Don Cherry just said, without laughing, that the Islanders’ Martin-Cizikas-Clutterbuck trio is the “best fourth line ever in hockey”. OK.

— John Matisz (@MatiszJohn) January 18, 2015

The New York Islanders agreed to terms Casey Cizikas to a five-year contract extension worth $3.35 million on average per year.

This extension sent shock waves throughout Twitter, Reddit, and discussion boards as it seemed to be a hefty price and term to pay for a member of the team’s fourth line. The Islanders were not without their defenders, though, with many pointing out the “best fourth line” label the trio of Casey Cizikas, Matt Martin, and Cal Clutterbuck are often given.

Prior to debating whether or not the Islander trio is actually the best fourth line ever (or even currently) in hockey, we should ask: How much is the best fourth line in hockey worth?

Just over two years ago, I broke up the league’s forwards into four separate buckets to look at the differences between first, second, third, and fourth lines.

There are distinct differences in impact and usage between the four lines and so this needs to be taken into account whenever evaluating them:

Note: each unit is a bucket of players, so the variance in on-ice percentages is artificially reduced in the above chart.

While there are other areas of impact, the difference in 5v5 goal differential between the top and bottom performing members in each line gives a solid estimate in the relative and marginal impact potentially gained when improving a team at one position.

Fellow Hockey-Graphs contributor, Carolyn Wilke, was kind enough to send me her NHL Cap data given in similar four-line-bucketing method and fashion to my past experiment:

The above data by Carolyn is for all situations, not just 5v5.

The summary of information here is not all that interesting on the surface level; top-line players are better on average and are in turn compensated as such (gasp).

I then took it to the next level, looking at the spread between the min and max in both goal differential and Cap hit for each line, but indexed relative to the spread we observe on the top line:

Here we start to see a bit more interesting results. The graph suggests that NHL General Managers are relatively overpaying for marginal improvements on fourth-line players, while underpaying for middle-six players.

The biggest gap is in third-line players. The potential marginal improvement towards a team’s goal differential is almost the same as second line forwards, yet the difference in compensation is nearly identical to fourth-line players.

Now, we are looking at spread over standard deviation intentionally to include outliers, because Casey Cizikas being paid more than 4 standard deviations above definitely is representative of an outlier type situation.

While NHL General Managers, statistical analysts, and (most) pundits and fans seem to recognize that fourth-line players have the smallest impact on the team and should be compensated appropriately, there is a noticeable flaw in the market.

Teams relatively over pay for improvements on their fourth-line players, giving too much to the best fourth-line players (due to ELCs, the minimum contracts are fairly similar for all 4 lines).

Extras and missing information:

There are two major factors that many will point out as missing in this analysis.

For one, there are the intangible characteristics to a player that could potentially improve the team in a manner not measured by this type of analysis. Fourth-line players are often lauded for their character. The hypothesis goes that fourth-line players can energize the top of the roster, create cohesiveness with their positive nature, and set an example with their dedication to the team and obedience to the coach.

Secondly, we are observing only differences in 5v5 impact, and not including potential impact on special teams, such as the penalty kill. Although, special teams likely spreads the impact relatively more for top-players than bottom.

These are both real and significant concerns, but I feel there are two other missing factors that balance out the equation.

There is another intangible that I often call the “intangible of being a good player.” A common adage trotted out by coaches and executives is that winning cures all ailments.

I’ve heard former NHL players state that everyone on the team is well aware of which skaters are hurt the team the most. I’ve long believed that while players have a latent variable impact through their intangibles, impacting the team even when not on the ice, but that a player’s impact also spreads via a feedback mechanism, whether indirectly through morale or directly due to garnering more opportunities (ex: better Corsi players provide more offensive zone faceoff starts).

In addition, I’ve actually over assumed the impact spread in talent and impact for fourth-line players in two ways. The goal impact data is including both regular fourth-line players and pressbox / call-up fourth-line players as well. As well, not all goals are equal. Fourth-line players, including the Islander trio, tend to play more in low-leverage situations, further reducing the actual impact difference fourth-line players could potentially have.

Addendum: Pretty visual showing population

Carolyn made an awesome visual and shared it in our Slack.

I thought some would like to see this.