APBRmetrics

[Nikos]

What happened to this guy after the 01-02 season? His PER dropped tremendously, and with an exception of a very solid 2003-04 season as a Nugget, he really has been just a very good PG instead of an upper tier PG.

Watching him last night against the Pistons throw flawless alleyoops reminded me for some reason of when he was considered a potential elite PG as a Cavalier when his PER was around 20-22 in the early 00's.

Why isn't this guy quite as good as his potential showed as a Cav?

http://basketball-reference.com/players/m/millean02.html

[Mark]

I see trend down in 3pt shooting % and # made dropoff after LA, I guess cutting back quantity offsets % dropoff to some degree, but maybe not enough.

Overall number of shots down modestly too.

Moving from basically a one passing guard situation in 01-02 in Cleveland (no other player with much more than 2 assists a game) to the Clips and Denver (2-3 others near or above 3 assists) probably hurt his assists as well.

Rebounding down some too if you compare three year blocks Cavs vs later elsewhere

[gabefarkas]

Is he really doing so poorly, in the three most recent complete seasons (2002-03 through 2004-05), he ranks 4th in total assists, behind only Nash, Marbury, and Kidd, and ahead of notable passers like J-Will, Cassell, Parker, Bibby, Arenas, Billups, Baron Davis, and Jamaal Tinsley, to name a few.

He may not be having an earth-shattering career, but he's still one of the top passing PGs in the league.

[mathayus]

He's made a bit of a comeback so that's good, but I was among those very disappointed in what happened after he left Cleveland.

I know look at him and Brevin Knight as the case in point for the idea that a pass-first point guard's impact is not necessarily all that proportional to his assists. Remember if you will that Knight also racked up nice stats for Cleveland, before being replaced by Miller. Both times the team went nowhere with them, and when they went to other team the stats didn't follow them. Knight has since started accumulating large numbers of assists in Charlotte, but again it has shown no correlation with actually moving the team forward.

[tmansback]

Agreed. I think the system allowed Andre to rack up some great assist numbers. Playing on teams with more talent and more ballhandlers may have hurt his numbers. He still a very good PG and probably underrated around the league.

[mateo82]

The Cavs were one of the worst teams of the decade at the time. This is what happens when you look at numbers and don't compare them to what you know from your subjective perspectives.

I'm guessing you were also perplexed when Damon Stoudamire stopped being an elite PG after he left Toronto?

[mateo82]

[Nikos]

I don't see too many PG's on bad teams with seasons as good as Andre's in 2002.

A lot of analysts were also claiming he could be the next Jason Kidd or better, despite knowing how bad the Cavs were. He certainly was an excellent post up PG back then, and was utilized a lot in that area. He was just more efficient on offense.

Damon Stoudamire never was even close to posting a 20+ PER even on a bad team.

Is it really Dre's fault the Cavs were horrible underneath and had nothing else aside from a couple of solid spot up shooters, and one energetic role player in Rickey Davis? Whose to say Steve Nash of today would make that Cavs team playoff bound? Why are Miller's assist numbers assumed as being meaningless just because his team was not successful?

[Kevin Pelton]

[tmansback]

[Mark]

Last year the Nuggets had 6 guys averaging 2+ assists a game. 2 assists isnt really enough for the individual to boast about but having 6 is a widely distributed passing game (about half the teams have 6 at 2+) and in this case adds up to #2 in the league on team assists and a 2.5 assist advantage. (They played at a pace about 2% higher than league average enhancing assist total.)

This year the Nuggets had only 4 guys averaging 2+ assists and traded one away (but havent really missed him from an assist standpoint in last few games). For the season they are still 3rd highest on assists but have a half an assist deficit (They play at a pace about 4% higher than league average, higher than last year. Offense efficiency has slipped from 12th last year to 15th this year. Dips in 3 of 4 factors with offensive rebounding being the biggest. Maybe too much running for a team lead by Andre Miller as suggested earlier? They seem ok on the assistometer in both cases, the other stuff should get most of the focus there.

(Note: I didnt calculate how much the pace affect would pull down their standardized assist ranking.)


Assists and assists from top guy I now give less importance to than I once did. With only a modest portion of assists creating something expecially easy or something decent out of nothing I am more interested in utilization rates and quality of shooting /scoring of key players.

Dallas' shift from premier PG passer to a very widely spread passing game is a stark contrast (4th lowest on team assist this year compared to 3rd highest last year with Nash) and I think Cuban must have felt it was statistically doable with lots of high utilization guys who shoot/score well enough. I wonder if he actually felt it was preferable? And what additional data he had for thinking about that.

82 games had some good research on overall passing quality and team passing leaders and how much or little it / they matter.
http://www.82games.com/chempass.htm

The conclusion: "... it appears that an exceptional passer in a five man unit (as judged by the "placeholder" 82games current passing rating) is a very good thing, although surprisingly having a combined unit passing rating that is far above average does not seem optimal compared to one that is average/slightly above average."

Where does Dallas fit in this analysis? Using the stated method for ranking teams it looks like Dallas is lowest /weakest tier on overall 5 man passing. They are just barely high enough to get in the second lowest tier based on top passer. They are middle of the pack on second passer rating and have several with a modest level of passing achievement below that.

Somewhat surprising it works ( I had my doubts at the beginnning of the season) but for their personnel / system it does.
Overall the team impacts based on these three passing measures seem pretty small. Other things are probably much more important. But maybe top teams and playoff data should be studied separately too before dismissing this area of research as minor. I just skimmed the surface and found that the 4 conference finalists ranked between 4th and 9th out of 16 on raw assist totals and Detroit had a 5 man passing ranking that put them at the very top of the third best passing tier, while San Antonio was in the middle of the second lowest tier out of 6 tiers. Neither excelled on 5 man starter ratings and the weaker team won. Very surface look but doesnt seem very important here either at first glance.

[mathayus]

[bchaikin]

Why isn't this guy quite as good as his potential showed as a Cav?

Rebounding down some too if you compare three year blocks Cavs vs later elsewhere..

This is what happens when you look at numbers and don't compare them to what you know from your subjective perspectives.

but I was among those very disappointed in what happened after he left Cleveland.

Playing on teams with more talent and more ballhandlers may have hurt his numbers.

As a criticism of Miller, he's not anything spectacular in transition, which is where he could rack up assists in Denver.

Well, if he had been able to continue to have impressive stats when he was traded to another team with another system they clearly would be useful. Bottom line is if you only produce eye popping stats on a particular bad team, that doesn't mean a whole lot.

Miller decline in assist are probably more a result of not having the ball as much as he did when he played in Cleveland.

off all the above comments about miller the last is the most pertinent. below are his year by year stats:

tm--year--min/g-ScFG%-touch-%sh--%fld--%to--%pass--reb/48
cle--9900----26---50.7----1.80----20-----7-----4-------69-----6.4
cle--0001----35---52.9----1.90----18-----7-----5-------70-----6.1
cle--0102----37---52.6----2.22----16-----6-----4-------74-----6.0
lac--0203----36---48.4----1.58----20-----8-----4-------68-----5.2
den-0304----35---52.9----1.58----21-----8-----5-------66-----6.2
den-0405----35---53.2----1.65----19-----6-----5-------70-----5.7
den-0506----37---52.2----1.93----16-----6-----4-------74-----6.0

"ScFG%" is Scoring FG% (combining 2pters, 3pters, and FTs for overall shooting %)
"touch" is touches per minute
"%sh" is shots per 100 touches
"%fld" is times fouled per 100 touches
"%to" is turnovers per 100 touches
"%pass" is passes thrown per 100 touches

i see basically the same player from year to year in terms of what he did once he got the ball - same turnovers per touch, same amount of passing per touch (2/3 or slightly more of all touches resulting in a pass thrown, 3/4 in his highest touches/min seasons), and he's gotten fouled with 6%-8% of his touches. his shots per touch have ranged from 1 shot every 6 touches (16%) to 1 shot every 5 touches (20%)...

even his ScFG% has been the same for all but one season (his clipper year) other than his rookie campaign and his rebounding has been relatively constant (down slightly with the clippers)...

the key difference i see in andre miller's annual numbers are his touches per minute. they were huge in 01-02 in cleveland when he lead the league, getting more touches/min than PGs like kidd and stockton. touches/min of PGs are often dictated not just by the player himself but also by the touches/min of his teammates...

take a close look at his 01-02 and current 05-06 seasons - what he did per touch is virtually identical, and even his rebounding and ScFG% are virtually the same. the only difference was he got about 13% more touches/min in 01-02...

also note his 02-03 and 03-04 seasons - two different teams yet the same touches/min and what he did per touch also pretty much the same...