Johnny Bball wrote:vini_vidi_vici wrote:DarkKnight wrote:No, you're ignoring what "net" means. Net takes on and off into account, not just on. If you put some scrub on the Jordan bulls, and he's only on the court with Jordan, he could have a +ve on rating. But when you take him off and the team has an even better +ve rating, then you end up with a NET negative.
I love watching people try to justify it. He's the only "star" in the recorded history of the league where his team is essentially always better (from an impact, more points than the other team standpoint) when he's off the court. It's not one year, it's not a coincidence, he's not one of a group of guys with the same consistent -ve net rating. He's the only guy. It's been 10 years. This is real.
This too is also false. Net doesnt take into account that, its saying while DD is on the court
the team is a positive/negative.
NET = NetRTG. On/off differential is its own thing.
Your argument actually indicates we had a great bench, which we did since 13/14 (except last yr).Using your logic, the last yr DD was here (
17-18), our best players were (in order)..
Bruno
Nigel Hayes
Lo Brown
FVV
OG
Pascal
Bebe
Jak
CJ
DD
KL
and they were our only "Net" positives. Bruno is clearly MJ (+55.6).
Suffice to say, its still false. DD was a positive 5 of 11 yrs, 1 yr he was 0.0, and 4 of those yrs the whole team (Raps) were awful, and this yr isnt finished.
Yep. But I’ve tried this before. People will continue to use no context about metrics they don’t understand. I just shake my head when read the term “net negative” since I know it means they understand little. Sorry but that’s true.
Neither of you understand this metric at all. I'll try to explain the on-off metric here slowly since apparently it's hard for both of you.
When a player is on the court, their team has an offensive rating - this is how many points per 100 possessions the team averages on offense during this time - and a defensive rating - this is how many points the opponents average per 100 possessions during this time. If you subtract these 2 numbers, you get a team's net rating for the minutes a player is on the court
When a player is off the court, you can do the exact same thing - get the PP100 they score and the PP100 they give up, and subtract to get the net rating for the team when a player is off the court.
If you subtract the net OFF from the net ON (On-Of, as shown in the chart) you get the overall net, i.e. how a player has impacted his team overall in terms of the score of the game. Here's a practical example:
A player plays 50 possessions and sits for 50 possessions. While ON the court, his team scores 50 points and gives up 45. This creates a net ON for the plays of +5.0. While OFF the court, his team scores 60 points and gives up 40 points. This creates a net OFF of +20.0. Subttracting these 2 numbers gives you -15.0, which is the on-off for this player. His ON looks fine, at +5, but it's clear his team was better in the minutes he was off, and this is reflected in the negative overall net.
Now that we've gone through that, we can look at your pet project, perennial negative Demar Derozan. Again, he remains the only "star" in recorded history of on-off to consistently (say, 80+% of his seasons) be a negative in this stat. But let's look at one of his seasons as an example. You used 2017-18, so let's look at that one.
When he was ON the court, the team averaged 115.7 PP100, and gave up 108.2 PP100, for an ON net of +7.5. That's good!
When he was OFF the court, the team averaged 111.4 PP100, and gave up 102.4 PP100, for an OFF net of 9.0. That's even better (for the team)!
However, the NET of these two numbers, i.e. ON-OFF, is -1.5. This means the team was more dominant when he DID NOT play than when he DID play. This is a fact, based on actual numbers, properly interpreted.
The 17-18 raptors were indeed an interesting team. Their starters played well but their bench blew the doors off teams. It's not surprising to me that 4 of the 5 starters (everyone but Siakam) was an overall net negative, while guys like Van Vleet and Onunoby were positive. Even Lowry was an overall net negative that year. The difference is that Lowry, across 13 seasons on 3 different teams with drastically differing circumstances, has only fallen into on-off negative territory twice - that 17/18 year and the year he was traded midseason. Derozan has mannaged to fall into that territory every season but 1.
No other good player has ever managed this. It's unlikely anyone will ever again, because (and I can't stress this enough) good players generally do well in this stat. You want your good players to play precisely BECAUSE they looks good in this stat, i.e. they make you better than the other team to a greater degree than you are if they don't play. Lebron has never been in the -ve, and averages +11.4 for his career. Too high a bar? Khris Middleton has never been a negative, and averages +7.2. Lowry averages +4.4. Marc averages +3.5. DANNY GREEN averages +4.3. Derozan? -3.0.
It's important to be able to understand the concept that a player being part of success when playing does not mean he isn't negatively affecting things. The only way to judge that is to take the player away and see if things get better or worse, and to do that over a large sample to help account for all the other factors that can affect it. One game isn't enough. One season isn't enough. 11 seasons is enough.