No Middle Ground for World Cup of Hockey
Photo: Gerry Angus/Icon Sportswire
There has been plenty of discussion on the relative success of the 2016 World Cup of Hockey. The answer usually depends on how you define success but based on the play-by-play numbers we can offer a more scientific answer.
At the start of the World Cup we looked at each roster and added up each player's average per game contribution from the 2015-16 season. This sum of the parts team score was the proverbial on paper ranking that did not yet have a coaching ingredient or factor in synergy of player combinations.
We now have the actual game results and can measure the coaching effect along with player synergies. But before we go that deep, we can also measure the average team performance in the World Cup as compared to NHL teams. This provides an indication of how the overall World Cup play compared to that of the NHL. The results are quite interesting.
Team TPI is the net sum of all 18 skater contributions during a game, which of course may be broken down into the smallest of details but is not necessary for our current purpose.
Over the course of two or more games each team establishes an average Team TPI (arithmetic mean). This Team TPI reflects the relative strength of each team (^^verified with a very strong correlation to winning).
Considering the World Cup is a much smaller sample size, here is how the World Cup teams compared to the NHL's 15/16 regular season teams.
The average Team TPI in the 15/16 NHL season was 221, and in the World Cup, 194.3. The average World Cup team performed 14% below the average NHL team. Not so far off at first glance but we would normally expect the average best-on-best tournament team to be stronger than the average team from the league providing the players.
The highest Team TPI in the 15/16 NHL regular season was 246.5. The highest Team TPI in the World Cup was 296.8, 20% higher and to no ones surprise was posted by Team Canada.
Even more interesting, the lowest Team TPI in the 2015-16 NHL season was 194.1 (almost identical to the World Cup average at 194.3). The bottom World Cup Team TPI was 128.3. The lowest performing NHL team was 51% better than the lowest performing World Cup team.
To put in context, there were four teams in the World Cup that performed below the lowest performing team in the NHL during the 15/16 season and all but two World Cup teams were below the average 15/16 NHL team.
On the plus side, two World Cup teams had a higher Team TPI than the highest NHL team during the regular season, which is what we would expect from all-star teams.
Sound like two different World Cups.... you bet it was. If you consider that high Team TPIs are the result of higher individual player contributions, including more good and great passes, shots, carrys, blocked shots, etc. you start to appreciate the roller coaster World Cup viewing experience for the fans. The highs were really high but the drops were deep and lacking the thrills of the usual roller coaster.
Let us take a quick look at Offense.
The average Team Off TPI in the 15/16 NHL regular season was 180.9, compared to 157.8 in the World Cup, a 15% advantage for the NHL. Again we would expect the opposite.
The top offensive team in the NHL in 15/16 was 205.2 vs 256 in the World Cup, a 25% advantage for the top offensive World Cup team, Team North America. This provided some of the best offensive hockey we have seen in years.
The bottom NHL offensive team in 15/16 was at 163, and in the World Cup 98.8, which is a 65% advantage for the bottom NHL offensive team over the bottom World Cup offensive team.
In summary, two World Cup teams had stronger offense than the leading offensive NHL team in 15/16. On the flip side, six World Cup teams were not only below the average NHL offensive team, but were below the lowest NHL offensive team in the 15/16 regular season. It seems clear two teams provided most of the offensive entertainment for the tournament.
Two World Cup stories were told..... was the glass half empty or half full, or was the Cup a quarter very full, and three quarters very empty?
^^TP hired an independent company to run a correlation analysis with data from the 15/16 NHL regular season. The analysis correlated the 30 NHL Team TPI's with their respective ROW (regulation and overtime wins). The result was a .82 correlation without goalie performance data, and a .91 correlation with goalie performance.
Learn more about OUR METHOD