Chris 🇨🇦🏈 Stats Junkie
banner
statsjunkie.bsky.social
Chris 🇨🇦🏈 Stats Junkie
@statsjunkie.bsky.social
Canadian football stats, history & rules junkie
One final note for now

In 2025, the PER listed in the post game stats packages for individuals was wrong more than 40% of the time 🤯

How is that for a stat?
November 29, 2025 at 5:35 AM
Reposted by Chris 🇨🇦🏈 Stats Junkie
Here is a list of all CFL QBs who achieved a Pass Efficiency Rating of 130 or better (based on a minimum of 20 pass attempts)

I also listed the PER provided by the CFL & highlighted the component of the formula where rating goes wrong

There is a reason I use 20 attempts but that is a long story
November 29, 2025 at 5:28 AM
Note 2: Tre Ford 157.8 (week 5)
Tre Ford failed to meet 2 of the 4 criteria required for a perfect rating

🔹️average gain of 12.47 is slightly below the 12.50 needed - 1 more yard would have done it

🔹️2 TD passes on 17 attempts also missed the threshold - 2 TD passes on 16.84 attempts was needed
November 29, 2025 at 5:28 AM
Note 1: MBT 156.7 (week 15)
Late in the game Ventrelli scored a TD on an OTFR which is officially scored as a passing TD. Genius is unable to apply to proper CFL statistical treatment to such a play.

A manual edit was done during the post game audit but the PER was not adjusted.
November 29, 2025 at 5:28 AM
In the post game stats, the CFL uses a slight variation of the 'uncapped' PER formula. If the result is greater than 158.3 then an arbitrary 158.3 is applied to the result

It is IMPOSSIBLE to connect the dots between the 'Uncapped' PER and a perfect rating of 158⅓
November 29, 2025 at 5:28 AM
Results that can skew the data on the high end are:

🔹️completion % > 77.5%
🔹️average gain per attempt > 12.50
🔹️TD frequency better than 1:8.42

0 Interceptions receives a maximum possible score of 2.375 which is why all categories are capped at the same value.
November 29, 2025 at 5:28 AM
For nearly 20 years, the CFL has used what is described as an 'Uncapped' version of the PER formula

While technically it is a correct description of the formula used, it also is a very wrong version to use

Without proper caps in place, a component of the formula can be overweight & skew the data
November 29, 2025 at 5:28 AM
PER is comprised of 4 EQUALLY weighted calculations.

Again 4 EQUALLY weighted calculations.

🔹️completion %
🔹️average gain per pass attempt
🔹️TD frequency
🔹️interception frequency
November 29, 2025 at 5:28 AM
Here is a list of all CFL QBs who achieved a Pass Efficiency Rating of 130 or better (based on a minimum of 20 pass attempts)

I also listed the PER provided by the CFL & highlighted the component of the formula where rating goes wrong

There is a reason I use 20 attempts but that is a long story
November 29, 2025 at 5:28 AM
This 🧵 is not an attack on @cflstatsgeek.bsky.social but rather a commentary on the formula used by the #CFL for the Pass Efficiency Rating (PER)

It is quite clear the PER data in the previous graphic comes from the post game stats packages provided by the CFL
November 29, 2025 at 5:28 AM
Also in the mix for the Blue Bombers OC position is Tommy Condell
#ForTheW #CFL
November 24, 2025 at 9:37 PM
November 19, 2025 at 7:43 AM
The ball was fielded at the -16 which would be outside the endzone in 2027, so no point would be awarded.

Chances are the returner would just watch it roll out the back of the endzone without making a play at the ball.
November 17, 2025 at 12:22 AM