Don't Know What To Make Of This

[HKSR]

Help... my head hurts trying to wrap my thoughts around how this all shakes down...

 

By the advanced stats, OEL-Myers are among the best defensive pairs in the league in terms of Expected Goals Against ....

 

[MaxVerstappen33]

What was last year for them 

[HKSR]

10th among pairings that played 800+ minutes together.  So again Top 10 in 2021-22.

[kanucks25]

Out of the 40 pairings that have played the minutes you've used as a filter, they are 31st in expected goals for %.

 

So yes they are top 10 in expected goals against, but bottom 10 in overall "impact" out of that group.

 

[HKSR]

2 minutes ago, kanucks25 said:

Out of the 40 pairings that have played the minutes you've used as a filter, they are 31st in expected goals for %.

 

So yes they are top 10 in expected goals against, but bottom 10 in overall "impact" out of that group.

 

From a defensive perspective though, they're top 10.  Would it help if they scored more?  Sure... but we're also Top 10 in GF in the entire league.  It's GA where we are #31OA.  From these stats anyways, it doesn't appear as though OEL-Myers are the issue.

[Ghostsof1915]

6 minutes ago, HKSR said:

From a defensive perspective though, they're top 10.  Would it help if they scored more?  Sure... but we're also Top 10 in GF in the entire league.  It's GA where we are #31OA.  From these stats anyways, it doesn't appear as though OEL-Myers are the issue.

But that goes against what the fans say is going on? 

[Strawbone]

This does not pass the eye test. They've looked terrible in half the games this year, directly responsible for a bunch of goals against!

 

But I'm not going to argue with math.

[kanucks25]

9 minutes ago, HKSR said:

From a defensive perspective though, they're top 10.  Would it help if they scored more?  Sure... but we're also Top 10 in GF in the entire league.  It's GA where we are #31OA.  From these stats anyways, it doesn't appear as though OEL-Myers are the issue.

Doesn't really make sense. Overall impact matters a lot more than one part of the game.

 

Their overall impact says that they are contributing a net negative.

 

Now, when they are paired together they are likely getting tough assignments but so are most of the pairings on that list.

 

Also when you sort by XGA/60 they fall down the list compared to the raw number you have it sorted by.

Edited November 16, 2022 by kanucks25

[HKSR]

1 minute ago, kanucks25 said:

Doesn't really make sense. Overall impact matters a lot more than one part of the game.

 

Their overall impact says that they are contributing a net negative.

 

Now, when they are paired together they are likely getting tough assignments but so are most of the pairings on that list.

 

Also when you sort by XGA/60 they fall down the list compared to the raw number you have it sorted by.

Yeah, but they're still near the top 20 in xGA/60.  Tied with Hanifin/Andersson out of Calgary.  That's not shabby at all.

 

But like I said, if we're looking to answers about defensive ability, ripping apart OEL and Myers is not the guys we should be going after.

[HKSR]

5 minutes ago, Ghostsof1915 said:

But that goes against what the fans say is going on? 

 

5 minutes ago, Strawbone said:

This does not pass the eye test. They've looked terrible in half the games this year, directly responsible for a bunch of goals against!

 

But I'm not going to argue with math.

I have no idea... that's why I don't know what to make of it. 

[Provost]

1 hour ago, HKSR said:

Help... my head hurts trying to wrap my thoughts around how this all shakes down...

 

By the advanced stats, OEL-Myers are among the best defensive pairs in the league in terms of Expected Goals Against ....

 

That is filtered and sorted wrong, you can't read and interpret it like you are suggesting.

You need to use the xGoals Against per 60 minutes to standardize it between the pairings, otherwise you are comparing apples to oranges depending on how many minutes they played and had goals for or against just using the raw totals.  The more minutes a pair played together the more likely they had goals scored against them.  A pair that played 158 minutes in total is likely going to have fewer goals scored against than a pair that played 250 minutes together.

If you adjust for per 60 minutes on the ice which is the only way you can compare one pairing vs another, they fall to 21 out of 38 pairings who have reached those minutes together.  If you look at xGoals% they are 30 of 38 pairs that have played that many minutes together.  If you reduce the filter to allow pairings with a few less minutes together you see that they keep falling and Hughes-Schenn are above them

You then have to look at deployment to see if they are comparable.  Are they always used in offensive zones, defensive zones, do they play the PP or PK together?  None of that is adjusted for just 5v5.  Hanafin and Tanev play double the number of PK minutes as either Myers or OEL play (not sure if they are paired together for those minutes)... so they would clearly be more likely to have goals scored against in PK minutes than 5v5.

You can't use partial pure stats to make a comparison at all it doesn't say what you are inferring it does.  They aren't top 10 or top 20 in defensive play in the league.

Naturalstattrick allows you to filter and get a better sense by comparing apples to apples.

 

Edited November 16, 2022 by Provost

[BigTramFan]

xGoals Against/60 is the measure. But also Demko currently sitting at Sv% of 0.874 this season in 11 games.

[Elias Pettersson]

Because you can't simply use analytics to evaluate how a player is performing.  It's pretty simple.  Analytics provide some of the information you need, but the eye test and watching them all game provides just as much.

[spook007]

...

Edited November 16, 2022 by spook007

[spook007]

...

Edited November 16, 2022 by spook007

[SID.IS.SID.ME.IS.ME]

Yeah, no. This isn’t it.

 

Only 40 pairings have even played >154 minutes together this season.
 

And OEL-Myers has the 3rd fewest minutes of that group.

 

You’d expect the pairings with the lowest TOI in the sample to have the lowest counts for expected goals against.

 

Once you get into the per minute rates, like xGA/60, OEL-Myers are 22nd of those 40 pairings. So middle of the pack.
 

But xGF/60 they are also 37th out of those 40. 
 

There are also different measures for expected goals. Natural Stat Trick’s numbers have OEL-Myers at 2.91 xGA/60 which ranks them 30th out of those 40 pairings (and 33rd out of 40 in xGF%).

[Bad_BOI_pete]

2 hours ago, HKSR said:

From a defensive perspective though, they're top 10.  Would it help if they scored more?  Sure... but we're also Top 10 in GF in the entire league.  It's GA where we are #31OA.  From these stats anyways, it doesn't appear as though OEL-Myers are the issue.

Goals in, goals out, matters alot.

 

when Beagle was defended i always pointed out if you cant break even, your not helping.

[HKSR]

1 hour ago, Provost said:

That is filtered and sorted wrong, you can't read and interpret it like you are suggesting.

You need to use the xGoals Against per 60 minutes to standardize it between the pairings, otherwise you are comparing apples to oranges depending on how many minutes they played and had goals for or against just using the raw totals.  The more minutes a pair played together the more likely they had goals scored against them.  A pair that played 158 minutes in total is likely going to have fewer goals scored against than a pair that played 250 minutes together.

If you adjust for per 60 minutes on the ice which is the only way you can compare one pairing vs another, they fall to 21 out of 38 pairings who have reached those minutes together.  If you look at xGoals% they are 30 of 38 pairs that have played that many minutes together.  If you reduce the filter to allow pairings with a few less minutes together you see that they keep falling and Hughes-Schenn are above them

You then have to look at deployment to see if they are comparable.  Are they always used in offensive zones, defensive zones, do they play the PP or PK together?  None of that is adjusted for just 5v5.  Hanafin and Tanev play double the number of PK minutes as either Myers or OEL play (not sure if they are paired together for those minutes)... so they would clearly be more likely to have goals scored against in PK minutes than 5v5.

You can't use partial pure stats to make a comparison at all it doesn't say what you are inferring it does.  They aren't top 10 or top 20 in defensive play in the league.

Naturalstattrick allows you to filter and get a better sense by comparing apples to apples.

 

 

24 minutes ago, SID.IS.SID.ME.IS.ME said:

Yeah, no. This isn’t it.

 

Only 40 pairings have even played >154 minutes together this season.
 

And OEL-Myers has the 3rd fewest minutes of that group.

 

You’d expect the pairings with the lowest TOI in the sample to have the lowest counts for expected goals against.

 

Once you get into the per minute rates, like xGA/60, OEL-Myers are 22nd of those 40 pairings. So middle of the pack.
 

But xGF/60 they are also 37th out of those 40. 
 

There are also different measures for expected goals. Natural Stat Trick’s numbers have OEL-Myers at 2.91 xGA/60 which ranks them 30th out of those 40 pairings (and 33rd out of 40 in xGF%).

I guess, but there's also theoretically 96+ pairings across the entire NHL.

[tas]

2 hours ago, HKSR said:

Help... my head hurts trying to wrap my thoughts around how this all shakes down...

 

By the advanced stats, OEL-Myers are among the best defensive pairs in the league in terms of Expected Goals Against ....

 

what means is it's a dumb stat, watch the game. 

[tas]

23 minutes ago, Petey_BOI said:

Goals in, goals out, matters alot.

 

when Beagle was defended i always pointed out if you cant break even, your not helping.

that's kinda dumb and oversimplified though because different types of players are deployed in different ways and different situations. defensive forwards that you line up against the oppositions' top lines are never going to break even, but they'll still do a better job of things than an offensive-minded forward would in the same role.