Great example of why pts/60 TOI is a basically useless comparison.

[CapnZin]

I agree points per 60 is ridiculous. All that counts are actual points.

If a player scores 90 points playing 18 minutes a game vs a player scores 100 points playing 23 minutes a games - give me the 100 points guy. 100 points is more useful than 90 points - and too bad for the 90 point guy for either:

1. Not being able to convince his coach to play him more minutes so he can contribute/produce more
2. Not being able to handle playing more minutes

The only time where i could maybe buy into looking at points per 60 metrics is if someone is able to show a clear example of how an increase in ice time led to a similar increase in points. Ie a player going from 15 minutes a game and 75 points to 20 minutes a game and 100 points. I'm thinking in 90% of cases, the player going from 15 minute a game and 75 points to 20 mins a game sees a few extra points, but nowhere near 25 extra points. ie - this is mostly useless. And this would *only* be relevant if we can bump the player's ice time up. If we leave him at 15mins a game - simply "knowing" he'd do more with 20mins is completely useless.

This example of logic is grossly incorrect.

The premise of using TOI/60 isn’t for a “X is better than Y” argument. It deals with efficiency on ice.

Your rates of 90/18 and 100/23 shows better production standards for the 90/18 player. Regardless of if that player is on the 2nd or 3rd line, they’re going to get a higher TOI in respect to their linemates because they’re more effective.

Basic statistics that you learn in high school or wherever you went to school can show a simple regression—> pace drops off and we can easily implement a -.02% in 1 min of TOI. Basic stuff right here. Increasing player X(90/18) and player Y(100/23) by the SAME STANDARD TOI/60, being 5 min, shows:

Again, people who use this in comparison as a standard of line delegation—>
A 4th liner producing well at a small TOI/60 to a struggling 1st miner (TOI/60)/82 also doesn’t understand the use of this metric... here’s the proof. These players have to be similar in play or else the error of the regression makes it irrelevant (comparing a 3rd liner to a 1st liner):

Player X: (90/18= 5..5*23=115.. 115/23)
...Player X = Player Y...
((115 + ([10*-.02]*82)) would be the drop off —> 115 - 8.2 = 98.6

Player Y:
((100 + ([5*-.02]*82)) would be the drop off
—> 100 - 8.2 = 91.8

Give me Player X that you said you don’t want.

Of course these numbers are malleable and aren’t concrete for every player, but the logic is sound.

I have a feeling you may not understand the use of TOI/60, but learning more can give you a better picture of its use.

It cannot be your argument, but it can help you make one. It really makes sense for saying hmmm...

For the Flyers: would using Wayne Simmonds or Nolan Patrick on the PP1 make a difference as a net front role.

(TOI/60)/PP says Simmonds. That can be a way to say Simmonds should, but it shouldn’t be your argument. Patrick works better as a PP1 net front role. I’m using this example to show that it can be wrong, but it has to be used with similar linemates.

 

[CantLoseWithMatthews]

Wow, there's still quite a few people who just don't understand the stat. Fortunately there are also a lot of people who do, which is nice to see.

 

[Korpse]

People actually use P/60?

 

[CantLoseWithMatthews]

Well that is obviously equally untrue.

McDavid has 4 points in 4 games where he has played under 20 minutes. 7 points in 6 games where he has played under 21 minutes and 33 points in 22 games where he has played more than 21 minutes.

Well, there goes the OP's brilliant argument. The weird part is he somehow suckered people into believing him

 

[Nithoniniel]

In most cases, probably not. An increase in icetime doesnt mean a for sure increase in production. 15 points is a pretty big number and actually production does matter. In that case I think you could maybe argue theyre closer than they appear but theres plenty of people on this website that would argue the 50 point guy is better because of his per 60 stats which is bs. P/60 doesnt make up for a tangible difference in production at the end of the day

There's a strong correlation between ice time and production. It should most definitely be the expectation that a significant increase in ice time will produce an increase in production. Maybe not at the same rate, but an increase is overwhelmingly likely. Of course, this is with the caveat that such an increase in ice time is not the result of increased PK duties.

Production matter, but in that case you can't just isolate it. The production you get is judged relative to the opportunity you got, and ice time is a limited currency. Meaning if you get PP time in the 90th percentile and produce in the 70th percentile, your returns are not as valuable to the team. Not if the alternative is to have two players sharing that ice time and giving better production returns.

However, in this case we are comparing two players. The scenario you describe would entail that if the two players switched place, with every other factor accounted for, the 55 point player would outperform the 70 point player in similar circumstances. For that to be true, per 60 numbers have to stay relatively constant.

From the cases I've read about and the work I've done myself, that has largely been the case. I'd also add that better linemates can have a huge effect as well, and that usually comes with an increased opportunity. Marchessault is a terrific example, going from replacement level production rate to elite level just by changing what kind of players he plays with.

The reason people have an issue with per 60 is the outliers, from my experience. And that's an area where you need to be wary. If a player is playing third line minutes with strong linemates beating up easy competition, an increase in minutes will likely mean tougher opposition and not much else to balance it. As a result, per 60 numbers in that case is bound to falter.

However, the issue there is not with per 60 numbers, but with people judging a stat by its outliers. If you instead look at someone who is killing it on an average third line and second PP time, he's bound to see a big bump in the numbers if you put him on the first line and first unit.

 

[WetcoastOrca]

It’s not a very helpful stat without context.
Someone posted in another thread that Conner Shearey was either the leader in PP 60 minutes or close to it one year.
Actual points are a better indicator. Good players get lots of ice time. The points reflect that.

 

[Machinehead]

Any stat sucks if it is used without context and other knowledge.

Just an example from my team: there are two forwards with an exact same amount of 5v5 points per 60. One of them is at 25 points, while the other has 9. Does the stat suck? Obviously not.

It just suggests that maybe, just maybe, the 9-point guy could do more than he currently does with 4th line usage and linemates.

If, however, you feel like ignoring /60 stats, which essentially add context to point totals, you might miss out on a good player that is surpassed by guys who play significantly more or get better linemates/PP time.

The story of Mathieu Perreault, really.

You get the stat. Awesome stuff.

 

[SwaggySpungo]

I will agree with OP that it is a really dumb and meaningless statistic.

 

[talitintti]

McDavid would produce more in Leafs than Oilers because Babcock gives every player under him career years by a long margin. Google it if you wanna.