keep us focused after we'd experienced the euphoria of scoring or the letdown of conceding. It turns out, there is some support for this notion from the 2014 MLS season. To the results! 

1) A goal is more likely to be scored in a five-minute segment if a goal was just scored during the last five-minute segment.

First, I should say that I controlled for the teams' abilities using season expected goals data, and I controlled for the gamestate as well, since there are fewer goals typically scored in zero gamestates. Once controls were in place, I found that if a goal had been scored by either team in last five-minute segment, the chances of another goal being scored in the next five-minute segment increased from about 15% to nearly 18%.  Put another way, after a recent goal the average goals scored in next five-minute segment increased by nearly 0.04, equivalent to about 0.70 goals over a whole game.

This isn't an obvious uptick in scoring. You probably wouldn't notice even if you watched a lot of games, but the effects of a recent goal are also not nothing. The game appears to open up a bit on average after a recent goal. 

2) The team that most recently scored is more likely to score again (than its typical scoring rate would suggest).

Breaking the first hypothesis down further, we actually see that the team most likely to score in the next five-minute segment is the team that just scored.* The chances of a team scoring in the next five minutes--whether it be the away team or the home team--are increased by 3 or 4 percent if that team scored recently. Chances increase from 6% to 9% for the away team and from 9% to 13% for the home team.

Typical sports fans may say "duh" because the existence of momentum in sports is a common belief. However, momentum is still very much a point of controversy among statisticians across all sports. I would say about the only thing we agree on is that the effects of psychological momentum are smaller than the common fan would believe, and perhaps even negligible in many scenarios. That's why this finding surprised me, especially when we consider that the team that just scored must then relinquish possession.

These results may not apply to a youth soccer team, or even a professional team from another league. But in MLS, there is an average effect on scoring, which is not necessarily negligible, that comes from a recent goal being scored.

Comments are welcome, especially if you can think of a way to further control the scenarios and weed out any biases in the observational data.

*Of course, the team that just scored is probably the better team. But that's why a control was put in place for overall team ability. What isn't controlled for is team ability on that day (due to injuries and what not).

Drew and Harrison on the Last Word Soccer Club Podcast

Here's our ASA plug for the week; go check out Last Word Soccer Club's 2014 year in review podcast from last night. Drew and Harrison went on and talked everything from the World Cup to MLS Cup, and most things in between. Go give it a listen!

A Best XI Of Possible Expansion Draft Targets

By Harrison Crow (@harrison_crow)

This started as a suggestion from twitter to make a team built from players left exposed in the expansion draft. I'll be honest, as much as I love writing these sort of things and doing the research; I'm rather terrible at these kind of things. My opinions are not often in line with teams and professional analysts that at this point I kind of figured I'd save myself the embarrassment and keep what little dignity I had.

I look at just two metrics (expected goals and duels won) coupled with the player's salary to make my decision using SBNation's mock expansion protection list, found here, to determine whether that player would be available or not. The bottom line here is simple, create a starting line-up that could conceivably make a run for the playoffs and abiding by the rules of the MLS expansion draft (e.g. not drafting multiple players off the same team). Insane? Can't be done? A complete waste of time? Probably all the above.

All the basics should be covered and anything else I'll mention on the fly. Let's get started.

GOALKEEPER  (~$49,000 spent)

Jeff Attinella, Real Salt Lake, Goalkeeper - $48,825

We don't have a ton of data on Attinella, but the best things to know are the following three: 1) he's cheap, 2) at 26 he's also young(ish), and third, with the two years of data and 82 total attempted shots against him, our expected goal model has seen him perform 4.50 goals better than the average keeper. He may not be the sexy pick considering that there are three or four more experienced keepers, but he's the most interesting from an advanced data/economic stand point.

DEFENDERS (~$380,000 spent)

Jair Beneitz, FC Dallas, Left Back - $97,875
I feel like Beneitz gets overlooked despite his ability for two reasons. First, Michele. Second, because he's not an overly athletic athlete and those players for tend to be overlooked in general. He's the top expected goals creator (0.19, per 90 mins) from a back line position and the top duel winners (8.18, per 90 mins). He may be the biggest no brainier of this whole process.

Jalil Anibaba, Seattle Sounders, Central Defender $159,620
David Horst, Houston Dynamo, Central Defender - $72,750
I choose both Anibaba and Horst as a team. Anibaba is a good possession/passer out of the back line and that shows from his .11 expected goals per 90. A long with that Horst is an awesome ball winner in central defense with over 6 duels won per 90.

If I learned much about central defending in the past year, it's this: you have to have a tag team with qualities that balance each other out. The two might not be enough win an MLS Cup but together should be enough to get you to the playoff spot.

Also, there wasn't much depth here... the pickings were slim.
 

Alvas Powell, Portland Timbers, Right Back - $48,828 (Drew's note: Sorry dude, there is a 100% chance that the Timbers protect him)

Okay, I cheated a tiny tiny bit on this selection. He was covered by the original mock expansion draft protection list submitted by Stumptown footy. However he was the last one chosen for on the roster and since the team just traded for Nat Borchers I feel that bumps him from the list and makes him a candidate here. Which is great because decent defensive depth pertaining to MLS is thin let a lone the bottom of the rosters.

The Jamaican is a poor man's DeAndre Yedlin in some ways. Young (only 20 years old), lots of speed down the right wing but also seems very capable to use his gifts to track down the ball and regaining possession (7 duels won per 90 minutes). But he also has the highest expected goal creation (.21 expected goals per 90 minutes) marks outside on the list. It was Benetiz before I cheated. But now... yeah. Exceptional young talent for the future.

MIDFIELDERS (~$384,000 spent)

Soony Saad, Sporting Kansas City, Left Midfielder - $51,500
Kansas City has a bit of attacking talent on the roster that won't all be protected. I really was drooling all over the nearly 9 duels won per 90 minutes by CJ Sapong but the fact is that he costs nearly twice as much as Saad ($112,000 vs $51,500)  and creates less on the attacking end (.33 compared to Saad's .39 expected goals created per 90 minutes).

Leo Fernandes, Philadelphia Union, Central Midfielder - $120,000
With exception to the one guy that I love, who also hasn't yet been mentioned, I wonder if Fernandes could be the steal of the draft for me. Fernandez was marvelous in the matches I watched.  His goal creation numbers (0.61  expected goals per 90 minutes) are very similar to Javier Morales (.56) and Brad Davis (.54) within the same filter. The biggest problem here is that it's an extremely small size of just under 1,000 minutes over two years. He could flop, he could blow away the league. Oh, he also averages over 5.68 duels won per 90 minutes. Which is also really good.

Khari Stephenson, San Jose Earthquakes,  Central Midfielder - $68,336
He's a veteran and while that kind of label draws tons of yawns and eye rolls there is something to having experience. He uses his speed to win balls, 5 duels per 90 minutes, and can aid the attack with .23 expected goals created per 90. Sure, kill me over this pick if you want. There aren't a lot of  central midfielder that are the right blend of xGoals, Duels and still cheap. He's the best bang for your buck here.

Sebastian Fernandez, Vancouver Whitecaps, Right Midfielder - $143,000
I liked Fernandez this past season. I liked him more when he came out of the midfield than at striker as he didn't seem to be the "scoring" type. He averages roughly .30 expected goals per 90 minutes and while that's okay you need more (really, a lot more) out of your striker. Though the near six duels won per 90 is really nice.

FORWARD (~$228,000 spent)

Patrick Mullins, New England Revolution, Midfielder/Forward - $100,000
Not sure if you've seen my tweet coverage in the build up to MLS Cup, but it's pretty much had bits of Mullins through out it. I can't get enough of this guy and while I was mostly on the fence about whether he could be a legit starter in this league, his performance in the MLS Cup pushed me over the line. I think Mullins is probably the best available player in this draft and he's my #1 overall pick here. .49 expected goals per 90 (just a touch higher than Lee Nguyen) and 5.56 duels won per 90 both lead the Revolution for individuals that saw more than 1,000 minutes on the pitch.

Luke Moore, Toronto FC, Forward - $128,000
True story, Luke Moore collected more expected goals (13.45) than the following:

Jermaine Defoe (13.14)
Quincy Amariwka (12.67)
Joao Plata (12.24)
Fabian Castillo (11.79)

Moore was one of the few reasons I thought Chivas could pin ball through some teams and cause some chaos by just playing a brand of wrecking ball. I was excited that they went after someone outside of their "brand" and was hardly disappointed when they traded him to Toronto just a few months into the year. Moore also grabbed 6.6 duels won per 90, which ranked him inside the top-20 of MLS starters.

FINAL TALLY (~$1,041,000 spent)

There are a lot of really interesting players on this list that just are too big of risks and unknowns to take a chance without knowing who A) your drafting for and B) how you plan to fill out your roster beyond the expansion draft.

A prime example is Krzysztof Król (Montreal Impact) who played 961 minutes produced .41 expected goals and won a little over 5 duels per 90 minutes. All solid numbers but I went with Horst because he's cheaper (kind of the point of this thing) and but also a more established commodity.

Also, you could have Jalil Anibaba easily enough because he's out of contract with Seattle. But trying to go into that much detail here defeats the purpose and really makes this whole process a bit more difficult than needed. I drafted because he's listed as available and he suits what I feel is the need in the situation.

Any thoughts or feed back on players you liked or wished I had mentioned? Leave your comments down below.

An ASA MLS Cup Preview

By Harrison Crow (@harrison_crow)

Here we are, the final match of the 2014 season. The MLS Cup, a game for the whooooooole enchilada, the whoooooole ball of wax, aaaaaaaall the marbles. Okay, let's end my John Candy moment and press with the real stuff.

It's been a long, fun and strange season when you look back on it. One club produced possibly the most dominating statistical season ever in MLS history. And here they stand at the last point in the season to bring home a trophy. The other team is one of the younger and more energetic clubs in MLS playing fun, attacking soccer. They are a band of lesser appreciated talents wearing a chip on their shoulder and seeking to prove themselves, led by Lee Nguyen.

TEAM	xGoals For	xGoals Against	xGoalDifference	xGoal Even Difference	Duels*
LA Galaxy	62.9	32.98	29.92	18.02	1534
New England	47.26	46.92	0.34	-3.74	1529

As you can see above, the LA Galaxy stand out far above any competition in the league. The Seattle Sounders are probably their closest competitor, and with them being dispatched in the Western Conference Finals, this becomes their cup to lose.

Most handicappers generally 'gift' the home club a half a goal going into the match, added to the top of what we would expect from their goal scoring talents. The LA Galaxy and their outrageous scoring ability and tremendous defense have a clear advantage in this situation because of those details.

That said, you're talking about what was a top-5 MLS defense last season, and one that held it's entire line together. While it's had its faults this season in terms of allowing opportunities, we've seen some very strong games from them that it wouldn't be outside the realm of possibility to keep LA off the board, or perhaps to allow a single goal in this match.

Right now our projection system has LA Galaxy as just over a 75% favorite to beat the New England Revolution. A very huge gap that is reflective of their season accomplishments.

Let's dig a bit deeper into these teams.

New England Revolution

#	POS	Player	Age	HT	WT	Min.	xG	xG/90	Duels Won	DW/90	Salary
13	M	Jermaine Jones	33	6'	170	613	1.49	0.22	57	8.37	$ 3,252,500.00
27	F	Jerry Bengtson	27	6' 1"	165	238	1.21	0.46	17	6.43	$    144,000.00
39	F	Geoffrey Castillion	23	6' 3"	170	14	0	0.00	1	6.43	$    329,033.00
11	M	Kelyn Rowe	22	5' 8"	150	1960	8.14	0.37	138	6.34	$    181,000.00
23	D	Jose Goncalves	29	6' 2"	180	2381	1.63	0.06	151	5.71	$    479,375.00
10	F	Teal Bunbury	24	6' 2"	185	2446	8.23	0.30	154	5.67	$    233,000.00
7	F	Patrick Mullins	22	6' 1"	175	1198	6.5	0.49	75	5.63	$    100,000.00
24	M	Lee Nguyen	28	5' 8"	150	2750	14.59	0.48	170	5.56	$    193,750.00
2	D	Andrew Farrell	22	5' 11"	185	2840	1.96	0.06	150	4.75	$    176,000.00
30	D	Kevin Alston	26	5' 9"	160	683	0.98	0.13	35	4.61	$    143,333.00
6	M	Scott Caldwell	23	5' 8"	150	1553	1.08	0.06	76	4.40	$      59,813.00
12	M	Andy Dorman	32	6' 1"	160	1237	1.07	0.08	59	4.29	$    135,000.00
5	D	A.J. Soares	25	6'	170	2806	2.68	0.09	118	3.78	$    159,180.00
25	D	Darrius Barnes	27	6' 1"	175	1628	2.15	0.12	67	3.70	$      86,666.00
16	M	Daigo Kobayashi	31	5' 10"	155	1867	3.49	0.17	74	3.57	$    136,666.00
4	M/F	Steve Neumann	23	6'	165	538	1.97	0.33	21	3.51	$      86,250.00
8	D/M	Chris Tierney	28	6'	170	1930	6.28	0.29	69	3.22	$    103,333.00
15	D	Stephen McCarthy	26	6' 5"	190	90	0	0.00	3	3.00	$    132,582.00
9	F	Charlie Davies	28	5' 10"	160	1143	4.39	0.35	37	2.91	$      78,940.00
92	F	Dimitry Imbongo	24	6'	160	32	0	0.00	1	2.81	$    122,375.00
14	M	Diego Fagundez	19	5' 8"	140	2142	9.45	0.40	55	2.31	$    137,200.00
33	M	Donnie Smith	23	5' 11"	160	45	0	0.00	1	2.00	$      36,500.00
17	F	Andre Akpan	26	6'		0	0	0.00	0.00	0.00	$      51,333.00
40	GK	Larry Jackson	24	6' 1"	180	0	0	0.00	0.00	0.00	$      36,504.00
21	M	Shalrie Joseph	36	6' 3"	195	0	0	0.00	0.00	0.00	$    294,000.00
44	D	Jossimar Sanchez	23	5' 9"	170	0	0	0.00	0.00	0.00	$      36,504.00
36	GK	Luis Soffner	24	6' 4"	210	0	0	0.00	0.00	0.00	$      36,500.00
19	M	Alec Sundly	22	6'	170	0	0	0.00	0.00	0.00	$      36,504.00
99	F	Tony Taylor	25	6'	170	9	0	0.00	0	0.00	$      79,371.00

It kind of feels like this whole team's attack, and possibly its existence to this point in the season, is based upon Lee Nguyen. It is... but then again, maybe not. Nguyen has turned into something beyond what the Revolution could have possibly imagined him being this season. and timely goals have had a lot to do with the Revolution's success. His 14.59 total expected goals (xGoals + xAssists) over the season indicates his imperative role in the Revolution's attack.

However, while many people want to wave the flag of Nguyen, the reality of the situation is that he doesn't rank in the top-10 in terms of creating scoring chances, that according to our shot metrics.

Nguyen has been the team leader, and as mentioned, scored some mighty crucial goals. That being said he hasn't been entirely alone, and had some help from key pieces a long the way.

An interesting story line for me is that which concerns Diego Fagundez. The narrative seems to be that he's regressed in terms of goal scoring and assists ability, and that this season was a step back overall. Yes, that is true, but Fagundez is still probably the second most dangerous piece in the Revolution arsenal. He's produced a fair share of scoring chances for the attack and finished second behind Nguyen in total expected goals created for the Revolution. 

The thing is, as Drew mentioned on the podcast, Fagundez didn't make an appearance in the Eastern Conference finals. Instead Jay Heaps chose to go the direction of Kelyn Rowe in the first leg and then Chris Tierney in the second for his defensive prowess and ability to mark out the Red Bull's Lloyd Sam. The question going into this weekend is how will Heaps negotiate the situation with what is likely New England's second best attacker on the bench, and needing to score goals against the best defense in MLS.

The problem isn't just slowing down the mighty Empire's fleet of Zardes, Keane and Donovan. It's the in the ability to produce goals too. Omar Gonzalez and AJ DeLaGarza are two of the most dominate centerback pairings in MLS, and even if Bruce Arena decides to push AJ out wide to a full back spot, Tommy Meyer or Leonardo both provide a superior ability to win duels and limit shots at the keeper.

If the Revolution want to win, they're obviously going to have to find a way to outscore the Galaxy. Duh, right? Well, that doesn't necessarily just start with their attack, which is potent enough, the responsibility for winning this game is going to have to lie with their defense to produce a gem of a performance.

LA Galaxy

#	POS	Player	Age	HT	WT	Min.	xG	xG/90	Duels Won	D/90	Salary
19	M	Juninho	25	5' 7"	145	2758	4.83	0.16	141	4.60	$    325,000.00
10	F	Landon Donovan	32	5' 8"	155	2720	20.66	0.68	109	3.61	$    260,000.00
7	F	Robbie Keane	34	5' 9"	160	2549	25.93	0.92	88	3.11	$ 4,500,000.00
20	D	A.J. DeLaGarza	27	5' 9"	150	2491	0.92	0.03	76	2.75	$    155,000.00
33	D	Dan Gargan	31	5' 11"	175	2400	2.97	0.11	116	4.35	$      48,500.00
11	F	Gyasi Zardes	23	6' 2"	175	2392	14.91	0.56	142	5.34	$    198,000.00
8	M	Marcelo Sarvas	33	5' 10"	155	2218	5.78	0.23	125	5.07	$    192,500.00
6	M	Baggio Husidic	27	6' 1"	172	2197	6.91	0.28	109	4.47	$      90,000.00
4	D	Omar Gonzalez	26	6' 5"	205	1902	2.65	0.13	120	5.68	$ 1,250,000.00
24	M	Stefan Ishizaki	32	5' 11"	165	1898	8.26	0.39	88	4.17	$    213,000.00
22	D	Leonardo	26	6' 2"	185	1837	0.51	0.02	95	4.65	$    105,000.00
14	M	Robbie Rogers	27	5' 10"	165	1372	2.71	0.18	103	6.76	$    167,500.00
21	D	Tommy Meyer	24	6' 2"	175	926	0.26	0.03	47	4.57	$      64,598.00
9	F	Alan Gordon	33	6' 1"	190	511	7.16	1.26	50	8.81	$    206,666.00
16	F	Rob Friend	33	6' 5"	205	376	2.22	0.53	70	16.76	$      91,000.00
2	D	Todd Dunivant	33	6'	175	370	0.37	0.09	11	2.68	$    160,750.00
34	M	Kenney Walker	25	5' 9"	170	329	0.36	0.10	17	4.65	$      48,825.00
26	D	James Riley	32	5' 10"	150	254	0.17	0.06	12	4.25	$      80,000.00
30	F	Chandler Hoffman	24	6'	160	61	0.8	1.18	3	4.43	$      48,500.00
40	F	Raul Mendiola	20	5' 8"	150	57	0.27	0.43	2	3.16	$      36,500.00
38	F	Bradford Jamieson IV	18	6' 1"	165	23	0.08	0.31	2	7.83	$      36,500.00
5	F	Jose Villarreal	21	5' 8"	160	19	0.16	0.76	5	23.68	$      50,700.00
25	M	Rafael Garcia	25	5' 6"	150	16	0.09	0.51	1	5.63	$      48,825.00
36	D	Oscar Sorto	20	5' 8"	155	15	0	0.00	2	12.00	$      50,700.00
31	D	Kyle Venter	23	6' 3"	190	0	0	0	0	0	$      61,000.00
32	F	Jack McBean	19	6'	175	0	0	0	0	0	$      48,500.00
27	F	Charlie Rugg	24	6'	175	0	0	0	0	0	$      48,500.00

It was hardly a far fetched idea that LA could return to an MLS Cup in 2014 after they bottomed out against RSL last year in Rio Tinto. Fast forward 12 months, and Robbie Keane has earned every bit of his MVP award. Not to mention the historic attacking force aided by the development of Gyasi Zardes and the resurgence of Stefan Ishizaki and Baggio Husidic. There was also this guy, Landon Donovan. Not sure if you heard of him. He was really good, too. I heard he's retiring, hopefully the league does something to honor him and what he's done for MLS.

Top to bottom, LA is the clear the favorite...mostly. While, yes, they have dominated in almost every facet of the game thus far into the season, the worrisome figure that I have is that while they don't allow a lot of shots the ones they do allow tend to be of a higher quality. Danny Page has shown that a few high quality shots are more dangerous than a lot of low quality shots.

Going into the game, with all the talk of having a dominant attack, the Galaxy do not have one player inside the top-30 for duels won. A lot of that could be due to their ability to hold possession and their high press, which creates bad passes and turnovers, rather than the need to win duels. Still it's something. Or maybe just half of a something to worry about with New England's monster in the middle.

Jermaine Jones is a beast in the midfield, and while I talked on the podcast about how he could have issues competing with both Marcelo Sarvas and Juhnino in the midfield, the truth is that he's bolstered the New England midfield with a one-man layer of protection. Since Jones signed with the Revs, the team has been 11-1-2. Through those 14 matches they allowed just 16 goals, which includes some high-scoring playoff games. Perhaps more importantly, during that same time the Revs have allowed only about one expected goal per game, a metric that means more in small sample sizes. 

The Rev's defense was pretty good last season, and while they returned all the same pieces, they've fallen away from where they we last year, allowing an additional 0.19 expected goals per game. That drop in performance is partially due to Jose Gonclaves not being the transcendent talent he was one year ago. Last season Goncalves was the team leader in duels won and made the top-10 in MLS, leading to his defender of the year award. Despite that, Goncalves has fallen away from where he was last year and it's distinguished by the fact he doesn't even lead his own team in duels won.

If the Galaxy want to win, they're going to have to continue to expose the opposing defense and play their usual dominant brand of soccer. This shouldn't be hard, playing at home. What may disrupt their dominance is New England's versatile midfield, featuring the attacking Nguyen, and the tougher, more defensive Jones.

It's the Empire's game to lose, and while that seems to slight the Revolution, it's more about the fact that LA has been one of the most dominating teams in MLS history. It's compounded by the fact that emotions are going to be running high, with it being Landon Donovan's last career match, a match in which the most decorated American soccer player is going to be looking to go out on top.

While I favor the Galaxy 3-1 in the end, my heart will be rooting the classic underdog in Lee Nguyen.*

*Drew favors the Galaxy 2 - 1, while Matthias thinks it will be 2 - 0. 

How Proactive is Your Favorite MLS Team?

By Jared Young (@JaredEYoung)

Jonathan Wilson, most notably the author of the soccer epic Inverting the Pyramid, wrote a piece  in 2012 for the Guardian called “The question: Position or Possession?”. In it, he discusses the merits of both possession and position and supposes that it’s difficult if not impossible to control both possession and position simultaneously. He cites Barcelona and their tiki-taka philosophy as a team that aspires to extreme levels of possession, while Chelsea prefers to maintain position on the pitch defensively in order to take advantage of space behind the offense.
The post got me thinking about how to measure whether or not a team was proactive in their use of the ball or reactive. I’m not convinced those are the correct terms to describe the two strategies, but I’m not sure there are two terms that define the extreme type of play adequately, primarily because I think the words mix two dimensions. The chart below illustrates the two dimensions on which a team builds its strategy.

The first is a passing dimension. On the one extreme, there is a direct passing philosophy which looks to push the ball up quickly to take advantage of space once a team gains possession. At the other end is an indirect passing philosophy which involves short passes that build possession up the field and eventually attempt break down the defense.

The other dimension is where a team begins their initial line of resistance; the pressing dimension. There are teams that give high pressure up the pitch hoping to get a turnover closer to the goal (e.g. Barcelona). There are also teams that allow opponents deeper in order to maintain defensive formation and positioning longer, and thus use a low pressure technique (e.g. Chelsea).

Teams tend to pair two ends of each dimension, so that there are two notable styles of play. Teams that press high also tend to pass more indirectly. This is mostly due to the physical demands of the high press. Teams cannot afford to take risky passes after working so hard to gain possession. Similarly, teams that sit back and offer low pressure tend to pass the ball more directly. Of course, as Wilson mentions, the best teams are able to do all of the above, and they can play any way needed based on the score of the game and what the opposition is doing.

To measure a team’s level of proactivity or reactivity we need to understand what statistics characterize games at the extremes. I located every MLS game in 2014 where one team held 65% or more of the possession. The thinking there is those are matches where one team is looking to cede possession and one team is looking to control possession. These possession levels are not necessarily due to dominant performances, but rather games where extreme tactics are carried out.
I deleted games that were decided by 2+ goals or more. The rationale there is game state can dramatically shift a team’s strategic intent during the course of the game. Closer games are more likely to have teams consistently employ their original strategic intent.

There were 16 such games played in MLS in 2014. Here is a breakdown the statistical results of the games. The team with higher possession is considered the proactive team and the team with lower possession is the reactive team. The average possession for the proactive teams is 68%.
Here are shooting results between the two styles of play.

Style	Shots Per Game	Finishing Rate (%)	Shots on Target (%)
Proactive	16.8	4.9	29.9
Reactive	9.2	12.2	44.9

Proactive teams shoot twice as often as reactive teams,  but they get fewer of those shots on target and score less frequently. This is because reactive teams are keeping their defensive shape and pressuring much closer to their goal, making it more difficult for the opponent's offense to get shots to the target or convert them.

There are similar differences in regards to passing.

Style	Passes	% of passes in final 3rd that were crosses	% Long Passes (25+ yds)	Pass Completion %
Proactive	530	18%	13%	83%
Reactive	259	13%	21%	68%

Extremely proactive teams average 530 passes per match, while extremely reactive teams average 259. The ratio of passes in the final third also holds up consistently, 152 to 86. However, the types of passes made by each team has some notable differences.

Proactive teams use 5% more of their passes in the final third on crosses. If defenses are truly keeping their compact shape, then offenses generally work the ball out wide to the available space.  More crosses make sense. Reactive teams strike 8% more of their passes 25 yards or longer (for our purposes, 25+ yard passes are called “long passes”). This is an indication of direct play. The longer passes leads to a lower completion rate. Don’t be fooled by the lower overall pass completion rate of reactive teams; the completion rate difference is due to risk-taking when they have possession. More direct passing comes with more risk as the passes are longer and are generally to players who are on the move. Proactive teams complete 15% more passes than their reactive counterparts.

While this analysis certainly isn't comprehensive or perfect, I do believe it allows accurate insight into how teams would operate at the extreme ends of the strategic spectrum.

Proactive Score

The goal is to develop a simple score that shows fans and analysts where on the spectrum a given team plays. To do that, I started with the variables that had the most spread in the analysis above. I didn't look at shooting statistics, as those measure an outcome of the style of play, not the style of play itself. The two strongest differences were total number of passes and long pass percentage. Of minor comfort is that a multivariate regression including total passes attempted, long passes attempted, and a home/away flag were all statistically significant in predicting possession of MLS teams at the game level with an Rsquared value of 70%.

That said, using passes to identify large spreads in possession is nothing short of obvious. But without Opta data, there are limited choices for fans to use to determine a team’s strategic intent.
After trying a combination of descriptive metrics and looking at the corresponding spread of results, I landed on using total passes attempted and two times long passes to create a score. While multiplying long passes times two adds complication to the formula, the larger spread does create more separation in the target metrics.

From there I scored a team’s performance on a scale of 1 to 7, with 7 being an extremely proactive team. The result when looking at the games in 2014 is a reasonably normal bell shaped curve.

The metrics that result from the modeled Proactive Score don't have as much spread as the small sample of 16 games at the extremes, but the model rank orders in line with the sample. In short, It isn't distributed as perfectly as our smaller sample size, but the trends remain

Pscore	SOG/Shots	Home Possession	Pass Completion %	Finishing %	Opp. Finishing %
7	35%	63%	85%	11%	13%
6	37%	59%	84%	10%	12%
5	36%	57%	81%	11%	11%
4	34%	51%	80%	11%	11%
3	40%	48%	78%	14%	11%
2	39%	46%	74%	12%	7%
1	43%	39%	69%	15%	10%

Proactive teams have more trouble finding the target due to the reactive team's defensive positioning and density around the goal. Reactive teams get more shots on target because they are taking their shots with more open space between the shooter and the goal, and are more likely to be on a breakaway. Shooting percentage for all teams is correlated with the shooting percentages in our smaller sample of the 16 lopsided possession games, though proactive teams shoot much better than the initial sample indicated. This could be because I chose to only look at close games which may indicate unlucky shooting by the proactive teams.

So what can this Proactive Score tell us about the 2014 MLS season? Here is a table of each team and their score on both the road and at home. Playoff teams are in color – orange if they were proactive, red if they were reactive.

Team	Total Pscore	Away Pscore	Away PPG	Home Pscore	Home PPG
LA Galaxy	4.9	4.3	1.2	5.6	2.3
Columbus	4.7	4.5	1.2	4.9	1.8
Seattle	4.7	4.3	1.6	5.2	2.2
Real Salt Lake	4.6	4.4	1.1	4.8	2.2
New York	4.4	4.1	0.9	4.6	2.0
Sporting KC	3.9	3.3	1.6	4.5	1.4
Vancouver	3.8	3.2	1.1	4.4	1.9
Portland	3.8	3.4	1.6	4.1	1.4
Houston	3.7	3.5	0.6	3.8	1.6
Colorado	3.6	3.6	0.5	3.5	1.4
San Jose	3.5	3.1	0.4	3.9	1.2
Toronto FC	3.5	3.4	1.1	3.6	1.4
Philadelphia	3.5	3.3	0.8	3.8	1.5
D.C. United	3.4	3.0	1.3	3.8	2.1
Montreal	3.3	3.0	0.3	3.6	1.3
FC Dallas	2.9	2.7	1.0	3.1	2.3
New England	2.8	2.3	1.3	3.4	1.9
Chicago	2.8	2.5	0.8	3.1	1.4
Chivas USA	2.7	2.1	0.8	3.3	1.3

A couple of things jump out. The first is that the seven most proactive teams all made the playoffs. Then there is a fairly noticeable gap before more playoff teams are found.  Only three reactive teams made the playoffs; the New England Revolution, FC Dallas and D.C. United.

Next, we can test the theory that the reactive teams were indeed reactive. A plot of shots taken by finishing rate will indicate if the reactive teams took less shots but finished at a higher level. The orange dots indicate proactive teams that made the playoffs. The red dots indicate the three reactive teams. Blue dots represent non-playoff teams.

Both FC Dallas and D.C. United show the shooting results of a reactive team. They ranked 17th and 18th in the league in shots taken but were 1st and 2nd in the league in finishing rate. Unlike the other two reactive teams, New England is in the middle of the pack from a shooting perspective. This is peculiar given the Revolution were the most reactive team that made the playoffs. More work would need to be done to understand the secret to the Revolution’s success from a number’s perspective. They attempted fewer passes with a higher percentage of long passes than other playoff teams, but were able to attempt an above average number of shots and finish them at an average rate. They were also slightly better than average defensively.

According to Jonathan Wilson, reactive teams prefer position to possession with a primary goal of keeping defensive position. The three reactive teams that made the playoffs were all ranked in the top eight in the league in fewest goals allowed, with D.C. United tying the L.A Galaxy for best in MLS.
Measuring a team’s tactical intent has a long way to go, and this is a very humble initial look. More detailed Opta data would allow for much deeper understanding of a team’s choice between possession and position. The importance of putting a measurable number to these tactics is so other statistics can be understood relative to that context. If we can understand what makes one reactive team successful while another fails, we can better understand the effectiveness of home and away strategies, in-game adjustments, and why (or why not) teams are successful.

My nagging issue with DC United and their Coach of the Year

By Harrison Crow (@Harrison_Crow)

I hate crapping all over off-season awards. Really, I do. The idea of me telling you the way MLS and voters determined something is wrong and/or different than how I would determine it, even though I have zero inside track on how any of them really defined the measurables of the award (outside of the vote conducted). It's all stupid. This is just the subjective nature of the award process.

Now that you know that little bit about me, let me now show you that I'm also a hypocrite as I do exactly the opposite here in talking about how the Coach of the Year award was distributed.

Ben Olsen is probably a pretty good coach, I don't know for certain. I don't know how to measure the set of instructions that he gives prior to the game or his half time adjustments relative to his counter parts. I don't know who he really wanted for his team in the last off-season and who he fought to acquire to give the United the best chance to compete.

What I know is very simple and logical; no one really knows any of that stuff. What happened in those important conversations and the metrics that we use to "quantify" the events are laden with noise.

It's subjective, and probably a bit bias, to think that Óscar Pareja or Greg Berhalter or Sigi Schmid deserved the coach of the year any more than Olsen. We are gauging the coaches accomplishments off the accomplishments of the players. Which means the teams talent level has an associated level of value in the determining how successful the coach was. Which seems a bit unsavory.

My counter point to this; if talent wasn't involved than why wouldn't we be talking about Wilmer Cabrera and the job he did as the proverbial caretaker for the club formerly known as Chivas USA?

Assuming that talent and accomplishment within the team isn't a real stretch for determining how those that voted applied some weight of value. We understand that Olsen helped usher his team towards the first place finish in the East and the end of season results convey importance to people.

The problem that I have is that I feel this result was a very luck driven.

Much of the praise for Olsen lies with the defense that tied for the fewest goals against in MLS. The problem that I have with that is that has more to do with their keeper and luck than it has to do with actual ability and "stalwartness". Overall DC by our measures of expected goals DC United had a pretty average defensive season finishing 13th overall with 1.43 expected goals against. Some people could be even a bit harsh and say it was a sub-par season.

The difference between them having 49 goals against versus the 37 that were actually scored can mostly be attributed to the outstanding play by Bill Hamid. I'm sure there are other factors that I'm missing here but considering that Bill Hamid was good for saving almost 11 goals above the average keeper (best in MLS) it speaks to the situation of the defense possibly being less than they really appear.

And if it wasn't for the run of luck with the defense their attack was paltry at best. Their expected goals scored was second worst (only to Chivas USA) for the second year in a row. This drove their expected goal differential from being 15 to being -10. A total 25 goal swing on the basis of luck and the development of one of the best young goal keepers in MLS.

I'm sure there are situations and arguments against Berhalter, Pareja and Schmid. But handing out this award is a reminder that the awards season is about narratives. Worst to First sticks out in everyone's mind and that's alright.

I'm not saying Olsen didn't deserve the award because his team in reality sucked. I'm saying the basis for which it appears most determined  his worthiness for the award may have come by way of a false premise. Olsen may very well have deserved the award based upon the merit of his skill, effort and dedication. But pinning it on his ability to "coach" defense and then backing it up by showing they allowed the least goals in the league seems hardly tenable to me.

Honestly, if you ask me, the league's BEST coach wasn't even on the ballot. I'm not even a Galaxy fan and I think that.

How Luck Integrates With Shots And Our Expected Goal Model

By Harrison Crow (@harrison_crow)

Let's talk shots and luck.

It's obvious that shots become goals by way of two distinct systems of measurement. First quality and then quantity. The better look you get and the frequency at which you take the chance to score implies the amount goals you might score.

This is the very premise by which we created our expected goals model. We've identified certain factors that we've found to create favorable shots. Likewise how often that type of quality shot would reasonable end in a goal.

Now the problem comes with things that we cannot account for in those factors of quality such as defender and keeper placement. A keeper being out of position would obviously improve the quality of the attempted chance and a defender in the right position would lower the quality.

Luck is inevitably part of the system that is active in quantity. Alan Gordon took a very high probability shot against Stefen Frei in the closing seconds of the Galaxy-Sounder match on Sunday evening. He used his most common foot, he took it from close range and he had a large percentage of the goal exposed to him. However, a bad touch created the time necessary for Chad Marshall to get into range to block the shot.

The likelihood of that opportunity becoming a goal is weighted by other scenarios in past instances of other attacks in similar situations and factors. Would Robbie Keane have had so many issues with that shot? How about Landon Donovan or Gyasi Zardes or even any other striker?

The gap between the elite and the good and the average has shown to be much smaller than most would think. Looking at our expected goal table you can see the "G -xG" category which is simply the amount of goals scored subtracted by the amount of expected goals with the variation between the two being around .02 probability per shot. Again, this is not a lot.

Past performance patterns show us G-xG is not necessarily a predictable skill. Eddie Johnson is a great example. With +4 goals in 2012, +1 in 2013 and -1 this year. Keane, a player I think most, if not all, would consider elite within MLS, has shown the same tendencies with 0 G-xG in 2012, +5 in 2013, +2 in 2014. MVP Candidate, Lee Nguyen, posted a -2 G-xG last year and this year is near +9 goals. This is a huge swing and cannot simple be about a jump in a players ability but more about skill WITH luck. We'll get more into this when we talk Jairo Arrieta in the coming days.

This isn't to say that I don't think that skill plays a role. It's obvious fair to say Gordon isn't at the same talent level as Keane, Zardes or Donovan. But that doesn't mean that Gordon is void of talent. I'm not going to say Gordon would make that shot 50 times out of 100. But I wouldn't say that about Keane either. I think there is an edge in Keane's favor, I just believe that edge to be much smaller than most.

It's reasonable to think that luck played a huge part of Gordon not scoring on that shot. I think it's also reasonable to consider this event when evaluating the end game scoreline from either teams perspective. Gordon was in great position to change the game state from a high probability shot. Accounting for that through expected goals gives us insight to how the the striker performed.

These are some things to consider when weighing luck and it's influences on a team's results in scoring goals and allowing them.