Days after someone watches a match, what do they remember? The decisive goals are recalled almost universally, as are the passes that led directly to them (especially if highlights are widely distributed). Other dramatic events like a goal-line clearance, violent tackle, amazing save, impressive piece of skill, or a good chance bungled away are many times recalled as well. An expert of the game would also remember more nuanced actions or non-actions as well. Still, how many people have accurate memory of more than 10 minutes of the actions in a match? Can the human brain even process all the details of the 11 versus 11 flow of the game in real time? How can fans (and even coaches and scouts) effectively evaluate a player when their brains are incapable of taking in the nuance of play?
Statistics can fill in some of the observational gaps in soccer. After all, one of statistics' primary purposes is to bring attention to trends which might otherwise escape our notice. Historically, the beautiful game hasn't been interested in quantitative commentary. It does seem odd that someone would try to break down into numbers a continuous-action sport which can be transcendentally unpredictable. If you can identify trends, though, you gain an edge on the stat-averse, but you have to dig deeper than you would in most sports.
Until recently, the only widely available soccer stats were goals, assists, shots, saves, and time of possession. These can be somewhat telling when summed up over months or seasons of play, but try picking three of a club's games and drawing larger conclusions from their stats. You would be hard-pressed because they are subject to swings based on tactics, luck, opponent strengths and weaknesses,, refereeing, and small sample sizes. We've all seen goals that had more to do with a fortunate bounce than skilled striking or passing. Chimu Solutions wrote a great article in Shin Guardian last season about time of possession's lack of relevance in MLS. Besides, most traditional stats detail the events that are most likely to be remembered by observers anyway.
So, which stats should we be paying attention to? Lots of very smart people have been working on that lately, and some very interesting conclusions have been made (and more are surely coming). Some of these breakthroughs are and will be closely guarded by clubs, as AC Milan has done with aspects of their lab that boosts player longevity. Hopefully, much more advanced data will become publicly available, as has occurred for basketball, baseball, hockey and football For now, let's discuss a couple relatively straight-forward advanced soccer statistics that some smart people have deemed to be quite meaningful.
As detailed in a Power of Goals post titled, "How We Know Who's "Winning" Before a Goal is scored,"duels have a clear correlation to wins, losses, and draws in the EPL. What are duels, you ask? The definition offered in the article is "as 50-50 contests between opposing players in which one players emerges successful." According to Opta, duels occur about once per minute in an EPL match. The importance of this stat makes perfect sense. If a defensive player wins a duel, he has effectively snuffed the attack (assuming he doesn't commit a turnover quickly), and an offensive duel won screws up the opposition's defensive shape as someone has to scramble to make up for the out-dueled player.
Opta provides team dueling data on stat sheets for individual matches. Records for individual would be a nice addition so we could objectively evaluate a player's performance, but the team duel stats have strong predictive power. If you see that your club does well in duels, you can be confident that results should follow.
Chances Created and Conversion Rate:
Some of the more unfortunate victims of memory are plays that led to a great scoring chance, which was then kicked into the stands by a "striker" who plays as though he's drunk (or in Maicon Santos' case, anesthetized). No goal, no assist, no matter who beautifully the plate was set. You can calculate chances created in a match by adding key passes and assists, both viewable in MLS matchcenter chalkboards. Chances created is essentially the measure of every pass that led directly to a shot.
Conversion rate, aka strike rate, is simply goals divided by shots. Very quickly, this metric shows you the quality of a club's finishing and/or propensity to settle for low percentage shots in the first place.
Why discuss these two stats together? Soccer Statistically analyzed EPL data of chances created, and found that:
The bottom line is that creating chances and conversion rates are the key to understanding goal scoring. A club can succeed with a high conversion rate (United) or by creating a lot of chances (Liverpool). A club can really dominate by doing both well (City).
The article's graph of chances created vs. conversion rate is a striking illustration of the offensive strengths and weaknesses of clubs as of the publication date (November 9).
FC Dallas Implications:
I pulled duel statistics from 10 of FC Dallas' matches last season, and found that they only won 49%. There was a marked decrease in duels won during FCD's late-season swoon. When watching a match, FC Dallas fans should pay particular attention to Shea and Ferriera's success rates when taking on defenders, and Loyd and Ihemelu's percentages winning the ball in 50-50 situations.
In 2011 FC Dallas were crap at finishing chances, but great at creating them. Dallas' conversion rate was 8.71%, which was 0.01% above last place San Jose, and far below the league average of 10.25%. I couldn't gather numbers for chance created over the entire 2011 MLS season, but shots can be a decent estimate. Chances created doesn't count missed shots that came from a long run (which should decrease markedly with Jackson's departure). Dallas was only 6 shots behind Kansas City for most shots in MLS with 482, which was 52 above the league average. Viewed together, you can see why the hoops were a slightly below average offense (42 goals vs league average of 43.9).
Unfortunately, finding useful statistics to analyze can be difficult if you don't work at a major media network or a club's front office. Most useful-sized samples of advanced stats are proprietary and unavailable to anyone that isn't willing or able to pay a substantial price. That makes sense when you consider the staffing and/or technology required to gather and organize data on location and action of everyone on the pitch. Even knowing that, it's still quite frustrating when one can't get data to verify the reliability of small-sample-size trends.
Reliable sources of publicly-available MLS statistics:
Opta - One of the premier statistics companies in the world, specializing in soccer. Their different twitter feeds (@OptaJack for MLS, @OptaJoe for EPL, etc.) offer a free glimpse at snippets of interesting advanced data. Also, MLS matchcenter chalkboards include Opta stats for that game which can be broken down by player and team.
Castrol MLS Ranking- Powered by Opta, these are based on "objective" rankings of every move a player makes near the ball. These can be a bit confusing on the surface, and this post is already a bit long. I'll give Castrol rankings a thorough breakdown at a later date.
Stats page on mlssoccer.com - Little advanced data, but that's made up by the ability to quickly get the basics on any team or individual 1996-present. Sorting and comparing different data points (like dividing goals by shots to get conversion rate) over the long term can lead to interesting conclusions.
If you know of other useful (and free) MLS data resources, please let me know either via my twitter handle @OptaHunt or in the comments below. Fans of other sports have gotten in the habit of checking box scores to see what they missed when watching a game. As a soccer fan I want to seek out data which will bring greater understanding of the game I love.