Hockey Analysis, War on Ice and the future of stats sites

The big news in the online Hockey Analytics community yesterday, aside from the Ottawa Hockey Analytics Conference, was the Andrew Thomas announcement that war-on-ice.com will be shutting down in the upcoming months. Although in many respects War on Ice was a competitor to my site (stats.hockeyanalysis.com and puckalytics.com) it was also a great site that was developed for different reasons and served a different purpose.

My sites were developed by myself over time largely as a hobbyist and a blogger. I started hockeyanalysis.com in 2005 after the ‘lost season’ lockout as a way to further enjoy my love of hockey and to participate in the newly growing online community of hockey fans like myself. Although I am not a statistician by trade (my background is in computer science) I have a fascination with digging into the numbers and so my intent was always to have a statistical/analytical viewpoint to my blog. It wasn’t long after starting my blog that I started writing code to pull data from the NHL.com game logs, play by play sheets and shift tables. Much of the code I wrote 8+ years ago still forms the core of my, now much larger, code base (that honestly probably needs a rewrite but it is stable and optimized so I dread the thought).

From the outset it was never my intent to build a stats website. Everything I was doing I was doing for my own interest. I wrote code to explore things that interested me. Over time though I realized that a lot of the data that I was creating would be of interest to others so over time I slowly started making that available and it developed into what we have today.

So, for the most part my stats sites were developed for selfish reasons. Sure, I was happy to share my work with others and am happy that so many people have found it useful over the years but what you see was largely created for selfish reasons. The stats you see on my sites are there because I, currently or at some point, had an interest in looking at them for one reason or another.

Although not that familiar with the details I can be fairly certain War on Ice developed somewhat differently. Not only was War on Ice developed from a statisticians and an academics point of view I am certain it developed with a much grander vision for their site than I have ever had for mine. To a great extent sharing my data was always a secondary project to mine while I get the sense than Andrew and Sam always had the goal of sharing anything and everything, including the kitchen sink if they could find a way to do it. While this is a big reason for its resounding success of the site and the impact on the community it may have also led to the eventual demise that was announced yesterday.

Although the open source, sharing model is often incredibly successful it isn’t unusual for projects to go by the way side when the original leaders move on. I have not spent much time looking at their code but based on the feature set they had offered on War on Ice I can only imagine how massively difficult it would be for an outsider to come in and pick up right where they left off in relatively short order. On top of that, rarely do the people who take over a project have the same level of enthusiasm, motivation, dedication, and emotional attachment to the project as the original owners.

So, while War on Ice will not continue to live on at some point, as with the ending of Extra Skater and CapGeek the hockey community will move on. There are other great sites out there and I am certain there will be new ones. Here are a few that you can check out:

There may be others that I have missed so let me know and I’ll add them to the list. If you like a site and want it to survive support the site in the following way:

  • If you use their sites in an article or tweet, link to it. The easiest way to support a site is to tell people about it. It’s easy and its free. It is also a great motivational tool to the site creator to see their work used, appreciated and acknowledged.
  • If the site has advertisements, support its advertisers. That will help offset web hosting costs and time and effort to build out the site.
  • If the site has donations, consider a donation. Every bit helps and it is another way to show appreciation for the effort involved in putting together the site.
  • This may be the most important: Provide feedback, good or bad (just be nice about it). If you notice a problem, let the site owner know. If you have a recommendation or request, let the site owner know. If you love a new feature, let the site owner know. The only way sites will get better or incorporate the features you want is with support and feedback.

The future of Stats.HockeyAnalysis.com, Puckalytics.com

I want to briefly discuss the status and future of my two stats sites. Ongoing development (new features) of stats.hockeyanalysis.com has pretty much stopped and I don’t expect it to return. I still use the site and it is certainly more usable (quicker) on a smart phone or tablet so I don’t see it going away any time soon but I doubt there will be any new features added to it. New development will mostly take place on Puckalytics.com.

I am in the process of evaluating what I want to do with Puckalytics.com. There are a few routes I could go including just operating it largely as is. Another option would be to put the effort to grow it immensely and adding a ton of new features that you all have been requesting over the years or features that will soon be absent with the loss of War on Ice. These could include:

  • More stats in similar format to what I already have (i.e. score adjusted stats)
  • New stats like War on Ice had including stats like scoring chances
  • Adding game by game data.
  • Where possible add date ranges to stats
  • Improving WOWY data including possibly looking at forward Trio’s as opposed to just pairs (a technical challenge but I think I can overcome it with creativity).
  • Adding visualizations (definite must).
  • Adding player biographical data and improving the player pages.
  • Adding salary data and capgeek style features (could possibly partner with other salary cap sites sites or leverage some of the work War on Ice did).
  • A gazillion other things that I aren’t coming to mind right now.

I don’t know yet where it will go but I can tell you that there will be a lot of effort required to put all of this together. With that said, I have a few questions for the community which you are welcome to answer in the comments or e-mail me privately at david (at) hockeyanalysis.com:

  1. If I started a kickstarter/gofundme campaign, how much would you be willing to contribute?
  2. Would you contribute more if I provided a guarantee that I would operate the site for at least 1 year? What about 3 years? 5 years?
  3. Would you pay for a subscription for advanced features such as stats by a date range?
  4. I will probably always keep the back end code (the scraping software, database structure, etc.) proprietary but if I opened up the website development would you be willing to contribute to its development? By that I mean contribute to the web site look and feel, user interface, visualizations of the data, etc.

I’d love to be able to add all of the features you want and I’d also love to guarantee you all that my sites will be around a year from now or even five years from now. The reality of all of this is that operating a hockey stats sites take time and effort and doesn’t pay the bills. This is why they disappear. People move on to bigger and better things that do pay the bills. Until someone finds a business model to make a a free, publicly available stats site profitable enough to live off of the reality is they will not last. I’d love to be able to make a 3 or 5 year promise to you all but I need to find that business model first and your feedback to the questions above would help greatly.

Thanks for reading and thanks for your feedback.

Update: I have started a GoFundMe project to get a Puckalytics.com enhancement project kicked off.

This article has 6 Comments

  1. As a former HS Hockey Coach
    I appreciate your time & effort with a
    ” new way ” of evaluating a Hockey Player’s
    performance on the ice.

    Visualization { eyes } is important with “in game adjustments”. Your expertise is great to give another perspective on game results

  2. Hi David,

    Were you to start a Kickstarter, I would gladly contribute. But it also does highly depend on the scope of your planned work.

    In any case, I actually would prefer an approach dissimilar from war-on-ice. Though I appreciate the breadth of their work, I found it incredibly dense with information and as a result poor with performance. Both of which made it hard for fans to consume. What I liked about ExtraSkater was it’s incredibly simple and clean user interface.

    Basically I’m ok with sacrificing features for user friendliness!

  3. First off, I love stats.hockeyanalysis.com, I use it all the time for research, along with war-on-ice.com. If you were willing to recreate some of the features of war-on-ice on puckalytics, I would gladly chip in 50 or 60$ to a kickstarter campaign.

    Those features would be: Shot location charts, adjusted save percentage, scoring chance data, and date-ranging for all of the above. Essentially, all that’s missing from stats.hockeyanalysis.com is insight into shot quality, and date ranges.

    A longer term guaranty that the site would stay running for three years would incentivize me to donate a bit more, and I’m not opposed to a subscription system for a 3-4$/month.

  4. Honestly, I have no idea how I got here, I was surfing through twitter being a button masher and found this article about WOI going the way of CapGeek, which is still mourned by my family to this day *pours out some diet coke for CapGeek*

    I am newish to advanced stats and WOI would sometimes leave me frustrated, I didn’t feel it was all that user friendly for the layperson. That would be one suggestion I have.

    Also, I think successful kickstarts should have some levels of “goods” that can entice a better contribution. for example (total spitball ideas here)

    $5 for a mention on the “contributors” page
    $12 for a 1 year subscription (if you go that route)
    $37 for a tweet about your site, blog, twitter handle, etc
    $66 for a banner to your site (appropriate to hockey audience)
    $80 Everything listed

    or something along those lines.

    I’ve bookmarked those sites you mentioned and agree 100% with your article. Thank you and look forward to seeing more.

  5. I’d probably give you $40 in a kickstarter, but would pay about $3/ mo for a subscription fee also.

  6. One request I would have is for a deeper look into data that adjusts for shot quality would be nice, since said data is always sparse and accounts for one thing but not another.

    I’ve seen some sites use game events, like rushes, rebounds, shot type (eg. wrist) etc. to estimate shot quality and I’ve seen some that use shot location to estimate shot quality.

    I don’t know of a site that uses both.

Comments are closed.