Hockey Analytics: 4v4 Play Performance

Gus Katsaros
·6 min read

It’s very common during broadcasts to hear comments about ‘open ice’ when both teams are down a player, skating at 4v4, and how conducive that is to potential goals. At the least, better scoring chance generation with all that space to work with in every zone.

Despite the extra space and the ability to put out a pair of forwards with scoring touch and complimentary blueliners, there aren’t many goals scored at 4v4.

WED

4/28

6:00 PM

9:30 PM

NHL

Blues at Wild

WED

4/28

9:30 PM

12:30 AM

NHL

Avalanche at Golden Knights

The NHL adopted the 3v3 overtime format in 2015-16. Using data from EvolvingHockey.com who have the category of 4v4 on their site, we can see the impact since 2015-16. We can begin that by determining the time on ice splits between forwards and defensemen. I think with the relative skill set entering the NHL, it wouldn’t surprise me to see teams experiment with three forwards and a defenseman at some point – as the concept of total hockey prepares developing kids into very versatile and well-rounded hockey players.

Defensemen have traditionally set a ice time and at 4v4 they led in time on ice by a 60-40 split, with forward pairs changing more often than the defense pairs. Timing is always a factor here, too since 4v4 play isn’t a sole feature of offsetting penalties. A penalty call while a team is on the power play will force 4v4 play, but for a short period than a full two minutes, so the weighting in ice time is skewed by the randomness of penalty calls and defensive pairings – an unwanted element that has emerged out of the darkness as a method of game management of late where officials try to balance out calls.

As the ice time has dropped over the seasons, the split remains.

4v4 TOI by Position NHL
4v4 TOI by Position NHL

In 2020-21, the New Jersey Devils lead the league in 4v4 goals with four (4). The Senators, Leafs and Lightning have three, and 11 teams with two goals. The only teams without a 4v4 marker are the Coyotes, Sabres, Blue Jackets, Flames, Predators, Sharks and Canucks. With the amount of ice acting a determinant in goal scoring rates, the disparity can be substantial. Florida has played 46.7 minutes at 4v4 while Vancouver only 19 minutes. When switching to rates, it’s clear to see some of the average time on ice have produced the best results at goals for per 60 minutes.

Generally, time on ice at 4v4 has been declining since ’15-16 pulling down absolute shots on goal and goals in its wake, with the more dramatic a downturn in the latter category. In ’19-20, 72 goals were scored in a shortened season due to COVID but the rates of the last two seasons have indicated a decline in general scoring. Time on ice is a distinct factor in the decline.

4v4 Goals Shots For and Time on Ice - Teams NHL
4v4 Goals Shots For and Time on Ice - Teams NHL

To get a better feel for scoring impacts, we can focus on using goals for per 60 minutes. Plotted along the black line, there’s an increase to a peak at 2018-19, right where expected goals was beginning to decline – the blue curve on the downslope. Even with a depressed goal scoring rate, teams seem to be generating more scoring chances according to xG than in ’19-20, closer to the ’16-17 rate before the massive spike. Goals may not be up in ’20-21, but expected goals are significantly higher.

4v4 Expected Goals & Goals Per 60 By Season
4v4 Expected Goals & Goals Per 60 By Season

Sticking with expected goals, we get a clearer picture of the peak in ’18-19, by splitting into positions. Defensemen have been impacting scoring chance generation at even pace, but forward contributions to expected goals peaked with 1.02 per 60 minutes, and 1.7 times the standard deviation from the average over the seasons (0.82). Forwards carrying the play from ’17-18 and through to the next season have facilitated the rise to peak. With the uptick in expected goals from the above chart, forwards are once again gaming the expected goals rate and causing an overall increase, despite the reduced ice time, and almost a 50-50 split in shooting percentages between forwards and defensemen.

Editor’s Note: Drafting is only half the battle! Get an edge on your competition with our Season Tools that are packed with rankings, projections, a trade evaluator, start/sit tools and much more. And don't forget to use promo code PUCK10 to get 10% off. Click here to learn more!

4v4 Expected Goals Per 60 between Forwards & Defensemen
4v4 Expected Goals Per 60 between Forwards & Defensemen

Summarizing forwards, rather than individual positions was an aim to cut down noise between positions. Evolved forward positions are ceremonial than functional. A center lines up for the face off, but may require a winger to take up that position in defensive support because on-ice conditions dictate the positional switches. In other words, the winger would become the ‘F1’ the lead forward in the defensive situation. After the puck is dropped, C, LW, RW turn to F1, F2, and F3, contingent on the situation on the ice. I’d rather consolidate the averages instead of individual positions.

Raw, absolute totals aren’t always the best way to analyze efficiency over time. Moving to a rates per 60 minutes normalizes effects over time.

The radar chart above shows the impacts between forwards and defensemen, while the chart below measures the average shooting percentage between positions over time since ’15-16. In ’20-21, that gap is much tighter than it’s been over the last couple of seasons as forwards scoring rates have been depressed and defensemen staying fairly constant.

4v4 Shooting Percentage by Position
4v4 Shooting Percentage by Position

Filtering the performances to a player level indicates ’18-19 having the biggest gap between goals for per 60 (0.98) and expected goals per 60 (0.72), while only ‘2016-17 exhibited a negative goals above expected value (0.56 to 0.63).

4v4 Expected Goals Shooting Percentage and Time on Ice - Teams NHL
4v4 Expected Goals Shooting Percentage and Time on Ice - Teams NHL

Below is a great example of the relationship between shot attempts and time on ice and its effects on shooting percentage over time. In ’20-21, as average ice time per team per season decreased, the average of individual shot attempts among all players increased closer to historical averages, with a corresponding dip in shooting percentage.

4v4 TOI, Shooting Percentage and Shot Attempts per 60
4v4 TOI, Shooting Percentage and Shot Attempts per 60

The disparity is similar to ’15-16, but in significantly less average time on ice over the course of the season, and that’s the key. The open ice and skill level entering the league should likely produce more results similar to the chart below.

While ice-time can fluctuate, the NHL is redeeming on an influx of talented young players that can freely move in the 4v4 space in the history of the game. With more ice, opting for skilled players just makes sense. Fourth line players are exceptionally skilled and could prove useful in 4v4 situations, so coaches have a plethora of options to use on the ice. As the game evolves further, I’d expect better performances from 4v4 play. Time will tell.

Editor’s Note: Don’t forget to download the NBC Sports EDGE app to receive real-time player news and updates. Plus, it allows you to easily track your favorite players. Get it here!