Ramblings: The Need to Blend Video With Analytics (Dec 8)

by steve laidlaw on December 8, 2016

Ranting about why simply using numbers in analytics is no longer enough, plus some fantasy hockey goodies!

 

Ever since the Gerard Gallant firing, I have been pondering the state of analytics in hockey. It seems we’ve reached an impasse where no one changes their mind any more. Every time someone publishes a new article or goes on a rant on TV, whatever side of the coin that you fall on, you just become more entrenched in your side.

Poor Kris Russell has somehow become the poster boy of this debate. Depending on who you ask, he is either the physical embodiment of heart or a trash defenseman who is ripping off the team who pays him. Those are over-simplifications of each sides’ argument but it’s all anyone hears. It’s just white noise. The discourse surrounding advanced stats has become hockey’s version of climate change.

Of course, every team in the league is using analytics to some degree. I think that’s why you see the standings so bunched this season. Everyone is smarter now. Ideas like Corsi aren’t even all that analytical or controversial. It is intuitive that shooting the puck more often than your opponent is a positive.

That hasn’t stopped certain from coaches talking about players “gaming” the system by taking shots from their own end, as though it’s fair debate to reduce an argument to an absurd conclusion and act like that’s what people talking about shot volume actually mean. It’s an effective distraction tactic because people can easily visualize someone flinging the puck from their own end, never mind that no player would actually do this.

Whenever I hear coaches talk like that, I see them as people who are being asked questions about strategy that they don’t want to answer. They don’t want to let people into their world and they certainly don’t want people telling them how to do their jobs. They want you to get off their lawn, even though without fan involvement, they wouldn’t get to make millions to coach a game. A more informed fan is a better fan, which helps to grow the game but it’s clear that no one in the league has much vision for that. How else do you get Gary Bettman spitting nonsense about how there is little interest among fans for salary cap info? Of course, I think that Bettman knows that there is interest but that the NHL owners would rather not have that information conveyed. For what reason, other than short-sightedness, I do not know.

But these sound-bites from coaches give people who aren’t buying analytics more ammunition to poo-poo these ideas. Since the folks in the industry don’t seem to want intelligent fans, it is on the outsiders to make us smarter. And we can’t get there without better means of communication. It’s no longer enough to explain the goal of shooting more, now there’s a need to start demonstrating how we get there.

Some folks have taken to that analysis by looking at passing data and zone entries/zone exits. My problem is that while all that analysis is good, it seems to just be leading to charts, upon charts, upon charts. If you are already sold on analytics, then more charts are a good thing. It’s more data to pore over. If you aren’t on board with analytics, all the charts do is push you further away because data just isn’t your language.

To start speaking the same language we can’t keep just using words or numbers. No more charts! Let’s get some real visuals. We need actual game footage. You’ve got data? Cool, now show me what it means on the ice. If Russell is a bad defenseman what are the plays he is (or isn’t) making that make him so bad. There’s a reason that “watch the games!” is still a criticism of the analytics. You aren’t proving critics otherwise.

There has been some dipping of the toes into this. Gus Katsaros has used footage in some of his articles, but surrounded by too much data for me to comprehend. Jack Han, who was recently profiled on TSN, has put together some interesting work with his One-Minute Tactics videos. But those all involve still photos, rather than film. Maybe I’m a dummy but I just don’t get much out of Han’s work, although I appreciate the sentiment.

I want something akin to what Zach Lowe has done for basketball. He effortlessly splices in analytics with real game footage to help explain a players’ tendencies, good and bad.

For instance, I have long lamented Evander Kane’s propensity to fling long range shots at the net on the rush:

Is this what coaches are talking about when discussing players “gaming” Corsi? This, to me, is about as bad a shot as there is. With no traffic, and no one heading to the net, it’s essentially a dump in.

Kane eventually corrals the rebound in the corner and the Sabres get an extended offensive zone chance. Perhaps there is some data that suggests Kane is a really good puck retriever. And perhaps the data even shows that these long range flings do lead to extended zone time and more subsequent chances

To me, this is why despite Kane’s tremendous skills he is merely an 8.7% career shooter. I also believe these sorts of shots are why teams have been reticent to use Kane on their top power play unit. I can’t say this for certain. But I know that blending analytis with clips like the above could answer answer that question and others like it. Maybe the data isn’t yet there for this sort of analysis and that’s why we haven’t gotten there. Or maybe the right person hasn’t come along to do the work of blending these stats.

Showing is better than telling. It’s time to stop telling, stop yelling, stop beating our heads against the wall and start showing. And if we can’t, what the hell is the point?

*

And if there are folks out there doing this sort of work, please tell me, because I haven’t seen nearly enough of it.

*

Matt Niskanen left last night’s game with a possible concussion and it damn near tanked the Capitals, who jumped out to a 3-0 only to give it all back as their defense wore down running with only five men. We’ll have to wait and see how long Niskanen is out but John Carlson had already grabbed his spot on the top PP unit so there won’t be a huge fantasy impact. Niskanen is a good puck mover though, so don’t be surprised if the Capitals as a whole are a weaker squad without him.

TJ Oshie made his return to the Capitals’ lineup and hopped back onto the top power play unit, and also on the second line with Nicklas Backstrom and Marcus Johansson. Here were the lines last night:

#1 24.5% KUZNETSOV,EVGENY – OVECHKIN,ALEXANDER – WILLIAMS,JUSTIN 

#2 24.5% BACKSTROM,NICKLAS – JOHANSSON,MARCUS – OSHIE,T.J. 

#3 19.2% BURAKOVSKY,ANDRE – ELLER,LARS – VRANA,JAKUB 

#4 18.8% BEAGLE,JAY – WILSON,TOM – WINNIK,DANIEL 

That third line is theoretically good, but Vrana has yet to score since jumping to the NHL, Andre Burakovsky is ice cold and Lars Eller left whatever scoring talent he had in Montreal.

Williams, despite being dropped from the top PP unit, found the net a couple of times. He’s spent much of the season buried on the third line and second PP unit. With this usage alongside Ovechkin, I think he could have some minor value. In most leagues, he won’t offer you much but I’ve used him as a streaming option this week. The Caps have the Friday/Sunday weekend slate, which is favourable for boosting your games.

Evgeny Kuznetsov notched a pair of assists and was dangerous all night. The buy-low window is shutting. Three assists in the past two games. He features in my latest article for Puck Daddy, which looks at some of the least fortunate players in the league in terms of Individual Point Percentage.

*

#16 on the year for David Pastrnak:

What a gem. Points for both Pastrnak and Brad Marchand tonight but none for Patrice Bergeron, who sits with just eight on the season. You won’t be shocked to find out he also appears in my latest article on IPP. I think that’s two straight weeks he’s appeared in a column discussing various metrics in which he is underperforming compared to career norms. I am sure his pre-season injury has played a role but he’s bound to turn things around. Considering the hole he has dug, it’s tough to see Bergeron getting to 55 points but 50 could be happen in things pick up for him.

Part of the funk for the Bruins is their power play, which has dipped all the way back to the bottom five of the league at just 14.4%. Do they really miss Loui Eriksson that much? David Backes has been okay, and has a pair of power-play goals already but Eriksson had 10 for them and was simply a more dynamic player. Without much of a power play, Bergeron and Ryan Spooner have seen their scoring figures tank while both David Krejci and Torey Krug got off to slow starts they are only recently coming out of.

I am not sure that the Bruins’ power play will pick up much. Other than last year’s jump they have mostly been mediocre.

*

Mike Babcock with some criticism of William Nylander:

This, along with the Leafs’ depth at center, is why Nylander is playing mostly wing so far. Frankly, I don’t see the rush to get Nylander to play center. Auston Matthews and Nazem Kadri offer a solid 1-2 punch and Tyler Bozak is just fine in a smaller role. Having flexibility is great so Nylander making strides is what we want but he’s 20 years old and playing in his first full NHL season. I am fine with him if he isn’t strong defensively.

I like that Babcock is still finding minutes for Nylander despite his defensive warts. He is still skating 16 minutes a night with top PP time. Only twice has his ice time slipped below 13 minutes in a single game.

Nylander’s production has fallen way off, with just five points in the last 15 games. I had to drop him in the one-year league that I had him in. Too many productive players on the waiver wire to be riding out this kind of cold streak. He’s now on a 54-point pace, which is right around what I’d expect for him.

I am sure there will come a time when I look to pick Nylander up again. This is all part of me using the waiver wire as an extension of my bench. And if someone in my league decides to scoop Nylander and ride the hot and cold streaks to a 54-point finish, I am okay with that too.

Check out the move by Bozak:

*

It sounds like the Leafs are gearing up to make Karri Ramo their backup in place of Jonas Enroth. I am fine with replacing Enroth but Ramo doesn’t look like the right option to replace him. Ramo was below average in three seasons for the Flames. Don’t expect Ramo to have much fantasy relevance or do enough to keep Frederik Andersen from starting 65+ games.

Instead, it’s going to Andersen’s annual injury that halts him before getting to 60 and then the Leafs are screwed.

*

Great timing for Pierre LeBrun’s profile on Devan Dubnyk as he pitched another gem to pad his league-leading .946 save percentage.

Before this season began, I was pointing out how Dubnyk had essentially the same season as Martin Jones did but did so for a team offering less offensive support. Other than wins, their numbers were virtually identical. Jones was a top-five fantasy guy thanks to 37 wins on top of his solid play in stopping the puck.

Dubnyk is getting more goal support this season but is still lagging behind Jones in wins, all while playing the kind of lights out hockey we haven’t seen since Dubnyk’s last run of brilliance. Odds are his play dips back into that league-average range for save percentage and the Wild’s scoring numbers are dipping. If I had to bet on who would be the better fantasy own for the rest of the season I would bet on Jones, but Dubnyk ain’t far behind.

*

The Sharks-Senators battle was full of my favourite delights. Erik Karlsson netted three points, while Brent Burns had one. The two best offensive defensemen in the league, genuinely unique players, going tit for tat. Just wonderful.

The Mike HoffmanMark Stone pairing remained hot as well, which is nice to see.

Joe Pavelski had an assist but couldn’t buy a goal all night, despite many a glorious chance. We expect him to always be buzzing but this is your reminder that he is a constant threat.

Logan Couture scored his 10th goal of the season. Six of those have come with the man-advantage. Sharks are still looking for depth scoring, and yes Mikkel Boedker still stinks.

*

Considering my rant about the need for better visuals, I have started to do some GIF work. Kudos to Jaccob Slavin’s wicked foot pass for the inspiration:

Slavin has grabbed Noah Hanifin’s spot on the top power play unit and has been up over 25 minutes of action in each of the past five games. Only two points for Slavin in those five games, but this ups his short-term value a little, although Justin Faulk remains the go-to option for fantasy value on the Hurricanes’ blue line.

I know Faulk has been trapped in an igloo but note the secondary assist on the Teuvo Teravainen goal above. Only a matter of time before Faulk start producing regularly.

*

With Rickard Rakell out the past couple of games with a lower-body injury it’s been Nick Ritchie’s chance on the top line with Corey Perry and Ryan Getzlaf. Those two haven’t been great shakes so it was a bit of a surprise that Ritchie scored last night. Perry scored as well so perhaps we are seeing the tide turn for the big boys.

The line that is really clicking is the Ryan Kesler, Jakob Silfverberg, Andrew Cogliano trio, which combined for two goals and six points last night. Cogliano doesn’t have much relevance but Kesler might be the multi-category MVP and Jakob Silfverberg is finally putting together a meaningful fantasy season.

Obviously Kesler won’t continue to score on 18.5% of his shots but his usage is tremendous, skating on this great secondary line and on the Ducks’ lethal top PP unit. He’s on pace for 70 points, which I have a tough time seeing him hit, but 60 could definitely happen.

Silfverberg, meanwhile, is seeing the most power play time since his rookie season. He’s only a second unit guy but it’s still meaningful that he’s getting 1:50 per game after averaging just 51 seconds and 35 seconds per game in the two previous seasons respectively. He only has two PPP thus far but his career high is four so it’s safe to say he’ll blow that out of the water. If he even gets to 40 points it’ll be a career year but he’s on pace for 58. I think he can get 50, if he gets the right bounces.

*

With the Penguins getting healthy, they sent down Derrick Pouliot and Jake Guentzel yesterday. Interesting to contrast these two as I know some are ready to paint Pouliot a bust, while fawning over Guentzel as a late bloomer. They are both 22 years old, with solid scoring numbers at the AHL level. But because Pouliot was a high draft pick who has been on our radar for four years now and hasn’t done much in the 57 games of NHL action he has played, we assume he’s a dud. That may prove true but I would suggest that we need more time to find out for sure.

Looking back on Pouliot’s draft year (2012), the two best fantasy defensemen were taken in round three with Colton Parayko and Shayne Gostisbehere. You can certainly make an argument for the best “real life” defenseman coming out of the lottery section with Morgan Rielly, Hampus Lindholm and Jacob Trouba all top-10 picks but for my fantasy roster, I want the third-round guys. Take from that what you will but, to me, this is further evidence that full on fantasy rebuilds are unnecessary.

*

 Thanks for reading! Follow me on Twitter @SteveLaidlaw.