Re: 2003 Questions
- --- In APBR_analysis@yahoogroups.com, "Mike G" <msg_53@h...> wrote:
> --- In APBR_analysis@yahoogroups.com, "aaronkoo" <deano@r...> wrote:that
> > Given the common perception that Walker is awesome (I can't
> > so many Boston fans think he's better than Pierce), I realize
> > there is quite a burden on me to show that he's not so good.No doubt about this. The common opinion among people I know who get
> The fact remains, that the Boston coach gives Walker 40 minutes
> every night, so he apparently doesn't consider him 'pitiful'.
paid in the biz is that Walker and Pierce are the only 2 real NBA
players on the C roster. That doesn't really help make Walker look
> As for assigning losses to basketball players, you probably know Iit
> think that's a sack of doodoo, and destined for the ashcan of
> analysis evolution.
> If Walker has 'lost' 15 games for Boston this year; and lacking
> Walker 'even worse players' would have lost even more games, then
> seems the assigning of losses cannot appropriately be done forAssigning wins and losses to individuals definitely has conceptual
> individuals, but only for teams.
difficulties. (I don't understand why assigning losses would be
tougher to assign conceptually than assigning wins.) It, like linear
weights, is just a summary of a bunch of stats relating to some form
of greatness. It has one benefit on linear weights in that it is
tied to reality -- individual wins and losses add up to team values.
It's definitely more complicated than linear weights and the
interpretation is more difficult (is a 5-1 record better than a 10-4
record?). As a consequence, I use it primarily as you guys use
linear weights -- rough measure of who is better than whom. In doing
real analysis of trades, free agent pickups, etc., I never use it.
Other things are much more predictive.
> In other words, individual won-loss records for basketball are at
> least as impractical a player-rating method as it is for pitchers
> baseball.Bill James called pitcher win-loss record a "very important"
statistic in his 1987 Abstract because it correlates well with
winning and is intelligible. It scored relatively mediocre with
regard to how well it reflected only a player's ability. "Some
people put a lot of stock in a won-lost record. Over a period of
years, you should. In one year, you can't," he says. I think this
is fair to say of the individual win-loss record I generate, too. It
is impacted by context and is not as good a predictor of future
success as the components that go into it (or, as James says, ERA is
a better predictor of future win-loss record than is current win-loss
>What about this year? Who is similar?
> sim player . . sco reb ast
> .00 Antoine Walker 20 9 4
> .32 Marques Johnson 20 8 4
> .33 George McGinnis 19 10 4
> .39 Shareef AbdurRahim 21 9 3
> .42 Brad Daugherty 20 10 4
> .45 Billy Cunningham 20 9 4
> .47 Mickey Johnson 17 9 4
> .48 Sidney Wicks 17 9 3
> .49 Bob Dandridge 19 6 3
> .50 Juwan Howard 18 8 3
> .50 Christian Laettner 17 9 3
> .52 Detlef Schrempf 17 8 4
> .52 Alvan Adams 17 9 5
Is Glenn Robinson similar? Their linear weights ratings end up
pretty similar. My stats have them surprisingly similar (similar
difference between offensive and defensive ratings, Big Dog's w-l 3.0-
10.3) and Robinson fortunately missed some games this year. Those
games show the Hawks being significantly better defensively at 99%
significance and significantly better winning, too (69% without and
38% with). The offense wasn't significantly different.
- I was forced to watch a huge number of Hawks games this year. One
thing nobody talked about was how bad this team's bench was. Nazr
Mohammed just flat out sucked, and Alan Henderson wasn't much better.
Darvin Ham can't make a shot if it isn't a dunk, and Dan Dickau was a
total flop. The only halfway decent bench player was Ira Newble, and
he started half the season.
As for the Terry point/not a point issue, the real problem was that
he was the only guy who could handle the ball, especially when Newble
was starting at off guard, and that was why they made so many TOs.
The Hawks just need somebody else who can dribble and take the
pressure off; which guard spot he plays is almost irrelevant.
As for Big Dog, he should be a bench dog. The team played better
without him, and at this point Newble is better anyway.
--- In APBR_analysis@yahoogroups.com, "lilnemoinslumber"
> --- In APBR_analysis@yahoogroups.com, "aaronkoo" <deano@r...> wrote:them?
> > 4. There are some good pieces in Atlanta. What to do to fix
> Take the ball out of Terry, Robinson, and Rahim's hands.
> Atlanta's "Big Three" are averaging a Big 3 TO's apiece.
> The Big Question is who exactly will be at the helm? Do they keep
> Stotts? Do they go after a Big guard to play point and keep Terry
> the 2? Do they undergo a Big overhaul in their offensive schemes tois
> try and eliminate the TO inducing iso's? Or do they punt?
> One at a time:
> We can't divine who will end up as coach, but we'll assume Stotts
> staying on. He brought the Hawks a bit of cohesion as their secondbe
> half finish can attest.
> I like the idea of a big guard playing point to Terry's 2. But it
> seems most tall ball-handling guards either look to score for
> themselves or are too easily harassed by quicker, smaller guards.
> Admittedly, getting Glenn to play D, or Theo to stay healthy would
> the answer. But playing better team D, and cutting down onturnovers
> really ought to be enough to get this crew into the playoffs. Inthe
> east at least.