2016 U.S. Women’s Soccer- Blame it on the Rio Olympics – The Grande Finale – Part 4 – Sweden vs USA Quarterfinal Game – “SNAFU” Analysis

facebooktwitterreddit

2016 U.S. Women’s Soccer- Blame it on the Rio Olympics- Part 4 Sweden vs USA – “SNAFU” Analysis. This is a statistical look at the players in this game.

Sweden won the 2016 Rio Olympics Soccer Quarterfinal over the U.S. Women’s National Team 1 – 1 as decided by penalty kicks. But how did the teams perform statistically? Was this meant to be the result, a close game decided by penalty kicks? Or was there something peculiar about the game statistics, and “luck” factored into play? We will look at the game statistics, and see how each player may have affected the outcome.We will be using the “SNAFU” analysis from Part 3 of this “Blame it on the Rio Olympics” series. Of the mathematical models evaluated so far, this analysis seems to be the most reliable indicator of overall player performance.

STAT GEEK WARNING! This article will be less painless to read than previous articles in this series. However, there may be a few bad jokes here. Warning! This article also contains over 4500 words and may take about 15-20 minutes to read.

Introduction

In the previous 3 parts of this series, we looked at various statistical analyses to determine whether the best players from the NWSL actually did go to the 2016 Rio Olympics.

  1. From Part 1, we used a modified player index based on a model proposed by McHale, Scarf, and Folker in 2012.
  2. Then, in Part 2, we simplified the statistics with a ranking system that we called “KISS” for “Keep It Simple Stupid”.
  3. However, in Part 3, we came up a 10 category model that has “fudge factors” that we called “SNAFU” for “Statistically Normalized- All Fudged Up”.

The first 2 systems may have their flaws as the McHale et al. system favored the attacking player, but severely underrated defensive players like Becky Sauerbrunn. With the “KISS” method, defensive players became the top ranking players overall. However, with “SNAFU”, it appears to be reasonable method of combining the 2 prior analyses that gives a better overview of player ranking, including goalkeepers. Also, SNAFU can be used to determine overall team performance over the course of a season or individual game.

Here are the other articles in case anyone wants to read these.

We will be using the “SNAFU”  method to analyze the Sweden versus USA Rio Olympic Quarterfinal.

So, in this study we will look at game data provided once again by Alfredo Martinez and the fine staff at WoSo Stats, and we will determine the best players of the game.  Also, data may be provided by official match results for verification of goals, assists, etc.  (links in sources below).

Can “SNAFU” actually be used for this study?

The overall calculations and methods are described fully in “Part 3” of this series. The only difference is that in Part 3, we were analyzing players over the first half of the 2016 NWSL season. Therefore, the calculations were based on player averages “per 90 minutes”. This means that all parameters were divided by total minutes played multiplied by 90 to give a sense of a “typical game” for that particular player.

Here, we will use the game data, and use the same exact calculation methods. In Part 3, I discussed that the categories of “Appearance” and “Bonus” may not be needed for team comparison. Therefore, in this study we will verify whether both categories are actually needed, not needed, or either Bonus or Appearance may be needed, but not the other in the overall team result.

“SNAFU Statistic reminders”

As a reminder from the “SNAFU” statistic, a Grand Total score of 50 is average.  A player earning 40.25 points was on the bench. 100 is the theoretical perfect score, but I do not cap all of the categories at 10 points, so it may be possible to be slightly above 100 points.

Remember, the theory behind the “SNAFU” scoring scheme is that the overall difference between the 2 team averages should show a difference in goals. So, in theory, if “Team A” has 52 overall points, and “Team B” has 50 overall points, the difference is 2. Therefore, theoretically, Team A won 2-0, or 3-1,  etc.

So, in this preliminary study to test “SNAFU” under game conditions, we will once again look at game data provided once again by Alfredo Martinez and the staff at WoSo Stats. To determine the “fitness” of ‘SNAFU”, I chose a friendly between USA and Switzerland played on October 19, 2016.

Table 1. USA vs Switzerland “Quick SNAFU Summary”.

To describe the below columns, the Grand Total is the original 10 Category “SNAFU” calculation. Then you have the next column with Grand Total Minus Both the Bonus Points and Appearance Points. Then there is a column just subtracting out the Bonus Points only, and the last column with Grand Total subtracting out Appearance Points only.

The rows are self-explanatory. The Average of Both Teams was the “SNAFU” average of all players who played in the game. Then the next rows are sorting out the Swiss and American teams. With the bottom row simply being the USA team average minus the Swiss team average.

To save space and time, I did not include the player data. But, I wanted to give the reader a “feel” for SNAFU” data before we jump into the Sweden – USA game.

Grand TotalTotal- Bonus- AppTotal- BonusTotal-App
Average Both Teams52.1236.3143.6544.78
Average Switzerland (SUI)50.9734.1242.5942.50
Average USA52.9937.9844.4746.51
Difference USA- SUI2.023.861.884.01

So, can anyone tell me the score of the Switzerland- USA game? Anyone – Anyone? Ferris Bueller? 🙂

The answer was 4-0. This game was memorable as Lynn Williams scored about 48 seconds after her international debut to prove “SNAFU”, and my other statistical methods were right, and she should have been at the Rio Olympics. I am kidding.  Well, sort of. 🙂

Are all the categories needed to evaluate the teams?

Back from the tangent, we look at the Grand Total itself. There are only 2 goals difference. The Swiss did put up a good fight, and in alternate universe in which Ramona Bachman  played that could have been the outcome.

When you subtract out the Appearance points, you get close to 4 points. This is logical and obvious, a substitute playing 1 minute will have an appearance  score close to zero, thereby losing 10 points to pull down the team average. Therefore, Appearance can never be used to look at teams side by side.

What about Bonus? Is that needed? My guess is that having the Bonus points might be misleading. The Bonus points may not be a factor if both teams used the same number of substitutions, but will be a factor if you ever need to bring in a goalie.

Although using one data point such as the Switzerland – USA game does not prove “SNAFU” is right, however, it  indicates that we might be able to make some statistical observations, and see where it leads us.

Let’s go the “punchline”, what you have waited for, the Sweden – USA Quarterfinal itself! But first, here are some game highlights, complete with robotic voice that has no idea how to properly pronounce Pia Sundhage’s name…

There are better videos below (in the “other articles and videos” section below, but those direct links may or may not not work.

This is copied from my spreadsheet, Here are the results!

Table 2. Team Summary Sweden vs. USA including Various Totals and Player Ranks.

PlayerTeamGrand TotalTotal (MinusTotal (MinusTotal (MinusPlayer RankPlayer RankPlayer RankPlayer Rank
all categoriesBonus + App)Bonus)Appearance)Grand TotalTotal-B+AppTotal – BonusTotal – App
AsllaniSWE52.533.543.542.516241822
BerglundSWE40.731.431.740.430293028
BlacksteniusSWE51.634.042.643.017211919
DahlkvistSWE54.635.645.644.610141212
ErikssonSWE48.635.239.644.222172315
FischerSWE53.934.944.943.913191517
JakobssonSWE50.333.941.342.919222020
LindahlSWE50.439.449.440.4185527
RolfoSWE42.632.233.641.228262824
RubenssonSWE49.935.340.944.320162114
SamuelssonSWE54.435.645.444.611131311
SchelinSWE53.934.944.943.912181416
SchoughSWE43.632.034.641.026272625
SegerSWE56.637.647.646.671089
SembrantSWE53.834.844.843.814201618
PlayerTeamGrand TotalTotal (MinusTotal (MinusTotal (MinusPlayer RankPlayer RankPlayer RankPlayer Rank
all categoriesBonus + App)Bonus)Appearance)Grand TotalTotal-B+AppTotal – BonusTotal – App
BrianUSA58.239.249.248.25665
DunnUSA45.631.636.640.624282426
HeathUSA64.845.855.854.81111
HoranUSA41.131.432.140.429292928
JohnstonUSA61.142.152.151.13333
KlingenbergUSA61.842.852.851.82222
LloydUSA57.038.048.047.06877
LongUSA49.635.540.644.521152213
MorganUSA56.537.547.546.5811910
O’HaraUSA53.438.744.447.7157176
PressUSA43.532.534.541.527252723
PughUSA56.137.847.146.899108
RapinoeUSA45.333.836.342.825232521
SauerbrunnUSA59.340.350.349.34444
SoloUSA47.136.146.137.123121130
PlayerTeamGrand TotalTotal (MinusTotal (MinusTotal (MinusPlayer RankPlayer RankPlayer RankPlayer Rank
all categoriesBonus + App)Bonus)Appearance)Grand TotalTotal-B+AppTotal – BonusTotal – App
Average (Both Teams)51.936.143.544.615.515.515.515.5
Sweden51.035.042.543.516.817.917.117.6
USA53.437.544.946.013.512.213.112.5
Difference USA – SWE2.42.62.42.5-3.3-5.7-4.0-5.2

Brief Discussion of Table 2.

Note that the the first four columns after Player and Country are the same ones as above comparing Switzerland versus USA. The last 4 columns  the players are ranked in each of those previous “Total” columns.  At the bottom are using the same calculations that were shown in Table 1 comparing the Swiss – USA “SNAFU” data. The players were then ranked based on the total points.

Team averages of Total and the USA minus Sweden Difference at the bottom.are listed below the player scores.

The theoretical score of this game should have been at least 2 to 0 in favor of the USA if you take all the differences in all the averages. The “real score” could have been 3-1 since Sweden’s Blackstenius and USA’s Morgan each scored a goal.

I left the ranking averages there for curiosity, but in all cases the USA was higher ranked.

Brief Overall Player Discussion

Regardless of how you do the math, the top 4 players in the game are very clear. Tobin Heath wins the imaginary “SNAFU MVP” award. First runner-up is Megan Klingenberg, followed by Julie Johnston, and Becky Sauerbrunn. It is kind of ironic that the USA  defensive backline from approximately the 72nd minute to the end of the game are the game’s top 4 most valuable players. The reason why I say “ironic”, I believe it is where the USA “lost” the game. That will be discussed in more detail below.

Caroline Seger was the only Swedish player in the top 10 in all the totals.  Another Swedish player is top 10 depending on the calculation (3 out of 4 times). But. there are 11 players per team. That is complete USA domination!  But I think goalkeepers would want to be ranked without the bonus points. Lindahl was 5 and Solo was 11 or 12th if you take away the bonus points.

Table 3. Individual player scores for each category.

These are the players’ individual scores in each of the categories. Of course, if you these add up all the columns you will have the Grand Total in Table 2 above. I also listed the “fudge factors” described in Part 3 of this series as a reminder of some of the math involved in each category.

PlayerAppearancePunishmentGoaliePossessionAttack PassPass compSet pieceDefenseBonusGoal Scoring
105.004.504.503.754.504.5094.50
Asllani10.004.884.504.504.495.465.1794.50
Berglund0.325.004.634.503.754.504.5094.50
Blackstenius8.634.483.565.504.084.504.3797.50
Dahlkvist10.004.885.577.354.424.634.2394.50
Eriksson4.354.885.176.513.814.905.4494.50
Fischer10.005.004.904.504.694.506.7894.50
Jakobsson7.344.754.776.423.824.505.1794.50
Lindahl10.005.008.004.774.943.214.504.5014.50
Rolfo1.455.004.505.343.824.504.5094.50
Rubensson5.655.004.905.094.375.445.9794.50
Samuelsson9.764.745.444.505.026.245.1794.50
Schelin10.003.985.576.364.694.505.3094.50
Schough2.665.005.174.503.684.504.6394.50
Seger10.004.884.906.186.094.846.3894.37
Sembrant10.004.885.044.503.884.847.1894.50
PlayerAppearancePunishmentGoaliePossessionAttack PassPass compSet pieceDefenseBonusGoal Scoring
105.004.504.503.754.504.5094.50
Brian10.005.005.844.647.844.505.9795.37
Dunn4.924.993.964.503.884.505.4494.37
Heath10.005.136.647.346.038.325.8496.50
Horan0.735.004.634.503.754.504.5094.50
Johnston10.005.256.244.787.735.188.3994.50
Klingenberg10.005.005.847.767.285.786.6494.50
Lloyd10.004.365.308.095.294.846.1193.96
Long5.085.135.304.646.094.845.0494.50
Morgan10.005.124.506.364.294.504.7797.96
O’Hara5.735.135.576.485.835.455.9794.23
Press2.025.004.635.343.884.504.9094.23
Pugh9.355.135.977.274.824.505.8494.23
Rapinoe2.424.885.174.503.606.554.7794.37
Sauerbrunn10.005.007.314.647.864.636.3894.50
Solo10.005.004.004.634.644.024.844.5014.50
PlayerAppearancePunishmentGoaliePossessionAttack PassPass compSet pieceDefenseBonusGoal Scoring
105.004.504.503.754.504.5094.50
Average (Both)7.434.926.005.145.514.834.985.458.484.74
Sweden7.664.848.004.935.284.444.795.288.534.72
USA7.355.014.005.445.705.485.165.678.474.81
Difference US-SW-0.30.2-4.00.50.41.00.40.4-0.10.1

So, why did not the USA win? Where are the missing 2 goals?
Sweden only led in Goalkeeping throughout the game. Hedvig Lindahl kept them in the game with some nice saves. USA’s Hope Solo did not see many shots, and her low save percentage resulted in a low Goalkeeping score for her.Brief Discussion of Table 3.

The statistics for Appearance and Bonus points are meaningless for team comparison. But USA was dominating overall in Passing, Attacking Passes, and Set Pieces, and Defensive Skills.

The “real game”  goal scoring brief summary:

The goals by Stina Blackstenius and Alex Morgan are “real”. There were 2 goals in the game that did not happen because of referees’ decisions. Lotta Schelin was declared to be offside for Sweden, in which replays show that it was very close, and that she was probably onside. Carli Lloyd had a goal called back. There are 2 possibilities, that she was either offside, or it was a foul. I believe a foul was called, but there was some confusion about this. So, if my assumptions are real, then Sweden should have won 2-1.

We are now back to “SNAFU fantasy statistical soccer”. So, if Sweden should have won 2 to 1 based on actual performances on the field, then why do I believe USA should have won 3 to 1, or perhaps 4 to 2?

Anyone? Anyone? Sorry, I used that joke already. This is not an “official SNAFU” statistic, but the USA out-shot Sweden 26 to 3. Shots on target were 5 to 1 in favor of USA. So, what can “SNAFU” tell us about the game?

“Missing Goal #1”

We will look at Tobin Heath as she was the best player statistically. She had over 64 points if you use the Grand Total. This means by herself, she should have accounted for more than a goal if you were to take the team average. She is not alone as Klingenberg,  Sauerbrunn, and Johnston all made contributions to each give the team a chance at goal. Statistically, Heath played a great game. She was above the average as far as Possession, Attacking Passes, and Defense. She was only  one of four players to be above 4.50 in the Goal Scoring Category. 4.50 implies you did not shoot all game, and anything less than 4.50 implies that you were shooting and was not on target. By rights, Heath should have had a goal or an assist.

“Missing Goal #2”

Klingenberg, Sauerbrunn, and Johnston all did a  good defensively, except the lapse where Blackstenius broke free and scored. But they had above passing scores and defensive skill scores.

The problem is that other than Klingenberg and her ability to cross, there are no goal scoring opportunities here. The only way to make the math work here is that the defenders “erased” a goal that would have normally happened otherwise. Since Sweden only had 3 shots, I will pretend that in an alternate universe that Sweden may have had 5 or 6 shots more, but the USA defense took them away, and reduced to 3 shots. Therefore, “eliminating” a goal defensively is the same statistically as scoring a goal, at least according to the “world of SNAFU”.

I know it’s a fairly lame explanation, but I am trying to make the statistics work to match what may have happened.

Other Options?

If you want a better explanation, how about in our “alternate SNAFU universe”, Morgan Brian is the attacking center midfielder and passes the ball to (anyone) and they score?

Maybe on a set piece by Tobin Heath, one of the defenders such as Johnston “heads” the ball in for a score?

All are theoretically possible in the SNAFU statistics with the way that they played.

Let us see why the USA did not score those 2 “imaginary” goals that SNAFU estimated.

Back to “Real Life”

1. Team formations

Sweden played a “bunker defense” with a 4-5-1 formation. Only the forward was in any attacking position much of the game. The U.S. had a 4-5-1 formation as well.

USA Starting lineup:

Goalkeeper- Solo

Defense (left to right)  Klingenberg, Sauerbrunn, Johnston , O’Hara

Defensive midfielders- Long, Brian

Wings – Heath, Pugh

Attacking center midfielder- Lloyd

Forward- Morgan

What is wrong with the above lineup?

Defense

  • Even though I would have had Krieger in the defense, we will not fully discuss her curious omission from the game. My speculation is that coach Jill Ellis was willing to sacrifice pure defense for attacking defenders. Although I could not tell, but at times, it appeared that both O’Hara and Klingenberg were in the attack, leaving only the center backs Johnston and Sauerbrunn to defend the goal. I am not sure if that is what happened, but I did not see O’Hara and Klingenberg back in time on defense try to stop the Swedish goal. That is not a good strategy to begin with as Johnston and Sauerbrunn are not the fastest players on the field.

Forward Position

  • But, the problem with the attacking position lineup is really two-fold. Morgan is a fine forward, but she cannot be on an “island” by herself. Years ago, when it was Morgan and Wambach, they had a chemistry. Morgan had more speed, and of course Wambach had height. Both could shoot. But with that formation, and their different skill sets meant that had to pass to one another. As I described in the previous editions of this series, Morgan is not great at passing. Therefore, you need someone to pass the ball to her. But also. a solo attacking forward should be the fastest player on the field., and be able to pass to an oncoming midfielder if they cannot break down a defense by themselves. I am not sure if Morgan fits that definition anymore, therefore, she needs a “partner” or two to compliment her.

Attacking Center Midfielder Position

  • That “complimentary” person is not Lloyd. Lloyd herself is a copy of Morgan, a”poacher”,  be at the right place and time to score a goal. But someone has to get the ball to a poacher. And the “number 10” attacking midfielder spot is the last place you want a “poacher”. But if you look at the statistics, Lloyd’s attacking passes were surprisingly quite good. She had a completed cross, was 3 out of 5 for through balls, and had a launched ball completion. The problem was, she would often choose to “shoot first, ask questions later”. According to the MLS site below and WoSo Stats.  she had 5 shots, none on target (with some missing badly), which is 5 wasted possessions.
  • A possible solution to the above issue might be to put Heath or maybe Morgan Brian in that role. With Brian you are getting  her 93% passing efficiency as shown by Brian’s passing completion score of 7.84 as compared to Lloyd’s 5.29. Could this be the 2 goal difference there? Theoretically, 1 point in a category score should lead to a chance at goal in the “SNAFU” system.

2. The substitution fiasco

Back to Tobin Heath, she gets moved to defensive line after Megan Rapinoe substitutes in for Kelly O’Hara. So, why is Rapinoe in the game? She comes in for set pieces, as you can see that is her highest score.. But that was Heath’s job, and she did it well!  Look at Heath’s 8.32 set piece skills score.

But Rapinoe has 2 issues in this game:

  •  First, it was estimated at the time that she was game fit for only 20 to 30 minutes. (I cannot find an article to verify this statement, but here is an article from a couple weeks before the Olympics – http://equalizersoccer.com/2016/07/26/megan-rapinoe-out-uswnt-opening-match-rio-olympics/.)  Historically, Rapinoe was known for her decent crossing ability. After all, her cross to Wamabach at the 2011 World Cup Quarterfinal victory over Brazil is still one of the most memorable goals in soccer history. So, in this game, how many crosses did Rapinoe complete? Zero out of 1 attempt.
  • Second – What was Rapinoe’s passing efficiency? 44%. This gave her a “SNAFU” passing completion score of 3.60, which is hurting your team, by a “theoretical SNAFU goals” of about of about 1.4. .So, is this where we “lost almost 2 imaginary SNAFU goals”?

Heath on defense?

So anyway, Ellis puts Heath in the defensive back line. By doing this, Ellis is taking the “game’s SNAFU MVP” and placing her on defense? So, from the 72 minute to 99 minute, Rapinoe is playing Heath’s role and Heath is defending. At the 99 minute, Rapinoe is subbed out for Christen Press, who is another “poacher”! So, let me get this straight, there are 3 “poachers” attacking, and the game’s best player is defending? So, there is no one else but Morgan Brian and maybe Pugh and Dunn to get them the ball, except Brian herself has defensive midfielder responsibilities. Unfortunately for the U.S., Dunn and Pugh were reasonably neutralized by the Swedes for much of the game.

3. Sweden’s Bunker Defense

How did the USA try to break down Sweden’s bunker defense? Shoot the ball, sometimes very badly about 26 times. If you are going to have 2 or 3 players who like to shoot, but not necessarily pass, that’s what you get. There was no one who could easily penetrate the penalty box area.

Judging by the statistics, it is apparent that Sweden’s bunker defense targeted the forwards and midfielders to prevent them from effectively passing to each other. Only Heath and Brian had decent passing completion scores for the forwards and midfielders.  Sweden also knew if you kept these players outside the penalty area and prevented anyone from making any decent crosses, then they would have a chance to win.

However, the bunker defense limited Sweden offensively. Sweden had a lower passing completion score, presumably they were kicking the ball “out of harm’s way” first. They had only 3 shots all game, which is obviously not counting the “almost goal” from the Schelin offside call.

Back to “SNAFU Unreality”

What would I have done if I were Jill Ellis? I am glad you asked! Sorry, I must have a bad joke now and then. If I was forced to use the 18 players on the roster, and not be creative like promote Sam Mewis from the alternate roster, or use someone like Lynn Williams who could not play at the Rio Olympics. I might have tried something like this:

1. Starting Formation (4-4-2 with a “diamond 4” midfield):

.Goalkeeper- Solo

Defense  (Left to Right)- Krieger, Johnston, Sauerbrunn, Klingenberg

Defensive Midfielder- Long

Wings- Heath, Pugh

Attacking Center Midfielder- Brian

Forwards- Dunn, Morgan

Starting Lineup comments

  • Say what? No Carli Lloyd?!

How dare you take off the 2015 FIFA The Best Women’s Football Player of The Year?! Think about it, she only played one full game in April for the NWSL, and a couple friendlies in July. She may have been physically fit, but her teamwork skills were non-existent during the Olympics and prior to that. I am not a psychologist, but I think  having an angry Carli Lloyd stew on the bench for 70 minutes means that she will try to prove everyone wrong, and will score 3 goals in 16 minutes like she did with Japan at the 2015 World Cup. I do not recall Lloyd being ever “benched” in 2015, but I had a sense she was trying to “prove herself” then.

  • Dunn/Pugh comment.

I think Dunn and Pugh are interchangeable, and maybe have them switch places now and then to try to off-balance Sweden’s backline. Heath is well known for “roving” as well. So, maybe have the 3 move about occasionally? I do not know “mileage” statistics for soccer players in a game, but I would look at using that strategy if it helps keeping 3 of your best players “fresh” the entire game.

  • Backline-

I think having one of the league’s best defenders in Sauerbrunn next to one of the league’s worst defenders in Klingenberg makes more sense to me. The inclusion of Krieger makes sense to me over O’Hara as Krieger is the better pure defender. O’Hara can play midfielder and defense, so she is the perfect substitute if you have an unexpected injury early in the game. The other option could have been starting O’Hara over Long at defensive midfielder. But I do not recall ever seeing O”Hara at that position, so I have no idea at her comfort level at that position. But honestly, Long did not look comfortable to me there at the Olympics.

2. Substitutions

  • 65- 70th minute, take out Morgan and put in an angry Lloyd. I was never convinced that Morgan can be fully effective for 90 minutes. She seems to “disappear” for long stretches of the game. So, I have no idea if that is fitness, or whether the defense has her figured out.
  • At 70-75 minutes, Put in O’Hara for Long. Krieger goes to defensive midfield. O’Hara goes to left fullback where Krieger was. If you are losing the game or tied at this point, then have O’Hara in an attacking position, and have Krieger make sure no one starts a counter attack. Otherwise, if you are leading, go into a bunker defense.
  • If game is tied or losing at 80-85 minutes, then bring in Press for someone like Klingenberg. Bring in an offensive player, but take out your worst defender.
  • In extra time, save Horan for any injured player as was the case with Pugh. Or bring in Engen for an injured defender if you need to hold a lead.

3. Possible Strategies to Break Down the Swedish Bunker Defense

  1. First Half strategy would be to  preach “patience”. Do not waste opportunities, just play the game to the style of the players on the field. However, once in a while have someone from the backline loft a launched ball toward Dunn or Morgan and try to see if they can score  while they have fresh legs.
  2. Second half strategy might be to “burn up” Morgan. Let her chase down some long balls, but rotate Pugh, Dunn, and Heath to keep them “fresh” as possible. But also look for ways to maintain possession where possible. No wild shots, no long balls into open space where there is no hope of getting the ball.

Would this really work?

Honestly, I am not sure if that would be the best strategy. Morgan is not the best at “chasing long balls”.  But having Morgan and Lloyd on the field together makes no sense to me, since they do not have complimentary skills.. Is this why coach Ellis has all these bizarre playing schemes, to keep them both happy? (But a coach is supposed to care about winning, and not making players happy, yes?)

  • Overtime strategy – Keep an eye on the players and be ready to put in Horan in just in case someone has “completely run out of gas”. But do not do it too early, in case you have an injury.

Team/ Player Conclusions

  1. The United States should have won 4-2 based on the performance of Tobin Heath mostly, and the other members of the back line.
  2. Sweden’s Bunker Defense did nothing statistical for their team as it limited the number of shots. However, Sweden did shut down Crystal Dunn, minimized Mallory Pugh’s effectiveness,  and forced Carli Lloyd, Alex Morgan, and others into making several bad shots. Swedish coach Pia Sundhage (and anyone who played the USA team this year) knew that this was the only way to stay in the game, and to win could be through penalty kick shootout.
  3. Megan Rapinoe’s substitution wasted Tobin Heath’s talents on the back line for too much of the game. And there is nothing more wasteful in soccer than a substitute being subbed out,.

Other Conclusions/Comments

  1. Overall, the “SNAFU” system did an okay job at looking at an individual game as compared to a half season statistics previously. It accurately “predicted” that the Switzerland – USA friendly back in October would be 4-0. It also showed that the USA should have defeated Sweden by 2 goals as the USA dominated in all categories but goalkeeping.
  2. I want to thank all the readers who took the time to go through this series with me. To be honest, it is fun for me to see how well a player actually played (statistically) in a game. I may do a few more articles like this in the future.
  3. I want to thank Alfredo Martinez Jr. and the people at WoSo Stats who made this statistical analysis possible.
  4. So, what would have you done differently if you were the coach of the U.S. Women’s National Team? Feel free to leave your comments. No trolls please unless you are the cute animated variety (mainly because they are good eating!) 🙂

Videos

But here is a video of the highlights.I cannot find the complete USA Sweden game on YouTube, so here is The USA-Switzerland 10/19/2016 game in 2 parts.

Other articles and videos, which includes some Sweden vs. USA game summaries and a couple other items:

http://matchcenter.mlssoccer.com/matchcenter/2016-08-12-us-womens-national-team-vs-sweden-womens-olympic-team/recap

http://www.nbcolympics.com/news/watch-top-soccer-moments-2016-rio-olympic-games

https://www.theguardian.com/football/2016/aug/13/usa-women-national-team-soccer-rio-2016-analysis

http://www.ussoccer.com/womens-national-team/tournaments/2016-olympic-games/160812-wntvswe#tab-1

http://www.ussoccer.com/stories/2016/10/20/03/12/161019-wnt-new-look-usa-routs-switzerland-4-0-in-sandy-utah

http://matchcenter.mlssoccer.com/matchcenter/2016-08-12-us-womens-national-team-vs-sweden-womens-olympic-team/lineup

http://www.oregonlive.com/olympics/index.ssf/2016/08/megan_rapinoe_takes_on_new_rol.html

Sources:

  1. McHale, Scarf, and Folker: Soccer Player Performance Rating System , 2012 “On the Development of a Soccer Player Performance Rating System for the English Premier League”  Interfaces Vol. 42, No. 4, July–August 2012, pp. 339–351 ISSN 0092-2102 (print)—ISSN 1526-551X (online). ©2012 INFORMS
  2. I have acquired about a half season’s worth of 2016 NWSL data from Alfredo Martinez, who has https://wosostats.wordpress.com and I also took the data from his report here.
  3. Some data is also taken from the NWSL site
  4. Here is a quick glossary for some of the terms above: https://www.whoscored.com/Glossary
  5. http://soccer.epicsports.com/soccer-glossary.html