Puck++ Playoff Prediction Challenge – Round 1 Results

With round 1 wrapped up and the division finals now underway, it’s time to take a look at how the entrants in the first ever Puck++ Playoff Prediction Challenge performed. The table below has each entrants total score (using the Brier Score, lower is better) as well as the number of series that each entrant correctly predicted.

In total, the field averaged 4.2 series correct, with the median number of correct picks coming in at 4, right in line with the SAP model over at NHL.com (admittedly, polling everyone here would likely have been cheaper than hiring SAP). The most surprising results, per the probabilities provided (I really apologize for all the alliteration), were Minnesota’s six game win over St. Louis and the Flames upset of their Pacific rivals Vancouver. On the opposite end of the spectrum, most poolies (myself excepted) had no trouble guessing Tampa’s victory over Detroit, with the Pens defeat at the hands of the Rangers also being easily predicted by the group.

A few other stray observations:

• I’ve included a set of predictions that I’ve called the “Naive” predictions as a comparison point – these were generated by assuming each team’s winning percentage from the regular season was their true talent winning percentage. Obviously, these picks did very well in spite of the fact that they only got 4 of the 8 series correct. It’ll be interesting to see if the method holds up as the playoffs continue: for round 2, the Naive picks are NYR (73.4%), T.B (55.6%), ANA (63.4%) and MIN (51.3%).
• Tampa’s victory over Detroit had a major impact on the standings, primarily because it dropped me from a potential 5th place position down to my current spot in 18th.
• Unfortunately, no one managed to predict each of the 8 series correctly: congratulations to Arik Parnass and Sam Ventura for topping out at 7 each!
• Fortunately, no one managed to get all 8 series wrong: condolences to Micah Blake McCurdy who ended up 1 for 8 (although Micah can take some consolation in knowing that he still beat out 6 other entrants with his probabilities in general – at least he wasn’t overconfident!)
• Sam Ventura wins the prize for most detailed player: Sam provided his probabilities down to the 10th decimal place.
• Of the 35 initial entrants, 23 returned for round 2, which is a pretty good retention rate I would think (although I don’t really have anything to compare it to).

I’ll be posting the remaining 23 entries over the next few days so you can decide who to cheer for/against (assuming that, like me, you’ve got no real rooting interest left in the playoffs other than this competition). Thanks again to everyone who entered, and to those who are following along!