Inferior Goalies

[Jon-Seattle]

I just wanted to add a few more numbers to hopefully help Glenn diagnose if anything seems off.

According to the stats we have access to Shesterkin saw 923 shots on goal giving up 87 goals.

Shesterkin allowed 36 goals from position 14 which is at least double every other position on the ice.
According to Glenn’s formula, without taking into account goalie v shooter advantages one way or the other that means with Glenn’s calculations (of .893 save%) Shesterkin would have seen more than 1/3 of his shots (336 shots) from this position alone though I don’t have a way to check this against what the shot totally actually has been. League average is 34% of the leagues goals are from this position. For Shesterkin this is 42%.

The next highest scoring position was 11, at which there were only 15 goals. Which again given Glenn’s calculation (.825 save%) would mean only 86 shots against from that position. This accounts for 17% of goals against compared to league average of 21%.

So is the team allowing more shots against Shesterkin down low resulting in an increase of nearly 10% from what is expected from position 14? Well I see three options: either my team is allowing a higher percentage of shots from position 14, Shesterkins save % from pos 14 is significantly below league average, or my team is allowing a significant higher number of shots in general with a slightly abnormal but not insane statistical anomaly .

Just so people know which option I believe is responsible I’ve added this additional stat:
Shesterkins shots against per game - 33.0
Jake Allen’s shot against per game - 34.4

Most of the rest of the numbers below seem to be within normal statistical limits but pos 14 specifically seems to appear outside statistical probability without more information.

In case you want the raw numbers:
Pos on ice / # of goals / % of total / GHL Ave %
1-3 / 0 / 0 / 0
4 / 2 / 2.3% / 3.8%
5 / 2 / 2.3% / 4.7%
6 / 2 / 2.3% / 3.2%
7 / 2 / 2.3% / 4.3%
8 / 7 / 8% / 5.8%
9 / 7 / 8% / 3.7%
10 / 3 / 3.4% / 7.7%
11 / 15 / 17% / 21%
12 / 8 / 9.2% / 8.3%
13 / 2 / 2.3% / 1.7%
14 / 36 / 42% / 34%
15 / 1 / 1% / 1.8%

[Glenn-Philadelphia]

Good stuff Jon.

I think I mentioned it a few days back in this thread or another that the ice position 11 values were higher than expected and were corrected. This would have impacted Shesterkin as well as all the other goalies of course.

I am still working through my year end activities with work but will create another extract within the next few days to see how the numbers are looking now that we probably doubled the original sample size.

[Jon-Seattle]

Jan 3, 2023 14:09:47 GMT -5 Glenn-Philadelphia said:

Good stuff Jon.

I think I mentioned it a few days back in this thread or another that the ice position 11 values were higher than expected and were corrected. This would have impacted Shesterkin as well as all the other goalies of course.

I am still working through my year end activities with work but will create another extract within the next few days to see how the numbers are looking now that we probably doubled the original sample size.

Much appreciated that you’re willing to dig into this. I understand there’s a multitude of factors involved and this is one goalies stats vs stats from a few weeks ago so maybe this is absolutely normal but 42%, even from zone 14, seems off. Whereas the 34% average seems about right though spreading a few percent out to the corners might be worthwhile. Some other possible causes I can think of after having some time to think about it is too small of an advantage for goalies who are better than shooter in specific zones, too small a drop off for subpar goalies or that portion not being accounted for properly for some reason. If that’s not it than something with my defense could be preventing long range shots while encouraging short range shots, would definitely require evaluating shot charts though, which would also help inform us about shot rates per position in total as well as if those save %s per position are working as intended. This second possibility does seem strange since my team doesn’t seem to be giving up a significant number of shots and seems to go against the thought that my defense could be the issue as if I’ve read things correctly the shot chance is based on an off stat v def stat with increasing likelihood of a shot as you move up in the offensive zone, is that correct or am I reading that wrong? If that is correct than if my defense is good I should (possibly) see a decreasing number of shots but from closer in. If my defense is bad I should see more shots from farther out, which would mean more goals farther out over time. Maybe I’m reading that incorrectly though. Either way this seems to be an odd mechanic of the sim which on the surface seems to make sense but if overdone could cause goalies (even bad ones) on bad teams to see an artificial increase in their save % as they see a larger number of shots from high save % areas. But please correct me if I’m wrong on that.