From the PowerScout Research Desk
Measuring the Impact of NHL Fights on Momentum and Scoring
Posted by: Terry_Appleby Sun, 2012-01-08 22:40
Terry Appleby, M.Sc., founder of PowerScout Hockey
Why use only shots on goal to calculate momentum? Because the rate of shots per 5 minute time period offers the best fit between what PowerScout shows graphically as momentum and what we actually observe to be the true momentum swings during the game. In our research we explored other statistics that would suggest a different fit but none gave us as strong a representation to what we consistently saw was happening on the ice.
Momentum was born out of PowerScout's individual player effort curves that use three player statistics – shots, hits, and blocked shots, and there is common agreement that to gain these statistics requires 'effort' on the part of a player. Measuring NHL players this way captures their effort both offensively and defensively. But what works for individual players does not portray momentum as well at the team level, and ultimately shots on goal by itself was determined to be the simplest and most accurate proxy for team momentum.
As an example, see the Momentum Meter (right) from the Buffalo-Philadelphia game on December 7, 2011 where a fight (black dot) between Ellis and Bourdon at the beginning of the second period sparked an immediate increase in momentum for Philadelphia and a subsequent goal by Matt Read.
Learn more about Momentum Meter and its companion Dominance Curve.
We analyzed 1,563 fights between October 2009 and December 21st, 2011. We measured the difference between each team’s momentum value at the time the fight took place and its value three minutes later. We then separated these groups into Home and Visitor and whether each team experienced an increase, decrease or no change in its momentum after a fight.
Table 1 shows a matrix of the different momentum shifts by Home and Visiting teams, creating nine momentum situations, where the number of fights in that scenario and the relative percentage of the total (1,563 fights) are displayed.
After analyzing 1,563 fights since October 2009, we found that 76% of the time (1,188 fights) one or both teams increased their momentum, and 24% of the time (375 fights) both teams either stayed the same or decreased their momentum.
Of the 1,563 fights total, 23% involved an increase in momentum by both teams, but in only 4% of the cases did both teams have a decrease in momentum.
Overall both teams tended to mirror the results and led us to conclude that at home or on the road fighting does not give any one team an advantage because of the location of the game.
Question #2: Does fighting increase goal scoring?
A second analysis performed on the data measured the impact upon goal scoring depending on the momentum swing by both teams after the fight.
If fighting changes the momentum of a game then we should find an increase in goal scoring when teams play with increased effort. That’s what our research has found, though fighting does not lead to significant scoring (powerplay goals were not netted out of the analysis).
The complete results by team and momentum situation are shown here in table 2.
When both teams increase their momentum after a fight, goal scoring increases by about 0.2 goals on average for both teams (visiting=0.19, home=0.21) with a small bias to the home team depending on how the visiting team responded to the fight.
If the visiting team shows a decline in momentum then home team scoring increases regardless of how they themselves responded. This might be attributable to home ice effects.
Overall the best scenario for fans is one where both teams respond because a more wide open game ensues with goal scoring increasing (albeit marginally) for both teams.
Question #3: What average impact do fights have on momentum?
The full effects of fighting can only be measured if we determine how many fights we need on a team to generate one win considering it takes 6 goals to achieve 1 win (this is accepted among the hockey analytics community).
If a team’s fights occur without any regard to team or opponent momentum, then the team will score at an average gain of about a 0.1 of a goal. If so, this would require about 60 fights to score a win.
But good coaches don’t do things randomly - preferring to pick the opportune time to change team momentum. In such cases scoring will double to about 0.2 of a goal and the team will require only about 30 fights to get a win.
Table 3 shows the average impact to each team's momentum for each scenario after a fight.
SUMMARY OF FINDINGS:
1. If a fight occurs there is a 50/50 chance your team will have an increase in momentum. If so you can expect to gain about 0.1 of a goal in scoring if this happens randomly. Hence it will take just over 60 fights to get a win.
2. If a fight happens when the team is playing poorly, team scoring can be boosted to about 0.2 of a goal because team momentum increases. In this case it will take about 30 fights to achieve a win.
3. There is no home ice advantage to fighting.
CONCLUSION: Overall these results suggest that fighting by itself does not significantly help a team score more goals or win more games, but it can often increase short-term momentum (i.e. the RATE at which they are getting shots) for one or both teams. Statistically speaking, if fights happen randomly it will take about 60 fights to equal one win, but if their timing is managed by the coach it could take as few as 30 fights to equal one win. PowerScout has uncovered many other factors that can provide a much greater contribution to winning than fighting, such as having a good penalty killing unit.
For interest sake, further study into the impacts of fighting on team momentum could add more context to these results. Other factors that could influence the fight/momentum analysis include 3 minutes after the fight versus other time ranges, who won the fight, the score, period, shots by line, powerplay opportunity, time in period, time of year, and quality of opposition. These will be explored with future research depending on availability of the data required.
PowerScout's focus remains on WINNING: Our mission is to analyze and research hockey's offensive, defensive, and situational elements to uncover the factors and strategies that have proven over time to consistently help NHL teams win. Fighting and momentum are interesting analysis, but the most relevant research done by PowerScout in respect to winning relates to 12 key player skills uncovered after analysing 14,000 games and 12,000 players over a 13 year period. We created the Point Shares player evaluation system (AKA, ‘The Moneyball of NHL’) as a result of these important and highly relevant findings into winning. Take a look at Point Shares to learn about how the system works and to see what players so far this year are contributing the most to their team's wins.
A note of thanks from PowerScout...