Camden wrote:Q, why wouldn't you use a stat like BPM that accounts for the quality of team these players are on and how much better or worse a player is compared to what's average? In my opinion, it's a better indicator of how good or bad a player is in comparison to his peers regardless of situation. And like your system it's easy to differentiate between superstars, stars, role players, bench players, etc.
+10.0 is an all-time season (think peak Jordan or LeBron)
+8.0 is an MVP season (think peak Dirk or peak Shaq)
+6.0 is an all-NBA season
+4.0 is in all-star consideration
+2.0 is a good starter
+0.0 is a decent starter or solid 6th man
-2.0 is a bench player (this is also defined as "replacement level")
Below -2.0 are many end-of-bench players
Edit: What BPM does not account for is playing time. I feel like that's also important to note.
The problem with BPM is mostly the same one you get with PER. It doesn't measure defensive impact adequately. It also can miss things on offense, like screen-setting, movement without the ball, spacing, etc. Anything not in the individual box score is not captured.
That's why On/Off and +/- are important, along with their more complicated derivatives. It looks beyond the box score and to overall team performance while the player is on or off the court. It's also more noisy than box score stats which is why I like to look at it over multiple years and thousands of minutes.
It should be no coincidence that the LeBrons, Currys, and Jokics do incredibly well in both types of stats. Generally speaking, great box score performers also positively impact team success.
I disagree with that primarily because PER does not include team performance into its equation whereas BPM does. Those little important things that don't show up in the box score are included regardless of it not having a specific numerical value. It's accounted for in Team Adjusted Efficiency per 100 possessions which is then used to find BPM.
On/Off and +/- are heavily influenced by quality of teammates and other factors that players can't control -- so much so that the values can be misleading. Take Anthony Davis for example. We'd all agree he's a top-five player in the league -- top-10 if we're being safe. Does anyone actually think the Lakers were better with him off the court this year? No shot, right? Well, that's what his -3.0 On/Off would indicate.
Meanwhile, his 8.0 BPM plants him sixth in the entire league -- accurately depicting his value on the court.
Stats are a matter of preference and there's no perfect one to be used. They can be used to debate whichever side of the argument depending on the person. I get all of that. To me, BPM is a much better indicator of a player's impact.
Cam, I don't think the team metrics used in the BPM calc includes actual opponent results. It uses purely the collective box score results of the player and the team he's on. It consistently undervalues guys who are strong positional defenders along with other more nuanced things. I think a guy like Paul Milsap for example is seriously undervalued by BPM.
On AD, I think you point out one of the weaknesses in the Net On/Off stat, which is that unique circumstances can drive some outlier years. In AD's case, I think one of the reasons his was negative this past year was because LeBron was almost always on the court when AD wasn't. I think you'll agree that LeBron + a group of league average players is better than AD + a group of league average players. And the times they were on the court together, they were typically playing against the opponents' best 5-man unit.
But this is why I looked at career Net On/Off. And AD does quite well, as one would expect.
apollotsg wrote:This might be one of the better arguments we have had - please keep going because I have changed my mind several times now.
LOL.
I'm not sure I am arguing that this is the "one stat to rule them all". I just think it needs to be in the mix when evaluating players.
When someone like Zach LaVine - year after year after year - is in a situation where his team performs better with him on the bench, I'm not sure how that can be easily explained away as totally driven by team mates. How'd he get so damn unlucky that he's either backed up by a superstar or his fellow starters are so bad that the bench makes they and him look bad? A season or two, sure, but not 5 out of 6 with two different teams.
OMG, how could I forget DeMar DeRozan!? He is definitely on the high volume scorer/low impact list and wins my Lifetime Achievement Award for this category.
11 seasons. He's had ONE where his team didn't perform better with him on the bench. In fact, the Spurs have never recovered from that Kawhi Leonard for DeMar DeRozan swap. They were winning 60+ games and titles and are now a sub-.500 team.
Q has persuaded me to give more credence to the net on/off stat that I have in the past. I think the key, as Q explained, is to focus solely on a player's CAREER on/off numbers over a period of years. Otherwise, there is lot of noise related to who a player played or didn't play with, etc that doesn't get adequately filtered in a smaller sample size of one or two years.
Like Cam, I've always put more stock in the BPM. I definitely think that's a better metric for evaluating a player's performance in a particular season or a young player who hasn't been in the League very long. For example, in evaluating a young player with only 1, 2 or even 3 years of NBA experience, I'd give more weight to a player's most recent BPM than I would his career net on/off. It gives you better sense of that player's individual contribution and how his development is trending over his first few seasons.
I'm persuaded now that net on/of provides some fairly reliable insight into the value of players like DLO and KAT who've logged five seasons in the NBA. But in DLO's case, I'd give more weight to his BPM the past few seasons. I think his situation in LA was so messed up that I tend not to consider those seasons in evaluating his numbers. So in my mind, he's more of a 3-year NBA player whose value to the Wolves is best measured with the BPM. But his net on/off is troubling and it's one of the reasons I had some misgivings about the trade that brought him here.
Of course trends can change, especially for guys still in early in their careers. Both Ingram and Randle (guys on my volume scorer/low impact list) are trending in the right direction if you look at their last couple of seasons. I'd argue DLO is too if you throw out his Golden State stint and included his Brooklyn and brief Wolves stint.
It's an insightful metric that I believe should be used as part of a comprehensive evaluation of veteran players.
