Here is my long owed fan post to explain why I remain skeptical about Murray as a PG, though it became more of a general look into how the PG's compare. If it needs to be made clear: this is not an attempt to discredit Murray or boost any of the other PG's - it's an attempt to use the tools we have available based on last season to understand what each guy brought to the table.

Some groundwork to understand the scope of the stats and why I've broken them out the way I have:

• There are various arguments that have been made to me to the effect that Murray didn't actually get to play PG until the last 5 games of the season. This is hogwash, IMO, and the first part of the post will be to explain why I think that. It's ultimately pointless to look at any of this stuff if you don't buy that Murray played PG when it was he and Barton together; I think that the stats clearly show that he did. It's important to remember that no PG or primary ballhandler takes all of the "PG" duties, this side of extreme cases like Harden and Westbrook.
• Based on the above, there is one date that plays prominently - 1/21 is the day that Mudiay went down with injury and permanently lost his spot in the rotation. The game on 1/22 was the first where Murray was given PG duties, and that's the cutoff I am using for Murray's numbers.
• The sources I've used are primarily nba.com/stats for tracking data and NBAWOWY for lineup data, which plays prominently in the second half of the analysis, where I try to dig into how the players fit with Jokic. Notably for the nba.com data, they don't provide per minute numbers on the tracking stuff, so I've had to calculate them out. I've chosen to use per 36, as I think it makes it numbers easier to understand in a "per game" type of feel.
• Ideally I would have numbers for a league average for much of this stuff, but with no easy way to do that I've included Chris Paul as an example of a ball dominant PG, to contrast Denver's situation to the rest of the league.

Section 1: Why I believe it's fair to use Murray's numbers post Mudiay injury to evaluate his work as a PG

Far too often I found any criticism I gave out to Murray rebutted with "well he's not really getting to play PG, because Barton handles the ball". While it's not an argument completely without merit, I find it falls flat pretty quickly once you dig into the numbers a bit. Barton certainly saw his fair share of opportunities on ball, and he took over more next to Murray than he had next to Nelson or Mudiay, but the net was still that Murray was functioning as the PG with Barton functioning as an on ball wing.

The following table is based on the touches tracking data from NBA.com, and shows a pretty massive change from 1/22 for Murray.

Murray's profile prior to Mudiay going down had been fairly standard off ball. Low time per touch, low time of possession, not a ton of front court touches. Barton had pretty clearly been more heavily featured on ball to that point - more dribbles per touch, more touches per minute, more time of possession. After Mudiay, both guys saw notable increase in touches and time of possession across the board, but Murray's numbers vaulted ahead of Barton's across the board - more dribbles per touch, more time per touch, more touches, more time of possession. So while it's fair to say that Barton took a portion of the responsibility, it also seems fair to say that Murray became the primary ballhandler between he and Barton. The question then becomes: how does that compare to Nelson and Mudiay, and the rest of the league?

Some things become pretty clear from this. First, Denver PG's handle and hold the ball considerably less than Chris Paul, which is expected. Second, within the grouping of Denver primary ballhandlers, Mudiay and Nelson were given more responsibility than Murray, who did share the role with Barton - but Murray still had the primary role between he and Barton. A couple other interesting tidbits here: Nelson was the most prone to ball pounding, as shown by his higher seconds/touch and dribble/touch numbers, but not by a large amount. Mudiay commanded the ball more than any of the others by a fair bit, but didn't hold it very long for a PG. Murray and Barton both maintained high points per touch numbers in that role, which may indicate that they look to score more than "traditional" PG's.

I believe it's pretty clearly fair to compare these guys, with the understanding that Barton was less of a traditional PG than Murray who was less of a traditional PG than Nelson and Mudiay who were less ball dominant than someone like Paul. Which brings us to...

Section 2: The basic stats

I've included standard stats per 36 with minutes (remember, I'm using post 1/22 data for Murray and Barton) as well as some relevant advanced numbers in the next table.

Advanced:

First major take away - Chris Paul is in another stratosphere than anyone we could put on the court. While we knew this, seeing it broken out statistically like that is still fairly staggering.

Among the Denver options, Jameer's numbers stick out as the best in point guard type of categories - assist rate above 25, A/TO ratio over 3. Barton remained a score first guy, but did a good job creating from that threat with over 5 assists per 36 and an A/TO ratio over 2; unfortunately his overall effectiveness in the second half of the season was undermined by an inability to make shots, as he saw his TS drop to 52.8, a pretty rough number for him (to contrast, he sat at 56.6 before 1/22). Murray's numbers show some good and some bad. He may already be the best scorer in the group - his 54.5 TS% edges Nelson for the top mark, and he did it on relatively high usage at 23%, leading to a respectable 18.5 PP36 - just a hair under Paul's number. His NET was the best in the group, something I'll touch on in more detail when I break out some lineup data. On the flip side, his assist numbers show something of a ghastly tale - 15.5% assist rate is really low for someone who played PG, even in a partial capacity. He does take care of the ball -almost a full turnover less than Mudiay per 36 - but the scarcity of assists keeps his A/TO number under Mudiay's anyway. Mudiay's overall numbers were a train wreck across the board - most notably a NET of -5.5 and a TS below 50, while not exceeding Nelson's fairly modest assist numbers despite the higher usage.

Section 3: Deeper dive on passing and creating

NBA.com tracking has some very interesting stuff on passing and driving which lets us dig a little deeper than assist rate.

Here, I've taken the NBA tracking data, which provides secondary and potential assist numbers, and laid them out on a rate basis to attempt to get a picture of how much playmaking someone is doing. It actually paints a slightly better picture for Murray. He still lags significantly compared to Barton and Jameer, who both create at a pretty good rate -roughly 1 assist opportunity every 4 passes. But this shows that Murray's rate of assisting is actually fairly close to Mudiay's when adjusted to passes made, meaning that Mudiay's advantage in assist rate is as much volume as it is effectiveness at this stage. Neither guy did a good job of consistently setting up teammates given the opportunities afforded them, which makes Murray's superior scoring ability stand out more.

Of course, the real competition is probably Nelson and Barton, and there the story isn't so rosy for Murray. Taking Barton specifically, you have a guy - labeled as selfish by many - who created more opportunities for his teammates in considerably fewer touches, time of possession, and passes than Murray did. The gap that Murray - and, for that matter, Mudiay - need to close is with someone like Barton, who actually forces the defense to play him which results in the passes he makes turning into scoring opportunities for his teammates. Neither of the young PG's have been able to do that with any consistency to date. In Murray's case, that evolution is probably necessary for him to become a star as opposed to a nice roleplayer, while in Mudiay's case it could be the difference between a long career as a starter and being out of the league in 4 years.

The next table is tracking drives. NBA.com considers a drive anything in a non fast-break situation where a ball handler dribbles from the perimeter to within 10 feet of the hoop. Note that PF% is the amount the player is fouled, not charges.

Here is one of the couple of things that give me hope for Mudiay. He gets into the paint at an elite rate, and despite relative struggles once he gets there, that ability still produces fouls and points at a high rate. Given his struggles finishing, you'd like to see him pass more and shoot less, but in total this is an area where Mudiay has a clear advantage on the field. Nelson drove more than I would have guessed and passed a majority of the time when he did, leading to a high assist rate but virtually no fouls and not many points. Barton was unsurprisingly a highly productive scorer on a reasonable number of drives, with fairly minimal passing. Murray rarely drove, and rarely passed once he did drive, but was effective scoring there for a rookie, and drew a good amount of fouls. Paul is an interesting case -he drove almost as rarely as Murray, but when he got there something good almost always happened.

Lots of interesting numbers above, but the question will fairly be asked - with Jokic as the point center, do we actually care how effective our PG is as a distributor? That brings us to...

Section 4 - Lineup data

For this section I used NBAWOWY, a tool that lets you pull lineup data with any players on or off. For the main section, I'm looking at numbers with Jokic and then numbers without Jokic, with each of Nelson, Mudiay and Murray as the primary PG. Barton is excluded because he played essentially 100% of his minutes with one of the aforementioned guys. To determine who was the PG, I basically said Nelson took precedence over Mudiay who took precedence over Murray - so for Nelson, him being on was enough to consider him the PG. For Mudiay Nelson had to be off, and for Murray both of the other two had to be off.

Jokic on:

Jokic off:

Fun stuff! The first thing that jumps out is that Murray is the clear NET Rating winner, pulling the best number both with and without Jokic - largely driven by better defensive numbers, but also notably by not letting the offense tank when Jokic sat. Jokic was good regardless, but surprisingly quite a bit worse with Nelson on the floor from an efficiency standpoint - perhaps an effect of not having a PG that threatens to cut making him easier to guard? Next surprise - the best offensive configuration was Mudiay/Jokic. Mudiay without Jokic, however, was an incredible disaster.

Now let's split out a couple of other interesting factors. First, this lets us see how Murray did in the limited opportunities to truly be the single lead ballhandler, without Barton sapping some of the possessions:

First off, Murray's numbers without Barton but with Jokic comprise most of his with Jokic sample - roughly 80% of it -and yet the NET jumps by 4, which is a pretty big deal. The numbers without Barton or Jokic are a little more concerning - small sample size, but his efficiency tanked while his usage and assist rate spiked. The team still did well in those minutes, but you're not going to be able to keep that up long term with a guy shooting under 50% TS on almost 26% usage. Still, overall the numbers here are encouraging for Murray's future as a lead guard, even if he's more likely to be usage and scoring heavy than people like to admit. When forced into playmaking by being essentially the only option on the floor, his playmaking jumped enough to sustain solid offensive production - very encouraging.

Changing gears: for Mudiay specifically, the numbers are heavily impacted by the disaster that was Jusuf Nurkic in a Nuggets jersey last year (and, to be fair, Mudiay's own inability to work with Nurk on the court). The Jokic/Nurkic lineup got 224 possessions; Mudiay was the PG for all but 32 of those, and on the court with Nelson for another handful. Those minutes were truly awful for everyone, and Mudiay/Nurkic lineups were quite bad in their own right -most of those chalked up when Malone made the ill-fated decision to bring Jokic on the bench.

Mudiay lineups were consistently bad defensively, but the offensive production shifts greatly. Remove the 192 possessions of Jurkic from his Jokic sample, and all of the sudden Mudiay + Jokic looks like the best offensive pairing by a mile, and one of the best combo's we've seen, period - trailing only Murray/Jokic with no Barton. Lineups were still blitzed with no Jokic or Nurkic, but the removing Nurkic from the equation takes Mudiay's individual numbers from abhorrent to fairly mediocre. I think it's fair to remain concerned that a player who is only good with Jokic just isn't that useful - but it's also fair to be intrigued by how Mudiay's presence seemed to turn Jokic centric lineups to their highest offensive gear. That combined with his unique (among Denver options) ability to get into the paint, and I have a hard time giving up on Mudiay at this stage, even given all the red flags that show up both on tape and in these numbers.

Section 5: Defense

Quick and dirty here, but the lineup data certainly screams "Jamal Murray is the best defender of the bunch by a mile", so it's worth seeing if other metrics agree, or not.

Short answer is that metrics generally thought all of them sucked. That's probably a fair point overall, but given the magnitude of the difference in lineups I would like to look at it deeper, using stuff like diff and breaking down the lineups further. If I wait to do that, this will never get posted, so...

Section 6: Conclusions

• If you were picking a guy to play next to Jokic solely on their context independent stat line, I don't think there's any question that you'd choose Nelson. Low usage PG who distributes reasonably well, doesn't dominate the ball, and scores with reasonable efficiency when he does shoot. He's not a slam dunk option by any means, but from a fit perspective that is the best profile.
• That said, we have context, so we can make smarter decisions! Murray lineups were better across the board despite him being a rookie and having relatively little high level experience as a PG, he is already the best scorer in the group, and it appears that he has a positive impact defensively, which no one else can say. Combine that with his incredible work ethic and a stroke that screams "40% high volume 3pt shooter", and Murray is your clear, no brainer choice to enter the season as the starting PG.
• Mudiay should be given a fair shake at the job, as well. He's been the victim of outsized expectations and responsibility, impossible lineup situations and questionable coaching - combined with a shaky shot, arguably poor offseason focus, and piss poor defensive work and he's understandably soured on many people. Despite that, there are some underlying things that suggest he retains a high ceiling, and I believe it's in the teams best interest to continue to work him into the rotation - potentially even as an open competition for the starting role with Murray.
• Murray's the best internal choice for the job, but that doesn't mean that his lackluster assist numbers aren't a concern, and something worth monitoring. Teams need multiple playmakers on the floor, and while Jokic means that the PG doesn't have to be the primary, he may have to be the secondary. Hopefully it's something he grows into. Currently he's a lot more like Kyrie than people are willing to admit - drives to score, adept at creating for himself but inconsistent creating for others. His positive NET ratings are largely a result of DRTG improving, not ORTG, so there's some indication that the fit there isn't as good as people tend to believe, and I wouldn't be shocked if he played himself out of the role by the end of the season.
• I'm surprisingly OK with going into the year with Murray as your high ceiling 2nd year starter, Mudiay in a reserve role and Nelson in an emergency role - I just have to hope that's how Malone plays it. If you were going at it from the standpoint of purely maximizing lineups, I think you would start Mudiay, taking advantage of the trend of Murray based, non Jokic lineups treading water; but given the long view, Murray should be starting even if it's non-ideal for the bench.