But what if we take a much larger look at the numbers? What if we looked at 46 statistics/factors that are tracked and readily available from CFL.ca and other sources? And what if we looked at them for each and every game that the Hamilton Tiger-Cats played in 2014?
Well, that is exactly what I decided to do and before you ask why I would engage in such a project, let me give you the short answer: I am a stats geek and these kinds of things intrigue me greatly! But more importantly, by looking at some 920 measurements and comparing them to wins and losses, we can begin to attribute which statistics or aspects of the game most impacted the 2014 season for the Hamilton Tiger-Cats!
So, how do we do this? How do we measure, say, a team’s number of penalties to their win and loss record? Or how do we measure which is more important: score a lot of points offensively or hold their opposition to very few points?
In one word: Correlations
I will not bore everyone with a complex definition of correlations but, put simply, it is a measure of the strength and direction of the linear relationship between two or more variables.
One of the simplest and most common ways of quickly explaining correlations is in the area of turnovers or, more specifically, turnover ratio. For example, if a team were to win every time they had a positive turnover ratio and lost every time they had a negative turnover ratio and tied every time the turnover ratio was a draw, then this would result in a perfect correlation of +1. Yet if the exact opposite was true and a team won every game despite losing the turnover battle and lost every game when they were on the plus side of giveaways and takeaways, then this would have a correlation of -1.
I would also like to point out that correlations do not imply causation because something else might be at play. For example, on hot days people buy ice cream. Also on hot days people go to the beach. Sometimes, sharks attack at the beach. There is a correlation between ice cream sales and shark attacks since they both go up as the temperature rises. But just because ice cream sales go up does not mean ice cream sales caused an increased number of shark attacks.
Now, getting back to our little study, positive numbers indicate that the specific factor has a positive corollary relationship to Ti-Cat wins throughout 2014. Negative numbers mean that the specific statistic has a negative or inverse relationship to the team’s wins during regular- and post-season play. Overall, the higher the number means the stronger the correlation is to results in Hamilton’s win and loss column.
Getting a little more specific, correlation coefficients greater than 0.5 per cent are considered to be strong in either direction, positive or negative. Measurements of 0.3 to 0.5 are considered to have a moderate corresponding relationship, once again in either direction. Results that span either side of zero between negative 0.3 and positive 0.3 are said to have no correlation or interrelation at all.
So what do we see when we look at this mass of numbers and correlations? Well it looks a little like this when you graph the results from strongest positive correlation coefficient to strongest negative or inverse dependence.
Black = Correlations between 0.5 & 1 or -0.5 & -1 = Strongly Correlated
Yellow = Correlations between 0.3 & 0.5 or -0.3 & -0.5 = Moderately Correlated
White = Correlations between 0.3 and -0.3. = Not Correlated
(1). Defense Was More Important Than Offense
As someone who does not believe in the idiom “Defense Wins Championships” – my distaste for it rivals Josh’s distaste for attributing wins to the quarterback position – it pains me to say that Defensive or Opponent factors had much stronger correlations when it comes to Tiger-Cats’ victories in 2014.
When you add everything up, what Hamilton’s opponents did with the football, or more specifically what they did not do, had a stronger interrelationship to Ti-Cats’ results in the win/loss column. Opponent factors had a negative 0.32 correlation to Hamilton victories throughout the 2014 regular and post season. This is a moderately strong correlation coefficient that illustrates it was what the Hamilton defense did to limit their opponents that had a stronger relation to Hamilton victories. Compare that to what Hamilton did do with the football, which only had an indifferently positive 0.09 correlation to Tiger-Cat wins.
Holding opponents to fewer points was more important than scoring a lot of points. Team Scoring Against had a correlation of -0.62 while Team Scoring For had a correlation of only 0.19. An easy way of looking at this would be in the three games Hamilton played Toronto. A 13-12 victory on Labour Day over the Argos at Tim Horton’s Field versus two losses at SkyDome in which Hamilton scored 33 and 24 points. They won the game in which they curtailed the Double Blue’s scoring while lost the games in which they scored a lot.
(2). Running Was More Important Than Passing
Running the ball had a stronger correlation to winning than passing the pigskin. Team Rushing Attempts and Team Rushing Yards had moderately strong positive correlations to Ti-Cats’ victories (0.49 each). Compare this to neutral or uncorrelated results in the area of Team Passing Yards (-0.06), Team Passing Yards Per Attempt (0.07) and Team Completion Percentage (0.09).
Simply put, when Hamilton ran the ball more and ate up more yards on the ground, this had more of a positive impact on their winning chances compared to when they put up big passing numbers.
(3). Special Teams Were Not as Important as We Would Think
Team Return Yardage had a -0.17 correlation to Hamilton’s wins and losses last year. This was somewhat surprising to me since, to the naked eye, the Ti-Cats, and a certain No. 16 from Kansas State, were pretty special!
Now, this figure is a little misleading since the numbers I utilized do not separate out kick-return yards from punt-return yards, nor did they denote returns for touchdowns. As we all know, having a lot of kick-return yardage is more likely due to having been scored upon a lot and therefore receiving more kickoffs from the opposing team.
I am sure when I separate out these figures and add in returns for touchdowns, this correlation number will undoubtedly turn positive and most likely have a moderate to strong relation to Hamilton wins.
All the other numbers look to be in order and fit what we all would expect. For example Sacks Made by Hamilton had a moderately positive correlation of 0.40, while Sacks Given had a correlation of -0.26. This makes sense, and when viewed together can be concluded that sacking the quarterback was a little more important than giving them up.
Further analysis of all these numbers is required. Are there any trends that developed throughout the 2014 season? We were all witness to a significant turnaround that began September 1 when the Tiger-Cats took up residency at Tim Horton’s Field. Do the numbers show such a change? And if so, what factors of Hamilton’s game changed the most?
We shall save that for a little later as we continue to peel away at the layers that was the 2014 Hamilton Tiger-Cats Season By The Numbers.