Quote:
Originally Posted by Bingo
I think advanced stats are going to evolve and get better.
But they're not useless as they stand right now.
They are counts of events. The formulas are based on history of results for these events.
It isn't perfect, but it's not flawed or useless as long as the conclusions and projections that come with them aren't stretched too far.
Teams that give up more shots from the slot will probably lose more often than not, but it's not a formula as players have different finishing skills, fortune, and are facing different goaltenders.
But that doesn't change the fact that a summary of events is interesting if team X out shot team Y by a big margin, but didn't get the same margin of events in the slot from passes, shots and deflections.
The Flames played a sit back game and were awful. The shot counts support that. The scoring chances (slot shots without a pass, deflection or rebound) were heavy in Vancouver's favour as well so they had their slot time. The margin was tighter in the high danger category which could mean Calgary was focusing on blocking those pass entry points to the slot, but as a result gave up the slot to the guy carrying the puck.
Those chances aren't as dangerous to a goaltender that stops what he sees from a set position, but they finally beat Markstrom when Myers shot hit a Flame in front of the net.
|
Or there was no need for plays that qualify as HDSC because they were given other avenues to create scoring chances. It's a variable that is not easily measured. A simple "why pass back and forth when I can skate right in and shoot". In a weird scenario you could have a team score 5 breakaway goals but register no HDSCs. Same for 2 on 1s that didn't have passing plays. To me those are major flaws. Two of the most dangerous plays in hockey, no?
Another poster made a great and detailed post in another thread where he said that there are too many possibilities and variables in each play to properly measure them and frame in any kind of a stat. I agree with that.
Every goal scored is a result of 11 guys (in general) doing or not doing something. How can we measure such endless stream of randomness? I dont think we can.
We can do a lot of other things. Some may be much better indicators that HDSC. But these are events that require eyeballs.
We can sort of define what a "clean" shot is.
We can count clean shots from the slot.
We can count clean shots with goalie screened.
We can count breakaways, odd man rushes etc.
We can count clean shots taken by certain players. Because that matters too.
Forehand/backhand etc.
Deflections in front of the net. And on and on. All of these can be dangerous chances. But I don't believe there is a stat that show all of these scenarios. They on their own tell a much better tale than corsi, fenwick, hdsc etc.
Question. If you listen to the game coverage they often mention scoring chance numbers. Are these commentators looking them up somewhere or do they mark them on a notepad as they happen?
A great exercise would be to mark them as you see them and then compare with what the websites collect. If you do it for your own sake you can remove the bias factor. Better yet, have someone else do the same and compare notes after.