Statistics

[vis]

Some stats I thought were interesting.

 

Team stats:

• They average 2.50 5v5GF/G this year compared to 1.56 5v5GF/G last year.  They average 1.17 5v5GA/G this year compared to 1.88 5v5GA/G last year.  
• The team shooting percentage is 12.5% this year (4th overall) compared to 8.2% last year (24th overall).  Team shooting percentage leaders are Filppula (40%), Giroux (36.4%), Simmonds (26.3%) and Couturier (25.0%).  Those percentages are not sustainable.  For reference, the Capitals led the league last year at 10.5% with TJ Oshie leading them at 23.1%.  I'd be happy if they settled around between 9.5-10%, which should put them in the upper 1/3 of the league based on last year's results.
• The team save percentage is 91.1% this year (15th overall) compared to 90.1% last year (26th overall).  Still would like the save percentage to come up a little bit (last year's average was 91.3%).
• PDO is 104.3 this year (tied for 2nd overall) compared to 98.5 last year (tied for 29th overall).
• PP% is 26.9% this year (7th overall) compared to 19.5% last year (14th overall).  Not feeling any ill effects from trading Schenn.
• PK% is this year (27th overall) compared to 79.8% last year (21st overall).  Perhaps PMB and VV are missed.

 

Individual stats:

• Of his 9 points, Giroux only has 2 on the PP.  He has to pick up his play with the man advantage.  
• Filppula has sneakily scored 4 goals.
• Jake has no goals but has has attempted the most shots among forwards at 5v5 (81 SAT For).
• Laughton's line has the highest percentage of 5v5 o-zone starts among forwards (Leier: 69.44%; Raffl: 67.57%; Laughton 64.10%).  Filppula's line has the lowest percentage of o-zone starts among forwards (Filppula: 42.86%; Weal: 36.11%; Simmonds: 34.21%).  Thought this was surprising given that the fourth line is generally regarded as a defensive line.  I suppose they feel they do not warrant defensive minutes, which explains why they weren't on the ice much in the third period of the Nashville game after Nashville cut the lead to 5-4.  The other forwards for reference: Couturier (51.61%), Giroux (50.00%) and Voracek (50.82%); Patrick (45.45%), Weise (55.26%) and Konecny (57.89%). **
• However, Filppula, Weal and Simmonds have a CF% of 50.85%, 44.93% and 44.96%, respectively, which is impressive given their zone starts.  By comparison, Laughton, Leier and Raffl have a CF% of 56.18%, 56.10% and 54.44%, respectively, which is not-so-impressive given their zone starts.  The other forwards for reference: Couturier (49.08%), Giroux (49.38%) and Voracek (51.27%) [not bad]; Patrick (40.69%), Weise (42.97%) and Konecny (43.24%) [not good]. **
• Ghost and Hagg have the highest percentage of o-zone starts among d-men at (Ghost: 59.68%; Hagg: 59.38%).  Provorov and MacDonald have the lowest percentage of o-zone starts for d-men at (Provorov: 42.59%; MacDonald: 44.44%).   None of that is surprising, but helps us put Hagg's d-zone abilities in perspective.  The other d-men for reference: Sanheim (52.94%), Gudas (52.54%), Manning (50.00%). **
• Sanheim and Ghost have a CF% of 53.23% and 51.38%, respectively.  Not bad for Sanheim, but not good for Ghost given his o-zone starts.  Hagg has a CF% of 48.84%, again not good given his zone starts.  Provorov and MacDonald have a CF% of 45.55% and 44.57%, respectively, which is fair given their o-zone starts.  Manning's CF% is 42.35%, which is not good given his o-zone starts. **
• Surprised that Patrick's PDO (shooting percentage plus save percentage) is 111.8%, particularly since no one on his line has a staggering shooting percentage.  (Couturier leads with 111.3%). **

 

** Based on NHL.com's SAT stats, which is basically Corsi.  Also, not clear whether NHL.com's 5v5 zone start percentage stats take into account neutral zone starts or whether the percentage merely reflects o-zone (meaning inside the offensive blueline) and d-zone (meaning inside the defensive blueline) and disregards the neutral zone altogether or perhaps designates the redline has the division between o-zone vs. d-zone.

 

Finally, really bummed that puckalytics.com and its sister sites are gone.  Tried Corsica hockey but the site won't load.  Anyone use other advanced stat sites?  I think NHL.com is limited as is hockey-reference.com.

[Podein25]

22 minutes ago, vis said:

Thought this was surprising given that the fourth line is generally regarded as a defensive line.  I suppose they feel they do not warrant defensive minutes, which explains why they weren't on the ice much in the third period of the Nashville game after Nashville cut the lead to 5-4.

 

Good post.

 

My first thoughts were on this bullet. Basically, I think it's more a reflection of how thinking about lines has changed. The reason that 4th line is getting those OZ starts is because they want to keep the pace up, keep bringing the energy, keep them hemmed in

[OccamsRazor]

3 hours ago, Podein25 said:

 

Good post.

 

My first thoughts were on this bullet. Basically, I think it's more a reflection of how thinking about lines has changed. The reason that 4th line is getting those OZ starts is because they want to keep the pace up, keep bringing the energy, keep them hemmed in

 

 

 

 

 

 

 

[vis]

20 hours ago, Podein25 said:

 

Good post.

 

My first thoughts were on this bullet. Basically, I think it's more a reflection of how thinking about lines has changed. The reason that 4th line is getting those OZ starts is because they want to keep the pace up, keep bringing the energy, keep them hemmed in

Agree.  Fourth lines used to be for guys who wouldn't make mistakes, could be trusted defensively or for stashing tough guys.  Now, the fourth line is used to keep pressure on the other team and out of your zone.  They are a "forward" defense, in a way.

[Podein25]

19 minutes ago, vis said:

They are a "forward" defense, in a way.

Yeah, that's as good a way of putting it. Speed and tenacity required. Some smarts also.

 

I like our 4th line. Just a matter of time before Leier and Raffl pop some goals....

[OccamsRazor]

 

 

Good news.

[AJgoal]

47 minutes ago, OccamsRazor said:

 

 

Good news.

 

No, not really. The variance between goals for and expected goals for means that the Flyers are scoring more often than they should, and given the differential, it's not likely to last. That probably explains the offense coming back down to earth over the past couple of games. What you really would prefer to see here is the Flyers offense continuing to play well with the two numbers much closer together. 

 

The fact that Toronto has the top offense in the league and their GF% is below the expected is truly frightening.

[Podein25]

16 hours ago, AJgoal said:

 

No, not really. The variance between goals for and expected goals for means that the Flyers are scoring more often than they should, and given the differential, it's not likely to last. That probably explains the offense coming back down to earth over the past couple of games. What you really would prefer to see here is the Flyers offense continuing to play well with the two numbers much closer together. 

 

The fact that Toronto has the top offense in the league and their GF% is below the expected is truly frightening.

 

Yeah, I'm not really getting this particular statistical measure. 

 

A stat you can improve simply by lowering expectations? 

[elmatus]

I haven’t looked into it much. Does it take CF into account or just collective SH%?

 

If it’s just SH%, then meh. We had plenty of SoGs last season and could barely score. Quality > Quantity. 

[AJgoal]

Just now, Podein25 said:

 

Yeah, I'm not really getting this particular statistical measure. 

 

A stat you can improve simply by lowering expectations? 

 

Not sure what the question is here, but I'll try to explain the two a bit more.

 

Goals for % is simple. It's the number of goals scored by one team divided by the sum of the goals scored by them and against them. So if a team scores 10 goals and gives up 5, it's 10 divided by 15, or 67%.

 

Expected goals for is more muddy, and there are several models out there. Essentially though, they take factors like shot location and shot type into account and assign a number of goals that could be expected to be scored based on those factors. So a team taking a lot of in-close shots but running into a hot goalie might see an xGF that is higher than their actual, while a team scoring on unscreened point shots against Annti Niemi would likely see a lower xGF. Expected goals for % is then calculated the same as for the goals for percentage, but using these numbers.

 

What we're seeing with the Flyers numbers posted above is that they are exceeding their expected goals for percentage by a significant amount. If you take the Washington game as an example, Grubbauer and the Caps' defense were pretty bad, and the Flyers scored on a high number of shots, some of which weren't particularly dangerous. This is an atypical result which would be reflected in a lower xGF%. However, their actual GF% was much better. Looking at the disparity between the two numbers would give you a sense of whether the Flyers' scoring offense was sustainable or not. And when you just consider the norms of the NHL, nobody expects a team to score 7-8 goals a game.

 

All that said, we're still looking at a small sample size. The Flyers' number will start to normalize at some point, and may very likely stay somewhat elevated. But the numbers suggest one shouldn't expect them to continue to put up 4-5 goals per game over an extended period.

[vis]

Yeah, I'm not sure what to make of the xGF% stat.  Not something I am familiar with.  Based on @AJgoal's post above, it seems to be too difficult of a thing to predict, but who knows.  Curious to what extent takes into consideration the type of shots allowed by the opposition. All that said, it's interesting to see how the teams compare to each other.  At the very least, you can see that the Flyers are off the charts.  Couple that with the absurd shooting percentage and you have to expect things will come back to earth.  It's only a matter of time (and has already started, imo).

[AJgoal]

Just now, vis said:

Yeah, I'm not sure what to make of the xGF% stat.  Not something I am familiar with.  Based on @AJgoal's post above, it seems to be too difficult of a thing to predict, but who knows.  Curious to what extent takes into consideration the type of shots allowed by the opposition. All that said, it's interesting to see how the teams compare to each other.  At the very least, you can see that the Flyers are off the charts.  Couple that with the absurd shooting percentage and you have to expect things will come back to earth.  It's only a matter of time (and has already started, imo).

 

Basically. It's hard to objectively quantify something that doesn't happen, so it's fair to question xGF as a useful stat. However, in the Flyers' case, it is fairly representative of what we've been seeing.

[Podein25]

1 hour ago, AJgoal said:

 

Not sure what the question is here, but I'll try to explain the two a bit more.

 

Goals for % is simple. It's the number of goals scored by one team divided by the sum of the goals scored by them and against them. So if a team scores 10 goals and gives up 5, it's 10 divided by 15, or 67%.

 

Expected goals for is more muddy, and there are several models out there. Essentially though, they take factors like shot location and shot type into account and assign a number of goals that could be expected to be scored based on those factors. So a team taking a lot of in-close shots but running into a hot goalie might see an xGF that is higher than their actual, while a team scoring on unscreened point shots against Annti Niemi would likely see a lower xGF. Expected goals for % is then calculated the same as for the goals for percentage, but using these numbers.

 

What we're seeing with the Flyers numbers posted above is that they are exceeding their expected goals for percentage by a significant amount. If you take the Washington game as an example, Grubbauer and the Caps' defense were pretty bad, and the Flyers scored on a high number of shots, some of which weren't particularly dangerous. This is an atypical result which would be reflected in a lower xGF%. However, their actual GF% was much better. Looking at the disparity between the two numbers would give you a sense of whether the Flyers' scoring offense was sustainable or not. And when you just consider the norms of the NHL, nobody expects a team to score 7-8 goals a game.

 

All that said, we're still looking at a small sample size. The Flyers' number will start to normalize at some point, and may very likely stay somewhat elevated. But the numbers suggest one shouldn't expect them to continue to put up 4-5 goals per game over an extended period.

 

Thanks for that considered response. I think I understand it now, but I'm still left with the question: so what? How does it inform coaching decisions or strategy?

[AJgoal]

10 minutes ago, Podein25 said:

 

Thanks for that considered response. I think I understand it now, but I'm still left with the question: so what? How does it inform coaching decisions or strategy?

 

Well, if your expected GF% is low, you can look to the underlying numbers to see why. Are you blasting away from the point? maybe you try to work down lower. Maybe modify your defensive coverage to push shots more to the outside. A ratio like this probably won't help inform coaching decisions, so much as validate what is being seen with the eye test. What those numbers really say to me, is that the Flyers have been lucky. There's really no way to coach that.

[Podein25]

2 minutes ago, AJgoal said:

 

Well, if your expected GF% is low, you can look to the underlying numbers to see why. Are you blasting away from the point? maybe you try to work down lower. Maybe modify your defensive coverage to push shots more to the outside. A ratio like this probably won't help inform coaching decisions, so much as validate what is being seen with the eye test. What those numbers really say to me, is that the Flyers have been lucky. There's really no way to coach that.

 

Yeah, I guess. 

 

The problem with luck is that it runs out. Or is replaced by bad luck.

 

On the eye test, I will say this: I like our neutral zone, on both sides of the puck. I think it's the key to success in the new NHL, so that's a good thing.

[AJgoal]

Just now, Podein25 said:

 

Yeah, I guess. 

 

The problem with luck is that it runs out. Or is replaced by bad luck.

 

Yup. That was kind of the story with the 10 game winning streak last year. The Flyers were surviving on some very good luck. Then it went away, and it they had worse bad luck after that (if that makes sense).

[Podein25]

3 minutes ago, AJgoal said:

 

Yup. That was kind of the story with the 10 game winning streak last year. The Flyers were surviving on some very good luck. Then it went away, and it they had worse bad luck after that (if that makes sense).

 

Worse bad luck is the worstest kind of luck.

[OccamsRazor]

2 hours ago, Podein25 said:

 

Worse bad luck is the worstest kind of luck.

 

Well one stat i like a lot is the goals scored differential and the Flyers are +10 (with 2 shutouts against) only one team in the East is better and that is Tampa with +12 and they have played one more game.

 

[vis]

3 hours ago, AJgoal said:

 

Basically. It's hard to objectively quantify something that doesn't happen, so it's fair to question xGF as a useful stat. However, in the Flyers' case, it is fairly representative of what we've been seeing.

Yeah, clearly they have performed better than expected.  A number of stats confirm that, though.  This one is interesting, however, and not one I've considered previously.  Still trying to wrestle with how useful it is.  Hard to tell if your xGF% is good because your offense is performing well or because your defense is (similar issue with PDO).  Likewise, I would like to know to what extent the opposition is taken into consideration.  I suppose it is one more tool in the box.