JACKSON CAGE

Jackson Houck has a big season ahead, one that could thrust him into the picture as a possible Edmonton Oiler in the coming years. Houck had a slow start but strong finish in 2013-14, and will play on an emerging team in 2014-15. Houck will have a new coach, as this story from Elliott Pap of the Vancouver Sun details. One quote from Houck is interesting:

  • “I think everyone is a little bit nervous and excited to see what it’s going to be like. I mean, we’re all going to have to prove something to the new coach because he hasn’t seen most of us play ever. I met him a couple of years ago when he ran my agency’s camp and he talked about what he likes, the details and stuff like that. But I don’t have a really good grasp on him because I’ve never really been under his coaching yet. So it’s going to be new and exciting and you never know what will happen. I’m pumped. It’s been a long summer. Expectations will be higher. We’re not a rebuilding team anymore.”

This is regard to Troy Ward, the new coach for Vancouver. Houck had been playing for Don Hay—once rumored to be a strong candidate for Oilers head coach—and the uncertainty of the new coach is a factor in Houck’s coming season.

JACKSON HOUCK, BY DISCIPLINE IN 2013-14 (DRAFT +1)

JACKSON HOUCK 2013-14 GP G A PTS PPG
EVEN STRENGTH 69 24 17 41 .594
POWER PLAY 69 9 10 19 .275
SHORTHANDED 69 1 0 1
OVERALL 69 34 27 61 .884

That is Houck’s draft year +1 and he certainly delivered well considered the quality of team and his draft number.

EV POINTS-PER-60 ESTIMATES FROM EXTRA SKATER

  1. Greg Chase 3.0
  2. Mitchell Moroz 2.6
  3. Marco Roy 2.1
  4. Jackson Houck 2.1
  5. Jujhar Khaira 1.9

Houck is 6.01, 193 and if he can establish himself as a two-way player once turning pro there may be a job for him. In recent years, Edmonton has drafted several players who could serve what I call the Pisani role, but most of them are more physical than skilled. That usually means they’re going to fail in that role. Tyler Pitlick is the current contender for the role, with Mitch Moroz and Jujhar Khaira turning pro this fall.

Houck should be right behind them. Moroz is a bigger player, but his draft year +1 numbers are not at the same level as Houck:

MITCH MOROZ, BY DISCIPLINE IN 2012-13  (DRAFT +1)

MITCH MOROZ 2013-14 GP G A PTS PPG
EVEN STRENGTH 69 10 17 27 .391
POWER PLAY 69 3 4 7 .101
SHORTHANDED 69 0 0 0
OVERALL 69 13 21 34 .493

We’re a year away, but Jackson Houck has a chance.

 

written by

The author didn‘t add any Information to his profile yet.
Related Posts

31 Responses to "JACKSON CAGE"

  1. Ryan says:

    Damn, I am going to miss extraskater. Hockey will never be the same.

  2. John Chambers says:

    Houck’s boxcars on the surface don’t seem all that impressive but reading them two things stand out in a big way:

    1) the year over year increase in goal totals
    2) the improved +/-

    Now the big question becomes – in what role was he used? If Houck played top-line minutes against 19-year olds the we can say that this was a fine ES player who drew even against the Dub’s toughest, scoring a 5×5 goal every 3rd game. If reality is anything other than that, I can’t see why this is a potential NHL player.

    I assume Houck will play the toughs on an improved team in ’14-’15, but anything less than a 100-point campaign probably means a pro career that is short on the skill factor needed to play in the show.

  3. Pouzar says:

    If any of the geeks here wanna pitch in….here is a post-extraskater.com movement in the making.

    http://www.russianmachineneverbreaks.com/2014/08/19/extraskater-is-dead-lets-build-a-new-stat-site/

  4. Lowetide says:

    John Chambers:
    Houck’s boxcars on the surface don’t seem all that impressive but reading them two things stand out in a big way:

    1) the year over year increase in goal totals
    2) the improved +/-

    Now the big question becomes – in what role was he used? If Houck played top-line minutes against 19-year olds the we can say that this was a fine ES player who drew even against the Dub’s toughest, scoring a 5×5 goal every 3rd game. If reality is anything other than that, I can’t see why this is a potential NHL player.

    I assume Houck will play the toughs on an improved team in ’14-’15, but anything less than a 100-point campaign probably means a pro career that is short on the skill factor needed to play in the show.

    Hmmm. 5 guys scored 100 points in his league a year ago. That’s a pretty tough line in the sand.

  5. RexLibris says:

    Pouzar:
    If any of the geeks here wanna pitch in….here is a post-extraskater.com movement in the making.

    http://www.russianmachineneverbreaks.com/2014/08/19/extraskater-is-dead-lets-build-a-new-stat-site/

    If i knew anything about coding I would be all over this. Unfortunately I may have to make my contributions after a pecuniary fashion. If they let me get ahold of the code it’d shut down NORAD or some crazy thing.

  6. Ca$h-Money! says:

    Pretty sure the Pisani role is locked up for the next 5 years at 4 million per. Interestingly that’s gotta be darn close to the same relative cap hit as Pisani post extension.

  7. Ryan says:

    RexLibris: If i knew anything about coding I would be all over this. Unfortunately I may have to make my contributions after a pecuniary fashion. If they let me get ahold of the code it’d shut down NORAD or some crazy thing.

    The key question here … Does anyone have Metcalf’s Qoc toi-based formula?

  8. Ryan says:

    At least when Dellow’s site went down, it was tolerable because it was for the good of the cause.

    If metcalf had joined Dellow, it would have softened the blow…

  9. Ryan says:

    If there are other interested stakeholders here–not too mention some people smart enough to code a site like extra skater, I’d cut a check in the $5-$20k range as a partial investor.

    Now all we need to do is blog-source a lawyer to draft the ownership documents, find a few rare programmer gems who’re smart enough to also understand how to calculate hockey stats, and a few more guys willing to kick in a few bucks as part of the ownership group.

  10. classicT says:

    Ryan,

    I’ve got some relevant programming skills and (I think) a pretty decent grasp of hockey analytics. It will be interesting to see if the idea over at the russian machine gets going. If anyone else here is interested I’d love to contribute.

  11. Ryan says:

    classicT:
    Ryan,

    I’ve got some relevant programming skills and (I think) a pretty decent grasp of hockey analytics. It will be interesting to see if the idea over at the russian machine gets going. If anyone else here is interested I’d love to contribute.

    Lowetide blog is the epicenter of hockey analytics…. :)

    I could go either way. I’d definitely throw a couple of hundred bucks at a crowd sourced replacement for extraskater.

    If there was more interest in doing this in house, I’d like to be apart of that too.

    Metcalf’s Qoc seem to pass the smell test better than others I’ve seen, but I regret not paying enough attention to hiw he calculated it,

  12. slopitch says:

    Ryan, I’ve been talking with Gabe about giving behind the net a facelift. It has a lot of the back end in place plus a reputable domain. Email slopitch AT gmail f your interested.

  13. Gordies Elbow says:

    FWIW, just completed a data science class @ Johns Hopkins. Math bent, and have enough professional programming experience in C, C++, Java, and other languages to worry that my resume dates me. (Yep, even seen COBOL & ISPF/TSO, get off my lawn)

    Lately, spending my time with Hadoop & R, crunching data, but like the “new hotness” that is HTML5. Canvas tags for everyone!

    I’d be happy to hack out a new, improved version of ES. Given the changes to the NHL sites terms of use, I might need a lawyer, though. Anyone know a guy? One of the bloggers, Tyler D., is a lawyer. Anyone know how to get ahold of him?
    :-)

  14. Lowetide says:

    This thread is the reason I love the internet. Don’t understand more than six words, but rock on gents!

  15. G Money says:

    In the last thread, I proposed that we should consider crowdsourcing our own ES2 (which is probably the right way to do it in the first place!):

    G Money: So … let’s build ES2 …
    Who’s on board?

    And ClassicT was the first to volunteer:

    classicT:

    Not sure how serious you were but if you’re actually interested I’d discuss working on that idea, or some variation of it, with you.

    And now Pouzar finds some Edmonton-hating bastard has already set the wheels in motion:

    Pouzar:
    If any of the geeks here wanna pitch in….here is a post-extraskater.com movement in the making.

    http://www.russianmachineneverbreaks.com/2014/08/19/extraskater-is-dead-lets-build-a-new-stat-site/

    And then Ryan volunteered, and Gordies Elbow also has some skills (same vintage as mine – all my commercial grade software was written in C and C++, but I too lived in COBOL purgatory for a semester, and now my thing is Python).

    Best of all, RL‘s gonna foot the bill!

    LET’S DO THIS!

    My only worry is the comment section over there is already starting to fill up with “no, we should use *this* technology.” Most techies I know are pretty skeptical and non-theistic, but can easily be sucked into technology holy wars!

  16. classicT says:

    Gordies Elbow,

    If what’s said in the link to the ‘russian machine’ blog is correct it actually looks like the new NHL ToS might not be targeted at stats sites. I believe the league stated the changes weren’t put in place to counter anything currently going on.

  17. classicT says:

    G Money,

    I guess that’s sometimes a problem with crowd sourcing the whole beginnings of an idea to a massive group? A small group with lots of focus might actually be able to agree on things and get the ball rolling a little more easily.

  18. Gordies Elbow says:

    classicT,

    Yep, was just looking to get a Tyler D. dig in. Not going to stop me, or hopefully any others, from putting together sites.

    Tonight, am looking at how to best scrape their site. API? Who needs an API? (NHL.com clearly does.)

  19. Gordies Elbow says:

    Sorry, LT, for the geek speak, but the loss of ExtraSkater is a huge loss to our analytic community.

    G Money,

    Dude, so with you. I’d vastly prefer Python. Packages like “Scikit-learn” rock. I’ve extended Python with C++ in the past, giving the raw power of C++ with the weakly typed system of Python. For a specific process, a local company (owned by a company listed on NASDAQ) made a mint.

    That said, the problem is the raw data to hack on. NHL.com’s getting better, but the sources are hard to find. Even harder now that ES has gone dark.

    If I could get the data, I could munge it, but how to get it? I could hack a regex parser from sources, but where to find data like the stuff from the CHL?

    I’d be interested in calculating Chi-squares on specific algo’s. I’m not sure that the models hold up.

    Damn, wish that ES hadn’t gone dark.

  20. Lowetide says:

    Gordies Elbow:
    Sorry, LT, for the geek speak, butthe loss of ExtraSkater is a huge loss to our analytic community.

    Not at all. The geek speak gets me my Corsi’s!! Ramble on! :-)

  21. G Money says:

    Gordies Elbow: Packages like “Scikit-learn” rock

    Funny you should say that, I’ve been experimenting with using scikit-learn on both LT-style “RE” predictions as well as to calculate league equivalencies. I’m not comfy with the statistical learning approaches, so am also looking at fiddling with something more neural-networky, like PyBrain.

    Gordies Elbow: That said, the problem is the raw data to hack on. NHL.com’s getting better, but the sources are hard to find. Even harder now that ES has gone dark.
    If I could get the data, I could munge it, but how to get it? I could hack a regex parser from sources, but where to find data like the stuff from the CHL?

    Not sure. First aim I think is the NHL data. I believe the starting point is to scrape the NHL game event sheets (e.g. http://www.nhl.com/scores/htmlreports/20132014/PL021219.HTM), which would give you every event needed to calculate shot-related metrics, TOI, competition, QoT, etc. That’s really the starting point, and the bedrock, for ES and really all the analytics sites.

    I took a crack at doing it a while back using Beautiful Soup, but didn’t get to more than being able to track goals consistently. The ability to properly pull and classify events and related data exceeded my ability – actually, my patience – to deconstruct the HTML.

  22. G Money says:

    classicT,

    Yeah, absolutely. The challenge with any crowd sourcing approach is that somebody has to be the organizing element. Looks like the russianmachine dude at least has something kicked off. I’ll see what/if I can contribute to that effort.

    Might be best to see if it works out … If it doesn’t succeed (not sure what the failure rate is for crowdsourcing efforts, hopefully it’s not 98%ish like software startups!), and depending on why it doesn’t succeed, we can always take another rapid foot motion at the feline then.

  23. G Money says:

    A little bit of digging and there are two broad avenues for data scraping available.

    Wow, this guy has written a scraper in Java:
    http://redlinestation.blogspot.ca/p/blog-page.html

    He’s a Caps fan like those Russianmachine guys. Maybe a stats hotbed over there in the US capital??

    And there is apparently an ‘nhlscrapr’ package available for R.

    The program and architectural flow becomes this:

    - Database and schema to support the raw data
    - Scraper and scripts to automatically update the data as they become available on NHL.com
    - Secondary scripts to use the raw data to calculate the useful data (Corsi, TOI, QoC, shifts…)
    - The user interface to allow the user to pick and choose and then display the calculated and raw data for any given team, player, season, etc.

  24. Gordies Elbow says:

    G Money,

    If you’re looking at neural net’s, lack of data is likely a problem. I went down this path, using neural nets with a dropout algo using bernoulli gating variables, and ran into a challenge were I didn’t have enough data to be statistically significant. Hacked on it for shallow neural networks, though, found some interesting results. Not a Flames fan, but the numbers like TJ Brodie.

    For the results from the NHL and the game sheets, might want to look at extending an Apache UIMA pipeline. I’ve used it for text processing in the past, and UIMA has a number of tools that could be leveraged for what we’re looking for. I’m using UIMA elsewhere, and have been pretty impressed with how I can add items to the pipeline. It’s the underpinnings of IBM’s Watson, and if it can win a Jeopardy, you’d think it could help pick players, right?

    Back to the data – what about the other leagues, e.g. CHL/KHL data,? ES was incorporating their stats, and I can’t find sources. Any thoughts around that?

  25. Ribs says:

    All hail Al Gore! This thread is giving me the warm and fuzzies. One day we’ll look back and laugh at how we giggled at the TV screen when Corsi was first mentioned on Hockey Night in Canada. Heck, I’m doing it now! Good times ahead.

  26. Gordies Elbow says:

    G Money,

    Requirements:
    - Database and schema to support the raw data – I’d recommend semi-structured textual data, shredded to relational, for this. Start with delimited text, such as XML, and only shred data that include slowly changing dimensions. (Kimball is kicking me in the shins as I write this, DW guys will hopefully get the reference.)

    - Scraper and scripts to automatically update the data as they become available on NHL.com – probably a service job that runs nightly, not a challenge.

    - Secondary scripts to use the raw data to calculate the useful data (Corsi, TOI, QoC, shifts…) – Yep.

    - The user interface to allow the user to pick and choose and then display the calculated and raw data for any given team, player, season, etc. – Yep.

    Difference between operational reporting and analytic reporting generally is trending over time. Any thoughts to what you’d like to see?

    - CORSI, player, game by game?
    - Line, by averge linemates, TOI?
    - FENWICK for 5 player segments?

    Anyone have any thoughts around this?

  27. Bruce McCurdy says:

    I like CF%, FF%, and SF%, all three. Sometimes shots get thrown out with the bath water, but that option was something I really liked at Extra Skater.

  28. Pouzar says:

    Guys I am a bit of a techie (98% Client-Server unfortunately, just getting into the .NETweb based world) but I would keep the scope as low as humanly possible and get SOMETHING out. Nothing worse in SD than promising the moon and stars with incessant feature creep on top of that.

    A lot of smart cats on here hopefully the LT community can something going on their own.

  29. Lowetide says:

    Bruce McCurdy:
    I like CF%, FF%, and SF%, all three. Sometimes shots get thrown out with the bath water, but that option was something I really liked at Extra Skater.

    I also like Corsi Rel %, points-per-60/evens and 5×4, zone starts and despite its limitations, qual comp. some of those aren’t as advanced, but they have value for me.

  30. slopitch says:

    Gordies Elbow/G Money, I do have NHL.com/ESPN.com parsers written that I’ve been debating porting to python. I’ve been running them for 5 years and supplies data for frozenpools.ca. It handles many of the weird data cases – ie the NHL updating a stat 2 days after the game, weather delayed games ect. I’ve been working with Gabe over the years and would prefer working on that then starting over however.

  31. Ca$h-Money! says:

    I think this thread is fantastic. As someone who has neither the time nor the skill to contribute, I will say:

    1. Keep going everyone, great job; and,
    2. I will happily contribute $50 to the site should it ever get up and running.

    Also, Lowetide, according to Wikipedia you are “the kind father figure of the Oilogosphere”. I thought that was neat. That being the case, can I borrow $50?

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

© Copyright - Lowetide.ca