Measuring USMNT Results

Introduction

International soccer matches are unique within the sport and The US Men’s National Team is no different. Rosters vary dramatically depending on player form, injuries, the importance of the match – and sometimes at the whim of a player’s club team or national team coach. US players participate in a wide range of tournaments, friendlies, and qualifiers on widely inconsistent field conditions and against opponents of fluctuating quality. Thus, fans, pundits, players, and coaches alike understandably have different opinions on the quality of individual results and the standing of the US program. It is not uncommon to hear statements like “What a great win”, or “that is a terrible loss” after USMNT matches. But what do the adjectives attached to the result truly mean?

I tried to find an answer. The factors that influence opinions concerning a national team result largely can be quantified—and in instances when it cannot, possible options are fairly limited and can be ranked and assigned numerical coefficients. Some logical assumptions are necessary for the model to produce reasonable results, including:

• Games types are ranked in order of importance (including the round of a tournament), with the World Cup followed by the Hex, other qualifying, Copa America, Gold Cups, and friendlies.
• Away results are more difficult, so victories and draws are rewarded while losses are punished less heavily. The opposite factors are applied for home matches. Since 2011, the USMNT has won around 60% of its home games, and only 35% of its away games.
• Results against better teams are more difficult (and therefore should be rewarded) than results against lesser teams, and vice versa.
• To limit bias on roster decisions, roster quality is ranked based on an assessment if the current coach has called in his best possible team in his opinion (excluding injuries).

Overall, the model includes ten variables: location, result, score, squad quality, opponent’s squad quality, winning/losing streak entering game, opponent’s FIFA ranking, game type, rivalry score, and associated tournament risk.

So, with that in mind, I logged every USMNT result back to the first game of Jurgen Klinsmann’s tenure as the head coach, in summer 2011. In analyzing over 100 individual matches spanning from friendlies to the World Cup, a wide variety trends and insights on the performance of the USMNT are evident.

Best/Worst Results

Big wins –and losses – to rivals in tournaments and qualifiers dominate the best and worst leaderboard

Game Types

Gold Cups have brought out the best the US, with good performances boosted by long tournament runs and higher pressure matches.

High Points/Low Points

2013 was a historically successful year for the USMNT, and the numbers bear evidence to that. Klinsmann’s firing represents a recent low for the team.

The 4-0 rout of Costa Rica in last summer’s Copa America receives the highest marks since 2011 thanks to great performances from Clint Dempsey and Jermaine Jones — and favorable metrics that increased the importance of the match. Costa Rica was ranked in the top 25, both teams played their best possible lineups, and the match was a must-win for the US following their opening night loss to Colombia. Those factors, coupled with a commanding 4-goal win boosted this match to the top of the rankings. The second highest scored match, a famous 2-0 win in Columbus versus Mexico that clinched our place in the 2014 World Cup received high scores in the rivalry index, game importance, and Mexico’s quality, but did not qualify as a must-win game or carry the same blowout scoreline (but maybe Dos a Cero should get a category of its own). On the flip side, the worst results combine poor performances against lesser teams in crucial situations. The Gold Cup semifinals in 2015, with a best-possible roster at home against an inferior opponent rightfully ranks as the worst result since 2011. Some unfortunately familiar results stand out — including the match that ended Jurgen Klinsmann’s reign in Costa Rica in November — a result that ranks as 5th worst since 2011.

Opinions come with every game — share yours below