- --- Mike Goodman <msg_53@...> wrote:
> I have updated my statistics base thru 2001, and

Just curious here, Mike. How did you estimate the

> for fun I have

> Lakers careers (and sub-careers) ranked in the same

> manner as my

> infamous all-time list. I will include brief

> descriptions of the

> categories.

> G: Games, Regular season + Playoffs

> RegSe: Season productivity, projected Points +

> Rebounds, etc., per

> 36 minutes.

> PO%: fraction of career minutes which were in

> postseason

> PlaOf: Playoff productivity, see above.

> RSpts: Total regular-season production,

> standardized, square

> rooted and multiplied by column 4 (RegSe). Brownie

> points.

> POpts: Playoff production, likewise.

> Total: sum of columns 7 and 8.

> These last 3 columns are not meant to represent

> any real

> accomplishment, but are a value assigned to one's

> regular season and

> playoff careers, for the purpose of comparison.

> Remember, this is all just for fun.

>

> G RegSe PO% PlaOf RSpts POpts Total

> 1 Magic 1096 40.4 .18 39.7 8505 3970 12474

> 2 Kareem 1273 39.3 .14 37.8 8665 3454 12119

> 3 West 1084 40.5 .15 41.0 8300 3488 11788

> 4 Shaq 382 48.0 .19 47.4 6648 3176 9824

> 5 Baylor 979 36.2 .14 35.5 7066 2543 9609

> 6 Worthy 1069 29.6 .15 31.0 5052 2285 7337

> 7 Wilt 419 36.4 .20 34.1 4924 2047 6971

> 8 Mikan 284 39.7 .15 41.8 4143 2274 6417

> 9 Goodrich 760 29.8 .10 28.7 4047 1335 5382

> 10 Kobe 400 33.2 .18 33.5 3611 1752 5363

> 11 Scott 996 25.2 .16 24.0 3701 1509 5210

> 12 Mikkelsen 634 27.7 .10 26.0 3523 1247 4770

> 13 Divac 571 30.4 .09 28.5 3737 1021 4757

> 14 Wilkes 648 27.3 .11 24.7 3492 1154 4646

> 15 Nixon 544 26.4 .11 26.8 3145 1252 4397

> 16 Cooper 1038 21.5 .17 22.3 2879 1437 4316

> 17 Green 861 24.1 .13 21.7 3164 1122 4286

> 18 Lovellett 310 32.1 .08 32.4 2984 1010 3994

> 19 Hairston 390 29.5 .12 25.2 2919 916 3835

> 20 LaRusso 661 23.9 .12 21.5 2874 846 3720

>

> Notes:

> The "square rooting" is a means of diminishing

> one's regular

> season totals in comparison to one's playoff totals.

> For example,

> Gail Goodrich played 10% of his Lakers minutes in

> postseason, which

> means 1/10, or a 9:1 ratio (RS/PO). Taking the

> square roots of his

> totals creates about a 3:1 ratio in his "pts"

> columns. (the square

> root of 9 is 3). This reflects my bias, that the

> playoffs are what

> it's all about.

> The Lakers are perhaps unique in never having a

> prolonged drought

> between playoff runs. I found no major player in

> franchise history

> with more than 1 year of service, and no playoff

> appearances.

> Not showing in this chart is "rings", the number

> of championships

> players partook in. For each ring, a player is

> credited another 10

> playoff games at their career playoff rate.

> Players are credited an estimated number of

> blocks, steals, and

> turnovers from times before such stats were

> recorded. I gave Wilt 4

> blocks per game.

blocks, steals and turnovers for each player?

And when is that all-time Bulls list coming? I'm dying

to know where Leon Benbow rates. : )

Ed Weiland>

__________________________________________________

>

Do You Yahoo!?

Get personalized email addresses from Yahoo! Mail

http://personal.mail.yahoo.com/ - --- In APBR_analysis@y..., Ed Weiland <weiland1029@y...> wrote:
> > Players are credited an estimated number of

Sorry to take so long to get back to you, Ed.

> > blocks, steals, and

> > turnovers from times before such stats were

> > recorded. I gave Wilt 4

> > blocks per game.

>

I don't claim to have a decent method of estimating steals,

blocks, or turnovers. However, I have come up with an average

correlation with other stats, that I find relatively inoffensive.

Steals seem to correlate most closely with assists, and I guess

that a player's steals are related to his normalized assist rate,

thus: Steals = SqRt(Ast)-0.4

SqRt means "square root of"

For example, Maurice Stokes averaged 5.4 assists (per 36 minutes,

standardized). SqRt(5.4) = 2.3, and 2.3 - .4 = 1.9. So I credit

Stokes with 1.9 steals per 36 minutes, and his 3 year

career "equivalent steals" total is 391.

For sure, there are major deviations, from player to player, in

this estimate, and there are a handful of cases in which I use

anecdotal evidence to adjust my statistical "guess". However, for

the purposes of career evaluation, a difference of even a steal a

game very seldom changes a player's career ranking more than one

place, usually none.

In other words, assigning Maurice Stokes .9 steals, or 1.9, or 2.9

does not make much difference in the career rankings scheme.

While both assists and steals are a result of "good hands", and

anticipation, and quickness, etc; blocked-shots are mostly correlated

to rebounds, being a big-man thing.

The formula for blocks is : Blk = SqRt(Reb) - 1.6. For Mr.

Stokes, his career standardized rebound rate is 13.5, and his

estimated shot-blocking is 2.1. This could be way off, but it seems

to be a good average when applied to modern players. Again, there

are major deviations.

My feeling is that an educated guess is better than nothing, and I

am constantly open to what I call "anecdotal evidence" (as opposed to

statistical), and I consider any first-hand witnessing to be better

than no evidence at all.

Turnovers can be predicted, too, and the factors are many. The

traditional belief is that assists are most closely related to

turnovers, but in fact scoring and rebounding are, also.

If you have any idea what I am referring to when I talk about

standardized rates (and I have posted a few), the turnover formula is

this: TO = .08(Sco)+.07(Reb)+.16(Ast)+.05(Stl)+.10(Blk)-.005(MPG).

Here, MPG is minutes per game, and the other factors are multiples of

the standardized per-36-minute scoring rate, etc.

Strangely, blocks are 2nd to assists in correlation with turnovers.

Assists, obviously, require passing, and some turnovers are

inevitable.

Scorers lose the ball more than non-scorers, in general.

Apparently, rebounding causes a player to have possession, and he

can then be stripped or throw a bad pass.

Steals may indicate a gambling mentality, thus turnovers on the

other end.

Minutes are a factor because the jitters go away after the first

few; and retrospectively, players with good numbers but few minutes

may have been turnover-prone.

I would love to know how to actually plug in a few hundred player

stats and have my computer generate these correlations; all I have

managed is to tinker with the numbers until a good average is

achieved. After a tinkering, I just check the extremes at either

end, trying to minimize.

As always, I am open to any suggestion, except to "stop trying"

or "give up".

> Just curious here, Mike. How did you estimate the

> blocks, steals and turnovers for each player?

>

> And when is that all-time Bulls list coming? I'm dying

> to know where Leon Benbow rates. : )

>

> Ed Weiland

> >

> >

>

>

> __________________________________________________

> Do You Yahoo!?

> Get personalized email addresses from Yahoo! Mail

> http://personal.mail.yahoo.com/ - --- In APBR_analysis@y..., Ed Weiland <weiland1029@y...> wrote:
> And when is that all-time Bulls list coming? I'm dying

Alright, Ed, read 'em and weep:

> to know where Leon Benbow rates. : )

>

> Ed Weiland

> >

> >

Alltime Chicago Bulls Lineup:

RegSea PlaOf Total

1 Michael Jordan 11157 5447 16604

2 Scottie Pippen 6430 3126 9556

3 Artis Gilmore 5105 632 5737

4 Horace Grant 3856 1762 5617

5 Toni Kukoc 3477 1371 4849

6 Chet Walker 3605 883 4487

7 Bob Love 3459 960 4419

8 Jerry Sloan 3106 747 3853

9 Norm Van Lier 2900 727 3626

10 Tom Boerwinkle 2892 733 3625

11 Reggie Theus 3025 407 3432

12 Mickey Johnson 2864 453 3317

13 David Greenwood 2832 444 3276

14 B.J. Armstrong 2345 897 3241

15 Clifford Ray 2383 645 3028

16 Charles Oakley 2369 588 2956

17 Dennis Rodman 1916 964 2880

18 Orlando Woolridge 2602 265 2867

19 Luc Longley 1858 960 2819

20 Bob Boozer 2344 454 2797

21 Ron Harper 1813 951 2764

22 Dave Corzine 2367 339 2706

23 Bill Cartwright 1804 840 2644

24 John Paxson 1890 713 2603

25 Elton Brand 2596 0 2596

Leon Benbow is not this side of the horizon.>

>

> __________________________________________________

> Do You Yahoo!?

> Get personalized email addresses from Yahoo! Mail

> http://personal.mail.yahoo.com/ On Mon, 23 Jul 2001, Mike Goodman wrote:

[...]

> Turnovers can be predicted, too, and the factors are many. The

> traditional belief is that assists are most closely related to

> turnovers, but in fact scoring and rebounding are, also.

> If you have any idea what I am referring to when I talk about

> standardized rates (and I have posted a few), the turnover formula is

> this: TO = .08(Sco)+.07(Reb)+.16(Ast)+.05(Stl)+.10(Blk)-.005(MPG).

[...]

> I would love to know how to actually plug in a few hundred player

> stats and have my computer generate these correlations; all I have

> managed is to tinker with the numbers until a good average is

> achieved. After a tinkering, I just check the extremes at either

> end, trying to minimize.

What you want to use is "multivariate regression analysis" also known as

"ordinary least squares regression". I believe that Excel will only do

univariate regression. There are however freeware regression programs

available; I don't use any of them because I've got paid-for programs but

I know they are out there ... I know there was a shareware or freeware

econometrics program available at Penn State University's website. Also

there is a package called "R" which is a shareware or freeware version of

"S", a package widely used by statisticians. However S, and I imagine R,

are aimed more at theoretical statisticians and people who need to develop

and program their own statistics, rather than being aimed at users who

simply want to crunch some numbers using standard techniques.

The technique you describe is a standard one for filling in missing data;

i.e. run regressions to come up with equations predicting what a player's

turnovers per minute will be.

Obviously the technique becomes shakier as the amount of missing data

increases, in particular for years prior to 197? when there are NO data at

all on turnovers. Then you have to make assumptions that the turnover

equations for, say, 1957, are the same as the ones for 197?-2001. In

other words, extrapolation is a lot more difficult than interpolation, and

for years with no turnover data whatsoever, we're extrapolating rather

than interpolating.

So the equations should be double-checked by, e.g. looking at

season-by-season data to see if there are time trends. E.g. I believe

that offensive rebounding percentages gradually increased during the

1970s and 1980s. I believe that turnover rates (certainly per minute, and

possibly relative to scoring, rebounding, etc.) declined in the 1980s and

1990s. And for sure, field goal percentages rose for decades, until some

time in the 1990s when they started declining.

So the equations for predicting turnovers in the "modern" NBA may not work

for predicting turnovers in the NBA of the 1950s.

On the bright side, OLS will be much much faster AND lead to better, more

accurate equations than fiddling around by hand.

--MKT