Everything Derozan

[raptors999]

A perfect example was Hansbro. He was very effective at times, but only when used for 10-12 minutes max.

Comparing bench vs starters. Normalizing stats is the only way to compare Apples to Apples. Per36 is still useful for bench guys, especially energy guys.

[Barolt]

A perfect example was Hansbro. He was very effective at times, but only when used for 10-12 minutes max.

Right, but if you take two players, one guy plays 30 minutes a game, scores an average of 20 points, 10 rebounds, and 6 assists. The other guy plays 20 minutes a game, scores 15 points, grabs 8 boards and gets 4 assists.

You can present this as:

Player A: 20pts/10rb/6assists per game.
Player B:15pts/8rb/4 assists per game.

By this metric, player A looks clearly better.

Or you can present this as:

Player A:24pts/12rb/7.2 assists per 36 minutes.
Player B:27pts/14.4rb/7.2 assists per 36 minutes.

Now Player B looks slightly better, and it's definitely a lot closer than the per game stats. Normalization /36min or /100possessions allows you to compare and contrast players more accurately.

[SkywalkerAC]

Right, but if you take two players, one guy plays 30 minutes a game, scores an average of 20 points, 10 rebounds, and 6 assists. The other guy plays 20 minutes a game, scores 15 points, grabs 8 boards and gets 4 assists.

You can present this as:

Player A: 20pts/10rb/6assists per game.
Player B:15pts/8rb/4 assists per game.

By this metric, player A looks clearly better.

Or you can present this as:

Player A:24pts/12rb/7.2 assists per 36 minutes.
Player B:27pts/14.4rb/7.2 assists per 36 minutes.

Now Player B looks slightly better, and it's definitely a lot closer than the per game stats. Normalization /36min or /100possessions allows you to compare and contrast players more accurately.

Everyone knows that. The point is that all stats have their flaws, some more egregious than others depending on the perspective of inquiry. So...

[JWash]

It only makes sense to use Per 36 stats for players who play a similar number of minutes in order to normalize them. Using per 36 stats to compare a guy who plays 20 minutes to one who plays 40 is inherently flawed. If you're out on the court more you're experiencing greater fatigue, likely spending more time playing against better players, and probably have a larger offensive role on your team.

Per 36, Javale McGee averages 16-10 and 4 blocks for his career. Is he a superstar player? No because he does that in role player minutes and against role players. It's not as simple as to say "Oh if he got the minutes he'd have that production" no he wouldn't.

Context always has to be applied to statistics.

[Barolt]

It only makes sense to use Per 36 stats for players who play a similar number of minutes in order to normalize them. Using per 36 stats to compare a guy who plays 20 minutes to one who plays 40 is inherently flawed. If you're out on the court more you're experiencing greater fatigue, likely spending more time playing against better players, and probably have a larger offensive role on your team.

Per 36, Javale McGee averages 16-10 and 4 blocks for his career. Is he a superstar player? No because he does that in role player minutes and against role players. It's not as simple as to say "Oh if he got the minutes he'd have that production" no he wouldn't.

Context always has to be applied to statistics.

Agreed, and like I said before, I'm not claiming that the fact that DeMar doesn't run as much as the article implies is even a bad thing, I don't know that it is. I'm just saying that the premise of that part of the article, that DeMar moves on offense as much as anyone else in the league, is inherently flawed.

[A.I]

I don't understand what you mean. Per 36 is calculated as what stats the player puts up in the average 36 minutes that they are on the floor.

per 36 doesn't take into account fatigue or basically just not play well. For example, JV is averaging 12 and 9 in 26 mins, and his per 36 is 16 and 12, but it isn't guaranteed Jonas will average 16 and 12 in 36 mins, he could average less due to fatigue or just missing shots.

[Barolt]

per 36 doesn't take into account fatigue or basically just not play well. For example, JV is averaging 12 and 9 in 26 mins, and his per 36 is 16 and 12, but it isn't guaranteed Jonas will average 16 and 12 in 36 mins, he could average less due to fatigue or just missing shots.

Right, sometimes a guy is going to overachieve and sometimes underachieve. The goal of per/36 stats is to average it out. Some guys certainly have a bigger bell curve than others in terms of best vs. worst games. We aren't trying to tell that story with per/36 numbers, or any normalized stat, just trying to tell the story of where their most typical 36 minutes sits.

[JWash]

Agreed, and like I said before, I'm not claiming that the fact that DeMar doesn't run as much as the article implies is even a bad thing, I don't know that it is. I'm just saying that the premise of that part of the article, that DeMar moves on offense as much as anyone else in the league, is inherently flawed.

What you're saying isn't making any sense.

He's near the top of the league in average distance traveled on offense per game, period. Yes he plays a ton of minutes, but almost everyone on that first page is playing around 35 minutes per game. It's easy to say there are players who get a lot less minutes but may move more in the minutes they get, but the reality is they AREN'T playing as many minutes as DeRozan.

Things like stamina, role and who you're playing against are going to affect those stats as well. So it's easy to say that player X who covers 1 mile/game in 20 minutes would beat out Player Y who covers 1.8 miles per game in 40 minutes if he got his playing time, but that's just completely ignoring the aforementioned factors.

It's the reality of trying to extrapolate a small sample within a game based purely on mathematics. If you want an up close and personal case study, look at Amir Johnson. He was getting almost 4 blocks per 36 in Detroit, then it dropped off to less than 2 in Toronto. Why? Did Amir forget how to block shots? No, it's because when you're playing less like he was in Detroit you're going to be out there flying around expending as much energy as possible, plus you're likely playing against lesser players.

[Eliza]

Oh boy, this devolved in a hurry, lol. Another argument about the validity of per36 stats, ha. Sooooo, how about that article and the premise of how greatly effective DeMar is without the 3 point shot :-)

[3inthekeon]

per 36 doesn't take into account fatigue or basically just not play well. For example, JV is averaging 12 and 9 in 26 mins, and his per 36 is 16 and 12, but it isn't guaranteed Jonas will average 16 and 12 in 36 mins, he could average less due to fatigue or just missing shots.

Its often on the defensive end that big minute increases (and usage increases also) have the biggest negative impact.

[Barolt]

What you're saying isn't making any sense.

He's near the top of the league in average distance traveled on offense per game, period. Yes he plays a ton of minutes, but almost everyone on that first page is playing around 35 minutes per game. It's easy to say there are players who get a lot less minutes but may move more in the minutes they get, but the reality is they AREN'T playing as many minutes as DeRozan.

Things like stamina, role and who you're playing against are going to affect those stats as well. So it's easy to say that player X who covers 1 mile/game in 20 minutes would beat out Player Y who covers 1.8 miles per game in 40 minutes if he got his playing time, but that's just completely ignoring the aforementioned factors.

It's the reality of trying to extrapolate a small sample within a game based purely on mathematics. If you want an up close and personal case study, look at Amir Johnson. He was getting almost 4 blocks per 36 in Detroit, then it dropped off to less than 2 in Toronto. Why? Did Amir forget how to block shots? No, it's because when you're playing less like he was in Detroit you're going to be out there flying around expending as much energy as possible, plus you're likely playing against lesser players.

I'm not arguing why it's a flawed stat, I simply don't know. But it IS a flawed stat. It's like Kobe is scoring 18 points per game, which sounds really good. But he's doing it in a ton of minutes on a ton of shots, which isn't as good.

DeMar is covering a lot of distance, yes. But he's doing it in a ton of minutes, and he's not covering as much per minute as a lot of other guys. Keep in mind, when I ranked him 78th out of 148 players, that was players playing 25+ minutes per game.

Among players playing 30+ minutes per game, DeMar ranks 44th out of 81 players in terms of offensive speed.
Among players playing 35+ minutes per game, DeMar ranks 12th out of 19 players in terms of offensive speed.

The stat is flawed.

[KHD]

how the fuck are you people still debating about this?

The article used that stat improperly. Full stop.

[Mess]

If I drive at 100 miles per hour for 1 hour I'll go further than you will driving 120 miles per hour for 45 minutes.

Demar plays the 7th most minutes in the league, 36.5 min/game.

The stat is flawed. The article may not be, and in fact the inference of "constant motion" may not even be, but that stat is.

I just don't see the point of calling out the stat as flawed like he based the whole article (or even that section) on it. And pointing out that his usage of the term "constant" is incorrect is just annoying (to me anyway - and not that you specifically did that either). Same with another avg. speed reference. I get that it was mentioned to refute the premise that DD is always running around like a maniac when he is on the court - but that's not what the author was implying. My reaction to comparing player's average speeds is like knowing that Scola can bench the most on the Raps. It's cool but ultimately meaningless.

Ok, so the stat is flawed. Does that change the premise of the article at all? Do the Raps not run DD off a lot of screens (when they're not posting him up)? Is he not effective at gaining the defense's attention and being a greater threat that way?

[3inthekeon]

Except the article did use the stat properly.

[special1]

Everything Demar Derozan

I'm not arguing why it's a flawed stat, I simply don't know. But it IS a flawed stat. It's like Kobe is scoring 18 points per game, which sounds really good. But he's doing it in a ton of minutes on a ton of shots, which isn't as good.

DeMar is covering a lot of distance, yes. But he's doing it in a ton of minutes, and he's not covering as much per minute as a lot of other guys. Keep in mind, when I ranked him 78th out of 148 players, that was players playing 25+ minutes per game.

Among players playing 30+ minutes per game, DeMar ranks 44th out of 81 players in terms of offensive speed.
Among players playing 35+ minutes per game, DeMar ranks 12th out of 19 players in terms of offensive speed.

The stat is flawed.

The stat isn't flawed. Your argument is flawed.

Demar is constantly moving. Demar is 4th in the league in terms of ground covered on offense PER GAME. That is a fact.

Your brilliant idea is to bring in guys who actually cover less ground because they only play 25 mins.

What sense does that make?

They don't cover more ground because they don't get the minutes? Who cares why they don't cover more ground. This is just another way to discredit the FACT that he covers more ground in the 36 minutes he plays.

Sent from my iPhone using Tapatalk