I truly believe that the vast majority of criticism on hockey’s public statistical community are heavily overstated. The major users of statistical analysis understand that the numbers are imperfect and do not tell the whole story. In fact, they are usually the ones that have a better grasp of how much of the story is being told, since we can and actively do test the relationships are between two variables.
This is not to say that we are perfect in these areas. We are human and still fail. One area that is not discussed often enough, in my own opinion, is the how’s behind the results. An example of this can be seen with intangibles.
Now before you all start jumping on me, let me talk about intangibles and one that’s not discussed, even by the traditionalists not involved in the statistical community.
Statisticians and analysts heavily focus on the outputs, as the initial goal in analysis was to better measure a player’s impact and therefore strengthen opinions extending from the evaluation process.
This has shifted slightly lately, with more focus on neutral zone statistics and passing projects, with the returns on output research diminishing rapidly due to the limits within the currently public data. Still, most analysis is still on the outputs and not the inputs.
Now this isn’t completely backwards, as you need to have a good handle of outputs to help you sniff out how to weigh each individual input.
If this discussion of inputs and outputs is confusing you, there is an infographic made by my friend Mike to help out:
The desired result is winning the game by outscoring the opposition; however, a player is not always on the ice. So, we instead look at the individual outputs exclusively when the player is indeed playing, or how the team and/or linemates perform with the player on the ice versus away. The numbers here (whether goal differentials, shot differentials, point production, etc.) help paint a picture of the overall impact a player has. The inputs are merely a means to those outputs of outshooting and outscoring, where outscoring is the goal and outshooting is a byproduct.
Being adept in any individual input, such as strength, skating, or effort, becomes inconsequential if the sum of the inputs adds-up to a inferior output. But, these inputs are still important as they help describe why a player contributes to the degree that they do.
In other words, there is no “X” level of grit or leadership or speed or size required to succeed. Instead, all these factors combine with different levels of impact to give an aggregate and singular whole value. None of these inputs exist as a separate entity from a player’s impact on the game.
This should make sense intuitively. For example: hard work can overcome natural skill, but it is the level of effort and skill combined that makes a team perform how they do, and any measure of performance contains both their effort and skill.
There is a bit of a pattern though when it comes to public discussions over intangibles and player value: the players where intangibles are often argued over are rarely the ones that carry positive player outputs.
We can use Andrew Ladd as a Jet example of this. Ladd is not exactly a player that carries the prototypical “skilled player” package. No one is confusing Andrew Ladd from Vladimir Tarasenko, but Ladd is very good player who helps his team and garners positive results.
Why? Because of the many intangibles of Ladd. While Ladd does have a deceptive release and is highly-skilled in deflecting the puck, much of Ladd’s impact comes from areas like being defensively responsible, grinding to the prime scoring areas, fighting puck battles, and wearing down his opponents. Most understand and agree that Ladd has value, even if they disagree on the amount.
The arguments over intangibles usually comes from players in the bottom of the roster, when discussing the value of Mark Stuart versus Paul Postma or Chris Thorburn versus Eric O`Dell.
Why is that? Well there are likely two primary reasons behind this.
For one, when we look at the bottom of the roster, the gaps in skill and impact become less and less apparent. We are dealing with marginal players who are only marginally better than the other. A coach may find players that demonstrate desirable traits as more appealing and easier to choose over a more “traditionally skilled” player when the gap in skill is not the size of the Grand Canyon.
And second, lower roster players play substantially fewer minutes, with the bulk of their minutes under lower leverage situations, than their higher roster counterparts. This further diminishes the on-ice impact between two individuals. Fewer minutes means less time to be outscored, and lower leverage minutes means being outscored in those minutes will have less of an impact on a team’s win percentage.
Intangibles which may improve the team’s outputs with the player simply sitting on the bench then become relatively more important. Whether the intangible helps a team’s cohesiveness, energy, or morale, a player who in theory helps the team even when not on the ice does not lose this impact with riding the pine.
This is the reasoning behind why an inferior players may be worth their position on the roster. The media often exalts these blue-collar, hard-working, below-replacement-level individuals for these very same reasons. But, there is an intangible not talked about that is pulling value in the opposite direction from these players.
Now, we have previously discussed the danger in assuming an unknown value outweighs a known commodity like a player’s measurable on-ice output. That is not the discussion here. This is a distinctly new discussion on a brand new intangible. Not new in existence, but newly discovered.
Now, I haven’t been able to perfectly measure my new intangible, but I think I have a good proxy in measuring it within some reason. Here are the top and bottom Jets in this new intangible:
Now don’t take the order as 100% accurate, as there are likely some imperfections, but I’m confident that this intangible is real, has a significant impact, and the general sense in value above is accurate.
What is this unnamed intangible? It’s the intangible of being the better team. It’s the intangible in winning games. It’s the intangible in outscoring the opposition. It’s the intangible of knowing which players are helping you win and outscore. It’s the intangible of not sucking.
The order above is simply taking the Jets who outscore and are outscored the most (so one issue is shooting percentage variance).
Sure, good characters may help gel by creating a positive atmosphere in the dressing room. You know what else does that? Winning. Anyone who has played some high-level of competitive sports knows that winning brings a team together. Paul Maurice once even said something along the lines of winning cures a lot of ailments (not direct quote as it could not be found).
One group wondered how good NHL players deal with a teammate who is a below-average player, and whether good players realize bad players hurt the team.
Players may love their teammates, and desperately want to remain loyal to them, but they need to take care of their family first.
Keeping that in mind, good players understand poor players hurt their performance.
You don’t tell a teammate face to face he’s bad (some know they’re not good); he’s most likely a friend, so you hear him out and nod at his warped version of his talent. This causes awkward moments, of course.
Some good players go to the coaches and ask to not be put on the ice with their friend, as they want to succeed out there.
Most won’t say, ‘Because he’s bad’; it’s more like, ‘I don’t think we play well together.’
Players know they can only control what happens on the ice, so they want to play with the best available on the roster. Some agents of good players repeat this message to the general manager.
Teammates often know which players are hurting them. They know the players who are flubbing with the puck and consistently causing the team to get out shot and outscored. And even when they do not know, they feel the impact of the loses, which impacts the team’s morale, cohesiveness, and the room.
Good players have an intangible impact on the roster. They help the team win more games. This feels good and has a positive impact. Poor players likewise have an intangible impact on the roster through their poor play, in garnering fewer wins.
So when you look at these marginal roster players, when you are trying to discern whether or not their intangibles make up for their negative poorer on-ice outputs versus an alternative roster player, remember that there is an additional unseen cost, the intangible of a statistically good player.
Feel free to call this intangible whatever you wish. I just request you not to follow hockey tradition and name it after the person who first talks about it, please.