It'd take way too long to challenge you on every objection you just raised, but here's one that really boggles my mind: you really don't think Steph Curry is better than Rondo at this point?
Well, I pondered this question and all the related ones (Is Deron Williams better than Rondo? Is Kyrie Irving better than Rondo? Etc., etc.), input some numbers and came out with the following answer:
http://www.celticsblog.com/2014/12/17/7406937/ranking-point-guards-at-the-quarter-pole-an-analytic-approach
The short answer would be: If you ignore defense, then right now, yep, Steph Curry is probably just barely slightly better than Rondo right now. I have his net point impact per 36 at the top of the list at 39.2 points per 36. But Rondo's impact is only slightly behind in 3rd at 37.1.
This rating gives credit for points scored, assists and rebounds and penalizes for turnovers and missed shots.
Just keep in mind that this evaluation is being done with Rondo shooting _way_ below his career norms, to which is likely to eventually regress upwards to. That could add 2-4 points to his per-36 average. That alone might push Rondo's net point impact rating up above Curry's.
And seriously? Rondo's defense is definitely worth more than Curry's.
Here are the resulting ratings as of Monday night's games. See the link up above for the math.
Player NetPointImpact/36
Stephen Curry 39.2
Chris Paul 38.5
Rajon Rondo 37.1
John Wall 36.8
James Harden 35.2
Kyle Lowry 34.9
Ty Lawson 33.3
Jeff Teague 32.3
Damian Lillard 31.9
Dwyane Wade 31.7
M. Carter-Williams 31.2
Mike Conley 29.4
Eric Bledsoe 29.1
Jrue Holiday 28.5
Brandon Knight 28.1
Derrick Rose 27.4
Deron Williams 27.2
Tyreke Evans 27.1
Tony Parker 27.0
Brandon Jennings 26.6
Darren Collison 26.4
Kyrie Irving 26.3
Tony Wroten 26.2
Reggie Jackson 26.1
Mo Williams 26.0
Kemba Walker 24.2
Trey Burke 20.2
If I were to take defense into consideration (other than rebounds, which are already accounted for) I would probably put Paul #1, followed by Rondo and then Curry. Especially if you factor in strength of schedule (Rondo has played against a much harder schedule so far than either Paul or Curry - especially Curry. GSW has played the 4th easiest schedule so far.).
But even ignoring defense, it is pretty clear that as measured by raw production stats, Rondo is clearly still among the top handful of point guards in the NBA and most of the spew in this thread claiming he is "not even top 30" in the NBA is just hyperbole.
This was a pretty straightforward analysis and I suspect that Danny & Brad's evaluations are at least as sophisticated. I would bet their sense of Rondo's value is a LOT different than that of several of the bloggers on this board.
TP for the analysis.
Your choice of stats and allocation of weight to each stat is biased to Rondo's style of play in numerous ways, two of which I have the intellect to explain at 4:30 A.M.:
It penalizes other players for missed shots more so than Rondo is penalized for not taking enough shots to be considered as highly as most of those players are. We'll continue the Steph Curry theme of this discussion. Rondo takes, and therefore misses, far less shots than Steph Curry does and yet Rondo does not shoot any more efficiently than Curry does. It is my opinion that you get around this by measuring total missed shots vs. shot efficiency.
It credits players for the entirety of the total points scored off their assist, which is typically nearly double and in some cases triple what you weigh a single turnover. This sort of analysis presumes that without the assist the points would not be scored in the same direct fashion that it is presumed without a player taking and making a shot that the shot's resulting points would not otherwise be scored. I do not assess an assist to have the same point value as a made basket and I am assuming most other advanced metrics don't either, which is why Rondo is ranked so poorly, particularly on offense, by typical advanced metrics this season.
Your analysis pegs Rondo as a top 3 PG while most other statistical analysis suggests he's fighting for a top 20 spot at the position. I'm guessing his value is somewhere in the middle (I rank him around 12th in the league). I also find it hard to believe Danny and Brad are as high on Rondo as you seem to think when Danny has tried to trade Rondo numerous times (CP3, Steph Curry, Sac-Town offer last year that Rondo killed) and Brad has been steadily decreasing his minutes. Within two months of Rondo's return last season, Brad played him about 36 minutes a game to close out the season. In the month of November this average fell to about 32.5 mpg. Now in December Rondo is averaging about 30 mpg. That doesn't sound like a player valued as a top 3 positional talent by his GM and coach.
Yeah, because how are guys like Bradley or Thornton going to score in the exact same fashion off of a pin down or whatever unless Rondo gets them the ball exactly when they come off of said pick? I know that the definition of an assist has changed over time, but still, how would guys get a layup in transition off of a left handed bounce pass from half court if Rondo doesn't thread that needle? Give the man his due. He's somehow leading the league in assists on this craptastic excuse for a team. And why am I still up, lol?
That's not what is indicated in his on-court/off-court numbers in addition to other statistics. A scorer as dominant as Rondo is a passer rarely, if ever posts a net negative on-court/off-court offensive split. Yet the Celtics this year are scoring almost two more points per 100 possessions per 48 minutes when Rondo leaves the court according to 82games: http://www.82games.com/1415/14BOS1.HTM
Rondo also has -0.2 offensive win shares this offseason, which, to summarize, is not good. His offensive rating is 95, the lowest of his career.
On-Off is heavily affected by teammates and opponents and thus is a stat that doesn't "settle" enough be used as an individual stat without a huge sample size and even then, you have to pay close attention to context, such as whether the minutes are predominately as a starter or bench and against starters or bench.
Even the most casual glance at the 5-man units on the very page you looked at:
http://www.82games.com/1415/14BOS1.HTM#5manshould be enough to make it clear just how much impact on an individuals' plus/minus the make up of his teammates on the floor can be.
The strength of schedule is also a factor here that folks don't seem to appreciate. The Celtics played one of the top 4 toughest schedules through November. It has gotten much easier in December, but still, overall, they have played a much tougher schedule than say, GSW, which has played the 4th easiest schedule to this date. When you play a tougher schedule, the team's stats overall will be impacted, correct? That is, the team will give up more points and score fewer points than if they had a played an easier schedule. By definition, winning requires scoring more points than the opposition so that has to follow. Do you disagree with that assertion? Along with relative suppression of scoring, all related stats for the team will also be impacted, such as assists, rebounds, etc., etc. All this is backed up by the correlations you see in team standings and statistics. Teams that win tend to have better overall team stats than teams that lose.
Now, team stats are ultimately the aggregate of individual stats. If the team stats are negatively impacted by the strength of schedule, it follows directly that so too will be the stats of the individuals on the team.
Going back to the above 5-man units, the unit Rondo has spent the most time with is the RR+AB+JG+JS+Olynyk line up that we started the season with as our starters. That unit has a net rating of -.01 per possession. Not great, but not bad either. His next most used lineup seemingly just swaps in Zeller in place of Olynyk and *presto*, they have a net rating of +.15 per possession! That's a huge, dramatic swing!
If Rondo is so bad, how can one unit post such great numbers?
Now, the simplistic conclusion from that is that swapping Zeller for Olynyk resulted in the huge swing - but its not that simple.
Looking further down the list, we see another configuration that seemingly does nothing more than swap Zeller and Olynyk (with the foursome of RR+AB+JG+Bass) that has the complete _opposite_ result. In that case, it is the _Olynyk_ configuration which is awesome compared to the Zeller one!
The truth is that this is the result of our skewed schedule. The Olynyk starters configuration was used mostly during our brutal run through the very top teams in the NBA in November. Primarily against the
starters on the top teams in the NBA. The Zeller starters configuration has been mostly since then against MUCH easier competition. Similarly, the two 'Bass" configurations also were used inversely as they represent the bench big substitution overlap. In November, Bass+Zeller would be the two bench bigs overlapping with RR+AB+JG. In December that has changed to Bass+Olynyk - against much easier competition.
Once you are aware of these things it becomes apparent that regardless of Olynyk or Zeller, that the Celtic starters basically played very close to even with the starters of the top teams in the NBA and have generally crushed the starters of the crappy teams of the NBA. So, the net point of all this is that the stats you have cited are really, truly heavily skewed due to small sample size and the front-loaded schedule. They don't really inform you of what you seem to think they are informing you. You have to look into them carefully and consider the contexts.
This understanding of what is going on with the schedule and the lineups and the impact they have on the various statistics is, again, a pretty straightforward analysis that I'm pretty sure Danny and Brad are aware of and take into account in their evaluations.