Jay Nolly vs. Joe Cannon: The Final Comparison

Team Nolly or Team Cannon? For much of the 2011 Major League Soccer season, this was one of the defining questions that split Vancouver Whitecaps fans. Did you want to see the veteran and former Goalkeeper of the Year Joe Cannon carry the Whitecaps forward or would you prefer to see younger, cheaper fan favourite Jay Nolly get the nod? Teitur Thordarson was Team Nolly, Tom Soehn was Team Cannon, and the argument was never really resolved.

Martin Rennie cut the Gordian knot earlier this week when the Whitecaps re-signed Cannon and traded Nolly to the Chicago Fire for a supplemental draft pick. Joe Cannon has won the war. Even Nolly's stoutest defenders aren't that upset, for everybody concedes Cannon and Nolly are fairly close and Cannon, like Nolly, is an absolute class act who's spared no effort to make himself popular with supporters.

The war has been decided but the battle goes on. As I showed earlier this morning, Nolly's numbers in Major League Soccer were clearly inferior to Cannon's. There remained, however, the circumstantial defense that Nolly was playing behind a worse defense than Cannon; one more ravaged by injury. Even if it no longer matters we can still ask the question: can we prove or disprove this theory?

Of course we can. After the jump, a comparison of the defenses faced by Jay Nolly and Joe Cannon in 2011.

First, a recap of the basic goalkeeping rate statistics, as I showed you this morning.

The Whitecaps defense had a number of injury problems throughout the season and there were very few points from First Kick through to the end of the year when the defense could be said to be healthy. There was also significant rotation, particularly at centre back and right back, which put both goalkeepers at a disadvantage. The 2011 Vancouver Whitecaps had serious problems defensively no matter who was in goal.

The three regular Whitecaps defenders through the course of the season, the automatic starters when healthy, were Jay DeMerit, Jonathan Leathers, and Alain Rochat. They were also three of the four best Whitecaps defenders this past season, along with Mouloud Akloul. Rochat played regularly but both DeMerit and Leathers suffered from injuries. Which one of Cannon or Nolly was most victimized?

These tables list the number of minutes, as well as the percentage of possible minutes, each of these three defenders played with Joe Cannon and Jay Nolly over the course of the MLS regular season as well as the Voyageurs Cup.

Joe Cannon suffered badly from Jonathan Leathers's lower body problems towards the end of the year, saddling him with the Jeb Brovsky Experience (and probably accounting for quite a bit of that extra 0.45 shots per game). However, Nolly saw less of Alain Rochat than Cannon did and significantly less of Jay DeMerit. DeMerit played only 29.23% of the time across all senior competitions when Nolly was on the field, but he got into 72.33% of Cannon's minutes. That vast gulf in DeMerit's availability explains why, total, the three defenders listed played 69.93% of the possible minutes with Joe Cannon but only 62.22% with Jay Nolly.

The difference in Major League Soccer competition is even more stark: Cannon got just over 10% more time with DeMerit, Rochat, and Leathers than Nolly. Nolly enjoyed quite good defender availability in the Voyageurs Cup and, while it's obviously hard to tell much from two games against a non-playoff NASL team and one against lowly Toronto FC, Nolly posted ridiculously good Voyageurs Cup statistics of 0.60 goals against per 90 minutes and a 0.875 save percentage. (Cannon's Voyageurs Cup statistics are pretty great too: in his one game against Toronto he faced twelve shots, stopping ten of them).

So what effect did this have? Theoretically, a better team defense should concede fewer shots on target. Therefore, a logical follow-up to the above research would be to see which of Nolly and Cannon faced fewer shots in each of their games.

In 2011 Major League Soccer competition, Jay Nolly made 14 appearances and faced 63 shots, or 4.5 shots per 90 minutes. Joe Cannon made 20 appearances and faced 99 shots, or 4.95 shots per 90 minutes. This metric actually implies Cannon was the victim of inferior defending, to the tune of almost half a shot per game. Given Cannon's .677 save percentage, if Cannon faced 0.45 shots per game more than Nolly then that would account for an extra goal every 3.282 games. Over the course of Cannon's 20 games, that's six goals.

This implies that Nolly's defending may not have been so bad. Obviously there are many factors, not just the back four, which impact how many shots on goal a team would allow. Nolly played more games with midfielder Terry Dunfield than Cannon did, and while Dunfield got it from Whitecaps supporters for his unambitious passing he was also a solid, reliable defensive midfielder who the Whitecaps were unable to replace.

These statistics offer no firm conclusions. Yes, Nolly got to play less with Vancouver's top defenders than Cannon did, but Cannon faced more shots per game. Given the Whitecaps' lack of depth at right back, Cannon losing Leathers for so many games may have been more devastating than Nolly getting almost no time with DeMerit, even though DeMerit is obviously the better defender.

My judgment is that it's true: Joe Cannon did get more defensive help than Jay Nolly. It narrows the gap between the two goalkeepers but it doesn't change the fact that Cannon had the better season. Looking strictly at player performance, Martin Rennie kept the better man.