Do Stats Such as P/60 Ignore the Athletic Aspect of Hockey?

[DitchMarner]

I see such stats used frequently, but at times it seems as though they're applied with the assumption that all players are essentially equal physically, athletically and in terms of conditioning. However, NHL players are not video game characters. They don't share identical physical characteristics, and in a sense stats like p/60 and p1/60 fail to give players who are able to perform at a relatively high level in greater ice time than others credit for superior fitness and conditioning.

Click to read more...

 

[GeeoffBrown]

pts/60 is mostly used in situations when it favours a player someone likes and is useful in very few other situations

 

[Sidney the Kidney]

pts/60 is mostly used in situations when it favours a player someone likes and is useful in very few other situations

This. It has its value if used properly, but too often on HF it's used as a replacement for when favorite players don't stack up with actual production and need to be propped up using other methods.

 

[Empoleon8771]

Points/60 is a garbage stat IMO. Per 60 stats are good in some cases, but those cases are things like shot attempts per 60, not points per 60.

 

[Cursed Lemon]

P/60 would be a great stat if it wasn't so unbelievably wrong all the time, e.g. repeatedly showing Tatar as the best Red Wing.

 

[Knave]

Points/60 is a garbage stat IMO. Per 60 stats are good in some cases, but those cases are things like shot attempts per 60, not points per 60.

So you have a problem with arbitrarily choosing to divide all stats by the length of an NHL game despite nobody outside of goalies playing the full length of the game?

 

[Empoleon8771]

So you have a problem with arbitrarily choosing to divide all stats by the length of an NHL game despite nobody outside of goalies playing the full length of the game?

I've read this like 3 times and I'm not sure if you're agreeing with me or not. The 2nd part of your sentence contradicts the 1st part.

 

[East Coast Icestyle]

The real issue is in people's assumptions that it scales perfectly with increased time. Everyone tries to use that in arguments.

 

[tucker3434]

It can be useful when comparing two players that are pretty similar statistically. It sucks when it’s used as a crutch for players who don’t have high actual point totals.

 

[Martin Skoula]

The real issue is in people's assumptions that it scales perfectly with increased time. Everyone tries to use that in arguments.

I've never seen anyone argue that.

The discussion is usually this:

Player 1 has 20 more raw points than Player 2
Player 1 plays more ES and PP minutes
Player 2 has a similar P/60 at ES/PP

If Player 2 were to get the same minutes as Player 1, the raw points gap would be less than 20.

It doesn't have to be a perfectly scaling improvement in raw totals, even if Player 2 only scores a single point in the additional minutes it still shrinks the raw total gap.

 

[DitchMarner]

I've never seen anyone argue that.

The discussion is usually this:

Player 1 has 20 more raw points than Player 2
Player 1 plays more ES and PP minutes
Player 2 has a similar P/60 at ES/PP

If Player 2 were to get the same minutes as Player 1, the raw points gap would be less than 20.

It doesn't have to be a perfectly scaling improvement in raw totals, even if Player 2 only scores a single point in the additional minutes it still shrinks the raw total gap.

The thing is: People assume that coaches just don't like giving certain players a lot of ice time and they could increase their ice time if they wanted to. There may be cases where increasing a player's ice time may be detrimental to his overall performance even if it would increase his raw output. Not every player has the same level of stamina or fatigue threshold. The raw points (at least at ES) you get out of certain players may be more or less what they can give you if they're playing at an optimal level.

For some players, perhaps the risk of injury would increase significantly with increased ice time as well.

 

[FerrisRox]

P/60 would be a great stat if it wasn't so unbelievably wrong all the time, e.g. repeatedly showing Tatar as the best Red Wing.

That's about as wrong as a stat can be.

I mean, really, any stat that shows a player who is not even on the roster or under contract to the organization as the Best player on the team is tremendously flawed.

 

[McDNicks17]

More importantly than the athletic factor, it ignores all context.

Some people will point to a third liner with a higher P/60 than a first liner and try to argue that the third liner is better despite them playing much easier minutes.

It's fine when comparing players in the same role, but it's almost never used that way.

 

[ESH]

P/60 assumes a player’s scoring will improve linearly with their ice time, which is impossible to really prove.

 

[CantLoseWithMatthews]

like with every stat, there are flaws that limit it from being the only tool one should use to evaluate players. What we should really be asking is if it's still more valuable than points per game or raw points. Personally, I would say yes. If you're going to evaluate players based on their talent, it's important to limit factors that they don't have much control over or don't directly relate to playing ability

 

[CantLoseWithMatthews]

P/60 assumes a player’s scoring will improve linearly with their ice time, which is impossible to really prove.

no it doesn't. you could easily use it as a descriptive stat, and I've still not seen anyone try to claim it's a perfectly linear relationship

 

[XX]

P/60 assumes a player’s scoring will improve linearly with their ice time, which is impossible to really prove.

It's useful for comparing players in similar situations, like a number of advanced stats. If you have a player in your top 6 that is being fed the best possible zone starts and the most 5v5 time - essentially being treated as the top weapon - they need to deliver relative to other players on the team and other top 6 players around the league. Not all lines and offenses are created equally, and a less used player cannot automatically produce more given more time (and likely facing better defenders as a result), but it's still a useful reference at times.

Most people do not understand the limits of advanced stats
Most people do not understand how numbers relative to teammates work
Most people do not understand the impact of deployment and usage on numbers

The above 3 factors have people constantly quoting numbers and saying shit that doesn't make sense

 

[DRW204]

of course it does. p/60 just measure how efficient a player is at putting up points

 

[TomasHertlsRooster]

Separating 5V5 number and PP numbers and judging players relative to the ice time they’re given is generally going to be a better way to assess performance than just ignoring all of that, but some fans go too far and completely ignore the effects of accumulated fatigue when it doesn’t fit the narrative.

 

[Sidney the Kidney]

like with every stat, there are flaws that limit it from being the only tool one should use to evaluate players. What we should really be asking is if it's still more valuable than points per game or raw points. Personally, I would say yes. If you're going to evaluate players based on their talent, it's important to limit factors that they don't have much control over or don't directly relate to playing ability

The problem with that is it leads to results like Conor Sheary leading the entire NHL in P/60 at 5on5 and thus, arguing he was "better" that season than McDavid and Crosby, but simply wasn't given the ice time those two were.

That, to me, is the biggest flaw in using these stats as some sort of "they're better than raw totals to judge players". It has the potential to have way more anomalous results than raw totals would. Thus, why I don't agree at all that it's "still more valuable than raw points".

Per 60 numbers can be useful when they're used to SUPPLEMENT things like raw totals. The problems arise when they're used as a REPLACEMENT for raw totals when judging players.

 

[Dekes For Days]

This is the top 0.0001% or something of the sport playing in the NHL, all with access to similar benefits, and we are talking about the 1% of that that are the elite star players. The discrepancy between them is less than you think, especially in terms of conditioning. This isn't 50 years ago where some players were smoking cartons of cigarettes a day. The people who play more minutes don't have some special unique ability; they are just often on teams where the coach has a certain preference or teams that aren't deep enough and need to play their stars a ton.

There is absolutely zero proven evidence that some players can effectively play more minutes than others, yet this myth is often perpetuated to discount players who are just as good and produce just as well, but get less opportunity, either due to coaches or team strength or circumstance. TOI heavily influences production, especially on the PP, and it's dishonest to ignore it.

P/60 is one of the best stats we have for production. Yes, it requires context and proper sample sizes, but this is the same as all other stats including points.

 

[nbwingsfan]

It really only became a big thing around here when Matthews entered their league because it’s the only thing some Leaf fans can cling onto “showing” Matthews is a top 5 player.

edit: See above for a perfect example.

 

[Tad Mikowsky]

P/60: the perfect cherry picking stat when your player doesn’t play enough minutes, games, or produce enough raw totals

 

[Bounces R Way]

Raw point production will always be the gold standard in the NHL. I don't think there is anything wrong with that.

P/60 is fine but it doesn't account for a lot of things that really should considered. QoC, QoT, ZS% among other things will all have an effect on it.

 

[Dekes For Days]

QoC, QoT, ZS% among other things will all have an effect on it.

Just as it will have an effect on raw production.