## Strength Of Schedule

Expand Messages
• I ve added a Strength of schedule page to my website. True to my preferences, it is largely in the form of graphs. See it here:
Message 1 of 15 , Dec 20, 2004
I've added a Strength of schedule page to my website. True to my
preferences, it is largely in the form of graphs. See it here:
http://members.rogers.com/edkupfer/nba/schedule.htm

It's in beta stage, and I'm taking requests for additions or subtractions.

boxplot you should be able to figure it out. Essentially the graphs display
the entire range of opponent strengths for each team, the tails showing the
minimums and maximums, and the boxes showing the second and third quartiles.
Opponent strength is calculated as 2 parts this season's Pythagorean
strength, 1 part last season's. The strengths are calculated in "real time,"
in that they are frozen the date of the game -- that means that I don't
change the strength of the opponent after the fact if the opponent goes on
to win twenty in a row. Their strength remains the same as it was on the
date that the two teams played.

To all those who emailed me over the last few weeks -- I've been ill, and am
only now getting to my email backlog. Hang on a few days.
--

ed
"I ain't the world's best writer, ain't the world's best speller
But when I believe in something I'm the loudest yeller."
- Woody Guthrie
• ... From: igor eduardo küpfer [mailto:edkupfer@rogers.com] Sent: Monday, December 20, 2004 1:45 PM ... [...] ... Does the strength-of-schedule calculation
Message 2 of 15 , Dec 21, 2004
-----Original Message-----
From: igor eduardo küpfer [mailto:edkupfer@...]
Sent: Monday, December 20, 2004 1:45 PM

>I've added a Strength of schedule page to my website. True to my
>preferences, it is largely in the form of graphs. See it here:
>http://members.rogers.com/edkupfer/nba/schedule.htm

[...]

>boxplot you should be able to figure it out. Essentially the graphs display
>the entire range of opponent strengths for each team, the tails showing the
>minimums and maximums, and the boxes showing the second and third quartiles.
>Opponent strength is calculated as 2 parts this season's Pythagorean
>strength, 1 part last season's. The strengths are calculated in "real time,"

Does the strength-of-schedule calculation include a home/away component?
Playing at home is worth 3 points right off the bat (according to Jeff
Sagarin's ranking page; he used to say it was worth 4 points, but I imagine
the reduced scoring in the NBA has also reduced the magnitude of most
effects, when measured in points).

--MKT
• ... Ed, the link led me to your wonderful website. I haven t explored it fully, but have found enough to raise some questions: On the PAR ratings page, do you
Message 3 of 15 , Dec 21, 2004
--- In APBR_analysis@yahoogroups.com, igor eduardo küpfer
<edkupfer@r...> wrote:
> I've added a Strength of schedule page to my website. ..

Ed, the link led me to your wonderful website. I haven't explored
it fully, but have found enough to raise some questions:

On the PAR ratings page, do you figure the Replacement Player of
comparison to be playing on an average-paced team? Reason I ask is
that you have all 5 starting Suns among the league's top 36 players;
and other high-scoring teams (Bos, Orl, Sea, Sac, Min, Was) with
players ranked noticable higher than I've ranked them.

It seems you might be exacerbating the problem of high-paced teams
inlating player stats, when you subtract for the average replacement
type. The remainder is even more disparate among teams, according
to their pace.

Are Ricky Davis and Mark Blount doing better than Brevin and Emeka?
Or do they just play for a faster-paced team?

Also, wondering how you come up with defensive ratings: I see Bruce
Bowen is the worst (by far) defender for the Spurs (in your list).
You clearly describe how you get offensive ratings, then say you do
the same for defensive. -- ???

Is Andre Miller really having a better year than Kenyon or Carmelo?
Is Speedy Claxton really better than Jason Richardson? Is Chucky
Atkins an above-average starter?

Is PJ Brown better than Magloire? Are Peja, Bibby, and Brad all
having a better year than Webber?
• ... [...] ... Yeah, I forgot to say there s a 20% bonus/deduction for the home/away team, based on the general 60/40 home away win pattern in the NBA. ... Hmm.
Message 4 of 15 , Dec 21, 2004
> -----Original Message-----
> From: igor eduardo küpfer [mailto:edkupfer@...]
> Sent: Monday, December 20, 2004 1:45 PM
>
>
>> I've added a Strength of schedule page to my website. True to my
>> preferences, it is largely in the form of graphs. See it here:
>> http://members.rogers.com/edkupfer/nba/schedule.htm
>
[...]

> Does the strength-of-schedule calculation include a home/away
> component?

Yeah, I forgot to say there's a 20% bonus/deduction for the home/away team,
based on the general 60/40 home away win pattern in the NBA.

> Playing at home is worth 3 points right off the bat
> (according to Jeff Sagarin's ranking page; he used to say it was
> worth 4 points, but I imagine the reduced scoring in the NBA has also
> reduced the magnitude of most effects, when measured in points).
>
> --MKT

Hmm. I've seen that 3 points thing before, but never bothered to check on
it. I wonder how it corresponds to, say, Pythagorean projections.

The average team scores about 103 pts per 100 poss. Give them a 3 point
bonus for being at home, and chuck it into the Pyth equation:

(106^14) / ((106^14) + (103^14)) = 0.599

Very nice. Nice to see one of Sagarin's pronouncements has a basis in
reality.

--ed
• ... The replacement level I use (15% below league average) I got by basically pulling it out of my ass. I ve found previously that the replacement level itself
Message 5 of 15 , Dec 21, 2004
Mike G wrote:
> --- In APBR_analysis@yahoogroups.com, igor eduardo küpfer
> <edkupfer@r...> wrote:
>> I've added a Strength of schedule page to my website. ..
>
> Ed, the link led me to your wonderful website. I haven't explored
> it fully, but have found enough to raise some questions:
>
> On the PAR ratings page, do you figure the Replacement Player of
> comparison to be playing on an average-paced team? Reason I ask is
> that you have all 5 starting Suns among the league's top 36 players;
> and other high-scoring teams (Bos, Orl, Sea, Sac, Min, Was) with
> players ranked noticable higher than I've ranked them.
>

The replacement level I use (15% below league average) I got by basically
pulling it out of my ass. I've found previously that the replacement level
itself doesn't matter so much as the fact that you use *a* replacement
level. Comparing players to zero -- basically using their absolute stats --
rewards players who just show up to work a lot. Giving them a level they
must exceed before they get bonus points emphasises more efficient players.

(Try not to take the word "replacement" too seriously -- "marginal" or
"baseline" would have served just as well, and only the tradition of

But to answer your question, the Points Produced Above Replacement does not
take team pace into account. Why not? Well, I don't think it would change
the numbers too much -- PAR also includes defense, so fast paced team with
below average defenses are rewarded on offense, and penalised on D.

Also, PAR is a blunt instrument, and not really intended to measure value
too accurately. I just needed a simple number to combine offense and defense
together. Making fine distinctions of pace would ruin the fun for me by
making me work.

> It seems you might be exacerbating the problem of high-paced teams
> inlating player stats, when you subtract for the average replacement
> type. The remainder is even more disparate among teams, according
> to their pace.
>
> Are Ricky Davis and Mark Blount doing better than Brevin and Emeka?
> Or do they just play for a faster-paced team?

There are two ways of measuring "better": ability and value. Ability is
measured by a rate -- FG%, REb%, ie successes/opportunity. Value is measured
by the absolute contributions of that player (or in my case, I measure value
by the total contributions of that player over and above a baseline). A
player who scored 1700 points contributed more than a player who scored 1000
points. Who's better? Depends, but the first player added more value to his
team than the second. That's what PAR measures -- total contributions:
value.

For ability measures, see the ORtg and DRtg on which the PAR numbers are
based. These are simple rate stats (well, not really simple) which aren't
biased by pace or opportunity.

>
> Also, wondering how you come up with defensive ratings: I see Bruce
> Bowen is the worst (by far) defender for the Spurs (in your list).
> You clearly describe how you get offensive ratings, then say you do
> the same for defensive. -- ???
>

I use ORtg and Drtg definitions from DeanO's book (with minor
modifications). I won't put them here because they wouldn't fit in a single
email, but I will say the DRtg is, once again, a blunt instrument. Don't
take it too seriously. In Bowen's case, it could be that he is in charge of
guarding the more offensively talented opponents.

> Is Andre Miller really having a better year than Kenyon or Carmelo?
> Is Speedy Claxton really better than Jason Richardson? Is Chucky
> Atkins an above-average starter?
>
> Is PJ Brown better than Magloire? Are Peja, Bibby, and Brad all
> having a better year than Webber?

Ah, I won't answer questions like that, counsellor. All I can say is, Dre
has added more to his team this season than Kenyon (and certainly Melo,
who's playing horribly). Speedy better than JRich? Who knows. That question
is too broad, too vague. Restrict it to something more manageable, like "is
Speedy more efficient than JRich?" and I can answer with an emphatic "yes."

--ed
• In the past I have been able to replicate Sagarin s computer ratings, both those that account for margin of victory and those that do not. (I have not tried
Message 6 of 15 , Dec 21, 2004
In the past I have been able to replicate Sagarin's computer
ratings, both those that account for margin of victory and those
that do not. (I have not tried this year. It was this replication
that gave me the idea how to compute DanVal.)

The home court advantage is an estimated parameter that comes
straight from the regression results. It is smaller this year than
in the last couple years, and it implies that the home court
advantage, for whatever reason, is smaller this season. Perhaps
another point of emphasis for the officials? One that was not
widely reported?

This smaller home court advantage this season cannot really be due
to reduced scoring, since scoring is up this year relative to the
last few seasons. Sagarin's estimated home court advantage pretty
much IS reality since it is estimated from the data. But I believe
it is estimated primarily from the point differential regression
equation, so it should better predict home court advantage in point

Other than some funny stuff that Sagarin does at the beginning of
seasons and the fact that he likely ignores key injuries, scheduling
issues, and the like, what Sagarin does is very sensible and very
good.

Best wishes,
Dan

--- In APBR_analysis@yahoogroups.com, igor eduardo küpfer
<edkupfer@r...> wrote:
> > Playing at home is worth 3 points right off the bat
> > (according to Jeff Sagarin's ranking page; he used to say it was
> > worth 4 points, but I imagine the reduced scoring in the NBA has
also
> > reduced the magnitude of most effects, when measured in points).
> >
> > --MKT
>
> Hmm. I've seen that 3 points thing before, but never bothered to
check on
> it. I wonder how it corresponds to, say, Pythagorean projections.
>
> The average team scores about 103 pts per 100 poss. Give them a 3
point
> bonus for being at home, and chuck it into the Pyth equation:
>
> (106^14) / ((106^14) + (103^14)) = 0.599
>
> Very nice. Nice to see one of Sagarin's pronouncements has a basis
in
> reality.
>
> --ed
• I think you re rating assists much more heavily than I do. I have the Nuggets rated at: K-Mart 17.21 Miller 17.02 Melo 15.65 and Golden State at: J-Rich
Message 7 of 15 , Dec 23, 2004
I think you're rating assists much more heavily than I do.

I have the Nuggets rated at:
K-Mart 17.21
Miller 17.02
Melo 15.65

and Golden State at:
J-Rich 17.44
Claxton 17.07

We agree on Brown vs. Magloire however, which makes me think it's an
assist thing.

>Ah, I won't answer questions like that, counsellor. All I can say
>is, Dre has added more to his team this season than Kenyon (and
>certainly Melo, who's playing horribly). Speedy better than JRich?
>Who knows. That question is too broad, too vague. Restrict it to
>something more manageable, like "is Speedy more efficient than
>JRich?" and I can answer with an emphatic "yes."

--ed
• ... Remember, Ed s from the half of what NBA players do is Bad camp. Missed shots Bad. Turnovers Real Bad. Points and assists Good, but only if you get
Message 8 of 15 , Dec 24, 2004
--- In APBR_analysis@yahoogroups.com, "John Hollinger"
<alleyoop2@y...> wrote:
>
> I think you're rating assists much more heavily than I do.
>
> I have the Nuggets rated at:
> K-Mart 17.21
> Miller 17.02
> Melo 15.65
>
> and Golden State at:
> J-Rich 17.44
> Claxton 17.07

Remember, Ed's from the "half of what NBA players do is Bad" camp.

only if you get them efficiently.

Rebounds are just in the air; some players collect them more than
others, no biggy either way.

Since there are rebounders and there are assisters, you might be
seeing the discount on rebounds as much or more than that on assists.

Give me Carmelo's extra 5 pts, 2 reb, and 1 TO. You can have
Andre's extra 2.5 ast.

>
>.."is Speedy more efficient than
> >JRich?" and I can answer with an emphatic "yes."

Has anyone done a study that correlates different stats with playing
time? It seems clear that many teams feature top scorers who are
not their teams' most-efficient scorers. But often they lead their
team in minutes.

It helps, of course, that JRich gets some reb and ast in his
minutes. But a Go-To Guy gets stuck with lots of bad shots, not by
design. One or two extra missed shots per game separate JRich from
Claxton, in their shooting efficiency.

Whether you call it "more efficient" or "adding more to the team",
why bother to give players an overall ranking, if it doesn't
indicate their Value, or How Good they are ?

Check this list: Walker, Pierce, Okafor, Hinrich, James, Rip, JRich,
Randolph, Ray, Bosh, Jamison.

All these guys lead their teams in minutes; all of them either lead
their team in TO/min, or shoot below the team's Eff%; or both.

Are 2/3 of the NBA's head coaches really wack in the way they dole
out the PT? What do these players offer over some guy who seldom
misses or gets a TO?

It seems proper and fitting (as stat analysts) to suggest that Coach
A or Coach B should play this guy more or that guy less. But when
there's pretty much universal acceptance of certain playing-time
criteria, what does that suggest? Maybe (for example) you've got to
have that player (or 2, or 3) who can create a shot?
• On the TO/Min thing, I would EXPECT the best player to lead the team in that, because he s the one making all the plays. A far more important metric is
Message 9 of 15 , Dec 24, 2004
On the TO/Min thing, I would EXPECT the best player to lead the team
in that, because he's the one making all the plays. A far more
important metric is turnover per possession used -- an equation such
as TO/(AST+FGA+[FTA*0.44]+TO)

> Check this list: Walker, Pierce, Okafor, Hinrich, James, Rip,
JRich,
> Randolph, Ray, Bosh, Jamison.
>
> All these guys lead their teams in minutes; all of them either lead
> their team in TO/min, or shoot below the team's Eff%; or both.
>
> Are 2/3 of the NBA's head coaches really wack in the way they dole
> out the PT? What do these players offer over some guy who seldom
> misses or gets a TO?
>
> It seems proper and fitting (as stat analysts) to suggest that
Coach
> A or Coach B should play this guy more or that guy less. But when
> there's pretty much universal acceptance of certain playing-time
> criteria, what does that suggest? Maybe (for example) you've got
to
> have that player (or 2, or 3) who can create a shot?
• ... team ... You ve replaced minutes leader with best player , and I can accept that. But maybe my question should be: With this player out of the game, is
Message 10 of 15 , Dec 25, 2004
--- In APBR_analysis@yahoogroups.com, "John Hollinger"
<alleyoop2@y...> wrote:
>
> On the TO/Min thing, I would EXPECT the best player to lead the
team
> in that, because he's the one making all the plays.

You've replaced "minutes leader" with "best player", and I can
accept that. But maybe my question should be: With this player out
of the game, is the TEAM more efficient? -- or more effective?

I guess the answer to both is NO.

In which case, assigning "efficiency" to individuals is about as
shaky as assigning defensive numbers that show a team's best
defenders as the Worst, since they are guarding higher-scoring
opponents.

Doing the high-risk jobs -- guarding high scorers on defense,
getting the ball to your best scorers on offense -- don't come with
team win, of course.

> A far more
> important metric is turnover per possession used -- an equation
such
> as TO/(AST+FGA+[FTA*0.44]+TO)

I think if you rank players in this way, you'll find big rebounders
getting "too many" TO, relative to weak rebounders. A rebound GAINS
a possession, and puts the rebounder at risk of a turnover.
• ... Ahh, Mike s semi-annual tis the season to be a cynic note. The camp to which you refer doesn t really call itself that, of course. We just say that if
Message 11 of 15 , Dec 28, 2004
--- In APBR_analysis@yahoogroups.com, "Mike G" <msg_53@h...> wrote:
>
> --- In APBR_analysis@yahoogroups.com, "John Hollinger"
> <alleyoop2@y...> wrote:
> >
> > I think you're rating assists much more heavily than I do.
> >
> > I have the Nuggets rated at:
> > K-Mart 17.21
> > Miller 17.02
> > Melo 15.65
> >
> > and Golden State at:
> > J-Rich 17.44
> > Claxton 17.07
>
> Remember, Ed's from the "half of what NBA players do is Bad" camp.
>

Ahh, Mike's semi-annual 'tis the season to be a cynic note. The camp
to which you refer doesn't really call itself that, of course. We
just say that if someone makes a basket, someone (or several people)
can be blamed for giving it up. It's just accounting. You call it

For JohnH to mention differences between his rankings and others is
not a problem. Ultimately, everyone is going to have some qualms over
every statistical evaluation (I'm sure JohnH has subjective problems
with some of his). There are going to be lots of reasons for those
qualms. Some of them can be traced back to a "weight" on a stat.
Some of them are just the structure of how a stat is put together.
Some of them are just the fact that stats don't capture everything. A
reason like being from some "camp" isn't very useful.

> >
> >.."is Speedy more efficient than
> > >JRich?" and I can answer with an emphatic "yes."
>
> Has anyone done a study that correlates different stats with playing
> time? It seems clear that many teams feature top scorers who are
> not their teams' most-efficient scorers. But often they lead their
> team in minutes.

Yup, it happens a lot where a top scorer isn't the most efficient.
Typically, there is some role player who is more efficient. But
analyses of those role players suggest that they can't ramp up how
much they shoot. Is that "good" or "bad"? Neither. It's a matter of
fitting the pieces together well. I look for pieces that I know we
can build a good team around. Pieces have different characteristics,

>
> It helps, of course, that JRich gets some reb and ast in his
> minutes. But a Go-To Guy gets stuck with lots of bad shots, not by
> design. One or two extra missed shots per game separate JRich from
> Claxton, in their shooting efficiency.
>

In this case, Claxton has taken 9% of his shots in the last 3s on the
clock and JRich has taken 10%. Dunleavy has taken 13%. Uncle Cliffy
has taken 12%. So everyone on that team is stuck taking some bad
shots. And, uh, Fred Hoiberg has taken 15% of his shots in the last
3s on the clock. So what about giving him some credit for those bad
shots?

> Whether you call it "more efficient" or "adding more to the team",
> why bother to give players an overall ranking, if it doesn't
> indicate their Value, or How Good they are ?

Because coaching and chemistry matter. Blending players of different
abilities is important in the NBA and I've tried to develop numbers
that quantify _that_, not just value. We're blending things a little
differently in Seattle and it makes all the players' numbers better.

>
> Check this list: Walker, Pierce, Okafor, Hinrich, James, Rip, JRich,
> Randolph, Ray, Bosh, Jamison.
>
> All these guys lead their teams in minutes; all of them either lead
> their team in TO/min, or shoot below the team's Eff%; or both.

Walker has historically always come out horribly inefficient. This
year, too. Note that 82games shows Atlanta's offense as 11 points
worse with him in the game.

Pierce is actually quite efficient and the team's O is better with him
on the floor.

Okafor is not and the team's offense is better with him off the court
than on.

Hinrich is efficient and the team's O is better with him on than off.

Lebron is hugely efficient and his team's O is hugely better with him
on the court.

>
> Are 2/3 of the NBA's head coaches really wack in the way they dole
> out the PT? What do these players offer over some guy who seldom
> misses or gets a TO?

Note that Minnesota's offense is much better with Hoiberg on the court
than off (since he seems to be a classic case of an efficient player
who is "bad" by many systems).

With all this, don't get me wrong. There is definitely a value to
creating shots, even if you aren't efficient. I went through that
pretty thoroughly in Basketball on Paper. But blindly saying that any
player who puts up a lot of shots is a good player is really making a
mistake. It takes a knowledge of how players fit together -- a
coaching or GM decision -- to see how to make a good team from
_different_ parts. Feel free to continue to rank Hoiberg as a "bad"
player. Feel free to continue to rank Antoine Walker as a "good"
player. To me, it's about building a "good team" and I look for ways
that players can be parts of that "good team", regardless of what some
overall value measure says by finding what they are efficient at.

DeanO

Dean Oliver
Consultant to the Seattle Supersonics
"Excellent writing. There are a lot of math guys who just rush from
the numbers to the conclusion. . .they'll tell you that Shaq is a real
good player but his team would win a couple more games a year if he
could hit a free throw. Dean is more than that; he's really
struggling to understand the actual problem, rather than the
statistical after-image of it. I learn a lot by reading him." Bill
James, author Baseball Abstract
• If I may build on that thought -- Another, very tricky, part of the equation is that some guys will go from being low efficiency on a bad team to high
Message 12 of 15 , Dec 29, 2004
If I may build on that thought --

Another, very tricky, part of the equation is that some guys will go from being low
efficiency on a bad team to high efficiency, albeit with much fewer shots, on a good team.
Michael Finley's career is a good example of that, and to an extent so is Hoiberg. (While
we're on the topic, my #s show Hoiberg to be more valuable than Walker this year -- PER
of 17.73 for the Mayor vs. 15.68 for Walker-- though it's easily a career-best for Hoiberg).

With all this, don't get me wrong. There is definitely a value to
creating shots, even if you aren't efficient. I went through that
pretty thoroughly in Basketball on Paper. But blindly saying that any
player who puts up a lot of shots is a good player is really making a
mistake. It takes a knowledge of how players fit together -- a
coaching or GM decision -- to see how to make a good team from
_different_ parts. Feel free to continue to rank Hoiberg as a "bad"
player. Feel free to continue to rank Antoine Walker as a "good"
player. To me, it's about building a "good team" and I look for ways
that players can be parts of that "good team", regardless of what some
overall value measure says by finding what they are efficient at.

DeanO
• The question, IMO, is this: how can we evaluate how Hoiberg would do if he was given Antoine s minutes and green light to shoot? ...
Message 13 of 15 , Dec 29, 2004
The question, IMO, is this:

how can we evaluate how Hoiberg would do if he was
given Antoine's minutes and green light to shoot?

--- John Hollinger <alleyoop2@...> wrote:

>
> If I may build on that thought --
>
> Another, very tricky, part of the equation is that
> some guys will go from being low
> efficiency on a bad team to high efficiency, albeit
> with much fewer shots, on a good team.
> Michael Finley's career is a good example of that,
> and to an extent so is Hoiberg. (While
> we're on the topic, my #s show Hoiberg to be more
> valuable than Walker this year -- PER
> of 17.73 for the Mayor vs. 15.68 for Walker-- though
> it's easily a career-best for Hoiberg).
>
>
>
> With all this, don't get me wrong. There is
> definitely a value to
> creating shots, even if you aren't efficient. I went
> through that
> pretty thoroughly in Basketball on Paper. But
> blindly saying that any
> player who puts up a lot of shots is a good player
> is really making a
> mistake. It takes a knowledge of how players fit
> together -- a
> coaching or GM decision -- to see how to make a good
> team from
> _different_ parts. Feel free to continue to rank
> player. Feel free to continue to rank Antoine Walker
> as a "good"
> player. To me, it's about building a "good team" and
> I look for ways
> that players can be parts of that "good team",
> regardless of what some
> overall value measure says by finding what they are
> efficient at.
>
> DeanO
>
>
>
>
>

__________________________________
Do you Yahoo!?
http://celebrity.mail.yahoo.com
• ... I can evaluate this, just not now. Skill curves have pretty good power that way. I can also look at how Antoine would do if he toned down his shots, but
Message 14 of 15 , Dec 30, 2004
--- In APBR_analysis@yahoogroups.com, Gabe Farkas <gabefark@y...> wrote:
> The question, IMO, is this:
>
> how can we evaluate how Hoiberg would do if he was
> given Antoine's minutes and green light to shoot?
>

I can evaluate this, just not now. Skill curves have pretty good
power that way. I can also look at how Antoine would do if he toned
down his shots, but the question there is whether he ever would.

DeanO

Dean Oliver
Consultant to the Seattle Supersonics
"Dean Oliver looks at basketball with a fresh perspective. If you
want a new way to analyze the game, this book is for you. You'll
never watch a game the same way again. We use his stuff and it helps
us." Yvan Kelly, Scout, Seattle Sonics
ty.mail.yahoo.com
• I can also look at how Antoine would do if he toned ... The question might actualy be could rather than would. Perhaps, some players lack the ability to
Message 15 of 15 , Dec 30, 2004
"I can also look at how Antoine would do if he toned
> down his shots, but the question there is whether he ever would."

The question might actualy be "could" rather than would. Perhaps,
some players lack the ability to ascertain what is a good and a bad
shot . . .'cause in Walker case I think it is clear (at least when he
was in Boston) that he wants to win and doesn't take bad shot for
personal gain but bcs (he thinks) he is trying to help his team win.
Your message has been successfully submitted and would be delivered to recipients shortly.