What terrible one-dimensional analysis. Even strength numbers can differ in a variety of ways, such as score effects. Simple numbers are nice when you are looking for a simple comparison, but ignoring the complexities of the Leafs goaltending and system infers you're just looking for a reason to support an oft-injured Reimer. 100 games played for a goaltender is generally the litmus test for the rest of their careers -- Andersen's adjusted even strength numbers are vastly superior to Bernier and Reimer's.
The ****ing Offseason Thread - 2016
- Thread starter MindzEye
- Start date
What terrible one-dimensional analysis. Even strength numbers can differ in a variety of ways, such as score effects. Simple numbers are nice when you are looking for a simple comparison, but ignoring the complexities of the Leafs goaltending and system infers you're just looking for a reason to support an oft-injured Reimer. 100 games played for a goaltender is generally the litmus test for the rest of their careers -- Andersen's adjusted even strength numbers are vastly superior to Bernier and Reimer's.
zeke
Well-known member
And when I say show your work, don't use blatantly skewed stats like these.
I don't know if you're far-sighted or something, but there's a link there just for you. Literally the best resource for analytics, including Sportsnet writer.
Oh, you want more supporting evidence? Oh okay. Here's another layer of zone adjusted Corsi ESSV% in close score effects in which Andersen ranks sixth overall among all goaltenders in the last three seasons with 100 games played or more. Bernier? 20th overall. Oh, you think it's too strict? Okay, let's try 2,000 minutes instead. Well, well, well. Looks like Mr. Andersen is at 7th. Greiss suddenly appeared at 2nd. Bernier? 26th. Reimer? Dead last.
Oh, you want more supporting evidence? Oh okay. Here's another layer of zone adjusted Corsi ESSV% in close score effects in which Andersen ranks sixth overall among all goaltenders in the last three seasons with 100 games played or more. Bernier? 20th overall. Oh, you think it's too strict? Okay, let's try 2,000 minutes instead. Well, well, well. Looks like Mr. Andersen is at 7th. Greiss suddenly appeared at 2nd. Bernier? 26th. Reimer? Dead last.
zeke
Well-known member
flipping between random splits (4yrs then 5yrs then 3yrs then whatever) is blatant slewing right off the bat.
AND a link to a Steve "I don't get why borderline 4th liner Bozak was better without Kessel in production ans possession this year" Burtch is not evidence, especially since the tweet refers to a bizarre 4 year split when Andersen has only been in the league 3yrs - the definition of manipulating stats to fit a narrative.
AND that's even allowing that fenwick save percentage - i.e. giving goalies full credit for missed shots - is even a useful stat, which is likely silly right off the bat.
AND even if we DO actually look at fenwick save% fairly we get this:
3yrs (i.e. since Andersen made the NHL) - min 4000 minutes (36 qualified goalies)
14. Andersen 125gms, .9477
17. Reimer 111gms, .9472
20. Bernier 151gms, .9466
22. Enroth 94gms, .9453
Career (72 goalies qualified)
14. Andersen 125gms .9477
16. Reimer 215gms .9476
23. Enroth 147gms .9466
26. Bernier 209gms .9462
AND now you're bringing in Corsi which means you're giving goalies credit for blocked shots which is even more bizarre.
AND you're using a 2500 minute cutoff for which Andersen just makes (2654) to skew the stats as far to his favor as humanly possible.
This type of blatant and silly stats manipulation is not gonna fly for a guy trying to talk down.
This guy knows Zeke better than Zeke knows Zeke.
MindzEye
Wayward Ditch Pig
How has a goalie who has never played more than 40 games in a season, proven anything about their ability as a starting goalie?
I had a feeling you wouldn't look at the link at all, so thanks for proving me right. There was no flipping at all. The numbers are all based on the last three years lol ... you wanted to use the experience argument -- which is just a not-so-subtle red herring to the argument of why Reimer was moved, why Bernier was traded, why the Leafs (and every other team without a starter or in need of an upgrade) wanted Andersen. So I lowered the threshold just for you -- because if you're going to use a weighted average based on the backs of experience, you might as well admit you're incapable of rational analysis instead of an appeal to emotions and loyalty.
1,500 minutes: Andersen is 7th on evens with zone adjusted Fenwick -- Reimer sits at 18th, Bernier 24th. This removes the weighted averages and goes directly towards recent performance while remaining neutral on recency biases by raising the upper limit for 1,500 minutes at evens. Also, shot-blocking has a generally positive effect on goaltenders performance, but Fenwick doesn't account for the impact that shot-blocking infers towards (not having possession), which can be a helpful barometre towards understanding a goaltenders excellence. So with that said, Andersen sits 3rd overall among all goaltenders with 1,500+ minutes played under the same criteria, with Reimer at 25th and Bernier at 27th.
So yea, your original argument still sucks. Whoops.
1,500 minutes: Andersen is 7th on evens with zone adjusted Fenwick -- Reimer sits at 18th, Bernier 24th. This removes the weighted averages and goes directly towards recent performance while remaining neutral on recency biases by raising the upper limit for 1,500 minutes at evens. Also, shot-blocking has a generally positive effect on goaltenders performance, but Fenwick doesn't account for the impact that shot-blocking infers towards (not having possession), which can be a helpful barometre towards understanding a goaltenders excellence. So with that said, Andersen sits 3rd overall among all goaltenders with 1,500+ minutes played under the same criteria, with Reimer at 25th and Bernier at 27th.
So yea, your original argument still sucks. Whoops.
Posters like him tend to use very one-dimensional arguments by effectively bullying and weighing down a discussion by using volume. It's like death by a thousand cuts -- it's a common internet discussion board tactic to make them feel superior to others who don't have the time nor inclination to argue with someone who is too simple minded to get it.
zeke
Well-known member
you're going to have to start reading your own posts.
"It's 6th overall in the last five years for goaltenders with 5,000 minutes played or more. Bernier, for context, is ranked 34th. In addition, among goaltenders with 100+ games played since 2012-13, Frederik Andersen ranks 11th in Adj.FenSV% in All Situations."
"Here's another layer of zone adjusted Corsi ESSV% in close score effects in which Andersen ranks sixth overall among all goaltenders in the last three seasons with 100 games played or more"
these were consecutive posts. first 5yrs, then 4yrs (i.e. since 12/13), then 3yrs. all with ever changing definitions of qualified.
you're not even aware of your own arguments.
And again, I am just humoring your notion that fenwick save% (giving goalies credit for missed shots) and corsi save% (giving them credit for blocked shots too) are actual things which we should care about. (p.s. spoiler - we shouldn't).
zeke
Well-known member
bullying?
feel free to reread our exchange today, and ask yourself which one of us is the a$$hole here.
I literally posted the tweet for you. I mean, I know clicking is hard, but you can't be that out of shape.
So you admit not to knowing what close score effect is, and the impact it has on leading or trailing by one, two, or more goals. Got it.
The argument never changed. You wanted more evidence, I gave it to you. You can deflect all you want, or you can continue being a big petulant little baby that you always do when numbers get too complicated to prorate via multiplication into parenthesis.
Ah, yes. A peer reviewed, and universally accepted measure of puck possession is something others shouldn't care about because it hurts your argument to defend Reimer and Bernier. I mean, 30 NHL teams are already using puck possession measures of shot counting, including missed, blocked, and deflected shots. But I guess we should just not care because we don't want you to have some sort of mid-life crisis over whether the last twenty years of learning to multiply up to 12 is useless.
zeke
Well-known member
when I use the Starter label I'm talking about usage - playing everyday and not just in a sheltered backup role. Reimer has had injury issues which I'm not arguing.
In his rookie year, he was the starting goalie in the AHL and then came up midway through the year and was used as the starter the rest of the way. 52 total games that year.
In his 3rd season, there was a lockout, and he started 33 of 48gms and 40/55 including playoffs.
Both those years he was used as an unsheltered everyday starting goalie.
Last year, after Bernie shat the bed again, Reimer was again handed the starting role and excelled while being unsheltered.
When i differentiate between starter and backup i'm talking about unsheltered vs sheltered, not health.
In his
zeke
Well-known member
Look, kiddo, I know you're embarassed but stop digging. We all here are much more well versed in analytics and their underlying factors like score effects than you are. You are trotting out numbers you clearly don't understand trying to impress people who can see you don't understand them.
This isn't HF.
Why you chose to go to war against me I'm not sure, but you'll have to do much better than this to justify your internet toughguy act.
I have explained to you why throwing a bunch of random numbers together is not evidence. you can ignore that if you want.
I'm trying to help you out here - fenwick and corsi have proven to be effective proxies for skaters' puck possession.
this doesn't in any way mean that it makes sense to give goalies credit for missed and blocked shots as the same as saves, and in fact using fenwick and corsi this way only taints the goalies' numbers with MORE team effects, not less - if you understood these numbers you wouldn't even try to use them this way.
zeke
Well-known member
now if you're interested in considering some fenwick factors in yoir save percentage, here's an excellent attempt at a new stat called xSV% and xGS: https://hockey-graphs.com/2015/10/2...g-performance-than-existing-models/#more-4317
unfortunately i can't sort the columns on my phone but I believe it shows both bernier and reimer well ahead of andersen.
unfortunately i can't sort the columns on my phone but I believe it shows both bernier and reimer well ahead of andersen.
You clearly aren't versed -- you wouldn't have a problem with zone adjusted numbers in the first place. If score effects are complicated, you can try reading up on Desjardin's articles on it on www.behindthenet.com -- you might learn something.
Solid evidence = Random numbers
You're an adorable old man.
And also effective proxies for goaltenders. You still don't understand it. By using CSv%, you're alluding to the fact that the goaltender's team does not have possession. The underlying theme of using CSv% is to allow for measurements of a goaltenders performance with shots being blocked. Whereas it is the opposite with Fenwick. How this is still a conscientious issue for you, I'm not sure, but I really wonder if you are just learning about analytics as I post, or you're still deflecting. Read more below.
Actually, it does. As it removes the effects of blocked shots using Fenwick or the other, you can view how a goaltender performs under duress under two conditions rather than one. Also, Burtch, Berkshire, and other well known sources of analytics, including Charron, the same person who was hired by the Leafs, are all in agreement that using Zone Adjusted Fenwick Sv% is the best measure of a goaltender's performance overall at even strength in close situations. Using CSv% isn't really questionable either -- as it adds some layer of context for FSv%. But this should be already known to you. Wait, it isn't. Please try again. Use more of those multiplication tables.
zeke
Well-known member
See, again, it's obvious you don't actually know what you're talking about.
Score Effects and Zone Starts are entirely different things, yet you think they are interchangeable.
pro-tip: the split you are looking to use on the hockeyanalysis website is "Close", not "ZS Adj". One is score effects, the other is not.
solid evidence = consistent apples to apples comparisons
Random numbers = random numbers
all this from a guy who doesn't know the difference between zone starts and score effects.
Ah, another red herring. If you clicked the link in the first place, the options have already been selected for zone adjusted starts and close situations. I'm not using them interchangeably -- you are creating it. This is another tactic I've noticed that you've used on HFBoards. You love to try and manipulate a conversation by putting it back on someone else rather than focusing on the argument itself. As I gave you two different options on two different sites using both and one or the other, I'm not certain what your issue is. I don't see where I implied or said that zone starts and score effects are the same thing. As you can't seem to come back with an argument that favours Reimer or Bernier other than qualification of experience, I think we're done here. I'll wait until next year when you come up with a new argument.
Last edited: