APBRmetrics

MVP: Westbrook
DPoY: Draymond
Executive: Myers
CoY: D'Antoni
MIP: Giannis
ROY: Brogdon
6MoY: Gordon

Anyone have any issues with these? Feels like the voters did a pretty good job this year, no glaring issues to me.

Among players starting less than half their games, Eric Gordon was 24th in the league in Win Shares -- and 4th on his own team! Lou Williams, Montrezl Harrell, and Nene` had more than 3.8.
With BPM of -1.4, his VORP was not top 50. PER of 13.1 was 60th.

But he did have highest Pts/G (16.2) of any non-starter for a good team; other than Lou Williams, who mostly played for LAL.

* * * *
Are these awards now meant to include postseason performance? Or was voting done earlier?

Mike G wrote:Among players starting less than half their games, Eric Gordon was 24th in the league -- and 4th on his own team! Lou Williams, Montrezl Harrell, and Nene` had more than 3.8.
With BPM of -1.4, his VORP was not top 50. PER of 13.1 was 60th.

But he did have highest Pts/G (16.2) of any non-starter for a good team; other than Lou Williams, who mostly played for LAL.

* * * *
Are these awards now meant to include postseason performance? Or was voting done earlier?

Regular season only.

Using the NPI RAPM J.E. posted for 2001-16 and my personal one for 2017, Westbrook (13th) is only the third MVP to finish outside the top 10 since 2001. The others were Iverson (44th) and Rose (23rd).

In 2017 regular season only RAPM Curry was the leader at +6.98, a full 1.02 points better than LeBron.

RAPM was a little more friendly to Eric Gordon. He was 17th among players who started less than half of their games. If you weigh the ratings by minutes played, he was 7th among those players.

Gobert led in DRAPM at +3.55. Draymond was third at +2.86.

Avery Bradley getting 12 first place votes for first team All-Defense is quite laughable. I know there are some limitations to DRAPM/DRPM, but Bradley was 417th in DRPM and 468th in DRAPM. So you can't even make the argument that his height or box score stats were dragging down his DRPM, because he ranked better in that metric. Those metrics aren't perfect, but I don't believe they are off by the several orders of magnitude necessary to consider Bradley one of the elite defenders this season. According to NBA.com tracking data, he was only in the 19th percentile in defending isolation plays, precisely the type of plays a lock down defender is supposed to excel against. Also, players shot 72% when Bradley was 6 feet or closer to them, which is in the bottom 50 of the league among players who played a reasonable number of games. This is a classic case of a guy benefiting from a reputation he earned from the one year he actually finished top 60 in DRAPM.

I know Bradley suffered this year from weird on-off issues with defense (like Kawhi), but no stats say he was even an elite one on one defender. I get respecting players opinions, but I know players overvalue one on one performance more than analytics say its worth. A single stand out game for AB while he's guarding one of these guys is enough for them to always say he's elite.

If I remember right, the main factor behind Kawhi's weird on/off splits this year was due to the opponent's 3P%. When Kawhi was on the floor opponents shot 40.3% on 3s and when he was off they shot 37.1% (lowest percentage for any Spur while off the court). The same doesn't hold true for Bradley. When Bradley was on the floor opponents shot 36.3% on 3s and when he was off they shot 35.6%. Kawhi's differential was 4.6 times as large as Bradley's.

JaVale and David Lee sporting (pretty sure) career high RAPM ratings.

Just shows how good Kerr and Pop are (and the surrounding talent) at mitigating a player's weaknesses.

JaVale only had a positive single season RAPM once in his career before this year, so +3.88 was by far a career high.

David Lee had a slightly higher single season RAPM in 2014 (2.74) and 2007 (2.85) than this year (2.28).

I'm impressed they voted Brogdon over Saric

Covington finishing 20th in points for All-Defense was disappointing. I would have put him over Giannis at least. Besides Covington and Embiid (who only played 31 games), practically everyone on the Sixers was a minus on defense, yet they still managed to finish middle of the pack defensively.

shadow wrote:Using the NPI RAPM J.E. posted for 2001-16 and my personal one for 2017, Westbrook (13th) is only the third MVP to finish outside the top 10 since 2001. The others were Iverson (44th) and Rose (23rd).

In 2017 regular season only RAPM Curry was the leader at +6.98, a full 1.02 points better than LeBron.

According to RPM (at espn), and including playoffs, the Wins leaders for this year are:Given that LeBron and Curry in playoffs boosted their RPM quite a bit, it's likely that Westbrook led the RS in RPM-Wins. Unless Butler did.

I don't know the exact formula used to calculate ESPN's RPM wins. However, the formula I typically use seems to be a close approximation and using the RS RPM ratings from the table linked below, LeBron still led the league in RS RPM wins with 17.34. Westbrook had 14.95. LeBron played in 7 fewer games than Westbrook in the regular season, but only played about 5 total minutes less since he averaged 3.2 more MPG.

https://public.tableau.com/profile/knar ... in1metrics

LeBron still led the league in RS RPM wins with 17.34. Westbrook had 14.95.

If those are correct, then LeBron added (20.43-17.34) 3.09 RPM wins in 18 playoff games, while Westbrook gained (17.34-14.95) 2.39 in just 5 games?

The knarsu table -- it's RS only, right? -- doesn't seem to have RPM wins, but these other things resemble some kind of total values added:One guy or the other led the league in all these but Win Shares (Harden 15.0, Gobert 14.3 ...)

Yeah, it appears the knarsu table is RS only for RPM. ESPN previously called the RPM wins column WAR, so maybe that's where WAR in the table is coming from? It says the stats come from BBR, ESPN and NBA.com and I'm not aware of a WAR metric at BBR or NBA.com so I guess it must be coming from ESPN.

I don't think my numbers are 100% correct, but logic dictates that LeBron had the higher RS RPM Wins because he had the higher RS RPM and they played approximately the same number of total minutes.