- You said, in part "says nothing at all about quality, it just means

that it gets dwarfed by other things in importance."

I disagree: I actually think assist-turnover ratio says nothing at

all about quality. It is dominated by players who never penetrate

like Jeff McInnis and Alvin Williams (or the previous example, Muggsy

Bogues) and as a result this particular stat is at least as much a

contrarian indicator as it is a positive one.

--- In APBR_analysis@y..., "HoopStudies" <deano@r...> wrote:

> --- In APBR_analysis@y..., "alleyoop2" <alleyoop2@y...> wrote:

> > That's what my #s show as well - offensive efficiency rose last

> year

> > after a four-year period that looks like a bottoming out, at

least

> > right now, with the lockout year being the worst.

> >

> > Interesting discussion on turnovers by the way, and I couldn't

> agree

> > more with whoever was ripping on Muggsy Bogues. Assist-Turnover

> ratio

> > is about the most idiotic statistic I've ever seen, and has

almost

> no

> > correlation with any kind of skill whatsoever.

> >

>

> Obviously assists are good and turnovers are bad, but the ratio is

> just one of so many indicators that the value of the two numbers

gets

> lost. I don't think anyone seriously ranks players by the ratio,

but

> it does tell you something -- which is the point. Kind of like

GIDP

> in baseball doesn't rank players, but it tells you something if

used

> properly.

>

> Basketball is so multidimensional -- with so many things going on --

> that simple stats often don't tell the true story. They tell part

of

> the story. Assist-to-turnover ratio is a fine example. Its "lack

of

> correlation" with skill doesn't mean that it says nothing at all

> about quality, it just means that it gets dwarfed by other things

in

> importance. I would actually say that coaching quality is

important,

> but its importance gets dwarfed by the talent that the coach has.

We

> try to give awards to coaches, but they are often pretty much

> guesswork because the tools for evaluating coaches aren't very

> refined to overcome the dwarfing impact of talent.

>

> DeanO

>

> >

> > --- In APBR_analysis@y..., "HoopStudies" <deano@r...> wrote:

> > > --- In APBR_analysis@y..., "John W. Craven" <john1974@u...>

wrote:

> > > > On Tue, 20 Aug 2002, HoopStudies wrote:

> > > > > Definitely turnovers have been coming down, just as they've

> > been

> > > > > coming down in the WNBA. Turnovers are much more plentiful

> in

> > > less

> > > > > mature leagues. I've always believed that the early high

> rates

> > > of

> > > > > turnovers were real. They are still coming down now as the

> > Mavs

> > > set

> > > > > a record last year for fewest turnovers. And Nowtizki had

> > about

> > > the

> > > > > lowest turnover rate I've seen.

> > > >

> > > >

> > > > Which raises a point: How has your overall offensive

efficiency

> > > rating changed over the years? The idea that shooting

percentages

> > may

> > > have fallen in part because teams are choosing less risky (in

> terms

> > > of committing TOs) yet less successful (in terms of made FGs

per

> > FGs

> > > attempted) options is one that I hadn't considered until this

> > thread.

> > >

> > > It went up sharply in the '70's, steady in the '80's, and down

in

> > > the '90's. It's at least flat or rising now.

> > >

> > > Season AvgRtg

> > > 1974 99

> > > 1975 99

> > > 1976 99

> > > 1977 101

> > > 1978 102

> > > 1979 105

> > > 1980 106

> > > 1981 107

> > > 1982 107

> > > 1983 105

> > > 1984 108

> > > 1985 108

> > > 1986 107

> > > 1987 108

> > > 1988 108

> > > 1989 108

> > > 1990 108

> > > 1991 108

> > > 1992 108

> > > 1993 108

> > > 1994 106

> > > 1995 108

> > > 1996 108

> > > 1997 107

> > > 1998 105

> > > 1999 102

> > > 2000 104

> > > 2001 104

> > > 2002 105 - --- In APBR_analysis@y..., "harlanzo" <harlanzo@y...> wrote:
> So how do you meaningfully interpret turnovers? I agree that

balance

> context is important (ie Rodman's low turnovers do not indicate his

> ballhandling ability). Can you estimate touches per game and then

> divide turnovers into this number? That would theoretically

> out the problems you mention. (of course determining what is a

real

> touch and tos in the half court v. fast break might make differnece

That's what Bob Chaikin does in his program. He defines touches per

> too.)

game as the sum of FGA, TO, the number of passes the player makes,

and the number of times he was fouled. It seems that the best

ballhandlers are usually the point guards, turning the ball over on 5-

6% of their touches. PF's and C's are on the other end of the

spectrum with TO/touch percentages of 10% or more.