Tag Archives: Hackathon

Oilers Hackathon: Season Wrap

With the Oilers season coming to an end, I thought I would take a look back at my Hackathon predictions. For questions 1, 2 and 3, I will wait for the Oilers to post the results on their website. They indicated they were going to do so, but have not as yet.

Related, someone from the Oilers contacted me in March about being interviewed for this piece on the contest. I never did get that follow up call which was too bad. One, I would have liked to brag to friends and get ripped on Twitter by strangers. Two, the only individual they interviewed was an academic who already has a proprietary player rating system. It would have been interesting to read about some of the amateurs who entered, be it me or someone else.

The fourth question was really a choose your own adventure. I opted to focus on special teams and made four predictions.

1. For the 2012-2013 season, 12 of the top 16 teams in the wPP%d metric will make the playoffs.

2. The Detroit Red Wings will finish out of the top 16 in wPP%d and will miss the playoffs.

3. The New York Islanders will finish in the top 16 in wPP%d and will make the playoffs.

4. The Edmonton Oilers will make the playoffs if they are around league average in PP shooting percentage (no more than 1% higher) and in the top 16 in wPP%d.

I think I wrote before that I wish I had simply used PP%d instead of weighting it with power play opportunities. I took a pretty simply concept and made it more complex for very little benefit.

Anyhoo, below is the data for the 2013 season. The ‘+’ in the yellow column denotes top 16 in wPP%d and made playoffs. ‘&&’ = made playoffs but not top 16 wPP%d. ‘-‘ = top 16 wPP%d but missed playoffs. ‘#’ = missed playoffs and not top 16 wPP%d.

2013 season

My predictions:

  1. 13 of the top 16 made the playoffs. Boom, it is 1-0.
  1. Detroit finished 13th in wPP%d and made the playoffs with a win on the season’s last day. A swing and a miss on this one.
  1. The Islanders finished 12th and made the playoffs. I am a genius.
  1. The Oilers were again a good special teams club, ranking 10th in wPP%d. I have not quite figured out the math, but per behindthenet.ca, the Oilers had the second worst PP shooing percentage at 5 on 4. That most certainly has them below league average. Swing and a miss, strike two.

So I ended up two for four with Detroit within smelling distance of being correct. The Oilers prediction was way off, in large part due to the team’s really poor results at 5 on 5.

That last sentence is really important to what I was hoping to find out with my entry. The idea was that we could totally ignore a team’s 5 on 5 play and infer team success solely from their special team’s play. Good special teams result in good teams.

Historically 12 of the top 16 special team performers make the playoffs and 13 of 16 did this past season. Each season there are examples of teams that make the playoffs despite poor special teams and teams that miss despite good special teams.

The Oilers clearly fall into that latter category. Generally, I view the 12 of 16 as the effect, not the cause. Good teams have good special teams. Good special teams does not necessarily mean good team.

I also wondered if special team performance was predictive to future years. Would a team with good special teams move into the playoffs? Would a team whose special teams decline be at risk for missing the playoffs the following year? I did not think this would be the case and nothing from the past season changes my mind. I went one for two in my predictions, but my post also identified San Jose and Colorado as teams whose fortunes might change. San Jose had the best special teams in the league while Colorado were near the basement. While they were not part of my submission, you could argue I went one for four. Special team performance is not predictive in terms of signaling a change in the standings the following year.

That about wraps up my thoughts on the Oilers Hackathon contest. If for some incredible reason I am one of the prize winners, perhaps a final post to try to understand why.

Advertisements

Leave a comment

Filed under Edmonton Oilers

Checking in on my Oilers Hackathon Predictions

As we hit the quarter pole of the abbreviate season, I thought I would take a look at my Oilers Hackathon question 4 predictions. We are only talking about 15% of a normal 82 game regular season, so I won’t get too excited or too despondent at this stage.

The predictions:

1. For the 2012-2013 season, 12 of the top 16 teams in the wPP%d metric will make the playoffs.

2. The Detroit Red Wings will finish out of the top 16 in wPP%d and will miss the playoffs.

3. The New York Islanders will finish in the top 16 in wPP%d and will make the playoffs.

4. The Edmonton Oilers will make the playoffs if they are around league average in PP shooting percentage (no more than 1% higher) and in the top 16 in wPP%d.

I will link you to a couple of posts that has all the work that went into my predictions. PP%d is the difference between the power play success percentage and 1 minus the penalty kill success rate. So if your PP runs at 20% and your PK is at 75%, your PP%d is 5%.

The PPd is the difference between total power plays and total penalty kill opportunities. If the Oilers took 150 penalties and had 100 power plays, the PPd would be -50.

I then convert those two into league rankings and blend them together, 25% of the PPd ranking and 75% of the PP%d ranking. That is the wPP%d figure which is then compared to the league average.

Jeez, as I write that I think I should have just stayed with PP%d. Easy to compute and almost as predictive as wPP%d.

Below is a bit of a nasty table with the special team results through Tuesday night. For point totals, I prorated to a 48 game schedule to accommodate for differing number of games played.

quarter pole

For prediction #1, currently 9 of the top 16 are in the playoffs, with four more teams just outside of the playoffs. Anaheim, Carolina and Detroit are three teams well in playoff position that have had middling special team results thus far. Columbus and the Islanders are two teams with good special team play but a familiar location in the standings. I don’t think it is necessary fair to assume that the position is the standings will regress based on the special team rankings. The Islanders aren’t going to score on 25% of the power plays, nor are they going to kill 90% of penalties over the season.

For predictions #2 and #3, I am correct for both the Islanders and Red Wings in terms of wPP%d. The Islanders lead the league, while the Red Wings are 24th. That has not yet translated to position in the standings, with the Red Wings up near the top and the Islanders right near the bottom.

As for the Oilers, their PP shooting percentage is 18.9% on the season, well above a typical league average of 12-13%. After a pretty good start, they are on their way down the standings. I don’t think they look like a playoff team most nights.

All in all, it isn’t looking so great for my submission, but it is very early.

I believe the Oilers will publish how entrants are doing on questions 1 through 3, so I will maybe comment on that next week.

Leave a comment

Filed under Edmonton Oilers

Oilers Hackathon: That’s a Wrap

I just submitted my submission to the Oilers Hackathon question. Rather than wait and use the early season information, I opted to simply be done with it. I broke down and made one edit from my previously published figures on goal differential. I originally had the Flyers with the highest goal differential and the Oilers with the lowest. I took 0.55 goals per game off the Philly estimate and gave it to the Oilers. What can I say, that Yakupov goal last night was fantastic!

Below is my actual submission for Question 4. The one page restriction made it impossible to demonstrate all the work that went into the predictions. Oh well.

I guess the Oilers will have some sort of leader board after the contest closes. Rather than use my usual jonathanjoyce37 username, I went with islandfever37. I’m sitting in the Cayman Islands currently and it is beautifully hot, hence the name. Continue reading

Leave a comment

Filed under Edmonton Oilers

Oilers Hackathon: Question 4: The Value of Special Teams Part Deux

In my last post, I looked at a few special team metrics over the past fifteen years. The wPP%d* metric indicated that a little over 12 of the top 16 each year would make the playoffs. For this post, I looked for trends that might suggest a team is about to jump into or fall out of the playoffs.

* wPP%d = weighted power play % differential = 25% PPd x 75% PP%d = 25% x(power plays – penalty kills) + 75% x(PP% – (1-PK%))

My attempt to gain insight from each team’s numbers is fraught with peril. I just finished reading Nate Silver’s “The Signal and the Noise” and I feel like I have super human signal detecting skills at the moment. The same thing happened when I watched “Spies Like Us” when I was a kid. I spent the next two days hiding out in my grandparents’ crawl space with my cousin waiting to uncover many of our family’s secrets. Overconfidence in my ability is right there, ready to go when something is so fresh in my mind.

With that self-awareness in mind, I looked at each team over the preceding fifteen years looking for one of two things: 1) teams shifting into the playoffs that had top 16 ranked power play metrics in seasons prior; and 2) teams shifting out of the playoffs after having their power play metrics fall out of the top 16. Was there some sort of substantive pattern to it all? Continue reading

Leave a comment

Filed under Edmonton Oilers

Oilers Hackathon: Question 4: The Value of Special Teams

The finish line to the Oilers Hackathon contest is within sight. This is the first of three posts on the fourth and final question.

Conduct a predictive analysis of your choice on some dimension of potential value to the Oilers. The analysis must be testable in the upcoming season and judged on its difficulty, accuracy, clarity, and value.

The fourth question will be submitted online, but will take the form of a PDF executive summary. The summary will be limited to one page of 10 pt font text that outlines the problem solved, data used, methodology, and predicted results. The methodology and predicted results must be sufficiently clear that the judges can replicate and/or verify the results. In addition to the one page executive summary, the Entrant may include up to four pages of annotated visuals or charts further describing their method and results. The source of any outside data must be clearly identified and publicly available.

The fourth question will be judged on four dimensions:

  • difficulty/novelty – how difficult and new is the analysis attempted by the Entrant

  • accuracy – how accurate were the predictions made by the Entrant

  • clarity – how clear is the reasoning, methodology, and results/implications. This includes both the written portion and any visuals in the submission

  • value – how valuable is the findings in helping the Oilers achieve their goal of becoming a perennial contender.

This post walks through the initial analysis I did to support my question #4 response to the Oilers. Today I look at the league on a year by year basis. Next up will be a post looking team by team. The third post will be my actual response to question #4. I should have that all done and on the site by the middle of next week. Continue reading

1 Comment

Filed under Edmonton Oilers

Oilers Hackathon: Question 3

I am now three quarters of the way through the Oilers Hackathon contest questions. I received an email last week letting me know that I am competing against almost 500 other entrants. The deadline of February 15th is subject to change pending the resolution of the CBA negotiations. Translation: we might cancel the contest!! That would be a nice piece of PR for the Oil.

One interesting piece to the email was an indication that the 10 winners would be required to sign the confidentiality agreement before receiving their prize. The implication is that we aren’t subject to the CA at this point in time. I am going to revise my post on question #1 and discuss the nature of the players in Appendix A and what I think the Oilers are doing. I will hold off on actually listing the players for now.

For questions #3, I will outlay exactly what I did and include my predictions for each team. Below is question 3.

Predict the goal differential per regular season game ((goals for less goals against) divided by games played) for all thirty teams for the upcoming season. Continue reading

2 Comments

Filed under Edmonton Oilers

Oilers Hackathon: Question 2

Yesterday, I spent a couple of hours answering question #2 of the Oilers Hackathon.

Predict next season’s even strength save percentage of the goaltenders listed in appendix A

My approach to this question was to keep my analysis fairly short and simple. I give the illusion of precision, but the reality is a bunch of guesswork. A goalie’s save percentage has a lot to do with luck and his teammates, and a bit to do with his skill and his age. Continue reading

1 Comment

Filed under Edmonton Oilers