Winnipeg Blue Bombers quarterback Matt Nichols (15) throws against the Saskatchewan Roughriders during first half CFL Banjo Bowl action, in Winnipeg on Saturday, September 10, 2016. THE CANADIAN PRESS/John Woods
There are a ton of numbers in football that just seem to make perfect sense, from the 10 yards necessary to move the chains to the six that go up on the scoreboard for a touchdown.
It’s also long been a given that 1,000 yards, in either rushing or receiving, represents excellence in a season and that eclipsing the 100-yard mark along the ground or in receiving in a single game is a dandy day at the office.
Then there’s this, as it relates to the most important position on the football field:
The three top-rated starting quarterbacks in the Canadian Football League in 2016 were Trevor Harris of the Ottawa REDBLACKS at 116.0, followed by Bo Levi Mitchell of the Calgary Stampeders at 107.9 and the Edmonton Eskimos’ Mike Reilly at 104.5.
And fully understanding the numbers behind the quarterback efficiency rating – a complicated formula that uses passing attempts, completions, yards, touchdowns and interceptions – would seem to require a PhD in mathematics.
That formula, after all, looks like this:
Adopted by the National Football League in 1973 and the CFL in 1996, the passer rating scale goes from 0 to 158.3. CFL fans have simply come to accept that anything in the 100-range is excellent and in the 90s is very good – even with the admission that our understanding of just what a 158.3 performance would look like is more than a little complicated.
It was during this past winter when Steve Daniel – the CFL’s head statistician – began discussions with a number of league personnel about creating a better formula for rating quarterbacks, one that included some of the other criteria coaches and GMs have come covet from their pivots.
He chatted with Kent Austin of the Hamilton Tiger-Cats, Calgary’s Dave Dickenson and Mike Petrie, Mike O’Shea of the Winnipeg Blue Bombers, Kavis Reed of the Montreal Alouettes, Ottawa’s Rick Campbell and Marcel Desjardins – among others – all in an effort to come up with a quarterback rating that will now encompass more of what a pivot brings to an offence.
“The whole idea was to come up with something people could understand and relate to the data,” explained Daniel in a chat with bluebombers.com this week. “And if I was going to do this, the first criteria had to be objective and so that’s why I made it out of 100 and not 158.3 – whatever that is.
“What fuelled me – and the league has supported me endlessly – is I want to be able to not just report our statistics, but to analyze them. And so, this is a merge of traditional statistics with innovation.”
What Daniel has come up with – after examining every play in every drive dating back to 2009 – is a new method to rate quarterbacks and it will be implemented by the CFL this season. He calls it ‘QUAR’ – quarterback rating – and it melds the current efficiency rating system with eight new metrics to provide a more comprehensive take on a QB’s performance.
“This set of data now exists for us and it’s powerful and we’ve done it at the team level,” said Daniel. “So for QUAR, we’ve taken the team stat, integrated it with the individual stats and come up with this measure that does both. The essence of QUAR, at its core, is to say it’s OK to combine individual and team statistics.
“We have become more of a passing game, and because of that, I believe we need to measure that more effectively. You can debate all these factors, the weightings for each, whatever you want. But you’ve got to start somewhere. This is a flexible system and we can change the weighting if we want, but it will react to the changes in the game itself.”
Here’s how QUAR will work, with 100 being a perfect score:
- The existing pass efficiency system that uses the variables of completion percentage, yards per attempt, touchdowns per attempt and interceptions per attempt, will now account for 40 percent of the new score.
- Win/loss – Does the quarterback’s team win the game? – will account 10 percent
- Offensive points per possession: 15 percent.
- Net offence per possession: 10 percent.
- First downs per possession: 5 percent
- Second-down conversion percentage: 5 percent
- Quarterback sacks taken: 5 percent
- Fumbles (can the QB take a sack/hit and not fumble): 5 percent
- Individual rushing yards: 5 percent.
Some examples, based on the formula above:
- Calgary’s Bo Levi Mitchell was 15-1-1 as a starter last season. That would earn him a 9.12 out of 10 in the new system; Winnipeg’s Matt Nichols, 10-3 in his starts, would earn a 7.69.
- Last year the number of offensive possessions averaged 15 per game. A quarterback and offence that averaged 2.2 points per drive would score 33 points – and win a lot of games.
- The impact of a quarterback who can extend drives with his legs won’t be under-valued in QUAR, as well. Consider that Mike Reilly led all CFL QBs with 406 yards rushing last year while B.C.’s Jonathon Jennings averaged 5.3 yards on the 68 times he tucked the ball away and ran in 2016.
“We consider all these attributes of a quarterback and that’s why you have them on the field for you,” said Daniel. “And the rankings, based on last year, were proof to me.”
Under the new QUAR system, the CFL’s quarterbacks would have graded out this way (last year’s league-wide average QUAR wold have been 62.4):
- Bo Levi Mitchell, Calgary – 91.6
- Mike Reilly, Edmonton – 88.8
- Jonathon Jennings, B.C. – 86.9
- Trevor Harris, Ottawa – 86.4
- Zach Collaros, Hamilton – 78.5
- Matt Nichols, Winnipeg – 78.0
- Ricky Ray, Toronto – 74.1
- Henry Burris, Ottawa – 73.3
- Jeremiah Masoli, Hamilton – 69.0
- Kevin Glenn, Montreal/Winnipeg – 62.7
- Darian Durant, Saskatchewan – 58.4
“The teams have told me they like this formula because it covers the attributes they are looking for in a quarterback,” said Daniel. “As an example, in our game there’s probably no more telling statistic, other than turnovers, than second down conversions. That’s why it’s been included in the new formula.
“QUAR is not a knee-jerk reaction to having to have another number. The league wanted this done right and they wanted to break down some boundaries with this kind of thinking.
“The point was to combine tradition with innovation, to be objective, and do it right.”
A bit of background on Daniel, for fans who might want to know the man behind QUAR. He began working full-time for the CFL in 2007, officially taking over the statistics for the league in 2008.
Prior to that, he worked for 10 yeas for the Vancouver/Memphis Grizzlies of the National Basketball Association as their Director of Operations and Scouting Statistics.
He has a Master’s Degree in Anthropology from UBC and is an archaeologist by trade whose specialty was ground-penetrating radar and had done work primarily for First Nations in studying burial grounds and locating markers that may be missing.
But at heart, he is a CFL diehard through and through.
“It’s so much fun to do this job and I enjoy it so much,” said Daniel. “And I hope it comes out in what we’re doing. I love this game, I love the CFL and doing statistical analyses like this only enhances my drive.”
