The truth about 'track meet' teams (Puck Momalytics)
(Jen Lute Costella is our new analytics writer, breaking down the fanciest of stats for you each week. She's a mom. She's writing for Puck Daddy. Hence, she calls this slice of stats heaven Puck Momalytics.)
Not all puck possession teams are created equally. Some teams control more of the play during a game but still give up a lot more shots than other teams. Likewise, some teams that may be very good at suppressing their opponent’s shots, but do not have an aggressive offense in terms of shot generation.
This is where Event Rates and rate based statistics are very helpful.
To illustrate this, think of a time you have heard an analyst or a broadcaster say “you don’t want to get into a track meet with this team” while you are watching a game. This is a favorite phrase of many broadcasters when discussing various teams throughout the league known to have a high-powered offense. A “track meet” type of game is technically one with lots of end to end action. Team X has an odd man rush, Team Y retrieves the puck off the rebound and speeds back to the other end of the ice and back and forth they go.
The odd thing about this? The commentators often apply this terminology to teams that are considered to be very good puck possession teams. Coaches, during their bench interviews or after the game, always say they do not want their team engaging in a track meet type of game. There is a very good reason for that. The best teams in the league do not play a track meet style of game. It seems that the people using “track meet” to describe teams are really just confusing a team that uses speed in transition with teams that allow lots of end to end action.
On the opposite side of this are teams that, when they are coming up on the schedule to play against your team, elicit an audible groan because you just know the game is going to be really boring. Both teams will be stuck in the neutral zone or trying to skate the puck out of their defensive zone on a breakout only to be stopped and go back and try again. There are very few breakaways or odd man rushes and overall, just not a lot action.
These examples are fairly easily translated by Event Rate statistics. These metrics also show us that the best teams in the league have far less “action” in their games than we may think. Metrics such as CF (Corsi For) and CA (Corsi Against) can be converted into rate stats to show the frequency of shot attempts for and against a team as the season goes on. The same can be done with Fenwick measures and Shots on Goal.
Below is a graph of the Corsi Event Rates for each team with last season’s numbers during 5 on 5 play with the score close. Teams on the left had a lower event rate. Teams on the right had a higher event rate. The vertical axis is the percentage of the play controlled by the team or CF%. Teams higher than 50% controlled more of the play than their opponents at 5 on 5 Score Close.
Teams that are successful usually don’t allow very many shots against and control the play more than their opponents. Rate stats help us visualize this. Last season, some of the best teams in the league, the Los Angeles Kings, the Chicago Blackhawks, the Boston Bruins and the St. Louis Blues were in the middle of the event rate spectrum and controlled a high percentage of the play. The San Jose Sharks and New York Rangers were on the higher end of the event rate, but also had good puck possession. Good possession teams that struggle with shot suppression, i.e. that give up a lot of shots, tend to face significant hurdles in the playoffs and that is consistent with what happened last season.
Teams such as the Ottawa Senators, Carolina Hurricanes, Dallas Stars and Montreal Canadiens were far more likely to engage in “track meet” types of games last season than others. Teams that had the reputation for being a bit boring to watch last season, namely the New Jersey Devils, Minnesota Wild and Nashville Predators, earned that label. They were among the lowest event rate teams in the league. The Detroit Red Wings snuck into the low end of the spectrum as well, but their games were a bit more exciting to watch because some of their forwards have amazing speed and creativity through the neutral zone.
This season some teams have shown some drastic changes in their game. Granted, we are early in the season, but the event rates still bear attention.
Chicago and Los Angeles have moved to the high event rate end of the spectrum early in the season. While Chicago is still maintaining good possession numbers, L.A. has struggled a bit out of the gate. Many of the usual suspects appear at the high end of the spectrum, namely San Jose, Dallas, Montreal and Ottawa. Minnesota has taken the spot held by L.A. last season as a lower event rate but high possession rate team.
When we look at the event rates from a Fenwick For and Against perspective, we can visualize exactly how these rates work out.
In order to keep the graph manageable and show how the teams line up against each other, I have calculated the current* league average for FF60 and FA60 and plotted the deviation from the average for each team. Teams below the horizontal axis give up shots at a lower rate than the current league average (FA60). Teams above the horizontal axis give up more. Teams to the right of the vertical axis attempt more shots than the current league average (FF60). Teams to the left of the vertical axis attempt shots less frequently.
The optimal place to be is in the lower right quadrant, i.e. having a higher shot rate and giving up shots at a lower rate. At this point in the season, Minnesota is far and away the leader in this regard. This is a massive change for the Wild. While they have been a good shot suppression team for the last few seasons, their offense has been stuck in the low shot rate category. A few other teams that have made big changes include Toronto and Edmonton. Both of these teams were high in the upper left quadrant last season, but this season have made big strides forward.
To get an idea of just how much change we can measure from last season to this, I calculated the deviation of each team’s current Fenwick rate stats from their rates last season. This is all during 5 on 5 play with the score close to cut out any noise from Score Effects.
Teams above the horizontal axis are currently giving up shots at a higher rate than they did last season (FA60). Teams below it are giving up shots at a lower rate. Teams to the right of the vertical axis are attempting shots at a higher rate than last season, while teams to the left are shooting more infrequently (FF60).
The circles for each team represent the total change in their rates. For example, the Minnesota Wild have increased their FF60 mark this season by 12.39 from last season and have decreased their FA60 mark by 7.52 giving us a total change of 19.91. This is a rather remarkable change for the Wild and is most certainly taking them in the right direction. It remains to be seen if they can sustain this type of change, but if they do, I would not be at all surprised to see them roll into the playoffs.
Teams making notable improvements in terms of increasing their shot attempt rates and their shot suppression marks include the Toronto Maple Leafs, Edmonton Oilers, New York Islanders, Washington Capitals and Pittsburgh Penguins. The Tampa Bay Lightning and some others have done a better job of suppressing shots this season which is a goal all teams share.
There are some teams heading in the wrong direction in the early going at least this season. The teams appearing in the upper left quadrant of the graph above currently have a lower shot attempt rate and a higher shots against rate than they did last season. The most notable team in that area is the defending Stanley Cup Champion L.A. Kings.
Except for Minnesota and Toronto, both of whom are showing large positive changes, the Kings have the highest total rate change but in the wrong direction. The Kings have been without several pieces of their defense in the early going and have had some key injuries to forwards who make a big difference to their team. L.A. is a talented team with an extremely smart coach, so it doesn’t seem likely that they will continue in the wrong direction for long.
Event rate stats can be very helpful in illuminating for us some of the things that teams are doing right and wrong or just plain differently as the season goes on. If you have noticed a change in how your team has been playing, comparing the rate stats can certainly shed some light on the things you are seeing. As the season goes on it should be interesting to see if some of the changes we have noted here continue on. If they do, they could result in some surprises at the end of the season.
*All data used herein is current through November 10, 2014 and was collected from www.war-on-ice.com
MORE NHL COVERAGE ON YAHOO SPORTS
