BinCookin's Defensive Scoring

[BinCookin]

Again I will be scoring our defensemen for the Red Wings this year. I will also enjoy anyone else who wants to help, and combine your scores with mine.

Here is how I have been scoring players:

Each player receives a + (plus) or a - (minus) for good plays and bad plays they make during the game. This is of course a subjective rating system, and I will include in these stats anyone else that wants to keep tabs on the defenders during a game.

Examples:

Examples of a Minus (-) = Turnovers, non-pressured icing, fumbles, bad loss of board battle, really bad pass, bad/careless penalty
Examples of a Plus (+) = Nice Break out pass (usually underpressure to infront of our net), great offensive play, great pass, great hit, good poke check etc etc.

I occasionally give -2 if a player screws up and virtually causes a goal by himself.

Note: The exact scoring system is not as important, as the fact that all the D-men are being viewed closely by the same standards)

I will average scores for a night if multiple people rate the D men.


Here are the current 2016-2017 Yearly Totals:

https://docs.google.com/spreadsheets/d/1q4VrVqT940wR2V_qhhCxR5h9HDQqcTmnwINkfIRtUcw/edit?usp=sharing

Special thanks to any others who contribute to this system (You will be listed here)

Also Henkka Feel Free to post your stats in this thread as well!!

Click to read more...

 

[TheMule93]

Isnt this similar to how PFF ranks football players? They watch each snap then give a player a + or a - or neither

This is kinda off topic, sorry

 

[Flowah]

So... basically confirming that Smith has been a dumpster fire....

As if we needed stats to know that. What a horrible start to the season he's had.

 

[skate skate skate]

Thank you for doing this!

 

[Henkka]

My stats are pure offensive with eye-test. I will bring my combined data from first games in here until I have time to watch that Montreal game replay.

 

[BinCookin]

Ok Henkka, Perfect, just update them in that post there, and they will be near the top for others.


Also I'd like to point out that my "eye-test" stats did show Smith to be the clear cut defender with benching for Dekeyser, and it was not even close for 2nd worst.

 

[Sheriff Bert]

I have been actually thinking about doing a similar system for the offense this season. I think I'll do a trial run next game. Of course it really won't be focused on defensive play, but overall will help us see what players have been good, and others who haven't.

I'll post here and hopefully people will give me some feedback or suggestions for improvement, and maybe we can collaborate for a master system.

Examples:

Plus: Good scoring chance or shot, good possession, good pass, blocked shot, under pressure moves.

Minus: Turnovers, bad pass, inability to carry into zone, etc.

 

[BinCookin]

So... basically confirming that Smith has been a dumpster fire....

As if we needed stats to know that. What a horrible start to the season he's had.

But I did need stats to show how strong Quincey has been (Similar scores to Green, Kronwall) and how Ericsson and Kindl are moderately respectable in their numbers.

So those calling for the heads of Quincey, Kindl and Smith recently...seem to be spouting vitriol but only one of the 3 names actually seems to be performing in the way he is being villified for.

I have been actually thinking about doing a similar system for the offense this season. I think I'll do a trial run next game. Of course it really won't be focused on defensive play, but overall will help us see what players have been good, and others who haven't.

I'll post here and hopefully people will give me some feedback or suggestions for improvement, and maybe we can collaborate for a master system.

Amazing Idea!! I look forward to it.

 

[Mijatovic]

Yeah I think I could have easily said Smith was by far the worst with honestly Quincey/DDK being the only pluses from the other 6.

 

[SpookyTsuki]

Ericsson must be like 1 point above bad.

 

[Martinez]

Quincey needs to change his name. People absolutely love to hate on that guy. Contract or not, saying he's garbage is just wrong.

 

[Fugu]

Each player receives a + (plus) or a - (minus) for good plays and bad plays they make during the game. This is of course a subjective rating system, and I will include in these stats anyone else that wants to keep tabs on the defenders during a game.

Examples:

Examples of a Minus (-) = Turnovers, non-pressured icing, fumbles, bad loss of board battle, really bad pass, bad/careless penalty
Examples of a Plus (+) = Nice Break out pass (usually underpressure to infront of our net), great offensive play, great pass, great hit, good poke check etc etc.

Yeah, just a wee bit subjective, but more importantly, you're scoring these things equally, for the most part, yet taking a bad penalty or a turnover in front of your net is a bit worse than say a turnover at center ice or losing a board battle somewhere.

I occasionally give -2 if a player screws up and virtually causes a goal by himself.

You really shouldn't.

Note: The exact scoring system is not as important, as the fact that all the D-men are being viewed closely by the same standards)

Only if you can claim absolutely that you have no biases on your own.




Can I also ask why you guys aren't comparing these to officially tabbed advanced stats?

 

[Mount Suribachi]

Quincey needs to change his name. People absolutely love to hate on that guy. Contract or not, saying he's garbage is just wrong.

Tell me about it. I think its down to two things

1) Holland giving up a 1st for a player he waived several years previously (ignoring the fact that they were two completely different players)

2) Seeing everything he does as if its still 2013 and he's paired with Brendan Smith, passing the puck to each other like its a live grenade.

Fact is the guy is a real solid veteran #4, and he has been for two seasons now. Its not his fault that we lack high end talent in #1 - #3.

 

[BinCookin]

1. Yeah, just a wee bit subjective, but more importantly, you're scoring these things equally, for the most part, yet taking a bad penalty or a turnover in front of your net is a bit worse than say a turnover at center ice or losing a board battle somewhere.

2. You really shouldn't.

3. Only if you can claim absolutely that you have no biases on your own.

4. Can I also ask why you guys aren't comparing these to officially tabbed advanced stats?

Answers to your questions:
1)Obviously I already stated its subjective. If I have multiple people "play" this game with me, I will at least have more "judges".

1+2) You mentioned certain mistakes are more costly then others (Correct) and mentioned that i shouldnt give a -2 for certain mistakes. I have only done this once this year (Green on his horrible turnover and goal against) (Mostly because there was very little reason that turnover should have happened) (Technically I am penalizing him for a "drastic" mistake) But overall its a small factor in the overall scores. I.E. for that game he was 3 and -2 (he could have been 3 and -1 if i only gave him -1 there, but i felt that one play was quite bad)

3) I have very little bias because I am mostly curious how my scores will come out. Kindl and Smith are our horrible pairing right? They both are horrible right? So I watched closely and Kindl is making all his passes, he is rarely if ever turning it over. He has been out of position 1/2 times which i dock him for, but I also don't mention he is out of position because smith is out of position. So the pairing looks like a cluster ____. (For the most part, watching the D men really closely is interesting, I almost ALWAYS watch our offense and just ignore the defense because they are boring). We seem to complain alot around here about certain players because of the one "Memorable" play they had during the game. I am trying to be very non-biased when doing this. Because it really is not too hard to do when you break it down...
"Did that guy just pass it to no one causing a turnover?" -1.
"Did that player just perform a nice set of passes to relieve pressure on our zone, when I imagined they couldnt pull that off? +1.
I will be honest, I find grading Ericsson is very hard. Because he surprisingly does NO risky plays all game. He doesnt try any offensive plays that I would ever give him a + for. He also rarely goes for anything other than the obvious pass. He is the very definition of a safe stay at home defender. But he racks up minus's for mistakes like everyone else. His game is different than the others by a wide margin. But feel free to play this game with me (try to do my system) and see how it works out. You might be surprised at the results.

4) Can I also ask why you guys aren't comparing these to officially tabbed advanced stats?

- So what is a good advanced stat?
- Shots on Goal?? Possession? That really doesnt seem associated with the D men directly. Blocks? Hits? I just feel there is not a good "stat" to actually compare the D men. But feel free to enlighten me.

Curious if my system actually works?? Try it, tell me the results, and I WILL average those results with mine. And post them here!!

 

[Flowah]

SOG/Possession is a fine stat for everyone.

Better defensemen find a way to allow less SOG for opponents and more for their own team. You can do that by limiting opportunities in your own zone, keeping them to the perimeter, take aways, and good breakout passes. You can do it by keeping the puck alive in the offensive zone and distributing the puck well from the back end while you're in there. You may not get the SOG personally, but you will create an environment where your team is generating them.

 

[BinCookin]

SOG/Possession is a fine stat for everyone.

Better defensemen find a way to allow less SOG for opponents and more for their own team. You can do that by limiting opportunities in your own zone, keeping them to the perimeter, take aways, and good breakout passes. You can do it by keeping the puck alive in the offensive zone and distributing the puck well from the back end while you're in there. You may not get the SOG personally, but you will create an environment where your team is generating them.

So from this page:

http://www.nhl.com/stats/player?rep...tersummaryshooting&season=20142015&gameType=2
(Change to Defensemen and Team Detroit)-I cant save that in the link apparently

(NHL advanced stats on Detroit Defenders)..
SAT (Corsi)
USAT (Fenwick)
Which of these SAT/USAT's Do you like, or you think really judges the quality of our defensemen accurately??

 

[Number1RedWingsFan52]

Quincey needs to change his name. People absolutely love to hate on that guy. Contract or not, saying he's garbage is just wrong.

Actually he's been pretty solid when he was partnered with DDK.

 

[Flowah]

So from this page:

http://www.nhl.com/stats/player?rep...tersummaryshooting&season=20142015&gameType=2
(Change to Defensemen and Team Detroit)-I cant save that in the link apparently

(NHL advanced stats on Detroit Defenders)..
SAT (Corsi)
USAT (Fenwick)
Which of these SAT/USAT's Do you like, or you think really judges the quality of our defensemen accurately??

Corsi can't be used in a vacuum. Need also QOC, ZS, and WOWY, imo. For example, we know Ericsson is a Corsi blackhole. Kronwall being paired with him drags him down. The few times Kronwall isn't with Ericsson his numbers go way up. He is also obviously playing against other teams' top lines most of the time.

Kindl has good numbers because he's got a decent offensive mind, but he was also very sheltered in terms of competition and minutes. I mean really, look at his numbers. He had more than double the OZ faceoffs relative to his DZ faceoffs. Compared to guys like DD, Kronwall, Ericsson who are almost even in their usage. That alone is going to pump his numbers a fair bit.

Stuff like that. But yes I think generally both are quite useful as long as you contextualize them. I think it shows our middle pairing was fairly good last year. Their Corsi certainly indicates that. And they weren't getting sheltered zone starts.

 

[redwingsphan]

I'm glad that the nhl has dumped the corsi and fenwick titles. It was seriously stupid.

 

[BinCookin]

Corsi can't be used in a vacuum. Need also QOC, ZS, and WOWY, imo. For example, we know Ericsson is a Corsi blackhole. Kronwall being paired with him drags him down. The few times Kronwall isn't with Ericsson his numbers go way up. He is also obviously playing against other teams' top lines most of the time.

Kindl has good numbers because he's got a decent offensive mind, but he was also very sheltered in terms of competition and minutes. I mean really, look at his numbers. He had more than double the OZ faceoffs relative to his DZ faceoffs. Compared to guys like DD, Kronwall, Ericsson who are almost even in their usage. That alone is going to pump his numbers a fair bit.

Stuff like that. But yes I think generally both are quite useful as long as you contextualize them. I think it shows our middle pairing was fairly good last year. Their Corsi certainly indicates that. And they weren't getting sheltered zone starts.

Well you seem to be much better at these stats than I am. I wonder if my system which is subjective, but actually much more specific can infact provide more information? I can give a player a + / - based on the exact context of the situation. Kindl causes a turnover because he is 2on1'ed behind the net is something I dont have to penalize him for, vs if he has possession of the puck, and passes it into a turnover. Maybe both plays end up with loss of possession and a shot on goal against, But one is clearly Kindl's fault and one is not (Maybe the forward caused a turnover or his partner to put him in that position in the first place).

Either way, I would love if a few other people would rate our D men my way just for some more data points. And Maybe Flowah can compare our results (net results) after 10 games or 20 games to the advanced stats. And you can see if these things all seem to line up, or what not?

 

[The Zetterberg Era]

Biases always exist to some extent.

Smith has been very bad so no surprise, but I remember a similar thing happening when it wasn't really the case last year. Along with DeKeyser getting almost no minuses.

The real key would be to get enough people going to hurt some biases. I just don't think it can happen with guys like Kindl and Smith, really now Ericsson too. Should they be the minuses leaders, yes Ericsson and Smith should be leading. But for instance I find some of the minuses landing on them even when they are a partner error.

We will see, I would like to do it for a couple games, I would probably have to do it while watching the tape though and not actually live so that could be a problem for the system.

 

[Actual Thought*]

What about when a defenceman is just in good position? Like clogging a shooting/passing lane? There are so many things a blueliner can do that have massive effect and won't show up on any stat sheet. Preventing plays from ever happening doesn't stand out like being caught out of position.

 

[odin1981]

What about when a defenceman is just in good position? Like clogging a shooting/passing lane? There are so many things a blueliner can do that have massive effect and won't show up on any stat sheet. Preventing plays from ever happening doesn't stand out like being caught out of position.

That is one of the main reasons why "advanced stats" that most people track should really be used with a liberal amount of salt from the viewers. D-guys should really have there own metric because there is way too many variables that the current advanced stats can't keep in mind.

 

[Rzombo4 prez]

Well you seem to be much better at these stats than I am. I wonder if my system which is subjective, but actually much more specific can infact provide more information? I can give a player a + / - based on the exact context of the situation. Kindl causes a turnover because he is 2on1'ed behind the net is something I dont have to penalize him for, vs if he has possession of the puck, and passes it into a turnover. Maybe both plays end up with loss of possession and a shot on goal against, But one is clearly Kindl's fault and one is not (Maybe the forward caused a turnover or his partner to put him in that position in the first place).

Either way, I would love if a few other people would rate our D men my way just for some more data points. And Maybe Flowah can compare our results (net results) after 10 games or 20 games to the advanced stats. And you can see if these things all seem to line up, or what not?

BinCookin, I think you are on to something but I do think your system is a bit subjective and susceptible to confirmation bias. Expanding the pool of people rating players was a smart idea.

I found myself having a hard time rating our defensemen using your methodology. When I tried your system, I found: (a) that I couldn't rate players in real time. I had to go back and watch a replay of the game, (b) I had a very hard time holding all players to the same standard. I felt like I was crediting bottom pair guys for stuff I just routinely expected top pair guys to make, (c) there were a lot of little plays that I didn't know what to make of. You might technically call them good plays or bad plays but they didn't necessarily have a meaningful impact on what was going on, (d) I had a hard time defining turnovers at times. Obvious, unforced turnovers are easy but there are plenty of times that a defensemen takes possession of the puck and his forwards aren't giving him options and the fore check is all but requiring that he move the puck somewhere. A turnover in that context may not be good but may also be the most likely outcome of the play or the least bad of two bad options (in terms of where the turnover occurs). Again the relative abilities of the players invariably factored into how I treated these, and (e) lots of small crap like gap control, anticipation, simple puck movement to the weak side, etc. didn't necessarily fit into the methodology all that well.

If I have more time this year, I may try a simplified version of your system that focuses exclusively on zone exit plays, zone entry plays and time required to transition (absent line changes).

 

[Martinez]

Actually he's been pretty solid when he was partnered with DDK.

Loved that line last year