So Brian Burke who is a loud-mouth contrarian who doesn't like analytics hires the contrarian of the hockey analytics movement. This should be interesting.
You do reslize that these sites are like the under grads who use the profs work and helps him/ her develop.
They do not have any origional thought or developed any theories.
Most of these sites exist cuse of theories i developed more than 10 years ago.
What is scary is most do not ask the correct questions or use the correct
1. situational averages and ranges for 516 situations of comp, team , ZS ( which includes comming off bench with or without pocession.
What becomes evident is you can finaly look at a players performance relative to an expected average.
Goaldif can have an expected average of + 25 to -30. You can acctualy have a player who is -27 in the toughest situation were he is expected to be -30. So relative to expected performance he is +3.
Volman,s player usage chart showed up after i presented the idea on lowetides site.
Volman,s does not present a 3 axis of data with smplitude bubble.
Which presents highly failed results.
2. Repeatability of performance. Failed eye test.
I watch video in partnership with analytic work. Loking for system play and player mechanics.
You do not measure a player by an expected cieling.
You look at the level of performance they can repeat while cobsistently following team system.
Develops player trust to read off each other.
A player who cannot be relied on being in the proper position cannot be trusted.
I have believed this for 35 years.
This is how belichek develops his teams.
Most fans have a failed eye.
They are influenced by one play rather than looking at the multitude of pkays that occur in 200 shifts.
Poster child for this was Sheldon Souray.
He was a product of NJD HD defenceman factory.
Faced 1st/2nd comp most of his career.
He would have one masive blowout that look awful.
Most people look at that and say what a terrible dman.
But one mistake is one mistake.
It does not matter if it is a massive defensive blowout or a failed cover near the net in a hd area.
Souray had one of the best mistakes per 200 shift rates in the game.
Were some of the best skating dmen in the league have a higher rate of mistakes per 200 shifts.
Guess who belichek and I would take.
Another classic example is the classic cycle pkays for 1-2 min.
People have pointed out some of these.
What do i watch for?
The low danger and high danger shot count.
Some of them did not have 1 high danger shot.
Meaning the skared alot in OZ.
But failed to penetrate HD area.
3. High danger theory
I developed this 12 years ago.
Observed that the average hd sh went in 17.5 % of the time
The average LD shot went in 3.5 % of the time a ratio of 5 to 1
From this you can get an expected ga for a given dpair.
Clearly you want to get the bst HD Corsi supression dmen in the game.
Dmen establish the avg save% a golie is asked to perform around.
Thier are 30 shots in a game.
Avg
10.5 hd shots x .825 save % = 1.8375 GA
19.5 ld shots x .965 = . 6825 GA
(30 - (1.8375 + .6825))/30 = .916
Worst
13.5 hd shots
16.5 ld shots
.902 save%
Best
7.5 hd shots
22.5 ld shots
.930 save%
4. Goalie game performance relative to cummulative hd/ld shot density
5. Open hole clsed hole shot theory.
Etc........
Thier are some theories that are so counter to current hockey belief. That old school is not willing to accept them.
Thou last summer i had a collection of div 2 and 3 coaches tell me they were changing thier shot system from high volume to hd penetration. Cause the5 times more success rate justified sacrificing volume for high hs shot count.