Eastern Edge: Defenceman Arbitrage and the Ottawa Senators

by Cam Metz on August 21, 2018
  • Eastern Edge
  • Eastern Edge: Defenceman Arbitrage and the Ottawa Senators

 

The goal of covering the Atlantic division before September has been accomplished. Unfortunately all of the digital ink that I am about to spew on the Ottawa Senators can promptly be thrown out the window once/if Erik Karlsson is traded.  I had hoped that by avoiding the Senators up until this point that something would have happened on that front. Sadly, I was wrong. 

While I promise to dive into Ottawa there really isn’t much there in terms of fantasy value, so I thought I would expand on a previous article on valuing defensemen.  In the article I throw out the value of goals scored, since the difference between 15 and 13 goals scored should not produce a large statistical advantage from one blueliner to the next. 

So what do blueliners do well? They get assists, power play points, and shots on goal.  So if we rank our defenseman specifically towards valuing those stats we can get a great idea on who is leading the way.  If you are in a bangers league – it would be wise to add in hits and blocks, but I would highly suggest you run the observations independent of each other.  Points are a lot different than counting stats; I have included SOGs simply because there is very high correlation between SOG and points.

In my mind I can’t even believe I haven’t been ranking D this way for the last couple of years. I think it is in your best interest to follow suit as well. As I write this I am cringing that fellow league mates will now have both my strategy and value ranking.

One little piece of information that I added to my defenseman model is primary assist rate on a per game basis.  Why did I choose this? Well P1 assists are correlated with consistent production rather than a player who stumbled into a rash of secondary assists.  I’ve weighted the value of this advanced stat so that it doesn’t completely overtake a PP QB1 from an efficiency standpoint (someone who only scored 15 total assists all primary should not create lopsided value comparisions).

The below plot shows the value or z-score of SOG on Y axis and z-score of assists on the X axis, far right quadrant is particularly valuable. Trust me I understand this is a terrible way to depict data as there are way too many data points, but I do want to show why certain players contain so much value, I think you’ll be able to ascertain who I am talking about here:

 

Z-score of average SOG/gm vs. z-score of average assists per game (size of data point = total value of player)

 

The fun part of this graph is Will Butcher sitting by himself at a positive z-score relative to his peers in assists, with the size of his dot factoring in the z-score power play points and primary assist rate. Also there are some players in here that when you take a look at their ADP and what other defenseman they are being drafted around is a testament to how bonkers the valuation of defensemen is in terms of assists and SOG, the two things that they are most valuable at producing.

View a fantasy draft like a stock market in that you can expect other players to have the same information by drafting based on ADP.  The ADP should reflect the overall market sentiment/value of a player.  Given this belief the ADPs of top 50 defensemen should be predictive of value and reflected with a statistically significant p-value when they placed in a linear regression.   

In fact, this is true: the ADP listed so far for the upcoming season on Yahoo! is predictive of player value and is also statistically significant. 

Given a 95% confidence interval we can plot the player value, and in this case I am going to use my player valuation (assists, SOG, A1 rate, PPP).  The plot below can show the player valuation compared to their current ADP.  Players above the line are values compared to their ADP and players below the line should be taken behind their current ADP.   I’m having a hard time getting this graph to blow up nice and big so you can find the tweet here, please hit follow as well; I will be doing this for all the other positions this summer.

It’ll be interesting to see if the market can correct as the season gets closer, but in the meantime start thinking about grabbing the guys above the lines as values compared to their peers.

 

 

Switching gears……..back to Ottawa

Like for the Wild West series, the ranks are based on a 12-team, head-to-head league, using the categories of goals, assists, power play points, shots, hits and blocks for forwards/defensemen and wins, saves, save percentage and goals against average for goalies. Player eligibility for this series is based on Yahoo, and draft ranks are based on average draft positions compiled from Yahoo, ESPN and CBS by FantasyPros.  2018-2019 ADPs are up if you couldn’t tell!

Check out yesterday’s column by Chris Kane on Minnesota.

 

Recap 

Ottawa managed to go from one goal away from the Stanley Cup Finals in 2016-2017 to one of the absolute worst teams in the league last year.  The power play was atrocious and led to Erik Karlsson’s worst power play assist rate since his first year in the league, when removing the year lost to injury in 2012-2013. 

Unfortunately the midseason trade for Matt Duchene did not produce the anticipated results and now the team is staring down the loss of a potential lottery pick in the 2019 entry draft.   Don’t expect this team to suddenly rocket to the power play elite and have multiple players in the top six worth owning.  The players worth owning on this roster for next year can be counted on one hand.

 

                              

 

Overvalued

Bobby Ryan – Enters the year as a 31-year old that has failed to produce anything greater than a 0.5 game pace for basically the last 4 years.  He was drafted at the 202 spot and never produced any fantasy value for either a points or bangers and mash league.  Don’t think this year with an increased opportunity will bring anything different for Ryan, draft a talented rookie rather than Ryan.

 

Games Played

Goals

Assists

Points

Points/ Game

Shots

Sht%

Hits

PPlay Points

Blocks

62

11

22

33

0.53

112

9.8%

97

4

52

 

Undervalued

Mark Stone – Enter Dylan Larkin 2.0. – I was sorting my spreadsheet for the last three years of production for Stone when I noticed a fairly large outlier.  He scored ONE PP goal! ONE.  He was expected to score roughly four goals, and he shot a whopping four percent on the power play.  For a guy who everyone had crowned as taking the next step last year it sure seems like Stone has another gear to find if he can find a couple more points on the power play, given the surge in his PP usage from only 56% of the available ice time I fully expect Stone to improve on his totals and rates this coming year.  This could be the opportunity share that pushes Stone into a league winner as he returns loads of value on his ADP (No ADP of value is currently available due to Stone being listed as injured).

 

Games Played

Goals

Assists

Points

Points/ Game

Shots

Sht%

Hits

PPlay Points

Blocks

58

20

42

62

1.07

132

15.2%

53

11

39

 

Matt Duchene – Duchene is currently listed as the 177th-ranked skater on Yahoo! He has been stripped of his RW tag that helped return some positional value. Duchene did turn it on for the second half of last year after the trade so it will be interesting to see what he is able to do in the upcoming campaign.  Given his skill level and the opportunity available I believe Duchene will provide value on his ADP this coming year, don’t sleep on him as a late-middle round selection (career avg point per game pace 0.68)

 

Games Played

Goals

Assists

Points

Points/ Game

Shots

Sht%

Hits

PPlay Points

Blocks

82

27

32

59

0.72

190

14.2%

60

13

49

 

Erik Karlsson –. Right now he is being slated as the 36th-best skater, when you compare his z-score to other players he will produce value that won’t be accurately captured in the player rankings since it does not factor in positional scarcity.  In most leagues Karlsson is still a first-round or possibly even second-round pick.  I’m not sure drafting Karlsson in the second round is the best options given the value that is available later in your draft even just a couple rounds later.  That being said it is tough to begrudge picking the perennial top fantasy defenseman.  Keep this in mind:

  1. His three year average 5v5 shooting percentage is 6.1% – last year he shot 2.9%
  2. He has averaged 19 power play assists the last three years – he has 13 PPA last year

 

Games Played

Goals

Assists

Points

Points/ Game

Shots

Sht%

Hits

PPlay Points

Blocks

71

9

53

62

0.87

196

4.6%

58

18

120

 

I hope you enjoyed the analysis of defensemen ADPs – if you’re interested into digging into the source files or have any question about utilizing z-scores to value the players in your league send me a note on twitter. 

Sources: Hockey-reference.com, Naturalstattrick, Corsica.hockey, Frozen Pool

 

Comments are closed.