The Statistics Behind Toronto FC Defenders in 2014

Four goals. That was the difference between Toronto FC’s goals against last season (54), and the highest number of goals allowed by a playoff team (50, both Seattle and NYRB). Defensively, TFC were, at least, 4 conceded goals away from a post-season appearance. The number of goals SCORED by TFC (44) was also poor relative to playoff-bound teams as it was higher than only Vancouver’s, but we’ll save the discussion on Toronto’s offence for another article.

How should we kick off the conversation about the state of Toronto FC’s defence? Beside the obvious answer- "with a good stiff drink in hand", my inclination would be to stay away from the hypothetical (who do we need / who are we going to bring in) and instead focus on the impact of our key players in 2014. To start off, let’s establish an assumption that I will be working with- there is NO debate as to who our three most consistent and useful defenders were, over the course of the 2014 season. In no particular order, Steven Caldwell, Justin Morrow, and Mark Bloom were the three pillars of our back line, who all came out and performed time and time again. These three were TFC’s bona fide starters, and while we can discuss and argue whether or not any of them could be upgraded (or are getting too old/wavey-armed to keep up their performance level), there is very little substance to any viewpoint that places Henry, Hagglund, and/or Orr’s 2014 contributions ahead of the aforementioned three Reds.

With that assumption in mind, I decided to use a simple (albeit narrow-minded) statistic to explore the effect of each of the three starters on TFC’s results- "number of goals against per 90 minutes". As I embarked upon this statistical journey, I recalled a brief back-and-forth in a Waking The Red comments section between yours truly and a former WTR manager who will remain unnamed. The latter’s initial comment discussed how Caldwell’s injury led to a steep decline in TFC’s performance and results over the last season, to which I replied that Morrow and Bloom were also hurt for some of those games and that TFC’s regression may not have solely been due to the captain’s absence (for the record, Dunc… I mean the commenter, agreed about that possibility).

With that context in mind, and the hypothesis that Caldwell’s absence impacted TFC’s goals against more than the absence of either Bloom or Morrow, I ran the numbers. Surprisingly, whoscored.com does NOT track goals against while on the pitch (at least not in any way that’s easy to find). So this was all manual work, including going back to actual game reports if a defender played partial minutes, to see if they were on-field for any of the goals scored against the Reds.

Let’s start off with the basic numbers. The following table shows how many minutes each of the three defenders played over the course of the season, and how many goals were scored against TFC during that defender’s time on the pitch. It then averages that out to per 90 minutes. Of course it doesn’t take into account who else was on the pitch, but we’ll try to address that shortcoming later.

Unsurprisingly, Justin Morrow, who played the most time out of all three defenders, had the worst goals against per 90. Playing so many games for a 7^th-place team will do that to a defender’s statistics. The surprise? Toronto FC conceded fewer goals, on average, in all of Bloom’s time on the pitch, than they did while Caldwell was playing. That’s right, the defender who TFC left unprotected in the expansion draft, the man who Orlando was given the opportunity to scoop up, had the fewest goals against per 90 of Toronto’s three bona fide starters. Hmmm… maybe those who were worried about losing Bloom weren’t "over-rating" him after all. Of course, one number in isolation does not a whole story tell (see Eric Hassli’s strike rate with Toronto FC), and therefore I decided to push forward with the analysis.

The second set of numbers is what really left an impression on me. I took each individual player, noted the games that they did NOT play in (or only played partial minutes in), and added up the goals conceded by Toronto FC while that player was absent from the pitch.

Remember when TFC "went downhill" because Caldwell was injured? Turns out the team conceded more goals, on average, when either Bloom or Morrow were absent, than when Caldwell was missing. Yes, this analysis is limited in that it is based on simple regression logic rather than multiple regression- for example, I’m not giving Caldwell any extra credit for when he was missing and one of the others were absent which would have inflated those goals against. However this statistical shortcoming applies for all three players, so it could be argued that it doesn’t necessarily lend a significant advantage to any one defender. The fact remains, the number of goals against per 90 is startlingly high when Bloom was watching from the sidelines versus the numbers for Caldwell and Morrow.

Before anyone gets upset, I would never even consider concluding that Mark Bloom is Toronto FC’s most important / best defender, as the goals conceded statistic is only a tiny piece of TFC’s defending equation. This simple analysis also ignored Bloom’s inability to contribute offensively, as well as Caldwell’s intangibles- his leadership traits, steadying presence, and how it influences the entire squad. I will, however, contend that the hypothesis that Caldwell’s absence was worse for TFC than the absence of Bloom or Morrow is potentially misinformed.

One thing is safe to say- the health and performance of Bloom, Caldwell, and Morrow are all vitally important to the squad’s success in 2015, and TFC would do well to find some suitable replacements (and another starting CB) should the injury bug decide to "Luis Suarez" the back line again in 2015.

-Of note, I also quickly ran the numbers for when two or more of the three were absent from the pitch. When Morrow and Caldwell were both missing, TFC average 2.38 goals against. When Bloom and Caldwell were both missing, TFC averaged 2.04 goals against per game. While I don’t think this nullifies any points made by the previous data in the article, it certainly does provide some food for thought- feel free to discuss in the comments!