NHL DFS Analytics Report: Volume 1

I am excited to bring this new piece to RG. As I type this, I am really not sure what would be valuable for you readers to read on a consistent basis, so each week I am going to bring out something that I feel is worth discussing to the table. This week, I want to look at a few things. First off, I want to look at some changing statistics from this year to last year. The sample size is still VERY small, but there have been a few interesting things I have noted.

In this article, I am going to discuss what I think of some new themes this year. Which teams should we be targeting? Which teams are not as bad as people think? Which players are looking great for their new teams and with their new linemates? Which teams are overrated? Vegas is always trying to use public perception to their benefit, so using their lines early in the season makes you a bit of a sheep, especially if you don’t watch hockey. There are much better things to look at in order to evaluate how a game will go.

A Look At Some Early Season Trends

Looking at statistics from last season can be beneficial, and it still is, as we must realize that a very small early season sample size is hard to dub as a new trend. However, teams change, and moves are made, which makes some teams better and some worse. It is early. Keep that in mind when you read the chart that I am about to post, but there are some really telling stats nonetheless.

This chart is from war-on-ice.com, and was custom made by me. For those unfamiliar with corsi, it measures any shot attempt. Whether the shot misses the net, hits the net, or is blocked, it counts as a corsi attempt. Corsi is a great indicator of volume of opportunity teams will get, and with volume of opportunity comes more chances to get goals on your roster. The higher up a team is on the list, the better their goals for and against is, as you can see in the Y-axis. On the X axis we see teams that are out “corsi-ing” their opponent further right, and teams getting badly outplayed on the left. The more blue a team’s circle is, the more corsi attempts per 60 minutes they allow, and the bigger the circle, indicates that they are lucky, which is measured in PDO.

Based on this chart, we can make a few observations. First of all, Colorado is still really bad at possessing the puck. I have voiced my displeasure on various outlets about how bad their offseason was, and it is showing. Colorado was a bottom three possession team last year, and this year, they have been completely outplayed by their opponents in all their games. Colorado is getting badly outplayed, and their current goal differential is unsustainable, especially considering that Reto Berra is in net (career .900 save percentage) for the immediate future.

Everyone has been stacking heavily against the Leafs. While that can be justified with their terrible penalty kill so far, they are clearly a league average team at even strength, with two solid goaltenders. Their new coaching staff, led by Mike Babcock, has already gotten the team to buy in, and you can see in their interviews that they are all on board. Jake Gardiner mentioned that the team last year worked hard, but didn’t work smart. In other words, they are working smarter now. The Leafs are playing really good hockey, and I would advise to not just autostack against them every night, especially in GPPs. For those of you wondering where the Leafs are in that chart, they are in the middle. If you can’t read a team’s name, it means they are around league average.

The Columbus Blue Jackets are getting ripped on by their fans and media. There were high hopes for them, and as you can see, they are not getting the results. But, it might not be a product of bad hockey. They are getting poor goaltending from Bobrovsky. Their small circle indicates a very low PDO, and that is simply team shooting % for added up by team save % for. If a team shoots 8% and their goalie has a .900 save percentage they have a 98 PDO. 100 is considered standard luck. Columbus sits at a PDO just over 90. This obviously will not sustain. For those of you who might argue that their corsi might not reflect how many good quality shots they get, the Jackets have been getting around 50% of the scoring chances as well. People will be quick to pick on the Jackets, but they are not that bad, and the inflated ownership based on early season raw results make stacking against Columbus a poor decision, especially in GPPs.

On the flip side of the PDO spectrum, the Coyotes have a PDO close to 110. They are a young team, with a lot more talent than last year, but Mike Smith is not a good enough goalie to sustain this play, and once he regresses, which he is already doing, Arizona will be well on their way to their expect lottery finish. Picking on the Coyotes could be a +EV move for the next week or two, as many people actually think this team is good.

The two best corsi differential teams are St. Louis and LA. This should surprise nobody. LA was my preseason pick to win the Stanley Cup, but I do think they need one more finisher on their top six. They are a tremendous systematic hockey team, but may need one goalscorer to allow them to go up on that chart. Regardless, the teams at the top right are ones you should be stacking right now.

Other teams that continue to look vulnerable on a carryover from last year are the Calgary Flames and Vancouver Canucks. Without Mike Babcock, the Red Wings have been outplayed in most games this year, but that should change once Pavel Datsyuk returns.

Applying to DFS

A look at Sunday, October 18th GPP ownership percentages indicates that people aren’t really digging into the numbers.

Conveniently for this article, last night, the overvalued Avalanche travelled to LA, to face the Kings. This was a three-game slate. When building my lineups, I really expected LA to be severely highly owned, which is why I took four of their skaters and a goalie. LA was not only low owned, but MORE PEOPLE had Reto Berra over Jonathan Quick. Nathan Mackinnon was 40% owned, and Gabriel Landeskog was 30% owned.

Here is a look at my lineup from last night to show you how low owned the LA Kings were. LA outshot Colorado 40-23, and they were the way better team. It was lucky for Colorado owners that the game ended 2-1. I think this really displays how weak the NHL playing field is right now. You have the clear cut top possession team taking on the clear cut worst possession team at home, against a backup goalie no less. If anyone used Colorado last night, I would love to hear a non-narrative driven reason as to why you did it.

So, there you have it. Do your homework and you can get a huge edge on the field.

Now that I have covered which teams look good, here is a more detailed look at two players. One has put up great numbers this year, and the other has not. Both results might not have been expected by many. Here is why they can sustain it. I have introduced the corsi statistic, and now I will introduce WOWY stats. I will use this to test the sustainability of certain players on certain lines. For some, this may be familiar in NBA, and it works the same in NHL.

Here are a few players that look really good with their linemates so far. All stats will be via stats.hockeyanalysis.ca.

Jaromir Jagr

Jagr has been tremendous with his linemates in Florida. In fact, going back to last year since the trade to Florida, he has 24 points in 24 games. Is this sustainable? What has changed?

Let’s take a look at his possession stats and shooting stats going back into last year, so we can compare how he does with Florida and New Jersey.

As you can see in these WOWY stats, Jagr scores way more with these guys, and that is a result of attempting more shots per 60 minutes and playing a much faster pace. It also helps to get out of New Jersey, who play a defensive shell, and Jagr’s linemates in New Jersey were nowhere near the talent of Barkov and Huberdeau. If Jagr’s age doesn’t catch up to him, I see no reason why he cannot continue this stretch. His style of play is also ageless, as he is a patient guy who does not need to be going at full speed to be effective.

Patrick Sharp

Patrick Sharp is very cheap in DFS right now, and not many people are using him. This suggests that people are really uninformed. Sharp only has two assists in five games so far, but you have to realize that at his price point, he is a must play every night, as long as he skates with Tyler Seguin and Jamie Benn. They’ve only had roughly 45 minutes of ice time together, but when this trio is on the ice, they predictably take over 70 shot attempts per 60 minutes, and they have scored an insane 5.5 goals per 60 minutes. While this sample is obviously very small, Patrick Sharp absolutely dominated when he skated with Toews and Hossa in Chicago, so we can expect those offensive numbers as a line to continue. If the line is scoring that much, Sharp will get his points, and he has just been a little bit unlucky to not have his name on the scoresheet more, as he looks incredible with this line.

This Week’s Games

Using the stats I have provided above, I have handpicked one team to really look to target on each night of the upcoming week.

Tuesday, October 20th

Washington at Calgary – The Capitals’ top line has insane possession numbers, and are lethal on the powerplay. Calgary is an overachieving team since last year, with average goaltending.

Wednesday, October 21st

Carolina at Colorado – The Hurricanes have been an elite possession team this year, and Colorado has not, as I have discussed. Not only are they really safe plays, their perception of being a bad team will likely leave Colorado as favorites, making Carolina a great GPP stack..

Thursday, October 22nd

Arizona at New York – The Rangers might be an underowned team on this slate, but they should dominate this game. Arizona’s average possession numbers cannot sustain themselves long term, with their thin talent and poor goaltending. The Rangers are a good team, and that top line should do work.

Friday, October 23rd

Washington at Edmonton – I know I already recommended these guys. Looks like a long week of cheering for OVY. You’ll see where Edmonton sits in possession stats, and not much has really changed since last year despite McDavid and Yakupov proving to be really important pieces.

Saturday, October 24th

Columbus at Colorado – I discussed over and undervaluing teams early in the season leading to whacky percentages. Once again, more people will be on COL than their better opponent in this one. It’s a great GPP play.

Sunday, October 25th

Los Angeles at Edmonton – My favorite team to take players from against a really weak defensive Oilers team. Though I will probably be watching football, I will definitely get some exposure to this slate for the sake of using some Kings in this great spot.

Thank you for reading guys! I hope I provided some good information. You can hit me up on twitter or in the comments if you have any questions.

About the Author

  • Warren Kosoy (billsfan777)

  • billsfan777 has had multiple live final appearances and has numerous tournament wins in MLB, NFL, NHL and NBA . Born and raised in Toronto, the Leafs fan can be found on twitter talking hoops, NFL and the NHL @billsfan777dfs.

  • Neiderman

    • 2016 DraftKings FBWC Finalist

    Love the write up !!

  • aka_the_goat

    Cool. Thanks Warren. Good stuff!

  • ydomes

    Excellent article and fantastic analysis.

    Two questions:

    1. Based on the chart above, Boston looks to have a goal differential of -5 and an average Corsi differential. Do you think this is just a by-product of sample size? Or has Boston really lost it from last year to this year? Will Chara’s return and Bergeron’s impending return drive this number up over the course of the season or are they really going to be a bottom feeder (especially considering they are starting rebuild mode)? One final point, would you recommend stacking against them? Personally, I find it pretty tough to do when Rask is between the pipes but obviously have zero issues when the Monster is in there.

    2. For DK GPP roster construction, what would you recommend (especially if you are trying to embrace the boom bust nature of GPP lineups)? Would you stack CWD, CWD and then stack WUTIL or would you go CWW, CWUTIL with a mixture of D? Sorry if this question is a little bit confusing but I can elaborate if need be.

    Thanks for your help!

  • gravity44

    Love this. I appreciate the work that was put in. Thank you.

  • dyro01676

    good shiet!

  • billsfan777

    • 349

      RG Overall Ranking

    • 2015 DraftKings FHWC Finalist

    • Blogger of the Month

    Thanks for the feedback.

    @ydomes 1. I think Chara returning will help. Horrible offseason for the bruins though…. 2, it depends on the slate and which guys you are targetting. no right way of answering that.

  • mtp8904

    Good stuff!

  • Danomite6

    Awesome write up. It is true that there have been a lot results early that don’t reflect which teams should be winning a lot of games and which should not. This is an excellent opportunity to take advantage of before reality sets in and the true contenders win regularly while the bottom of the barrel show their true colors.

  • Tank_91

    • 2017 DraftKings FHWC Finalist

    Great work! Love seeing hockey related content on RG.

  • chrisp200

    thumbs up

  • bbmoen82

    Great content. Looking forward to future articles

  • Dnews76

    Wow, great article! In addition to the topics you covered, I’d also be interested in a players/lines to avoid list – for example, Bergeron and Toews have great defensive reputations based on Corsi, etc. Kind of an “Avoid playing this guy because he is going against Revis” type list. :D

  • maxeernst

    Great article, Warren. I actually was asked this question in the Grind Down regarding GA60 RelTM —- maybe you can help explain it for me.

    “If GA60 RelTM measures a player’s on-ice goals against per 60 minutes minus the average goals against of his linemates, couldn’t that number be significantly swayed by situational play?

    Couple of examples:
    1) You have an offensive minded player who is often substituted for on the defensive end. (How does that stat assess linemates? If Kreider plays most shifts with players X and Y, but misses some shifts stemming from faceoffs in the D zone in which players x and Y get scored on, wouldn’t that drastically affect kreiders GA60 RelTM?)
    2. You have a linemate or linemates who are atrocious defensively. Naturally, their poor play and numerous goals against would drastically affect a 3rd player’s GA60 RelTM.

    Maybe I’m way off base here and just don’t understand the stat that well. I like to think that analytics have to always be considered in a certain context and just wondering how the strategic use / deployment of a player affects GA60 RelTM.”

  • proshares

    Great article, much appreciated.

  • mdej1819

    Fantastic article here! I’m new to DFS and NHL DFS this year, so I’ve been reading a lot and trying to understand all of the different statistics to review when choosing players. This article, by far, is one of the best I’ve read. Thanks!!

  • Zieg30

    • 384

      RG Overall Ranking

    • 2018 DraftKings FGWC Finalist

    Thanks for the article!

  • Americano4pm

    Sweet read!

  • adam1xavier

    Well done Warren! Always appreciate your insight. This year I have made it a point to focus on analytics more. These are the items I find that give an edge. Thanks again.

  • mikedallas23

    Thanks billsfan, great writeup. This sort of goes to a more simple concept I’ve been using in hockey, which is that your #1 goal with skaters should be to rack up a ton of shots (assuming you’re not talking a bunch of 3% goals/shots rate shooting defenseman). You can mainly accomplish this by a) targeting opposing teams that give up a bunch of shots and b) targeting players that have been unlucky in goals/assists lately so their prices are depressed.

  • iancurtin

    As someone who never really understood the whole corsi concept, this was a great and informative article. Keep up the good work!

  • degenerate777

    great write up sir, appreciate the info

  • natroslezka

    Good job here. Also important are offensive zone start and finish percentages. Just a thought.

  • billsfan777

    • 349

      RG Overall Ranking

    • 2015 DraftKings FHWC Finalist

    • Blogger of the Month

    @maxeernst said...

    Great article, Warren. I actually was asked this question in the Grind Down regarding GA60 RelTM —- maybe you can help explain it for me.

    “If GA60 RelTM measures a player’s on-ice goals against per 60 minutes minus the average goals against of his linemates, couldn’t that number be significantly swayed by situational play?

    Couple of examples:
    1) You have an offensive minded player who is often substituted for on the defensive end. (How does that stat assess linemates? If Kreider plays most shifts with players X and Y, but misses some shifts stemming from faceoffs in the D zone in which players x and Y get scored on, wouldn’t that drastically affect kreiders GA60 RelTM?)
    2. You have a linemate or linemates who are atrocious defensively. Naturally, their poor play and numerous goals against would drastically affect a 3rd player’s GA60 RelTM.

    Maybe I’m way off base here and just don’t understand the stat that well. I like to think that analytics have to always be considered in a certain context and just wondering how the strategic use / deployment of a player affects GA60 RelTM.”

    @maxeernst Yes to both answers. I don’t really go that deep in my research. I think that is getting to be a waste of time. Many of those stats are hard to measure whether they are a reflection of talent or luck. I prefer seeing how many SCA/60 and CF/60 a player has and allows. Relative corsi and relative scf/60 are nice too. There are a lot of stats like that which I can try to get to.

  • Jburgio31

    Nice call on Ovetchkin and Co. last night. Of course I switched off of Halak for Schneider last minute. I figured it was such an easy play he’d have 25% ownership….oops. Good think it only cost me a buck.

  • Kopitar11

    Nice article, great information…thanks

  • X Unread Thread
  • X Thread with New Replies*
  • *Jumps to your first unread reply

New to DFS?

Be sure to click through our links and use our exclusive promo codes to receive the industry's best sign-up bonuses, including free access to our premium content.