P/60 stats: How useful and reliable are they? Does production increase linearly with ice time?

[Master P]

Can't wait to see what next years overused advanced stat will be. Last year it was expected goals for (lol). Who knows what next year will bring.

 

[Aceboogie]

Pts/60 is good if you dont buy the assumption that NHL coaches are infallible and always play the best line up/combo possible. I believe coaches have a ton of biases and this shows up in how they deploy players. Certain players play more than they should, and other play less than they should.

P/60 measures effectiveness, which is great because it cuts thru a coaches BS if they are overplaying a certain player and underplaying another

 

[Bank Shot]

You seem to again miss the point, because none of this address my point. I didn't say that everybody regress to the same OISH. I never claimed that it will automatically correct based on what happened before.

Yes, Matthews numbers last season was likely inflated. But the opposite is true for his first season. That means that his true talent level probably lies in between these two. That true talent level is what is the most likely outcome going forward, as you note above. And how he performs at that level is illustrated by using the two year sample, where he has a 10.5 OISH%, within the limits you outlined yourself. And in that sample, he compares well to McDavid.

What you are doing here is picking and choosing. You're arguing that the numbers get inflated in Auston's second year, but you don't apply the same standard to his first year.

Matthews first year doesn't matter. If you adjusted him up to the average he would still lag behind Mcdavid.

There is still a substantial gap no matter which way you group the stats.

 

[Volica]

Is that anything wrong or weird though? Your 15/15 third line winger might be better if given a bigger opportunity. And he very well could hit 40 with luck, nothing odd about that either. Difference being that now we know a bit about who we can expect something like that from, and we know when it is actually luck and not just skill. Is that bad?

We've always wanted to figure out who can be better than they are, it's just that now we measure it.

I don't think it's weird, it's funny, for sure. It's because as fans we simply can't put away our biases when evaluating players; and we simply don't watch every player.

My issue has always been cherry picking of stats in terms of these forum debates and totally excluding any eye tests as well. Also, how do we tell what's luck and what's skill if we just look at numbers.

Here's an example. I'm not going to name either player, one is clearly the more skilled between the two, here are their stat lines:

Player 1: 1.82 P/60, 49.81 CF%, 1.042 PDO, 52.49 xGF%, TOI% QoT:31.17, TOI% QoC 29.58
Player 2: 1.75P/60, 49.54 CF%, 1.035 PDO, 48.66 xGF%, TOI% QoT: 32.06, TOI% QoC: 29.65

These two players are almost identical in terms of production, and most stats right? They product at the same clip, player 1 is actually the better player here; he spends less time with great teammates, he produces more on a 60 minute basis, he's expected to score at a higher clip... like wow.

Now what if I said Player 1 is a 26 year old player, 3 years pro experience, the most he's every scored was 40 points. Player 2 is a 25 year old player, 7 years pro experience, has been around that 60 point mark 3 times in his career, scored over 50 in his rookie season, +50 point pace for 6/7 of his seasons in the NHL. Wait, what?

I'm not saying stats like P/60 are bad/useless, it's just like every advanced stat in hockey; they need context.

 

[stampedingviking]

i've never seen anyone suggest that

That exact argument is used in every Matthews thread, but you've never seen it?

 

[zeke]

Doesn't matter.

Luck doesn't have a cumulative effect. It just happens.

Both are most likely to be between 10-11% this season.

That would cause a substantial drop in Matthews ES production.

which is why I included this in that post:

1st 2yrs

McD: ES 2.90 / 2.30 (11.2oish%) ---- PP 6.39 / 3.51
AM: ES 2.59 / 2.34 (10.5oish%) ----- PP 6.29 / 4.49

 

[zeke]

And how many of each type of game were the Oilers involved in?

depends on which season you're talking about, of course.

but mcdavid had identical p1/60 this year and last year, despite having vastly difference team winning/losing situations effecting his icetime.

 

[Fataldogg]

It can't be too useful. World class coach AV clearly ignored it with some of his athletes like Buchnevic.

 

[AKL]

That exact argument is used in every Matthews thread, but you've never seen it?

I’ve seen it used countless times in the last few days, just for Matthews alone

 

[stampedingviking]

I think the clear and obvious thing that should be gathered from the stats of last season is that there's a very good chance next season McDavid, Matthews, and Mackinnon should establish themselves as the Top 3 Centers in the league by a decent margin.

They all have very comparable advanced metrics across the most important categories(P/60, P1/60, QoC, QoT, etc) and they seem to be the leaders in those categories by a decent margin.

I'd say its splitting hairs to argue who will be better. I'm sure for Oilers/McDavid fans they dont like to hear that but the advanced stats certainly suggest it to be true. Of course if McDavid continues getting the significantly higher minutes he'll continue to produce more points as well. But if the ice time evens up I think these 3 will separate themselves as the class of the field offensively as Crosby, Malkin, Stamkos, etc. slow down a little as they get older. At least at 5 on 5.

There's also definitely still something to be said about both Matthews and Mackinnon as well. For Matthews, the stats all suggest he's due for a big jump in production but as we've seen countless times hockey is played on the stat sheets and Matthews still has to go out and actually take that next step and perform at that high level first. And for Mackinnon, last year was the first year where he really blew up and put everything together over a full 82 game season the way he did. So he still has to go out and prove that last year wasn't just a 1 year peak for him and that he's capable of doing it again.

Until Mackinnon repeats, and until Matthews on ice product finally catches up to the statistics, I think its fair to say McDavid is still in a class of his own. And even if Matthews and Mackinnon both take that next step next year its very likely McDavid still ends up a small step above the other two.

In what context? Or, do we just ignore Crosby/Malkin/Bergeron/Kopitar/Barkov?

 

[blinds]

I'm guessing this thread was spawned by a Matthews debate? Lots of Leafs fans with an apparent axe to grind in here.

 

[Nithoniniel]

Matthews first year doesn't matter. If you adjusted him up to the average he would still lag behind Mcdavid.

There is still a substantial gap no matter which way you group the stats.

No offense, but it's kind of hard to take your position seriously when you make a big deal about needing to adjust for unsustainable factors in Matthews' second season, and then says the bolded above his first.

If we want to compare the two by how they have done so far in their career, which was the premise you responded to, then you can't just choose to adjust one of the numbers and ignore another.

I'm not disagreeing that there is a gap. I'm just pointing out the logical inconsistencies in your argument.

 

[Saltcreek]

Are people really trying to compare Matthews with McDavid again? Are we doing the whole "loose puck retrievals" again? While you can compare their numbers they are not comparable players at this point. McDavid is on a whole different level to Matthews. p/60 can be an useful stat but as demonstrated in this thread but also multiple other threads people misuse it to suit their narrative. They are so many other stats that prove that McDavid is on another tier and to ignore all the data we have is just plain being a homer.

As it has been stated many times an increase in ice time will not have a linear increase in point production. If that was the case then why not just play McDavid 30 minutes a game so he could get 140+ points a season. There are so many other factors that come into consideration.

 

[stampedingviking]

Except, the people still working with the Cave Man mentality refuse to acknowledge the importance of using stats to show a players worth.


While on the flip side the analytical crowd for the most part are very open to the idea that the Numbers aren't the be all and end all and are just a tool that can be used. And that obviously watching the games themselves and seeing how the players are actually playing on the ice adds significant value as well.

They are just as blinkered (perhaps moreso) than the "Cave Man" mentality. Just read HF a while and you'd see that clearly.

 

[Pierce Hawthorne]

In what context? Or, do we just ignore Crosby/Malkin/Bergeron/Kopitar/Barkov?

In terms of offensive production, Mackinnon, Matthews, and McDavid are the top 3 in pretty much all categories at even strength when it comes to production.


Malkin benefited a tonne from a crazy strong PP, Crosby simply wasn't on the level of those 3 offensively last year and shouldn't be going forward.


At even strength those 3 Centers should separate themselves from the pack as the best producing Centers in the league next year, assuming they build on the numbers they put up last year.

 

[biotk]

Pts/60 is good if you dont buy the assumption that NHL coaches are infallible and always play the best line up/combo possible. I believe coaches have a ton of biases and this shows up in how they deploy players. Certain players play more than they should, and other play less than they should.

P/60 measures effectiveness, which is great because it cuts thru a coaches BS if they are overplaying a certain player and underplaying another

I don't think that people should need to say that they believe coaches have a ton of biases anymore. Vegas' performance this year pretty much confirms how fallible and biased coaches are. The teams who take that most seriously and do the most to overcome those biases will be the most successful going forward.

 

[CantLoseWithMatthews]

They are just as blinkered (perhaps moreso) than the "Cave Man" mentality. Just read HF a while and you'd see that clearly.

Nope. The people who deliberately try to utilize as much data as possible aren't as closed off as people who try to ignore as much data as possible

 

[stampedingviking]

In terms of offensive production, Mackinnon, Matthews, and McDavid are the top 3 in pretty much all categories at even strength when it comes to production.


Malkin benefited a tonne from a crazy strong PP, Crosby simply wasn't on the level of those 3 offensively last year and shouldn't be going forward.


At even strength those 3 Centers should separate themselves from the pack as the best producing Centers in the league next year, assuming they build on the numbers they put up last year.

Ah, best producing. So we're ignoring brilliant defensively oriented centres because they don't score so many points. Right, that's a brilliant idea

 

[stampedingviking]

Nope. The people who deliberately try to utilize as much data as possible aren't as closed off as people who try to ignore as much data as possible

Yep.

 

[Bank Shot]

Kind of hard to take your position seriously when you make a big deal about needing to adjust for unsustainable factors in Matthews' second season, and then says the bolded above his first.

If we want to compare the two by how they have done so far in their career, which was the premise you responded to, then you can't just choose to adjust one of the numbers and ignore another.

I'm not disagreeing that there is a gap. I'm just pointing out the logical inconsistencies in your argument.

I'm not adjusting anything.

You can't retroactively adjust someone's point totals. What is done is done.

I'm just talking about projecting things forward. It appears some Toronto fans feel that Matthews is closing the gap at ES, but they should be mindful that that is just an illusion brought about by unsustainable shooting percentage.

 

[Nithoniniel]

I'm not adjusting anything.

You can't retroactively adjust someone's point totals. What is done is done.

I'm just talking about projecting things forward. It appears some Toronto fans feel that Matthews is closing the gap at ES, but they should be mindful that that is just an illusion brought about by unsustainable shooting percentage.

And projecting things forward with the best possible sample size includes unsustainable shooting percentage both ways, both seasons. If you want to say that Matthews' last season was unsustainable and should be taken with a grain of salt, then that's fine. We agree. But the position you argued took that into account by looking at a larger sample too, one that is not unsustainable and fills our purpose. You seem to continually dodge that.

 

[Bank Shot]

And projecting things forward with the best possible sample size includes unsustainable shooting percentage both ways, both seasons. If you want to say that Matthews' last season was unsustainable and should be taken with a grain of salt, then that's fine. We agree. But the position you argued took that into account by looking at a larger sample too, one that is not unsustainable and fills our purpose. You seem to continually dodge that.

What's to dodge?

One guy lags the other consistently.

 

[olli]

And how many of each type of game were the Oilers involved in?

Mainly type A, but for the most part my point still stands.

 

[Aceboogie]

Ok jesus, Can this McDavid - Matthews BS get moved to a different thread. Its absolutely painful to read. They are both uber elite players, just enjoy them

 

[Machinehead]

The stat is used to judge players equally when it isn't accumulated under different circumstances, quality of line mates, quality of competition, usage, etc.
It's useful but it's also extremely variable.

But we give more credence to raw points which don't have the context of ice time.