<![CDATA[BEER LEAGUE HEROES - Hockey Analytics with Walt!]]>

<![CDATA[BEER LEAGUE HEROES - Hockey Analytics with Walt!]]>Thu, 19 Mar 2020 03:52:11 -0700Weebly<![CDATA[Breakdown of the Wild/Blues Series: Why the Wild won]]>Tue, 28 Apr 2015 13:15:24 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/breakdown-of-the-wildblues-series-why-the-wild-won

In the first round of the playoffs, St. Louis not only lead Minnesota, but all 16 teams in score-adjusted shot attempt differential and scoring chance differential (as defined by war-on-ice.com). I assumed that in a perfect possession world in which shooting% and save% are held constant, St. Louis easily wins the series. But my assumption was wrong! Both teams had nearly identical 5v5 shooting (close to 8%) & save percentages (approximately 92%). What gives? I decided to go deeper and probe the underlying numbers. Thankfully, war-on-ice.com provides a tonne of data to do this kind of in-depth analysis.

Special Teams?

In an online discussion, someone suggested that perhaps special teams made a difference. I investigated this angle first. St. Louis took 8 more penalties than Minnesota, so of course that didn't help. Fortunately, though, that didn't burn St. Louis while they were short-handed. However, it did mean 16 fewer minutes--relative to Minnesota--of 5v5 time, which they controlled (as measured by shot attempt & scoring chance differential) in 5 out of 6 games. Overall, though, differences in special teams didn't appear to explain this unusual situation.

Did Dubnyk win the Series?

In games 1,3, & 5--games Minnesota won--St. Louis suffered in both shooting% and save%. In particular, in games 3 and 5, Dubnyk shut them out at 5v5. St. Louis did not lack in scoring chances as they were either tied or ahead in chances in each of these games. It seemed, then, like St. Louis had the scoring opportunities, but that Dubnyk closed the door. But I still wasn't satisfied with that explanation.

I wanted to see if Dubnyk was really the key factor, or was there something else going on. For that, I went deeper into each game to see if St. Louis' scoring chances in games 1,3, & 5 were "High Danger Zone" (i.e., the slot). Because if the Blues had relatively more shots in this area, then no question that Dubnyk was a major factor in the series.

Shot Quality?

Using the shot diagrams for 5v5 situations, I compared the number of High Danger Zone shots (S) and goals (G) in each game that Minnesota won.

Game 1: Blues = 4 S (1 G), Wild = 6 S (1 G)
Game 3: Blues = 0 S (0 G), Wild = 6 S (2 G)
Game 5: Blues = 5 S (0 G), Wild = 5 S (1 G)
Game 6: Blues = 4 S (1 G), Wild = 7 S (1 G)

Now the picture becomes clearer. In terms of the highest quality shots, the Wild out-shot the Blues 24 to 13. Moreover, the Wild outscored the Blues 5 to 2 from this area. Why were the Wild able to get more quality shots? Comparing blocked shots I think provides the answer.

Defense!

The Wild's defense blocked shots like there was no tomorrow. The shot diagrams were so crowded with Bs (for blocked shots), that I was unable to count with any kind of accuracy. An overall number, though, tells the story. Over these 4 games at 5v5, the Wild blocked 74 shots to St. Louis' 39; almost a 2-to-1 ratio. This helps explain a lot. St. Louis had more shot attempts, even in the High Danger Zone, but because of Minnesota's blocking ability, they hampered the Blues from getting high quality shots. This makes Dubnyk's job easier. During the regular season, Dubnyk was letting in shots from the High Danger Zone at a 13.5% scoring rate (5th best in the league), thus the 2 goals on 13 shots was on par with his regular season.

Who were the primary shot blockers for Minnesota? I figured a few players might stand out, but as it turns out, it was a team effort. The defense, centers, and wingers all contribute. In these 4 winning games, among the defense, Spurgeon and Scandella lead with 8, Suter & Leopold with 7, Brodin had 6, and Dumba had 5. Among the centers, Coyle lead with 7, Parise was next at 5, and Brodziak had 3.

Offense too

Defense and goaltending are important, but a team still needs to score goals. I won't note the High Danger Zone shots by the Wild, but I'll note who scored. They include Parise (2 G), Pominville (1 G), Niederrieter (1 G), and Zucker (1 G). Similar to the goal scoring, noting who took the shots also showed several different players. As with shot blocking being a team effort, so was the Wild's effort to create high quality shots. Now I don't know if St. Louis were a low shot-blocking team all season, but I think it's sufficient to argue that they didn't do enough of it against the Wild. Chicago will have to figure out how to overcome this blocking tendency, or they too might get thwarted of excellent scoring chances by the Wild.

Overall, although Dubnyk was solid (except for Game 4), he did not have to overachieve for Minnesota to win the series. The team in front of Dubnyk took care of shot quality at both ends of the ice by creating more High Danger Zone scoring opportunities and by limiting St. Louis' high quality chances through a team effort.

Walter
Twitter: @Waltlaw69

]]><![CDATA[Yakupov - A Bright Light in a Dim Season]]>Mon, 23 Mar 2015 01:31:13 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/yakupov-a-bright-light-in-a-dim-season

As the end of the season approaches, we as Oilers’ fans can only shake our heads at another under-achieving season. In terms of points (62 points as predicted by Micah McCurdy), the 2014/15 season may turn out to be one of the worst seasons since 2010/11 (also 62 points). Naturally, fans are looking ahead at the draft as media pundits are saying it's a deep draft with the top 2, Connor McDavid and Jack Eichel, considered franchise players. Some (unfortunate) fans are even hoping that the Oilers luck out with the #1 pick overall, but I would argue that a 11-13% probability are not meaningful odds to pin your hopes on. I won’t belabor on this point, or the reasons why the Oilers have done poorly during their 2010-to-present rebuild. Instead, I want to focus on a positive: Nail Yakupov’s rediscovered offense!  

Fans and media seem to agree that there are at least two positives this season: Yakupov's improvement and the Oilers winning a little more under Todd Nelson. There is no doubt in my mind that Derek Roy taking Yakupov under his wing—through no encouragement from coaching or manager--has helped. If you watch games, you can see that Yakupov is visually better (hat-tip to Craig MacTavish). He enters the zone with more control, sometimes driving to the net if there’s a lane, or creating space to make a pass. Earlier in the season, and in past seasons, Yakupov's zone entries tended to be somewhat out of control with an occasional 1-on-4 mission-impossible entry. Many puck possessions were lost. But thanks to Roy's mentoring, and for a time, Benoit Pouliot's guidance, Yakupov has learned to think about where he should go and then use his speed, instead of rushing headlong into the zone and hope that it somehow works out.

  I thought it would be useful to quantify what are eyes are telling us by looking at Yakupov’s metrics. First, I'll examine his metrics at even-strength (5v5) and then on the power-play (PP). Quantifying his improvement can help us identify if what we're seeing is truly improvement, or possibly just luck finally going his way. For instance, is the fact he's scoring more merely the statistical quirk known as regression to the mean (i.e., his bad puck-luck has finally turned around), or is he actually generating more shots on net?   

For each 5v5 metric, I'm going to compare Yakupov's numbers playing with Roy versus his previous centers.

Even-Strength (5x5) Possession Metrics

The team’s possession differential with Yakupov and Roy on the ice are worse compared to Yakupov’s time with his other main centers, Draisaitl and Arcobello. Historically, Yakupov has consistently had weak possession numbers. In particular, as a broad approximation of defensive performance, his 3-year average Shot Attempts Against relative-to-teammates (per 60 min) is +4.38, which is 2nd worst on the team. (Ference’s metric is the worst at +4.67.) What this figure means is that for every 60 minutes Yakupov is on the ice, compared to other teammates, the team is allowing 4.38 more shots against. Comparing to other forwards in the league, Yakupov ranks 276 out of 288. (The Oilers best forward on this metric is Eberle, whose 3-year average of -1.55 puts him at 91st.)   

Here are some factors to consider, though, in Yakupov’s time with Roy. Looking at their zone starts, they are not getting cushy offensive zone-starts (54.5% offensive zone%). This is much less than Yakupov’s time with Draisaitl (80.5%) and Arcobello (70.7%). Also, with Hall out for an extended time with injury, Yakupov and Roy have essentially become cornerstones of the Oilers 2nd forward line. If I had to guess, their quality of competition is probably tougher than when Yakupov played the 3rd line minutes with Draisaitl and Arcobello. Moreover, the entire team’s 5v5 possession metrics have fallen since Nelson took over. This doesn’t necessarily implicate Nelson’s coaching system. During that time, the Oilers have lost David Perron and Jeff Petry (to trade), as well as Taylor Hall and Benoit Pouliot (to injury). Tough for the Oilers to maintain (or improve) their possession numbers when two of their best possession forwards—Hall and Perron—are not on the ice.   

Still, using the “eye-test” and relative possession metrics, Yakupov’s defense is something he needs to work. At this point of his career, given his struggles, I’m all in favour of helping him building his confidence on offense. Yakupov has repeatedly stated he loves to score goals. Speaking of which, how is his 5v5 offense with Roy when compared to his previous centers? Let’s take a look.

Even-Strength (5x5) Production

Looking at his time with Roy, a few numbers stuck out: His 16.62 shot attempts per 60 minutes and his 1.74 points per 60 minutes. Prior to Roy, Yakupov’s shot attempt rate (12.66 per 60) ranked him among the middle of the pack and points (0.72 per 60) ranked among the bottom-fourth of NHL forwards. Obviously, we all expect much more from the Oilers’ 2012 #1 overall draft pick. Thanks to Roy (and I cannot thank him enough), we’re now seeing a glimpse of Yakupov’s offensive potential. Where does his 1.74 points/60 rank, if he were able to sustain this production for a full season? At this level, we see players like Logan Couture, Paul Statsny, and Gabriel Landeskog (i.e., ranking 120-125 out of 400 forwards), which situates him as a slightly above-average 2nd line forward. That’s some pretty solid company. Where does his 16.62 shot attempts/60 place him? This is even more impressive: The top 40! Specifically, players who shoot at this rate include players such as Patrice Bergeron, Phil Kessel, and Corey Perry.  

This is all well and good. Yakupov is generating more shots and more frequently setting up teammates. When I looked at his shot attempt shooting percentage, though, I did have concerns. What we know about this figure, however, that you need about 3 seasons of data to have a reliable metric of ability. This season, with or without Roy, Yakupov’s shot attempt efficiency of 2.9% is below the league average among forwards of 4.2%. However, his 3-year career average is 4.8%. In other words, his puck luck is off this season, but in time, his shot efficiency should regress back to his “true” ability. 

 If Yakupov can maintain his shot and production level with Roy (presumably) next season, he should be able to put up 15-17 goals and 35-40 points. Now you may be thinking that this isn’t a lot of points, but keep in mind this is even-strength only. Forwards at this range include the likes of Scott Hartnell and Justin Williams.

  Overall, at even-strength, Yakupov’s production is at respectable level with Roy, comparable to solid 2nd line forwards. Assuming 13 minutes/game at 5v5, his phenomenal shot attempt rate could help him produce 15-17 goals and 35-40 points over a full season. Although he has work to do on his two-way game, I think he needs to build his confidence on offense first to feel like he belongs in the NHL. Being mentored by Roy has helped his confidence tremendously.

Power-Play: Possession & Production Metrics

I examined Yakupov’s metrics on the power-play by comparing his numbers between coaches. I used coaches as a comparison for a few reasons. One, as I and other writers have noted, the Oilers’ power-play (and penalty kill) have improved under Nelson. Second, Yakupov sometimes plays the first-line PP with Nugent-Hopkins, but other times on the 2nd line with Roy. Simpler, and more logical then, to compare Yakupov’s performance between coaches then between centers. I had a few questions in mind in doing this analysis. Is Yakupov playing more PP under Nelson (as Eakins didn’t seem to trust Yakupov as much)? Similar to improved production at 5v5, has his PP output noticeably improved? Finally, when Yakupov is on the ice, is the Oilers’ PP unit still as potent as the 1st line unit? This table answers the first two questions.

Looking at proportional time-on-ice (TOI%), Yakupov is actually playing slightly less under Nelson (36.5%) than Eakins (38.3%), which surprised me. It’s not much less, so perhaps over a full season it might not be much different at all. Where there is a difference between coaches, though, are Yakupov’s improved scoring chance and shot attempt rates. Scoring chances, as defined by war-on-ice.com writers, includes:

Unblocked rebounds and rush shots only in the low danger zone;
All unblocked shots in the medium danger zone; &
In the high danger zone, all shot attempts (assuming blocked shots taken here may be more representative of more “wide-open nets”).

What are these different levels of danger zones? War-on-ice provides a useful diagram in which the yellow is “low,” light-purple is “medium,” and light-blue is the “high” danger zone. (The numbers are a partial season sample from Los Angeles.)

Scoring chances are a useful metric because they account for shot quality, unlike shot attempts. Yakupov’s scoring chance rate of 20.34/60 and shot attempt rate of 35.03/60 both improved dramatically under Nelson. In fact, we see Yakupov improving even under the Mactavish/Nelson tandem. Where would Yakupov’s PP shooting rates rank him? This was hard for me to believe, but his scoring chance rate ranks among the top 15 to 20 with the likes of Alexander Ovechkin, Joe Pavelski, and Patrice Bergeron. Similarly spectacular is Yakupov’s shot attempt rate, which ranks him 25 to 30 with players like Phil Kessel and Corey Perry.

How about Yakupov’s scoring rate?

His shooting efficiency of 16.7% is slightly above average, which bodes well for being sustainable (i.e., not due to variance or “puck luck”). If his proportional time-on-ice remains at 35% to 37%, Yakupov should be able to score 7-8 PP goals over a full season. Add to that about a similar number of assists and that’s 15 PP points.  

Above, I asked the question: When Yakupov is on the ice, is the Oilers’ PP unit still as potent as the 1st line unit? Here is how the PP performs with Yakupov (and compared between coaches).

There is very little difference between the two PP units. At the level of shot attempts, the difference is negligible. At the level of scoring chances, the unit Yakupov appears to be generating a higher proportion of chances, but unlikely this is statistically meaningful. The main point is that both units are pulling their weight. Shot Attempts-For per 60 minutes (SAT/60) is the best predictor of goal differential on the PP. On this metric, since Nelson took over, the Oilers’ PP ranks 7th.  

The big question is: Are these numbers sustainable? Can Yakupov, and the Oilers as a team, keep producing and generating shots at this pace? After 31 games under Nelson, the PP has not slowed down. To watch it in motion, it is a thing of beauty. We see a well-oiled machine in which players the puck incredibly well, especially side-to-side, which inevitably opens a scoring lane for one-timers and/or open nets. And even more surprisingly, they have done it mostly without their top playmaker and most productive forward, Taylor Hall.

Summary

At even-strength, Yakupov continues to struggle on defense, but his offense has turned around. For that, we can thank Derek Roy, who has probably rescued Yakupov’s career. Playing with Roy, Yakupov’s shot attempt rate (16.62 SAT/60 min) ranks him among the top 35 to 40 forwards. His 1.74 points/60 minutes ranks him solidly among 2nd line forwards. Due to bad luck (i.e., statistical variance), his goal scoring is a bit low. Still, if can maintain these levels of shot generation and point production at even-strength next season, he could produce 35 to 40 points, with 15 to 17 goals. Under Coach Nelson’s power-play system, Yakupov’s offensive has also flourished. His shot attempt rate (35.03/60 min) places him among the top 25 to 30 forwards. Many of his shot attempts have been of higher quality. His scoring chance rate of 20.34/60 minutes ranks him among the top 15 to 20 forwards. If he continues to play 2nd line PP minutes next season, he should be able to produce about 15 points. Overall, Roy’s mentorship and Nelson’s PP system has reinvigorated Yakupov’s career. If Yakupov can maintain these shot generation and production numbers next season, he could reach realistically reach 50 to 55 points. For some, this may still seem low, but given his struggles over the last 3 years, for me, this kind of production would be a huge relief. At 21 years of age, Yakupov is still young. Forwards tend not to reach their prime until they are 24 years old, so he has time to improve even further.

Hope you enjoyed reading this post. It was fun to write, and the Oilers’ season has not been that fun to write about this season. Please leave any comments or questions below.  

Walter 
Twitter: @Waltlaw69

  Data courtesy of war-on-ice and stats.hockeyanalysis.

Notes: Instead of the term Corsi, I have used the descriptive term, shot attempts, which if you’re unfamiliar, includes blocked shots, missed shots, and shots on goal. Sometimes this makes for cumbersome wording. However, with the NHL now using shot attempts, and other descriptive terms, instead of the terms used in the analytics community, I’m going to try keep my blog posts consistent with the NHL’s terminology.

]]><![CDATA[Have the Oilers Improved Under Todd Nelson?]]>Fri, 20 Feb 2015 01:12:00 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/have-the-oilers-improved-under-todd-nelsonThe analytics suggest mostly no.

I'm as happy as any fan that the Oilers are winning more, but let's not give Nelson too much credit. Correlation is not causation. Since Todd Nelson has been coaching, the Oilers winning record has improved, but that doesn't mean they are winning because of Nelson. There may be another ("3rd") variable involved in causing the improved win/loss record, which I explain below.

First Reason for Losing: Poor Save%
What are some the reasons the Oilers were losing before Nelson and during Dallas Eakins time? The main problem was horrible goaltending. And before the reader blames the poor goals against on defense, this season's defensive stats--Corsi-Against and scoring chances against--do not show that the Oilers were allowing more shot attempts or scoring chances than previous seasons. In fact, under Eakins, the Oilers were better in these metrics. Thus, the terrible goals against can be squarely blamed on the goaltenders.

Has the Oilers' goal tending improved since Nelson? Scrivens' adjusted save% (5v5) before Nelson was 90.9% and Fasth's was 90.2%. After Nelson (excluding the shared coaching time with MacTavish), Scrivens is 91.7% and Fasth is 91.2%. With an improvement of 0.9%, this translates into 5 fewer goals against over 23 games (547 shots against) . In 1-goal games, which the Oilers have had many, that's the difference between a loss and a tie, or a win. The reason, then, for improvement in goals against is most simply explained as regression to the mean. Most goalies regress toward a 92% save% (except for the few elite goaltenders like Lundqvist). In other words, due to variance or "back luck," the goalies were under-performing under Eakins, but under Nelson they are playing closer to their "true" ability. Is it logical to attribute an improved save% to Nelson's coaching? I prefer the simpler explanation: regression to the mean.

Second Reason for Losing: Poor Shooting%
The other problem the Oilers were having under Eakins was scoring. Have they generated more even-strength (5v5) shot attempts and scoring chances with Nelson compared to Eakins? With Eakins, the Oilers 5v5 Corsi differential was 50% and the scoring chance differential was 49.4%. Under Nelson, the 5v5 Corsi differential is 46.6% & scoring chance differential is 41.8%. Thus, at 5v5, the Oilers offense generation has worsened under Nelson. No doubt the permanent loss of Perron, in addition to the injury-loss games of Hall and Pouilot have contributed to this drop in offense.

But what about shooting%? The Oilers shooting% has been poor, but this is a tricky metric.It doesn't necessarily reflect team ability to score. It tends to regress toward 8% regardless of team. For that reason, it's considered a "luck" or "variance" variable much like save%. Before Nelson, shooting% was 6.7%. After Nelson, it's 7.6%. This translates into 5 more goals over 23 games. Again, similar to improved save%, this is a case of regression to the mean.

Now one could say the Oilers shooting% has improved because they are crashing the net more, deflecting more shots, and screening the goalie more. If this is true, we would see an increase in scoring chances. But as we saw above, scoring chance differential has declined under Nelson, not improved.

It's All About Regression to the Mean
To summarize, the Oilers goaltending has regressed toward the mean as has their shooting%. Both these statistical factors have boosted the goal differential by 10 goals (5 for & 5 against) compared to these metrics during Eakins' period. These statistics happen on their own as more games are played, specifically, short-term unusually high or low numbers in shooting and save percentage gradually level out to their respective averages over the long-term.

All that being said, one sets of factors has improved under Nelson and that is special teams. Before Eakins was fired, the Oilers' special teams ranked 14th to 18th, which dipped even further under MacTavish and Nelson. (The power-play is reliably measured with Corsi-For per 60 min & the penalty-kill is reliably measured with Corsi-Against per 60 min.) Based on my possession-based ranking system, under Nelson, their PP moved up to 7th and the PK improved to 12th. When separating Eakins' from Nelson's period, the differences are even more pronounced. As noted by Jonathan Willis, under Nelson's time only, the Oilers' PP ranks 4th and the PK ranks 3rd. I will credit Nelson for improving the special teams.

Win/loss records and goal differentials can be deceiving over short periods of time. They are not reliable indicators of what's actually happening due to the randomness in goals in smaller samples. Thus, what we're seeing as improvement under Nelson at even-strength is somewhat illusory. What's occurring is regression to the mean. Nonetheless, I will acknowledge Nelson has made improvements to the team--based on measurable evidence--with respect to special teams, which is good. But the drop in even-strength performance is a concern and shows that the team really isn't all that better overall.

I'll also acknowledge, though, that losing Perron (traded to Pittsburgh) and losing Hall to injury (the Oilers' best possession and offensive player) is contributing to the team's weakened Corsi and scoring chance metrics. Thus, I am not implying Nelson is making the team worse. The loss of top-6 forwards like Perron, Pouliot, and Hall has undoubtedly figured strongly into the Oilers reduced offensive generation.

Thanks for reading! I welcome your feedback and criticism. If you don't agree, let me know (with your reasons, of course).

Walter Foddis
You can follow me on Twitter @Waltlaw69

Notes. Data courtesy of war-on-ice.com.

Addendum: I didn't realize Jonathan Willis had provided a similar analysis about 5 days ago. We pretty much come to the same conclusions. One thing Willis did was break down who was scoring more to improve the team's shooting%. According to Willis, the improvement can be completely attributed to Eberle's scoring. Over the last 5 games, I think you can also add Yakupov's scoring, whose goal-scoring is long overdue. Thanks to Derek Roy's self-appointed mentoring, Yakupov's scoring is finally "regressing to the mean."

]]><![CDATA[Fancy Stats Summary!]]>Tue, 10 Feb 2015 16:30:01 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/fancy-stats-summaryOilers vs. Devils (Feb.9/15)

From my possession-based rankings, we see a Dr. Jekyl/Mr. Hyde duality for the Devils' even-strength metrics. Although their offensive is terrible--only Buffalo is worse--their defense is stellar, ranking 3rd overall. What this amounts to on the ice is a low event game, that is, boring hockey. The first period was certainly boring as the Devils really took it to the Oilers by out-shooting them 6-4 (Corsi: 13-9). That's one shot on goal every 2 min for both teams!

At 5:47 of the 1st, Elias, who was honoured in pre-game for earning 1,000 points (400 G, 600 A), made a smooth play to back off from Petry in front of the net, received a nice pass from across the net, and put it up high short-side to make it 1-0.

Toward the end of the period, though, the Oilers began to press. At 19:24, Klefbom snapped a goal low from Eberle, who fed him from behind the net. Going tied 1-1 into the 2nd, the Oilers found their mojo and took it to the Devils and out-shot them 12-3 (Corsi: 24-11). The only goal came from a Yakupov one-timer on the power-play. Like the first Oilers' goal, it was a sweet pass from Eberle, who was at the half-wall. Yak was pretty excited and who can blame him? It was only his 6th goal of the season. Speaking of the Oilers' PP, it looked very good tonight. In 8 minutes of PP, the Oilers unloaded 12 shot attempts with 6 scoring chances (including the goal). That's the equivalent of 90 shot attempts per 60 minutes, which is comparable to Winnipeg's 6th ranked PP. Our PK was also particularly good by only allowing 2 shot attempts (1 blocked, 1 missed) over 4 minutes.

The 3rd period remained scoreless, despite a Devils 6-5 for the last minute or so, in which Cammalieri couldn't connect on an open net. Thank you, merciful gods of PDO! The Oiler still came ahead in the Corsi game 14-10 and out-shooting the Devils 8-5. In terms of scoring chances, the Oilers also came out ahead 15-12 (courtesy of war-on-ice ).

Scrivens looked solid when he had to be. Hard for a goalie when they are not busy. He had one misadventure behind the net mishandling the puck, but luckily no shot on goal.

By observation, the top line of Eberle, Nugent-Hopkins, and Purcell looked the strongest out there. I would have expected them to have the strongest Corsi differentials. Checking out hockeystats.ca, lo and behold my memory didn't fail me. The top averaged about a 69% Corsi differential; the best among the Oilers. Klefbom (68% Corsi) and Schultz (63%) were our best possession defensemen. Marincin (53%) and Fayne (55%) also did OK. That's two decent possession games (out of 2) for Marincin since being recalled.

Unfortunately, despite the game being low in Corsi events, Pakarinen (22%) and Gazdic--with only 7:55 of ice-time---(30%) were crushed, as was Ference (30%). Can't wait for Pouliot to return from injury is all I can say. I read that Pouliot is skating with the team, so he should be back soon. Hall, though, was sent to Edmonton for an MRI on his ankle. Let's hope it's nothing serious.

Realistically, the Oilers will likely finish 2nd last with or without Hall, but at least they'll win a few more games with him than without him. Players feel better about themselves when they win. I implore to the injury gods to at least not rob of them of a few more wins; in an ideal world, a .500 record for the last 28 games.

Thanks for reading. Feel free to comment below.

Walter

]]><![CDATA[Fancy Stats Game Summary - McDavid Bowl!]]>Sat, 31 Jan 2015 04:34:29 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/fancy-stats-game-summary-mcdavid-bowlWalter Foddis has our fancy stats summary for the Oiler/Sabres game on Jan.29th, 2015.

Similar to the Oilers last game vs. Buffalo, the Oil thoroughly dominated the Sabres in possession 5v5: 63 to 34 shot attempts (Corsi). They also out-chanced the Sabres by a slightly more than 2-to-1 margin: 26 to 12 scoring chances.

Our 4th line looked great out there; Lander getting 3 points (1 G, 2 A) with his first goal in 55 games, along with Fraser getting 3 points (1 G, 2 A). Interesting to see that our marginal-NHL, but successful AHL players were the most productive line. Moreover, it's a rare game that our 4L all earn a positive Corsi: Gazdic +3, Lander +4, & Fraser +5.

Speaking of possession, it was a feeding frenzy for Schultz and Klefbom, who had Corsi +/- scores of +20 & +24, respectively. The word pylon comes to mind. Don't know why.

The top line of Nugent-Hopkins, Pouliot, & Eberle had Corsi +/- scores of +12, +15, & +13. Surprisingly, Ference and Petry only had +1 each. Checking who they played against, the forward line of Gionta-Flynn-Mitchell did much of the damage to their Corsi.

Visually (a word that may never die in the Oilers' lexicon) Scrivens made some pretty good stops to keep the Oil in the game. At 12 scoring chances against, he was not tested often. He finished with a 89.5% save% 5v5, which is bad, but with the Oilers suppressing shots quite well tonight (or perhaps it's because the Oilers has the puck in the o-zone so often), Scrivens did a good enough job to earn the win.

Compared to Eakins, Nelson has had better PDO-luck (see the upward trending line in January) on his side, in terms of both a better save% and shooting%. But the way I see it, the Oilers were statistically due to win more games. The time for this miraculous process called regression to the mean just happens to be in January.

Edmonton Oilers PDO: 10-Game Moving Average

Next game is Calgary on Saturday (Jan. 31/15). Again, I'd like to see a strong possession (Calgary is a weaker possession team) and a win.

Data courtesy of war-on-ice.com & hockeystats.ca.

Thanks for reading. Any questions or comments, please drop me a line.

- Walter

PS - If you want to follow me on Twitter @Waltlaw69. In the next few weeks, I'll be posting results and links from some interesting hockey analytics projects I've undertaken. One involves a thorough analytical breakdown and ranking of defensemen. In the other, I get down and dirty to assess the impact of Quality of Competition and Quality of Teammates on certain players, in particular, Corsi superstars. Is Bergeron the best possession player in the NHL as his Corsi differential and dCorsi Impact score suggests? Stay tuned to hear the answers to this and other questions.

]]><![CDATA[Small Miracle at Rexall!]]>Sun, 11 Jan 2015 16:00:02 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/small-miracle-at-rexallThe Oilers beat the Blackhawks and Yakupov busts out! Its the Fancy Stats Wrap-up for Jan.9th, 2015!

The Oilers pull off a small miracle against a tired Hawks team and win a close game 5-2. (The 5th goal was an empty-netter by Hall.) The Hawks had played the night before against Minnesota, although it did sound like they coasted for the last 2 periods getting only 10 shots on goal.

I'm so proud of Yakupov! Nelson even called it in the pre-game presser in saying something like, "I hope he pots a few tonight." What a fun game to watch! I was a little on the edge of my seat in the 3rd period, but the Oilers did look to be in control before Yakupov scored that beauty of a goal; a backhander over Raanta's left shoulder. His goal and total game has got to help his confidence. He needed a confidence boost more than any other Oiler.

Yakupov did almost everything right. He made one mistake resulting in a turnover, but skated like mad to recover and take away the pass in front of Scrivens. The Roy-Yakupov-Pouliot line was magic tonight. In the interview with Prinicipe, Pouliot was still expressing some shame about his bad penalty vs. Detroit (which essentially cost the Oilers the game), but I'm sure his goal and the win have put those feelings to rest. I was hoping that the 2-game experiment with Fraser-Lander-Yakupov would end. They were getting absolutely crushed in possession. Nelson's decision to switch Purcell with Yakupov pleased me before the game. After this game, with Roy getting 3 points, Pouliot with a goal, and Yakupov with a goal and an assist, I expect to see them together next game.

Gordon-Hendricks-Klinkhammer continue to impress. Combined, they were just below 50% Corsi, but with heavy d-zone starts, this is just as good as 50%+. Col. Klink took command of the physical game with 12 hits. Given his size and speed, they aren't soft hits. But what impresses me more is his smart and aggressive forechecking, puck protection ability, and ability to cycle the puck down low with Gordon and Hendricks. This line is golden. With this super-solid checking line, I really cannot see a reason for Gazdic to return to the line up.

The game was exciting and emotional for me, as I'm sure it was for many long suffering fans. On the negative side, Nitikin and Ference pissed me me off. I yelled at each of them a few times. Returning to the positive side, Scrivens was solid once again with a few excellent pad-kicking saves. If I were MacT, I would say visually he's played well his last 10 games. As an analytics blogger, though, I would note that his score-adjusted 5v5 (even-strength) save% over that span is 92.14%, which is the trend we'd all like for him to continue.

Top Corsi forwards: Lander (63%), Purcell (60%), & Fraser (60%).
Top Corsi d-men: Petry (60%) & Ference (58%)

Bottom Corsi forwards: Nugent-Hopkins (40%), Hall (38%), Eberle (38%)
Bottom Corsi d-men: Schultz (40%) & Nitikin (42%)

Chicago ended slightly ahead in the score-adjusted Corsi: 45 to 42. Chicago did out chance the Oilers 20 to 15, according to war-on-ice. The shot chart below shows that many of their scoring chances were up close and personal (i.e., the slot). Indeed, they have the 2nd highest shot rate in the league from the slot; shooting 19% above league average. (Tampa Bay is shooting gangbusters from this area: 27% above league average!)

The Hawks executed the rebound-as-a-pass move to perfection with Saad getting 2 easy goals that way. Toews and Hossa were the rebound passers. Hossa got away with one, though. His trip on Nitikin went unnoticed, which resulted in his 2-on-1 with Saad and the Hawks 2nd goal.

Now I'm not going to overplay the significance of this game; it was the Hawks 2nd game of a back-to-back. Still, winning a game the right way against a tired, yet elite team--arguably the best team in the NHL--feels good. The boys should be proud of themselves tonight. I know the fans are.

Your feedback and questions are always welcome.

- Walter Foddis (@waltlaw69)

Game statistics courtesy of hockeystats.ca and war-on-ice.com.

]]><![CDATA[Analytics of a Trade: Klinkhammer vs. Perron]]>Mon, 05 Jan 2015 16:00:01 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/analytics-of-a-trade-klinkhammer-vs-perron

I like David Perron. He spoke his mind and didn’t often resort to clichés. He also sounded knowledgeable about hockey. I recall him debating analytics with Tyler Dellow—the Oilers current analytics director--on Twitter last season. In my mind, he stood out as a leader. We had heard that he was on the trading block back in December, so it was only a matter of time.  

What fans saw in Perron was a highly skilled player who had a penchant for being an agitator. Occasionally, he would take ill-advised penalties, but his aggressive style made him prone to such penalties. If he was on the ice during a scrum, you can bet he was part of it. Not a very big player, but he played with a huge heart. I’ll miss him, but I’m glad he’s gone to my 2nd favourite team, the Pittsburgh Penguins.  

In a previous post, I discussed some of the metrics of Marc Fraser and Derek Roy in which I suggested what we can realistically expect from them as Oilers. I also discussed whether the Oilers came out on top in the trade for Roy. (So far, so good.)   

In this post, I examine the metrics of Klinkhammer vs. Perron. Klinkhammer came to the Oilers, along with Pittburgh’s 2015 1st round draft pick, for Perron. Straight up, the Oilers lost offensive production. Perron has consistently produced above league average over the last several year. Similar to other Oilers, though, his output has diminished this season (1.72 pts/60 min). But even with this drop, he produced more than Klinkhammer (1.1 pts/60 min).   

But what do the other metrics tell us? To be honest, I was surprised. I had heard that Klinkhammer had strong possession metrics, but I did not realize how good. Here is the proof. (Note that these metrics are derived from even-strength (5v5) situations. Green indicates positive or above-average values. Red indicates negative or below-average values. )

Delta Corsi (dCorsi) indicates the difference the expected and the observed Corsi of a team while a player is on the ice. dCorsi is computed using a statistically derived equation, which is explained here. The dCorsi Impact score shows how many more, or fewer, shot attempts occurred than expected while the player is on the ice. Klinkhammer’s dCorsi Impact was excellent last season (+117.15) as well as this season (+47.17). To illustrate, in 2013/14 when Klinkhammer was on the ice, the Coyotes had 117 more shot attempts (compared to their opposition) than expected.   

When Klinkhammer’s dCorsi Impact is broken down, his influence on team Corsi appears to happen at both ends of the ice. (Please note that the following values are standardized to 60 min. intervals.) In 2014/15, his dCorsi-For score was +3.9, which means that while on the ice, his teams--Arizona & Pittsburgh--generated 3.9 more than expected shot attempts. At the same time, his teams also suppressed 21.13 more than expected shot attempts. His apparent defensive strength in suppressing shots is also seen in his Corsi-Against relative to teammates. So far this season, Arizona and Pittsburgh allowed 6.49 fewer shot attempts per 60 minutes while he is on the ice compared to the rest of his teammates. We also see this favorable possession pattern for Klinkhammer in 2013/14.   

In contrast, Perron’s possession metrics have not been that great during his time with the Oilers. To qualify this statement, none of the regular Oilers have had particularly strong possession numbers this season, except for Petry, Hendricks, & Gordon, especially the latter two. Still, Perron’s Corsi-Against relative to his teammates was good (-4.59). His actual vs. expected dCorsi measures were close to zero with a slightly negative dCorsi Impact score (-10.49). Overall, then, I would say Klinkhammer easily wins in the battle of possession metrics.  

I’ve also included a comparison of scoring chance metrics, which are interesting, but there hasn’t been any analysis, to my knowledge, showing that this measure is a repeatable for individuals. In other words, is the value too random or--because it is repeatable--does it suggest a player skill or ability? More work needs to be done here.

  Finally, I’ve thrown in individual points percentage (IPP) and zone starts. IPP indicates how involved a player is when the team scores. Perron contributes much more than Klinkhammer, as expected. Zone starts affect possession, but from what I can tell, knowing this doesn’t add much information. For instance, Perron has more offensive zone starts this season, which would slightly inflate his Corsi-For, but we already know that Klinkhammer has the better possession metrics.  

In the midst of writing this post, I watched the Oilers win 5-2 against the Islanders in regulation! I have the Islanders ranked 2nd in my possession-based ranking system. They’ve been near the top 5 or 6 all season, so I’m beginning to consider them a potential elite team. I may write about this game in another post. I mention the game because I was able to get a good look at Klinkhammer. He played on the 4L with Gordon and Hendricks, which is where I thought Fraser would play. Give Todd Nelson credit: Klinkhammer looked fantastic with the Oilers other two most reliable, defensive forwards. Klinkhammer’s strengths appear to be speed, puck protection, and an aggressive fore-check. Indeed, one of his fore-checks lead to New York’s goalie, Halak, turning over the puck behind the net. Klinkhammer quickly passed it to Hendricks for an easy goal.   

The Corsi differential of the Oilers’ 4L tonight: Klinkhammer = 62%, Hendricks = 64%, and Gordon = 55%. Add to the fact that they were often up against the John Tavares line, and mostly starting in the defensive zone, these numbers are even more impressive. Looking to the future of the 4L (a little premature, but I’ll go out on a limb here), it’s hard to imagine Gazdic getting much ice time with Klinkhammer on the roster. I’m sure Gazdic will get some ice time—he always does--, but I’m hopeful it’s mainly in the case of injury or other emergency.

I’ve focused on what we lost in Perron and what we gained in Klinkhammer, but the Oilers also acquired Pittsburgh’s 2015 first round pick. According to Canada’s TSN, “the first-round pick is the bigger asset for the Oilers. Even if, as expected, the Penguins finish near the top of the standings, a pick in the 25-30 range offers a little less than 15% chance of turning into a top-six forward or top-four defenceman, but a 55% chance of landing an NHL player.”

If that’s true, and presuming Klinkhammer continues to impress as a reliable 4L or 3L, this trade looks like it has more upside than fans, or I, anticipated. In short, the Oilers lost some offense, but gained a defensively responsible & puck-protecting bottom-6 forward, along with a late 1st round draft pick.

Here are a few thoughts about playing for the Edmonton Oilers by Klinkhammer himself (skip to the 1:58 mark if you don't want to listen to Derek Roy's comments about the game).

Look forward to hearing your feedback, opinions or questions.

-Walter Foddis (@Waltlaw69)

]]><![CDATA[A Waive and a Trade]]>Tue, 30 Dec 2014 14:32:44 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/a-waive-and-a-tradePossession metrics for Matt Fraser and Derek Roy.

Matt Fraser

Matt Fraser was acquired off waivers on Dec. 28, 2014. He was part of the blockbuster trade that sent Tyler Seguin to Dallas. His NHL even-strength productivity is on the low side at 1.1 points/60 minutes. Last season, playing for Boston's AHL affiliate, the Providence Bruins, he had 30 points in 44 games.

Because Fraser's NHL production numbers are pretty thin, I thought some of his possession metrics could provide additional useful information.

Matt Fraser

Corsi-For Relative to Teammates/60 min: -8.6 (bad)
Corsi-Against Relative to Teammates/60 min: -4.3 (good)
dCorsi Impact: 7.18 (good)

The Corsi-For (CF) relative-to-teammates metric (-8.6) indicates that when Fraser was on the ice, the team was relatively weaker in generating shot attempts. However, his Corsi-Against (CA) relative-to-teammates metric (-4.3) suggests that the team was relatively stronger at suppressing shot attempts when he was on the ice. In other words, Fraser appears to be stronger defensively than offensively. Fraser's dCorsi Impact score supports this conclusion. Specifically, his dCorsi Impact of 7.18 was primarily driven by his observed CA exceeding his expected CA. With Fraser on the ice, this stat tells us that the team allowed 7.2 fewer shot attempts than would be expected. Granted, his dCorsi Impact is not a huge number, but at least it's something positive. For those unfamiliar with dCorsi, this individual player metric incorporates the main factors that influence a player's Corsi such as his team's possession strength, zone starts, and player position. To read more on the dCorsi metric, go here.

In general, then, Fraser's Corsi metrics suggest he is stronger on defensive than he is on offense. This is important given how often the Gordon-Hendricks pairing--presumably Fraser's new linemates--starts in the defensive zone. The Oilers appear to be betting Fraser will be better than Pinizzotto. Probably a fair bet to make. Fraser is younger; has potentially more upside than 30-year old Pinizzotto, plus he has more NHL experience.

While I'm discussing the 4th line, I still don't understand why Joensuu's time-on-ice was suddenly reduced, and then even worse, Eakins scratched him to make way for Gazdic and Pinizzotto, who are inferior players. This is kind of move that leaves the thinking fan very puzzled. I wouldn't be surprised if Joensuu had asked to be released given this situation.

Derek Roy

Returning to Roy, he is 31 years old and has shown a sharp decline in offensive production in recent years. This is a typical pattern NHL forwards. Eric T., of SB Nation, notes: "On average, players retain about 90% of their scoring through age 29, but the drop from there is pretty sharp -- they hit 80% at age 31, 70% at age 32-33, and 60% at age 35." Including last season, Roy's production has already dropped to 55% of his prime years (0.96 pts/game; 2007-2011), so he's way ahead of the normal decline. Not a good sign. Let's at his metrics more closely, including possession.

Derek Roy

Points/60 min (5v5) = 1.66 (Forward avg = 1.52)
dCorsi Impact = -18.55
Corsi-For Relative Teammates/60 min = -0.98
Corsi-Against Relative Teammates/60 min = +1.15
iCorsi/60 min (5v5) = 10.8 (Forward avg = 11.86)
Face-off % = 48%

To summarize, Roy is slightly above average among NHL forwards for point production and slightly below average for shot attempt generation. His relative-to-teammate Corsi metrics are near zero. Finally, his face-off% ranks below the top 60 centers. To put it simply, Roy appears to be an average 3rd line center.

Are these metrics an upgrade on Arcobello? Let's take a look.

Marc Arcobello

Points/60 min (5v5) = 1.04 (+1 Roy)
dCorsi Impact = -46.35 (+1 Roy)
Corsi-For Relative Teammates/60 min = -2.11 (+1 Roy)
Corsi-Against Relative Teammates/60 min = -3.20 (+1 Arcobello)
iCorsi/60 min (5v5) = 11.14 (+1 Arcobello)
Face-off % = 47.9% (no difference)

Points-wise, Roy is an upgrade (Points/60 = 1.5 vs 1.1). He also appears to have an advantage in helping his teammates generate more shot attempts, although Arcobello takes slightly more shot attempts ( iCorsi/60 min = 11.14 vs. 10.8). Finally, their respective face-off percentages are nearly identical. Overall, Roy has a small-to-moderate advantage in offensive production, but nothing beyond that. To summarize, Roy could be an upgrade, if we're lucky, but in the long run, he may not add more to the team than what Arcobello was contributing. One thing Arcobello was not afraid of doing was checking (54 hits), which, if nothing else, can get fans into the game. Roy tends to shy away from the physical game (3 hits).

Please feel free to leave comments, criticisms, or questions. Hope you found this informative.
Data courtesy of Puckalytics, war-on-ice, Sporting Charts, BoysOnTheBus, and HockeyDB.

]]><![CDATA[Game Summary - EDM vs. ARI, Dec. 16/14: “Meet the New Boss. Same as the Old Boss.”]]>Thu, 18 Dec 2014 17:05:18 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/game-summary-edm-vs-ari-dec-1614-meet-the-new-boss-same-as-the-old-bossBefore I launch into my statistical review, I want to say a few things about the Oilers coaching direction.

Eakins was fired because the Oilers were losing. MacTavish said as much in his announcement. Eakins had made questionable decisions, for sure, but as we all in Oilersland know, the problems start at the top with management’s inability to roster a full NHL-caliber team and continue down to player performance, including a lack of scoring, defensive lapses, and poor goal tending. With a new coach, I wanted to see Eakins’ mistakes corrected (assuming Eakins was making the final roster decisions). I did not see that last night.

Instead, I saw the opposite of what I expected: Schultz having the most time-on-ice among all the defensemen. Petry and Fayne are the Oilers’ two best defensemen, but their TOI’s were 4th and 5th, respectively, behind Schultz, Hunt, and Ference. Hard for me to believe that Tyler Dellow, their analytics consultant, would endorse any such player deployment. MacTavish thinks the world of Schultz, unfortunately, so I guess it’s not a surprise. But Hunt with the 2nd most TOI? Simply makes no sense to me. Hunt is an AHL-quality player trying to break into the NHL and he’s getting top-2 minutes? I’m amazed the Oilers did as well as they did!  

Then there is Gazdic and Pinizotto, two of the Oilers worst possession players, getting about 7 min TOI. Admittedly, they played OK last night (except for a few shifts, one in which resulted in a goal against), but over the long-haul these two are hurting the team. First, they barely create any good scoring opportunities. When Gazdic gets the puck in the offensive zone, the play usually dies with him. In addition, the opposing team usually gets a Corsi uptick. Then in only playing them 7 min, the top-6 forwards are getting considerably more ice-time. Is this such a good thing? Teams like Los Angeles and Chicago play their 4th liners at least 9 min./game and usually more than 10 min. Why? Because players are not robots. They need to rest. The schedule is tough. A more rested team plays better. I can’t wait for Gordon to return. Hendricks, Gordon, and an AHLer is much preferred to Hendricks with 2 AHLers.  

With these kinds of deployment decisions, I see dark days ahead. On to my game review…  

The Oilers and Coyotes were evenly matched possession-wise in the first period. With a little help from Purcell holding a d-man back, Nugent-Hopkins scored with a wrist shot on Dubynk. But then the 2nd period? That was a disaster with a tornado touching down during a hurricane: 27 shots against! Scrivens surely deserves an honorary doctoral degree for that period alone. The 2nd period onslaught started at around 8 min., which was quickly followed by a Coyotes' goal--thanks to a defensive miscue by Gazdic--by some guy who hadn't scored in over 40 games. The Coyotes’ machine-gun target practice on Scrivens continued until the end of the period.

In the 3rd period, the Oilers got their act together and held the Coyotes to one shot attempt during the first 8 min. The Oilers won the 3rd period Corsi possession game 33 to 14.

The timing of the Coyote's overtime goal with under a second to go was unbelievable. The clock stopped at .3 seconds, but really it was .8 seconds. Huge difference! I yelled, "You've gotta be kidding!" The goal was also an excellent slap-shot by Ekman-Larsson. According to the Arizona commentators, Ekman-Larsson is becoming known for his OT heroics this season.

Due to the lopsided debacle in the 2nd period, the Coyotes won the even-strength Corsi game 61 to 49. In the 5v5 shot chart below, you’ll see the Coyotes also created more scoring chances (shots within 35 feet) with 10, whereas the Oilers had 4 (5 if you include Perron’s shot parallel to the net). Being out-chanced has been a chronic pattern for the Oil playing against western teams. In particular, Oilers have been out-chanced 251 to 190.

Our top 5v5 Corsi (Corsi +/-, Corsi%) players were Perron (+9, 65%), Arcobello (+5, 59%), and Pitlick (+4, 58%), with the latter 2 taking 4 shot attempts each. Fayne (+2, 54%) and Davidson (+2, 52%) were best possession defensemen. Our bottom Corsi players were our top line: Hall (-13, 37%), Eberle (-10, 37%), & Nugent-Hopkins (-9, 40%). The Nuge, though, fired at will with 12 shot attempts, which is probably one of his highest totals of the season. In contrast, Hall only managed 2 shot attempts. As an aside, former Oiler, Gagner, along with Chipchura, and Erat dominated possession against the Oiler’s top line.

  Our bottom 2 possession defensemen were Ference (-8, 42%) and Hunt (-6, 44%) which is expected. Ference’s defensive partner, Petry (-4, 46%), did a bit better despite being paired with Ference. I will note this especially for Petry-haters: Petry has the strongest possession metrics among our defensemen, easily, with a Corsi% of 53.3% along with a composite Corsi WOWY (with or without you) of 4%. I find David Johnson’s (owner of puckalytics.com) composite Corsi WOWY score more useful than relative Corsi. Specifically, Petry’s composite WOWY means that when he is on the ice, the team, on average, generates almost 6 more shot attempts (Corsi-For = +5.86) and allows 3 fewer shot attempts (Corsi-Against = -2.94) for every 60 minutes.

How about special teams? They looked good. Despite not scoring on the power-play (PP), the Oilers’ had 11 shot attempts over 4 min of PP (165 Corsi-For/60 min.), which is exceptional. Take note, though, Arizona’s penalty-kill (Corsi-Against/60 = 120) is the worst in the league. Although I liked seeing the Oilers generate a tonne of shots attempts, their shot quality was limited. There was only 1 shot on goal from the slot, which was a rebound (see power-play shot chart below).

The Oilers’ penalty-kill was solid by only allowing 8 shot attempts, including 2 shots on goal from the slot, over 8 min. This amount of PK shot suppression is a positive sign because the Coyotes have the 4th best PP (Corsi-For/60 = 93.5). If Arizona’s PP was firing normally, they would have had about 15 shot attempts. The Oilers’ special teams, then, executed well, but they still need more net traffic on the PP for screens, rebounds, and deflections. The Oilers did try to park a guy in front with Purcell and Draisaitl having opportunities in front of Dubnyk. Draisaitl had a nice deflection, but it went over the cross-bar. I think Draisaitl is an excellent choice because of his size and soft hands. He is also a great passer, but I think we have enough strong passers on the PP that he can perch himself in front. Now I’m wondering why they haven’t tried Draisaitl sooner.  

Thanks for reading and please leave your comments below. I will answer your questions and address any concerns.

Even-Strength (5v5) Shot Attempts (S=shot on goal; M=missed shot; B = blocked shot; enlarged S = rebounds)

Power-Play Shot Attempts (S=shot on goal; M=missed shot; B = blocked shot; enlarged S = rebounds)

Data and charts courtesy of war-on-ice.com , hockeystats.ca, & puckalytics.

]]><![CDATA[Fancy Stats Ranking Update! Wild Take over 1st; Oilers Plummet to 19th & Possibly Why]]>Sat, 13 Dec 2014 01:34:40 GMThttp://beerleagueheroes.weebly.com/hockey-analytics-with-walt/fancy-stats-ranking-update-wild-take-over-1st-oilers-plummet-to-19th-possibly-whyFancy stats ranking is updated to December 10, 2014. Bad news, Oilers fans, these last 10 days the Oilers have not only been losing games, but their overall possession metrics have also dropped...

Particularly, possession numbers have taken a hit in even-strength offense, the power-play (PP), and in penalty killing (PK). On the plus side, their even-strength defense actually improved slightly. In any case, the Oilers dropped the most of any team by 6 places to 19th. Their possession metrics are no longer in playoff territory.What has changed in the last 10 days? I'll throw out a few roster changes.

Out: Gordon (2 games), Pouliot, Joensuu, Marincin, & Nikitin (thankfully).

In: Gazdic, Pinizzotto, Aulie, and Klefbom (thankfully).

As much as some fans like the toughness of Gazdic and Pinizotto, they are significant possession liabilities. Gazdic's Corsi% and Relative Corsi%: 44.6% & - 6%. Pinizotto's respective Corsi metrics: 42.4% & -10%. Not a very wise coaching decision, if you ask me, no matter how many guys these two put at risk for developing a degenerative brain disease.

I'm also becoming more convinced (thanks to a new Jonathan Willis' analysis) that the Oilers goal-tending is indeed a problem. The team's shot suppression has improved from all shooting locations, yet the opposing teams' shooting% from all locations are above league average. Sure, you could argue that screens by opponents are contributing this higher scoring rate, and that the defense has to do a better job of clearing the net, but all goalies have to work around screens.

Looking at the rest of the league, 4 teams moved up 4 places: Winnipeg (3), Detroit (10), Carolina (13), and Florida (15). Aside from Edmonton, other teams that have dropped significantly are Vancouver, who dropped 5 places to 14th and Toronto, who dropped 4 places to 27th.

The top 4 teams changed slightly. Minnesota and Chicago switched places, and Winnipeg surged ahead of the New York Islanders, which displaced San Jose to 5th place.

1. Minnesota (If the Wild's PP was top 10, they would easily be the most dominant possession team. It is currently 18th.)

2. Chicago (The Blackhawks' even-strength defense is a weak point; ranked 10th. Their shot suppression is not as good as last season. Not terrible, mind you, still better than 20 other teams.)

3. Winnipeg (The Jets are quickly losing their underdog status in my books. What's helped them break into the top 3? A #1 ranked PK.)

4. NY Islanders (The Isles continue to have a very strong even-strength offense, which is 3rd, but only a satisfactory defense at 12th. What's really hurting them is their special teams with a 17th ranked PP and 26th ranked PK.)

I urge readers not to be overly results-focused with how Minnesota and Winnipeg are doing in the standings. These teams are much better than their win/loss records indicate.

Data courtesy of puckalytics.com & war-on-ice.com.

]]>