Journeyman Point Guard trend

[Mike G]

The last few years, the same backup point guards keep turning up with new teams.
Here, ranked by total minutes, are players who have been with 6 teams in the last 5 seasons:

Code: Select all

6+ teams        Pos    MP      WS   WS/48   VORP
D.J. Augustin    1    8977    18.9   .101    1.6
Ramon Sessions   1    8579    15.2   .085     .3
Mo Williams      1    8143    10.2   .060     .8
Shaun Livingston 1    6980    13.2   .091    1.4
Nate Robinson    1    5726    11.8   .099    2.8
Aaron Brooks     1    5334     6.5   .058    -.4
Toney Douglas    1    4657     7.0   .072    1.5
Steve Novak      3    3546    10.5   .142    2.0
Ish Smith        1    2667      .8   .014   -1.4
Louis Amundson   4    2651     2.8   .051    -.5
. total        1.26  57260    96.9   .081    8.1

These players, traded or signed more than once per year are indeed dominated by PG.
Ish and Amundson have been with 8 teams in 5 years, near as I can tell.
The proper term may be "stops", rather than "teams". Some of these have returned to a former team, and that counts twice.

If a player is listed in 2 positions, it's usually more one than the other. I've given these guys a fractional position number -- 1.3 for a PG/SG, 1.7 for SG/PG, etc.

Among players with exactly 5 teams in 5 years, the minutes-weighted position is 2.75 -- probably not much below the league norm for the interval, what with smaller lineups being the trend.
5-team PG's include Ray Felton, Jerryd Bayless; plus PG-SG Jarrett Jack and Beno Udrih.

Among players with 4 teams in 5 years, the position avg is 2.25
PG include Collison, Calderon, Nelson, Ridnour, A Miller, D Harris, Lin; plus PG-SG Dragic and Vasquez.

The 4+ teams-in-5-yrs position avg is 2.32
PG and PG-SG number 27 of the 74 players, or 36% of the sample.

[bchaikin]

ish smith in college (4 years, 3668 min, wake forest) shot awful - 45.6% on 2s, 30.1% on 3s, 48.4% on FTs - and couldn't draw a foul (just 3.0 FTA/40min). not surprisingly he wasn't drafted...

in the nba from the ages of 22-26, among the 261 different PGs to have played 2000+ minutes since 1977-78, he has the 2nd lowest/worst overall shooting - just a 41.7% ScFG% - 41.2% on 2s, 24.0% on 3s, 57.5% on FTs...

despite this the sixers recently picked him up - the team that traded away michael carter-williams, who for the sixers was shooting 41.2% on 2s, 25.6% on 3s, and 64.3% on FTs...

go figure...

[Mike G]

I assume such a player remains in and around the league in part due to the trend toward small-ball: Spreading the floor, moving the ball, etc. Success requires having at least one distributor on the floor at all times.

This means having at least 3 PG's on your roster. If you have only 2, you are on thin ice. And a 4th is ideal, even if he's an inefficient scorer and generally unproductive.

A concurrent 'trend' (maybe) is a tendency to stockpile PG's to the point that you don't derive full benefit from any of them. The Suns and the Mavs started the year with multiple starting-quality PG's. They've let some go, but they've been replaced by more PG's!

This strategy might be great in a smaller league -- If there were 8 teams, you could gain an advantage by just depriving opponents of a vital position, by acquiring them all.

PG and PG-SG number 27 of the 74 players, or 36% of the sample.

The 27 include Barbosa and Jason Terry, who may be more SG/PG

These 27 played 42% of the minutes, with 43% of the Win Shares, of the 74 most-traded players of the last 5 years.
The remaining 64% (non PG) of players therefore avg 16% per position.
The remaining 58% of minutes were represented by 14 or 15% per position.

Relative to other positions, PG are 2 to 3 times as likely to be traded?

[Crow]

Of the 8 PGs on your 6 plus teams in 5 years, what where the win% for their teams before, during and after their tenure and from 82games what was the win% specifically for their minutes on the court? Was the acquisition rewarded or not? Was the replacement rewarded or not? What are the demographic characteristics of the PGs who had those minutes before and after (age, contract, minute experience, height, speed, "point guard-ness")? What were the demographics of the other PGs before, during and after? There are more details that could reveal more about what is going on. How does avg. yrs of contract compare between PGs, wings and bigs? For just backups? Is it an oversupply of possible PGs issue deflating contract length or an undersupply of PGs that coaches trust to run the offense relatively team first issue? Do diva starting PGs have bad relationships with these guys, the reverse or both? Are these guys constantly on hunt for more minutes and money and just not accepting the market conditions?

[Mike G]

Maybe a disappointing season by a backup PG tends to lead to his dismissal; and then a very successful season leads to irritation with the starting PG. It may be a position where minutes are seldom doubled-up; and any encroachment by the backup is to the detriment of the starter.

[Mike G]

Mo Williams is now at his 7th stop in 6 years.
Nate Robinson has gotten 23 minutes in NO, his 8th place in 6 yrs.
Ish Smith, at his 10th stop in his 6 seasons, is well on his way to being the most traveled player ever.

Code: Select all

2011  Hou, Mem
2012  GSW, Orl
2013  Orl, Mil
2014  Phx
2015  OKC, Phl
2016  NOP, Phl

He's averaged 430 minutes per stop.

[Crow]

Looking at DJ Augustin, his metrics would justify being let go every time except with Chicago. Hinrich kept the backup spot, being more stylistically compatible with Thibeau.

Sessions probably should have been retained in up to 5 cases where he moved on. Will the Wiz keep him? They may decide he is too expensive even though he was done very well there and others on team are worse performers and values.

[Mike G]

Looking at DJ Augustin, his metrics would justify being let go every time except with Chicago. ...

Last year, of players with >500 minutes, 74 are listed as primarily PG.
Ranking these by several metrics, we can suppose the top 30 should qualify as 'starters' and the next 30 as 'backups'.
The median of starters should be ranked 15 or 16, and median backup should be #45 or 46.

Code: Select all

PG rk   RPM   BPM  WS/48  e484   PER
15/16   1.9   1.5  .125   1.37   17.6
45/46  -2.1  -1.5  .065    .78   12.8

D.J. career  -1.5  .101    .90   14.0
2015   -4.0  -2.5  .083    .79   13.9									

Don't know Augustin's career RPM (if there is one). By these other metrics, he's been an avg or better backup PG.
It does seem as though RPM doesn't value him nearly as well; BPM picks up some of the disparity.

[Crow]

I was looking at season level data.

[J.E.]

15.5-year RAPM has Augustin at -1.6, ranked 1,047th of 1,521

Single year RPM's, starting in 2009, are
-2, -1.9, -2.1, -3.1, -4, -0.8, -3.9, and, for 15/16, -3

He's had a positive NET rating (=ON-OFF) twice in his career: In his rookie season, when it was +0.6, and for the historically awful 11-12 Bobcats, where his NET was +0.5 - hardly an accomplishment

But, of course, as a bench player, he doesn't have to be great, or even good. He seems to be right on the brink of "OK". The right coach, e.g. like Thibodeau in 13-14, can make him look servicable

[Mike G]

15.5-year RAPM has Augustin at -1.6, ranked 1,047th of 1,521

Single year RPM's, starting in 2009, are
-2, -1.9, -2.1, -3.1, -4, -0.8, -3.9, and, for 15/16, -3

Those 7.5 season RPM add up to about -2.6 -- a bit harsher than his career BPM, and well below the boxscore assessment that he's been average or nearly so.

RAPM of -1.6 is at 31 percentile of 1521 players.
Last year, -1.6 RPM, among some 475 players, would rank about #270, or 43 percentile.

[bchaikin]

Don't know Augustin's career RPM (if there is one). By these other metrics, he's been an avg or better backup PG.

15.5-year RAPM has Augustin at -1.6, ranked 1,047th of 1,521.

RAPM of -1.6 is at 31 percentile of 1521 players.

augustin came into the league in 08-09, has played 13193 minutes. among the 90 PGs that played 3000+ minutes since 08-09, augustin:

- ranks just 76th in 2pt FG% at 42.3%...
- ranks just 89th (2nd worst) in rebounding at 3.6 reb/48min...
- ranks just 83rd (8th worst) in steals at 2.5 ST/100min...
- ranks just 76th (5th worst) in blocked shots at 0.1% BS (just 22 career blocked shots)...

that's way worse than 31 percentile in 4 key statistical categories just among PGs, not even among all players...

his Synergy defensive numbers since 08-09 are worse than those of just the league average PG, and his SportVu shot defense data at http://www.nba.com/stats (only available since 13-14) shows a 49.3% (404/820) 2pt FG% allowed and a 36.1% (163/452) 3pt FG% allowed, that's a 51.0% eFG% allowed over 2+ seasons (the league average PG shot just a 48.2% eFG% since 13-14)...

with what much of what he does ranking well below the 31st percentile, and his man/shot defense certainly worse than that of just the league average PG, just what praytell did he do to rank that high in RAPM (better than about 1/3 of 1521 players)?...

[sndesai1]

well, he's also a decent to good 3 point shooter, great ft shooter, and ok passer...i mean his career ts% and per 100 poss offensive numbers are pretty damn similar to mike conley's

also, that long term rapm has non-nba caliber players like nemanja nedovic, charles smith, and jeremy pargo in it - i don't think that it's comparable to your minutes restricted group of 90 pgs

[bchaikin]

well, he's also a decent to good 3 point shooter, great ft shooter, and ok passer...i mean his career ts% and per 100 poss offensive numbers are pretty damn similar to mike conley's

he's also a much better shooter/scorer than say ricky rubio, but rubio is the far better man defender, almost twice the rebounder, and forces alot more turnovers...

is his career RAPM better than say that of rubio?...

also, that long term rapm has non-nba caliber players like nemanja nedovic, charles smith, and jeremy pargo in it - i don't think that it's comparable to your minutes restricted group of 90 pgs

fine - where does augustin rank in career RAPM among those PGs that have played the most minutes the past 15 years? i count approximately 340 PGs that have played the past 15 years, but just 148 PGs with 3000+ minutes played (and he's 52nd in minutes played)...

[sndesai1]

you provided some categories where he ranks in the 5th-15th percentile. i provided some other categories where i assume he ranks higher (though i didn't actually check the rankings). all together, he has probably been somebody that ranks in the 15th-20th percentile range among those who get >5 minutes a game. given that, i don't think being in the 30th percentile in rapm among all players that have played in the past 15 years is all that crazy.