2 January 2014

To think about: are the Sens inconsistent, or just bad?

For perhaps the entirety of the 2013-14 season's first half, followers of Ottawa Senators hockey have tossed the word "inconsistency" and its derivatives around like lettuce in a Caesar salad, yours truly included. And just like overtossed salad, the catchphrase has become somewhat limp and dry, with less substance than the no-nonsense carrots and tomatoes garnishing the salad (which, in this analogy, represent the fancy stats). But it seemed relatively accurate; we've reached our first winning streak since November 9th, and somedays, it seems as if our beloved Sens simply don't show up to play.

Yet the stats may say something different.post from new fancy-stats blog Hockey Graphs by @garrethohl has, in fact, the audacity to suggest that inconsistency isn't real; rather, "inconsistent" teams are just bad.

It sucks to think one's team's so-called inconsistency might mean that they're simply a poor-quality team, but it sucks even more to launch oneself on an emotional rollercoaster without being prepped with realistic #RealTalk on how well the team can be expected to do. Hopeless optimism can sometimes be a synonym for helpless masochism. Therefore, it seems worth the time and pain to consider the possibility that the Ottawa Senators might be wibbling-wobbling all over the place due to a deficiency in quality.

The post compares teams' Corsi close numbers' means and standard deviations, on the theory that the mean represents a team's average quality, and the standard deviation represents a team's consistency. With numbers drawn at both 5-vs-5 close play and 5-vs-5 even-strength play, the post illustrates that the means vary greatly over a wide spread, but the standard deviations' spread is over a much smaller range. In essence, the suggested conclusion is that variation in a team's Corsi close numbers is nowhere near as great as the overall variation in all teams' mean Corsi close numbers. This implies that one team could be of much greater or lower quality than the next, but no team is very much more consistent or inconsistent than the next.

Basically, consistency seems to be a non-factor.

After you've processed this conclusion, the first thing to note is that only even-strength Corsi numbers are calculated, to ensure most indicative possession numbers. Therefore, it stands to reason that if a certain team has had more non-even-strength playing time than others, then the numbers are less valid or true for those teams (although this does not negate the conclusion). Taking a look at Extra Skater's team total stats, we find that the Sens are third in 5-on-4 TOI, third in 4-on-5 TOI and first in 3-on-5 TOI. (The Flyers are similarly highly-ranked, which is another interesting thing to consider: do teams draw penalties at the same rate as they take them?) Therefore, it's possible that the consistency-is-non-existant theory does not apply... but in that case, the source of our inconsistency would be actually our penalties, and not the inconsistency itself.

It might not be exactly a relief to hear that the Ottawa Senators' inconsistency is really just a manifestation-illusion of their lack of skill. But look at it this way: if there's one thing worse to find a solution to than a forward's scoring drought, it's inconsistency. A team's quality is at least somewhat quantifiable and, more importantly, resolvable: hire better players. Asking for consistent players will only get you nowhere.


Post a Comment