Melo a GAME HIGH +27 vs Memphis and a game high +22 vs Brooklyn. Kanter a +20 tonight. And yet there are many on the game threads that constantly shit on these guys. Does anyone know what their +/- is together on the season? I can’t imagine it’s that bad.
It's a pairing that can feast at times on bad teams, or for spurts vs good matchups. Problem is is that pairing isn't being used as selectively as it should be. Still think they both present big issues for this team in the playoffs.
What’s great about Stotts is that he finds ways to get winning basketball out of them. No one wanted Melo. Kanter didn’t play for Boston. But Stotts takes what’s great about those players and uses it to win games. The 2018-19 Rockets absolutely could’ve had use for Melo vs KD-less Warriors but they gave up on him 10 games into the season. Last year’s Celtics could’ve used Kanter’s rebounding. This is why Stotts is a better coach than pretenders like D’Antoni and Stevens. He gets the most out of the roster in front of him.
The problem with Melo and Kanter is they give up as much points as they give us on the other end. I do love the hell out of Kanter's work ethic. The way he eats offensive rebounds for breakfast. Sometimes he plays really good defense, blocking the shit out of opponents and what not, but he often leaves his man to help to our detriment or fouls at the basket too much. I love the guy though. I enjoy when Melo is hot and hitting his shots. His ISO ball, refusal to pass alot of the time, and lack of effort on defense frustrate me.
But if they're giving it up on the other end, how is Melo a +47 in the last two games? I think his D and his passing is underrated.
Last ten games (including last night): You'll see that ACTUALLY the Blazers are best when Anfernee and NOBODY ELSE is on the floor. (That big Derrick Jones signing is looking INSPIRED.)
+/- stats for a single game may be the dumbest stat out there. I think even 10 game sample sizes are too small, although that may at least point to trends for the season: Kanter +5.4 Simons -3.0 Melo -5.6 (worst among rotational players) Leaf -75.0 (couldn't resist that one. You wouldn't think that's possible in 13 minutes) at that rate, if Leaf played 30 minutes, Portland would lose by 173 points....RHOF
Full season sample sizes are too small. While the full-season stat usually works for most players, you will have a sizable % of players that are way overrated or way underrated. That's why RPM stats using a single season is problematic. Good players will generally float to the top but you will have spattered in there shitty scrubs ranked higher than Lebron. Plus it doesn't take into account your opponents or team-mates. Starters play vs Starters etc.
Melo is +47 in the last two games and -131 in the rest of his games. Variance. There are 9 other players on the floor that can get hot or cold while Melo is in.
I don't disagree, especially on the rotation factor for instance Dame has a bad defensive rating and a bad DBPM. Some of that is on him, but a lot of it is due to a consistent rotational pattern. About 6 minutes into the 1st and 3rd quarters, the bench has been filtering into the games. By and large, Dame plays the entire quarters, but for nearly half the time he plays with the combo of Kanter and Melo. And often it has also been with Simons getting minutes. Those have been the 3 worst defenders on the team and that will drag down anybody's defensive numbers. This season he has both the worst defensive rating of his career, and the worst DBPM since his rookie season. That's not him somehow playing worse defense than his previous 8 seasons. That's due to who he has been playing with....IMO
Actually Dame hasn't been playing good defense who ever he on the floor with and yes he makes a good play here and there but overall his defense hasn't been good this and don't needs stats to show me that just watching the game tells me that.
My understanding of DBPM is that it does not track the defensive PPP of the team while he’s on the floor vs off the floor. It uses the team’s overall defensive rating in conjunction with a the player’s weighted box score stats. And it rates each player on the team in a way such that the combined ratings weighted by minutes matches up with the team’s overall defensive rating. So if the team has a bad defensive rating it will drag each player’s DBPM down but it doesn’t matter what happens while he’s on the floor vs off the floor. That’s my understanding anyway. I don’t put to much stock in defensive ratings. Box scores miss too much. Plus minus captures almost everything but is high variance so it and too much data... though I do look at that over multiple seasons.
I agree that defensive stats can have lots of noise. And I'd say that if a single defensive stat is the opposite of what your eyeballs see you can probably be allright trusting your eyeballs. But if 3 or 4 or 5 defensive metrics agree and are the opposite of what your eyeballs say, you might need to adjust your eyeballs, at least by a little as far as defensive rating, the only fashion I will put much weight in individual defensive ratings is when I gauge a player compared to his team. If he has a better rating than his team than I'd say the chances are he's above average...for his team. Opposite is true for worse ratings. But even then, as you've mentioned, there is still plenty of context to consider like average minutes, starter vs bench, and rotation patterns
