Friend of Leafs Hub @DestoPJ (Leafs in Bali) uses the eye test and the numbers test to analyze whether or not the Leafs defense is stronger this season with the additions of Romak Polak and Stephane Robidas.
I reach for my phone, shut off the alarm and check the time… 6:50am. Ugh, it’s another Sunday morning. After a Saturday night of Jack/Cokes, Jager Bombs, Tequila shots and who knows what else, the last thing I should be doing is waking up at 6:50am. So why am I, you ask? Well, I’m a Toronto Maple Leafs fanatical nut job who about 6 years ago decided to pick up and move to a tropical island on the other side of the world… AND, in order for me to watch ‘Saturday Night Leafs Hockey’ at 7pm (Eastern Standard), I have to haul my ass up at 6:50am Sunday morning here in Bali, Indonesia… that’s why. You can take the Blue and White fanatic out of Toronto but you can never take the Blue and White out of the fanatic.
Well, that’s the story on me, now let’s talk Leafs.
This offseason saw the Leafs’ brass bring in a number of new faces, effectively replacing our entire bottom six forward grouping while also injecting one new defensive pairing. Debates rang loud through social media as to whether any of these changes would make a significant difference in the team’s play. Being a member of Leafs Nation, the only thing I’m sure we can all agree on for certain, is that we as a Nation disagree on everything. Simple friendly Leafs Twitter conversations can turn to caps lock induced arguments (gasp!) quite quickly.
The escalation at times is followed by several tweets RT’d and MT’d to ‘holla’ at your twitter ‘posse’ for back-up in order to gang-up or, when out numbered, even-up the ‘fight odds’. Admittedly, I’ve been involved in a few (well, maybe more than a few) of these Twitter turf wars and to be strictly sincere, I don’t condemn them. Initially I planned to write how childish and annoying they were, but when pondering on it further, I’ve actually come to respect them. Why? Because nowhere else will you find a fan base more emotionally involved with their team than within Leaf Nation… and it makes me proud to be a part of that.
Of the many twitter turf fights in Leafs land this offseason, one trumpeted to the forefront:
Do Roman Polak and Stephane Robidas sans Carl Gunnarson make this defensive core better?
Considering the Leafs defensive struggles over the last several years, this debate/division in Leaf land was certainly understandable. One camp believed Polak and Robidas brought a complementary skill set to both the dressing room and the ice that the previous group lacked. These new additions were accompanied by catch terms like, ‘grit’, ‘leadership’, ‘sandpaper’ and ‘work ethic’. The other camp thought Robidas too old to contribute and Polak to be version 1.3 of the big, slow and dumb defenseman; previous versions being that of Komi, ROB and Gleason. Most who sided with this camp used analytics to support the claims made. Some with an inherently better understanding of analytics than most others, I should add. Both sides presented their arguments relatively well and now we get our first glimpse at what has actually transpired.
The “Eye Test”
Roman Polak is noticeably a big body out on the ice and logs quite a bit of ice time 5v5 and on the PK. He comes as advertised from STL with a commitment to taking the body and clearing the crease. I’ve always been a fan of a big, rough and tough, bearded beast on the back end and this Polak guy fits the bill quite well so far. He smashes anything that dares go into his corner of ice, shovels bodies from the front of the net and stands up guys at the red line more than I’ve seen any Leafs defenseman do in a very long time. He’s been paired with a rookie defenseman in Percy, who in my view was the latest victim of GM contract politics and Reilly, a defenseman with mass potential but still very young and just in his sophomore year. In my view he has well exceeded expectations and has not only become a shutdown defenseman for the Leafs but also a valuable mentor. Still, there are many that believe Polak is slow, horrible on the breakout and a giveaway machine. I don’t know how some have come to this conclusion, but just to be sure I’ll cross reference this ‘eye test’ with some statistical analysis shortly.
Stephance Robidas got the top-pairing gig due to a Franson injury to start the season. This probably wasn’t the best way to ease him back into the swing of things after coming off an injury that caused him to miss most of last season… and it showed. He had a tough go of it in the first several games and as a result spent a night in the press box. However, since then he has made some marked improvements in his play and is slowly getting to the level that many expected. Paired with Jake ‘The Snake’ Gardiner, the last 3 games have seen this pairing handle the 5-6 role quite well. More importantly, Gardiner looks to be improving his defensive IQ, or more aptly put, eliminating the ‘killer giveaway’ play from his repertoire. We all know what Jake can do offensively and if Robidas can mentor him on the defensive end, he’ll be well worth the cost, acquisition wise.
In summary, the Leafs back end has made a marked improvement over last year’s core, particularly in these last 3 games, but is it really due to these recent acquisitions or is it because of guys like Phaneuf, Franson, Reilly and Gardiner upping their game? Maybe Dion has found the “right fit” in Franson? The only way to be sure is to look at some numbers and verify what our eyes tell us.
The “Numbers Test”
I know many have a bad taste in their mouths when it comes to analytics. Whether it be the arrogant/rude delivery of it from some or the ill context use of it from others, in the end, more info is better. One of the reasons I’m a sports junkie largely has to do with looking over and analyzing stats and I am confident you the reader can relate to this at a certain level. Analytics can be quite useful but, and this is a big BUT, it needs to be put in proper context. Even though 2 players may have very similar data it doesn’t necessarily mean the same thing for both players. Players are used differently according to a coaching system and this needs to be taken into account when an analyzing any number.
So how can we know for sure whether the Leafs back end has improved over last year? We compare October ’13 to October ’14… apples to apples. Note: This season the Leafs only played 10 games in October compared to last October in where they played 14. As such I’ve included the CHI game on November 1 of this season to help even out the analysis.
The chart below depicts all Leafs defensemen (5v5) for October of last season.
The chart below depicts all Leafs defensemen (5v5) for October of this season.
Comparing one defenseman to another without any context can be misleading. Different defenseman serve different roles and should be recognized accordingly when evaluated. There are PK specialists, PP specialists, shutdown defensemen and rushing defensemen. But how do we use analytics to distinguish between them?
First, we evaluate them using 5v5 data. By eliminating any PP and PK minutes the unfair advantage that would be gained by defenseman who play more PP minutes vs PK minutes is eliminated. More shots will find the opponents net on the PP and vice versa on the PK. Shot attempt data is represented using a color scheme. The darker the blue, the more the shots ‘for’ and the darker the red, the more the shots ‘against’ whilst white means equal shots for and against.
Second, we account for the quality of player they play against. Shutdown defensemen typically play against higher quality opponents and as a result will likely face more shots against. The quality of your opponent can be estimated by the % of ice time they log. Logically, the higher a player’s TOI, the higher the quality of player. The higher a Leafs defensemen is on the chart, the higher the level of his competition. This is represented on the vertical axis and labeled ‘TOI competition %’.
Third, we account for where a defenseman will typically start his shift. A shutdown defenseman will typically be used in the Dzone and as such will start most of his shifts there. It’s fairly obvious the disadvantage this plays when analyzing shot data. We use faceoff data in the Ozone vs the Dzone to account for this. A Leafs defenseman who typically starts in the Dzone more than the Ozone relative to his teammates, will be further to the left on the chart. This is represented on the horizontal axis and labeled ‘Off ZS % Rel’. For example, Jake Gardiner will start in the Ozone approximately 12-14% more than his teammates do.
Last, we also account for average TOI/game; the larger the circle the greater the average TOI/game. I used a minimum of 50 minutes of total ice time to eliminate players who were of little contribution.
Let’s first look at this season. Phaneuf, Franson, Reilly and Gardiner all appear to be on the ice when the Leafs are ‘out-possessing’ their opponents (depicted by the whitish/bluish color of their circles), while Robidas, Polak and Percy appear to be ‘out-possessed’ when they are on the ice (whitish/reddish colour). However, this doesn’t quite tell the whole story. Polak and Percy were a pairing for 7 of the 11 games in question and typically started in the Dzone more than any of their teammates. They also played tougher competition than Reilly or Gardiner. Combine those two elements and it’s actually quite impressive they are at the light shade of red they are. Moreover, Polak logs more even strength TOI than any other Leafs player, even Phaneuf. This would indicate he is sharing a large portion of the tougher minutes with Phaneuf so far, something that didn’t happen with other defensive pairings last season. Since the more a player lands to the upper left in this chart the more likely he plays a defensive role within the team, it would appear Polak is relied upon in this way and since Reilly is his new found line mate, so too will he.
Last season, typically didn’t look good from a possession perspective, as depicted by the reddish color across all circles. What’s more concerning than that though is the disparity between the Leafs top pairing’s competition compared to the other pairings. Phaneuf and Gunnarson played most of the toughest minutes with little help. Yes, the Leafs were scoring at will, but their defense was questionable and were it not for some great goaltending would have been in trouble on the score sheet.
Basically, the addition of Polak has lightened the load for Phaneuf TOI wise and competition wise. Dion played an average TOI/game of 24.5 minutes last October compared to 22.0 minutes this October. The fact that Polak can carry another defensive partner on a different pairing and log big minutes, 21.1 minutes to be exact, has lightened Dion’s load considerably. Since Polak can also be counted on for Dzone face offs, this frees up Dion to pursue more Ozone starts. All these positives for Dion are illustrated in greatly improved possession numbers. Last season, October saw Dion have a 42.5% rate of possession 5v5. This season, he had a 51.1% rate of possession. These improved numbers could very well be the result of our forwards driving possession and with recent changes in forward lines I will delve into this more closely in the future. However, there is no denying that Polak has made a decent impact in the distribution of tough minutes to this defensive core. Hopefully, Robidas will continue to improve and do the same with the likes of Gardiner further benefiting this Leafs defensive core.
Both the ‘Eye Test’ and the ‘Numbers Test’ tend to favor this year’s defensive core over last year’s. We all wondered whether grit and sandpaper could be quantified, well look at Polak’s numbers and find the definition. All the following stats are 5v5.
- He has 41 hits this year. The next closest is Franson with 27 hits.
- He’s blocked 22 shots. Dion trails him with 12.
- He has started in the Dzone 80 times. Next closest is Dion with 61.
Despite a disproportionately few possession favourable opportunities, he still posts a respectable 45.5% rate of possession.
If this doesn’t scream grit, sandpaper or toughness to you, I don’t know what will.
Leafs in Bali
Data source: war-on-ice.com (numbers as of Nov.3rd)