Reading Penguins’ Goalies Stats And What Do They Really Mean

For those who don’t know that much about me, I once worked clinically for a diagnostic group that served over a hundred clinics. During that time, there was a technician that worked for me who once sent a test report back with tech notes that read, “The patient gave good effort, but it was raining outside.”

Confused? So was I.

Now there is a point to this little ditty.

I was going to write a reply to my friend Rick B. in way of an explanation as to our differing opinions, particularly where the Pittsburgh Penguins‘ Goalies. However, as I thought about how to write my reply in a cogent way, I realized it would be rather lengthy. So, I have turned it into a full post.

Apples and Oranges

This is America, anyone can have any opinion they like and point to any evidence they want, freedom of thought is inherent to freedom of speech. Now it is not going to sway me to accept any conclusion a person has, just because they have a right to think and express it. To get me to accept an opinion as at least equivocal to mine, let alone change my mind. After reading through literally hundreds of thousands of test reports (electronic records allowed me to easily keep track) while working clinically and sifting through even more research articles while plying my trade both clinically and in the strength and conditioning fields, points have to be empirical, objective, and logically derived for me.

It is well known and widely accepted that the Goals Against average (GAA) statistic is a measure of team defense, not of Goalie effectiveness. Whether we are talking about actual GAA or some made up statistic, expected Goals Against average (eGAA), or the difference between the two numbers we are not evaluating a team and not a Goalie, we are evaluating team defense. To hail a Goalie because he has a very high difference between his actual GA and his eGA is like sending a test report back to the office with the note, “The patient gave good effort, but it was raining outside”, it is non sequitur, one has little to nothing to do with the other.

So what good is the eGA or the diff between GA and eGA?

Good question! It is a statistic. It is a measure of an empirical fact. So, it has to have some value, but since it is a team statistic and can either represent a positive or a negative it requires greater attention to detail to understand it effectively. A cough can be a symptom of anything from a respiratory disease to a gastro-intestinal issue to atrial fibrillation. To get to the root cause of the observed cough, a clinician has to dig a lot deeper. So, to do that end we have to dig deeper to get to the root cause of the observed statistic. It requires a level of detective work, more detective work than I originally wanted to dive into to grade our Penguins’ Goalies.

How to Evaluate Statistics

Before getting too much deeper into statistics, it may be wise to talk about how to evaluate them. The first thing that you have to do is equalize them to a standard, while it is true that Arturs Silovs has given up 29 actual GA doesn’t mean much and it cannot be compared to any other goalie in the league. The number has to be standardized to a specific time frame, a GAA, not just GA. Even the GAA doesn’t mean much unless it is viewed in relationship to the League Average (LA)

	LA	Jarry	Silovs
GAA	2.49	2.37	2.45
eGAA	2.59	2.54	3.2

Since GAA and eGAA are measures of team defense If Goalie’s numbers are below the LA, then that team’s defense is playing really well. If those numbers are roughly the same as the LA, then team’s defense is relatively average. If the GAA and eGAA are above the LA, then the team’s overall defense is not playing all that well.

Both Tristan Jarry and Silovs have GAA that are lower than the LA. Jarry’s GAA is only 95.2% of the LA while Silovs is 98.3% of the LA. Just looking at these numbers, my first thought is that the team’s defense has been pretty good. Both Goalies GAA and eGAA are a little below the LA; Jarry’s is a little lower than Silovs but the difference is small and may not represent any significance.

Looking at the eGAAs a different story starts to unfold. Jarry’s eGAA is a fraction below the LA, 98.1% of the LA while Silovs eGAA was above the LA at 124%; Wow what a change! Same Forwards, same Defensemen, the only difference – different Goalies.

Hmm…

I guess it is time to dig a little deeper.

	LA	Jarry	Silovs
SA	26.55	26.94	30.02
HDSA	6.98	6.44	8.29
MDSA	6.67	7.63	8.41
LDSA	11.60	12.03	12/03
RAA	5.37	5.25	7.59

Jarry’s Shots Against (SA) are slightly higher than the LA which would be in line with his eGAA. Silovs SA are significantly above the LA so I expect his eGAA to be significantly higher.

Looking at the SA breakdown I can start to really see why Silovs eGAA is so high compared to his GAA and it is not because he is a good Goalie, quite the opposite, it is because he is really a bad Goalie.

Jarry and Silovs both have the same LDSA, 12.03. That number is slightly elevated over the LA of 11.58, which is in-line with the number of games they won with opponents trying to get back in the mix and just trying to throw anything they have at the net.

Both Goalies MDSA are also elevated in comparison to the LA (LA: 6.67, Jarry: 7.63, Silovs: 8.41). We are starting to see a split here. Both are above the LA but Silovs is starting to really split here. Silovs’ MDSA is 126% above the LA while Jarry’s MDSA is only 114% above the LA.

Looking at the HDSA we see the trend continue even more; Silvos HDSA is 8.29 or 1.19% of the LA while Jarry’s HDSA is actually below the LA ,6.44 (92.2%).

The elevated HDSA and MDSA does explain the difference in SA discrepancy as well as the eGAA differences. However, we need to look even deeper to really appreciate what is going on here. Both Goalies back stopped the same group of Forwards and Defensemen, so why does one have to face more HDSA and MDSA? The answer lies in the Rebound Attempts Against (RAA). Silovs rebound control is beyond atrocious. Silovs has to face 7.59 RAA per 60 minutes of 5-on-5 hockey while the LA is 5.37. Silvos faces 141% of the LA in RAA. That is why his HDSA and MDSA are so high and why his eGAA is so high. Silovs is his own worst enemy.

Rebound control is a must for a Goalie to be considered a good Goalie and earn anything other than a failing grade. Silovs fails horribly.

Final Thoughts

Rick, you are going to hold your own opinion, possibly regardless of what I have just written. However, if you continue to hold your opinions on the Penguins’ Goalies we will continue to be on opposite sides of the discussion.

Save Percentage Sv% is universally held as the standard by which Goalies need to be evaluated. That is why I use this stat. I also realize the not every shooter is equal. If a Goalie routinely faces a bunch of Dominik Simon s he should be expected to have a much higher Sv%. If a Goalie usually faces a bunch of Alex Ovechkin s, I expect him to have a lower Sv%, so I standardize Sv% by applying a correction factor based on the Shooting Percentage S% of the players the Goalie faced vs the LA S%.

Your discussion of GAA vs eGAA only strengthens my negative opinion of Silovs. It just supports my thoughts by directing attention to the why s that Goalie is so bad.

If I have the time I will go into the same detail over the Skaters on the team, but that is a lot of work.