30 Nov 2004 04:46:24
Tom Adams
I'm seeing Orange

I noticed that Syracuse is ranking in the top four in the AP poll.

My question is, do they have a tough enough schedule? Will they be
tested? I'd hate to see them take a one seed from our home boys
without some quality wins.

(I would not have to ask if the Borg had managed to ACCimilate them on
schedule.)


30 Nov 2004 09:01:36
Edward M. Kennedy
Re: I'm seeing Orange

"Tom Adams" <tadamsmar@yahoo.com > wrote:

> I noticed that Syracuse is ranking in the top four in the AP poll.
>
> My question is, do they have a tough enough schedule?

What year is it? Where are my geese?

> Will they be tested?

No, but if you happen to get enough talent you'll
win a national championship despite yourself.

> I'd hate to see them take a one seed from our home boys
> without some quality wins.

Being a Duke fan I'd like to see some weak one
seeds, preferably in our bracket.

--Tedward

Weak tit monkey shit




30 Nov 2004 16:36:41
Ar Q
Re: I'm seeing Orange


"Tom Adams" <tadamsmar@yahoo.com > wrote in message
news:ea44f5a1.0411300446.6e002ac8@posting.google.com...
> I noticed that Syracuse is ranking in the top four in the AP poll.
>
> My question is, do they have a tough enough schedule? Will they be
> tested? I'd hate to see them take a one seed from our home boys
> without some quality wins.
>
> (I would not have to ask if the Borg had managed to ACCimilate them on
> schedule.)

There is only one tough team left on their schedule: Oklahoma State. The
game will be played on Syracuse's semi home court-- Madison Square Garden.

Orange has beaten Princeton, Memphis and Mississippi State. Besides OSU,
only the winless Rice is notable. The other seven are all cupcakes. I expect
Syracuse to enter the conference play with 13-1 or 14-0 record.




30 Nov 2004 10:29:22
Tom Adams
Re: I'm seeing Orange

"Edward M. Kennedy" <nospam@baconburger.com > wrote in message news:<cohugs$4o6$1@gargoyle.oit.duke.edu>...
> "Tom Adams" <tadamsmar@yahoo.com> wrote:
>
> > I noticed that Syracuse is ranking in the top four in the AP poll.
> >
> > My question is, do they have a tough enough schedule?
>
> What year is it? Where are my geese?
>
> > Will they be tested?
>
> No, but if you happen to get enough talent you'll
> win a national championship despite yourself.
>
> > I'd hate to see them take a one seed from our home boys
> > without some quality wins.
>
> Being a Duke fan I'd like to see some weak one
> seeds, preferably in our bracket.

No! we want them. I wanna see the Tarheels as a 2-seed in Syracuse's
bracket. Then we can march all the way to the national championship
suffering nothing worse than Krzyzewski's scowl in the semis. UCONN
showed the way in '04.


30 Nov 2004 13:01:11
Clockwork Orange
Re: I'm seeing Orange

If you want me to go on arguing, you'll have to pay for another
five minutes, "Ar Q" <ArthurQ283@hottmail.com >!

>> I noticed that Syracuse is ranking in the top four in the AP
>> poll.
>>
>> My question is, do they have a tough enough schedule? Will they
>> be tested? I'd hate to see them take a one seed from our home
>> boys without some quality wins.
>>
>> (I would not have to ask if the Borg had managed to ACCimilate
>> them on schedule.)
>
> There is only one tough team left on their schedule: Oklahoma
> State. The game will be played on Syracuse's semi home court--
> Madison Square Garden.
>
> Orange has beaten Princeton, Memphis and Mississippi State.
> Besides OSU, only the winless Rice is notable. The other seven
> are all cupcakes. I expect Syracuse to enter the conference play
> with 13-1 or 14-0 record.

Princeton, Memphis, Mississippi State and the 'Boys are all stout
tests -- and only one in the friendly confines of the Carrier Dome
(Princeton). So far the Oranges are 3-0 against them. Okie State
is a good squad, led by that Lucas kid.

Oranges really don't need a lot of OOC tests to get ready for
March. A Big East schedule is plenty testing enough. They get
Notre Dame, UConn and Pitt twice each (not counting the BE
tourney), and don't laugh but WVU is pretty decent this year. And
Providence won at Michigan.

I should think when all's said and done and March rolls around,
the Orange will have played about a dozen games against Top 25
ranked teams. That ought to be enough, y'think?

The other thing is, this team has good veteran leadership in
Warrick, Pace, McNamara and Forth. All were major components of
the title team in '02-'03; three started and Pace played a lot. To
go along with them, they have several talented sophomores who all
played significant time last year. They are deep, experienced and
talented. They are fast and long, they can shoot, and they are to
be feared.

--
Cheers,
--Jeff
Let's Go Orange!


30 Nov 2004 19:14:27
Jack
Re: I'm seeing Orange

On 30 Nov 2004 13:01:11 -0600, Clockwork Orange
<no_this_isnt@my_email.com > wrote:

>>
>> There is only one tough team left on their schedule: Oklahoma
>> State. The game will be played on Syracuse's semi home court--
>> Madison Square Garden.
>>
>> Orange has beaten Princeton, Memphis and Mississippi State.
>> Besides OSU, only the winless Rice is notable. The other seven
>> are all cupcakes. I expect Syracuse to enter the conference play
>> with 13-1 or 14-0 record.
>
>Princeton, Memphis, Mississippi State and the 'Boys are all stout
>tests -- and only one in the friendly confines of the Carrier Dome
>(Princeton). So far the Oranges are 3-0 against them. Okie State
>is a good squad, led by that Lucas kid.
>
>Oranges really don't need a lot of OOC tests to get ready for
>March. A Big East schedule is plenty testing enough. They get
>Notre Dame, UConn and Pitt twice each (not counting the BE
>tourney), and don't laugh but WVU is pretty decent this year. And
>Providence won at Michigan.
>
The annual "Syracuse doesn't need out of conference challenges" claim
is noted. Besides, mini-picking gets you a good overall record. Never
leave your home state, the kids can get more study hall time in :)

Jack


30 Nov 2004 19:48:15
Ar Q
Re: I'm seeing Orange

> Oranges really don't need a lot of OOC tests to get ready for
> March. A Big East schedule is plenty testing enough. They get
> Notre Dame, UConn and Pitt twice each (not counting the BE
> tourney

ACC is a lot tougher than Big East. But look at Duke, NC, Wake and
Maryland's schedules. They still play a lot of big boys.

By the way, to be fair. They should drop NC State when Virginia entered the
poll. Seven points above Purdue is a shame. It ruins ACC's good name.




30 Nov 2004 13:48:29
Clockwork Orange
Re: I'm seeing Orange

If you want me to go on arguing, you'll have to pay for another
five minutes, bogro11@cox.net (Jack)!

>>Oranges really don't need a lot of OOC tests to get ready for
>>March. A Big East schedule is plenty testing enough. They get
>>Notre Dame, UConn and Pitt twice each (not counting the BE
>>tourney), and don't laugh but WVU is pretty decent this year.
>>And Providence won at Michigan.
>>
> The annual "Syracuse doesn't need out of conference challenges"
> claim is noted. Besides, mini-picking gets you a good overall
> record. Never leave your home state, the kids can get more study
> hall time in :)

I just looked at Duke's sked, and I only see one ranked team on there
OOC.

Mmmmm, cupcakes.

--
Cheers,
--Jeff
Let's Go Orange!


30 Nov 2004 18:02:46
Diamondback
Re: I'm seeing Orange

Sagarin sez it's Maryland #1 ... but we're only three games in.

"Edward M. Kennedy" <nospam@baconburger.com > wrote in message
news:cohugs$4o6$1@gargoyle.oit.duke.edu...
> "Tom Adams" <tadamsmar@yahoo.com> wrote:
>
>> I noticed that Syracuse is ranking in the top four in the AP poll.
>>
>> My question is, do they have a tough enough schedule?
>
> What year is it? Where are my geese?
>
>> Will they be tested?
>
> No, but if you happen to get enough talent you'll
> win a national championship despite yourself.
>
>> I'd hate to see them take a one seed from our home boys
>> without some quality wins.
>
> Being a Duke fan I'd like to see some weak one
> seeds, preferably in our bracket.
>
> --Tedward
>
> Weak tit monkey shit
>
>




30 Nov 2004 18:38:47
Edward M. Kennedy
A good laugh, was Re: I'm seeing Orange

"Diamondback" <diamondback@hotmail.com > wrote:

> Sagarin sez it's Maryland #1 ... but we're only three games in.

Actually they are #11.

For a good laugh: http://www.kiva.net/~jsagarin/sports/cbsend.htm

Look at who's in the top 10 -- heck, even the top 25.
#21 Duke is just three spots above Vermont! When
I think of basketball, Burlington always comes to
mind. Actually, I forget where UVM is. Might be in
Canada for all I know.

Yeah, I know, a few of those teams actually belong
there. It's still funny.

1 Kansas
Selfish bastids.
2 Old Dominion
A very capelable team, apparently.
3 Arkansas
Who resurrected these guys?
4 Texas A&M-CorpusChristi
A good SOS can do that!
5 Kentucky
Don't feel bad -- Duke is #21 -- but hey Texas A&M-CorpusChristi is better!
6 Kent State
Might actually belong up here in the rarified air, but not this high.
7 Syracuse
See Texas A&M-CorpusChristi
8 Virginia
It's just going to make the fall that much worse.
9 Butler
Yup, they did it.
10 Eastern Kentucky
You'd be real proud of this if it weren't for the other 4 mid majors ahead of you.

--Tedward





30 Nov 2004 21:30:40
Diamondback
Re: A good laugh, was Re: I'm seeing Orange

Well, that's what I get for relying on the Washington Post and not doing my
fact-checking on my own. Hell, if I don't do anything else in this life I
guess I could still be Secretary of State.

"Edward M. Kennedy" <nospam@baconburger.com > wrote in message
news:coj0b2$gb0$1@gargoyle.oit.duke.edu...
> "Diamondback" <diamondback@hotmail.com> wrote:
>
>> Sagarin sez it's Maryland #1 ... but we're only three games in.
>
> Actually they are #11.
>
> For a good laugh: http://www.kiva.net/~jsagarin/sports/cbsend.htm
>
> Look at who's in the top 10 -- heck, even the top 25.
> #21 Duke is just three spots above Vermont! When
> I think of basketball, Burlington always comes to
> mind. Actually, I forget where UVM is. Might be in
> Canada for all I know.
>
> Yeah, I know, a few of those teams actually belong
> there. It's still funny.
>
> 1 Kansas
> Selfish bastids.
> 2 Old Dominion
> A very capelable team, apparently.
> 3 Arkansas
> Who resurrected these guys?
> 4 Texas A&M-CorpusChristi
> A good SOS can do that!
> 5 Kentucky
> Don't feel bad -- Duke is #21 -- but hey Texas A&M-CorpusChristi is
> better!
> 6 Kent State
> Might actually belong up here in the rarified air, but not this high.
> 7 Syracuse
> See Texas A&M-CorpusChristi
> 8 Virginia
> It's just going to make the fall that much worse.
> 9 Butler
> Yup, they did it.
> 10 Eastern Kentucky
> You'd be real proud of this if it weren't for the other 4 mid majors ahead
> of you.
>
> --Tedward
>
>
>




30 Nov 2004 20:55:21
Burnted Orange
Re: I'm seeing Orange

"Ar Q" <ArthurQ283@hottmail.com > wrote:

> > Oranges really don't need a lot of OOC tests to get ready for
> > March. A Big East schedule is plenty testing enough. They get
> > Notre Dame, UConn and Pitt twice each (not counting the BE
> > tourney
>
> ACC is a lot tougher than Big East. But look at Duke, NC, Wake and
> Maryland's schedules. They still play a lot of big boys.

I just looked at Duke's sked, and I only see one ranked team on their
OOC schedule, Michigan State. Doesn't sound like "a lot of big boys" to me.

Syracuse has three.

Cheers,
--Jeff


01 Dec 2004 07:12:10
Ky58
Re: A good laugh, was Re: I'm seeing Orange

"Edward M. Kennedy" <nospam@baconburger.com > wrote:

> 10. Eastern Kentucky

> You'd be real proud of this if it weren't for the other 4 mid
> majors ahead of you.

EKU? Someone is delusional!

;-P




01 Dec 2004 10:53:30
Edward M. Kennedy
Re: A good laugh, was Re: I'm seeing Orange

"Ky58" <seventitles@bigblue.org > wrote:

> > 10. Eastern Kentucky
>
> > You'd be real proud of this if it weren't for the other 4 mid
> > majors ahead of you.
>
> EKU? Someone is delusional!

I find Sagarin's ratings are wacky before ~January,
when the anomolies start to get overwhelmed by
normal performance. I'm sure the RPI would have
equal if not greater bozoticy.

--Tedward




01 Dec 2004 11:02:49
Michael Sullivan
Re: I'm seeing Orange

Ar Q <ArthurQ283@hottmail.com > wrote:

> > Oranges really don't need a lot of OOC tests to get ready for
> > March. A Big East schedule is plenty testing enough. They get
> > Notre Dame, UConn and Pitt twice each (not counting the BE
> > tourney
>
> ACC is a lot tougher than Big East. But look at Duke, NC, Wake and
> Maryland's schedules. They still play a lot of big boys.

Duke? Who are they playing OOC this year?

The ACC may be tougher, but beyond a certain point you don't need a
hefty OOC schedule to be ready for the tourney or to deserve a #1. Any
of the Big 6 conferences will usually do, although occasionally one lays
an egg like the Pac-10 last year.

A weaker schedule than most ACC teams didn't much hurt UConn last year,
or Syracuse the year before.

The fact is, Syracuse OOC schedule this year actually looks pretty
reasonable. 5-6 cupcakes perhaps, but 4 solid teams is a lot stronger
than they have often been in the past. I recall many a year when the
Orange have strolled to an undefeated record late in the non-conf
season, but nobody really knew how good they were yet because they had
played no team of any consequence. This year, they will have played at
least 4 teams we'll probably be seeing in March before they get to the
Big East.

The difference between playing 8-10 games against top 25 competition vs.
1-2 is huge. The difference between playing 12-15 vs. 8-10 is not so
big. It will definitely affect where you'll be seeded if you have the
same record, but that's most of it. Readiness and the ability to
determine strength don't seem to suffer much. Despite the ACC usually
being the strongest conference, and the Big East often getting
prematurely reported dead, and rapped for weak OOC scheduling, the BE's
record in the tourney isn't that much worse than the ACC's.

Playing only a couple really good teams all year is clearly a problem
come March, but playing 7-8 doesn't seem to be, AFAICT.


Michael


01 Dec 2004 11:51:44
John Thacker
Re: A good laugh, was Re: I'm seeing Orange

On Wed, 01 Dec 2004 10:53:30 -0500, Edward M. Kennedy wrote:

> I find Sagarin's ratings are wacky before ~January,
> when the anomalies start to get overwhelmed by
> normal performance. I'm sure the RPI would have
> equal if not greater bozoticy.

Yeah, usually. However, I think that they're pretty decent today:

1 Kansas
2 Syracuse
3 Georgia Tech
4 Duke
5 Oklahoma State
6 Virginia
7 Cincinnati
8 Kentucky
9 Wake Forest
10 North Carolina
11 Illinois
12 Florida
13 Texas
14 Maryland
15 Iowa
16 Arkansas
17 Louisville
18 Southern Illinois
19 Rutgers
20 Creighton
21 NC State
22 Wisconsin
23 Mississippi State
24 Saint Joseph's
25 Marquette

John Thacker


01 Dec 2004 12:11:59
John Thacker
Re: I'm seeing Orange

On Wed, 01 Dec 2004 11:02:49 -0500, Michael Sullivan wrote:

> Ar Q <ArthurQ283@hottmail.com> wrote:

>> ACC is a lot tougher than Big East. But look at Duke, NC, Wake and
>> Maryland's schedules. They still play a lot of big boys.
>
> Duke? Who are they playing OOC this year?

Two cupcakes: Tenn-Martin, UNC-Greensboro
Six fairly decent teams expected to do well in their non top 6 conference:
Davidson, Valparaiso, Toledo, UIC, Princeton, Temple
One bad team from a top 6 conference: St. John's (long series)
One decent team from a top 6 conference: Oklahoma
One really good team: MSU

It's not a "lot of big boys," and it is a somewhat weaker schedule than in
the past. However, it is fewer cupcakes than the Orangemen are playing
as well. Playing those decent mid-majors can really help the SOS compared
to cupcakes, even though a good team should still win. However, Davidson
and UIC, et al. have enough talent to at least make for scares, as we've
seen, unlike UNC-G who simply wasn't talented enough at all to hang with
Duke, so it is somewhat of a risk.

> The ACC may be tougher, but beyond a certain point you don't need a
> hefty OOC schedule to be ready for the tourney or to deserve a #1. Any
> of the Big 6 conferences will usually do, although occasionally one lays
> an egg like the Pac-10 last year.
>
> A weaker schedule than most ACC teams didn't much hurt UConn last year,
> or Syracuse the year before.

Fair enough.

> The fact is, Syracuse OOC schedule this year actually looks pretty
> reasonable. 5-6 cupcakes perhaps, but 4 solid teams is a lot stronger
> than they have often been in the past.

Yeah, it definitely is stronger.

It is fair that playing a bunch of #100-#150 schools is so much better for
the SOS? Maybe. At least you're playing a lot of games where the other
team probably has the talent to at least stay on the same floor. It's not
precisely the same as playing top 25 teams, though.

John Thacker


01 Dec 2004 18:55:49
Ar Q
Re: A good laugh, was Re: I'm seeing Orange


1 Duke
2 Wake Forest
3 Kansa
4 Virginia
5 Syracuse
6 Oklahoma State
7 Kentucky
8 Illinois
9 Cincinnati
10 North Carolina
11 UConn
12 Florida
13 Texas
14 West Virginia
15 Wisconsin
16 Georgia Tech
17 Maryland
18 Louisville
19 Mississippi State
20 Saint Joseph's
21 Auburn
22 Iowa
23 Notre Dame
24 Michigan State
25 Marquette





01 Dec 2004 12:18:51
Tom Adams
Re: I'm seeing Orange

Clockwork Orange <no_this_isnt@my_email.com > wrote in message news:<Xns95B18E9A7D363Clockwork@67.19.196.138>...
> If you want me to go on arguing, you'll have to pay for another
> five minutes, "Ar Q" <ArthurQ283@hottmail.com>!
>
> >> I noticed that Syracuse is ranking in the top four in the AP
> >> poll.
> >>
> >> My question is, do they have a tough enough schedule? Will they
> >> be tested? I'd hate to see them take a one seed from our home
> >> boys without some quality wins.
> >>
> >> (I would not have to ask if the Borg had managed to ACCimilate
> >> them on schedule.)
> >
> > There is only one tough team left on their schedule: Oklahoma
> > State. The game will be played on Syracuse's semi home court--
> > Madison Square Garden.
> >
> > Orange has beaten Princeton, Memphis and Mississippi State.
> > Besides OSU, only the winless Rice is notable. The other seven
> > are all cupcakes. I expect Syracuse to enter the conference play
> > with 13-1 or 14-0 record.
>
> Princeton, Memphis, Mississippi State and the 'Boys are all stout
> tests -- and only one in the friendly confines of the Carrier Dome
> (Princeton). So far the Oranges are 3-0 against them. Okie State
> is a good squad, led by that Lucas kid.
>
> Oranges really don't need a lot of OOC tests to get ready for
> March. A Big East schedule is plenty testing enough. They get
> Notre Dame, UConn and Pitt twice each (not counting the BE
> tourney), and don't laugh but WVU is pretty decent this year. And
> Providence won at Michigan.
>
> I should think when all's said and done and March rolls around,
> the Orange will have played about a dozen games against Top 25
> ranked teams. That ought to be enough, y'think?
>
> The other thing is, this team has good veteran leadership in
> Warrick, Pace, McNamara and Forth. All were major components of
> the title team in '02-'03; three started and Pace played a lot. To
> go along with them, they have several talented sophomores who all
> played significant time last year. They are deep, experienced and
> talented. They are fast and long, they can shoot, and they are to
> be feared.

You should sent this speech to the members of the Selection Committee.

It sounds so convincing. I'm almost ready to grant that the ACC
should get at most one 1-seed, as I am sure they will after they
finish beating themselves up.


01 Dec 2004 15:38:29
Edward M. Kennedy
Re: I'm seeing Orange

"Tom Adams" <tadamsmar@yahoo.com > wrote

> It sounds so convincing. I'm almost ready to grant that the ACC
> should get at most one 1-seed, as I am sure they will after they
> finish beating themselves up.

warACC...?

--Tedward




01 Dec 2004 20:50:56
Re: A good laugh, was Re: I'm seeing Orange

John Thacker writes:

> Edward M. Kennedy wrote:

>> I find Sagarin's ratings are wacky before ~January,
>> when the anomalies start to get overwhelmed by
>> normal performance. I'm sure the RPI would have
>> equal if not greater bozoticy.

> Yeah, usually. However, I think that they're pretty decent today:
>
> 1 Kansas
> 2 Syracuse
> 3 Georgia Tech
> 4 Duke
> 5 Oklahoma State

You have to read the fine print. "Today", Sagarin is back to using
Bayesian ratings. The teams became connected earlier than usual,
and once connected, here is what Sagarin himself says is supposed
to happen:

"For the first few weeks of the season, the starting ratings
have weight in the process(BAYESIAN), but once the teams are
all CONNECTED, then the starting ratings are no longer used
and all teams are started equal and the ratings are then done
in an UNBIASED manner from that point on."

So, when the teams became connected after Saturday's games, Sagarin
started using unbiased ratings on Sunday, Monday, and Tuesday. That
led to some pretty wacky results, with Old Dominion, Eastern Kentucky,
James Madison, and Texas A&M Corpus Christi all being highly ranked.

So, how did Sagarin deal with these wacky results? Well, as of
Wednesday, he decided to contradict himself in the very next
sentence:

"The teams are now poorly CONNECTED and the ratings are
still BAYESIAN."

So, the ratings are not being done in an unbiased manner from that
point (the time all teams become connected) on. How long he'll
stick with the weighted results is a good question.

Usually it takes seven or eight games before all teams are connected.
Once the last connection occurs, lots of teams already have multiple
connections, which helps to average out the noise, and the ratings
are fairly decent. However, it is theoretically possible to connect
all teams after just two games (number ten teams 1 through 10, then
if the first game pits 1 against 2, 3 against 4, and so on, then if
the second game pits 2 against 3, 4 against 5, and so on, with
10 against 1, then all teams are connected). This year, I was
surprised to see all the teams become connected as of November 27;
many teams had played only one game (the connections coming from
their opponents, who had played more than one game). So it didn't
surprise me to see wacky ratings.

Incidentally, it appears as though Sagarin forgot to turn the
weight on for the first few ratings after the release of his
starting ratings. Memphis and Syracuse both appeared at #1 until
Sagarin turned the weight back on, at which point Duke reappeared
in the top spot, where they had started.



01 Dec 2004 22:24:09
John Thacker
Re: A good laugh, was Re: I'm seeing Orange

On Wed, 01 Dec 2004 20:50:56 +0000, tholen wrote:

> You have to read the fine print. "Today", Sagarin is back to using
> Bayesian ratings. The teams became connected earlier than usual,
> and once connected, here is what Sagarin himself says is supposed
> to happen:
>
> So, how did Sagarin deal with these wacky results? Well, as of
> Wednesday, he decided to contradict himself in the very next
> sentence:
>
> "The teams are now poorly CONNECTED and the ratings are
> still BAYESIAN."

Ah, good catch. I wasn't expecting him to go back to Bayesian, and missed
that.

John Thacker



02 Dec 2004 13:29:32
Tom Adams
Re: A good laugh, was Re: I'm seeing Orange

"Edward M. Kennedy" <nospam@baconburger.com > wrote in message news:<cokpdq$9r4$1@gargoyle.oit.duke.edu>...
> "Ky58" <seventitles@bigblue.org> wrote:
>
> > > 10. Eastern Kentucky
>
> > > You'd be real proud of this if it weren't for the other 4 mid
> > > majors ahead of you.
> >
> > EKU? Someone is delusional!
>
> I find Sagarin's ratings are wacky before ~January,
> when the anomolies start to get overwhelmed by
> normal performance. I'm sure the RPI would have
> equal if not greater bozoticy.
>

They stay wacky. As the season proceeds, these computer rating system
become a mishmash of the performance teams with injury-pocked season
histories.

By far, the best rankings come from the betting markets, like
www.tradesports.com.

This is one of the reasons that the current BCS is a joke. Don't get
me started.

I finally came up with a good model for www.poologic.com that
completely avoids computer estimates based on the injury-pocked season
history of the teams, the new Futures model. I still have the old
models there, but mainly for historical reason, and because they are
published.


03 Dec 2004 09:29:04
Edward M. Kennedy
Re: A good laugh, was Re: I'm seeing Orange

"Tom Adams" <tadamsmar@yahoo.com > wrote:

> > > > 10. Eastern Kentucky
> >
> > > > You'd be real proud of this if it weren't for the other 4 mid
> > > > majors ahead of you.
> > >
> > > EKU? Someone is delusional!
> >
> > I find Sagarin's ratings are wacky before ~January,
> > when the anomolies start to get overwhelmed by
> > normal performance. I'm sure the RPI would have
> > equal if not greater bozoticy.
>
> They stay wacky. As the season proceeds, these computer rating system
> become a mishmash of the performance teams with injury-pocked season
> histories.

"Wacky" is not accurate. Sagarin beats humans over the long
haul consistently -- ask James Armstrong if you want. That said,
his average picks also do very well.

Betting lines have their own flaws, they just tend to average
out.

> By far, the best rankings come from the betting markets, like
> www.tradesports.com.
>
> This is one of the reasons that the current BCS is a joke. Don't get
> me started.

In terms of predictive power you may be right but sports
isn't, in general, about expected performance. Past perfomance
with injuries or not is what you are judged by (in most
sports -- some start with a fresh slate if you make a cut).

As far as an evaluation of a team's entire season goes, the
Sagarin ratings would be better. When seeding brackets,
betting lines would be useful (but by definition, they are
not sufficient).

There's actually an interesting analogy to things like the
stock market and algorythmic investing, but I doubt
either of us know enough to be useful.

--Tedward




03 Dec 2004 10:42:57
John Thacker
Re: A good laugh, was Re: I'm seeing Orange

On Fri, 03 Dec 2004 09:29:04 -0500, Edward M. Kennedy wrote:

> "Tom Adams" <tadamsmar@yahoo.com> wrote:

>> They stay wacky. As the season proceeds, these computer rating system
>> become a mishmash of the performance teams with injury-pocked season
>> histories.
>
> "Wacky" is not accurate. Sagarin beats humans over the long
> haul consistently -- ask James Armstrong if you want. That said,
> his average picks also do very well.
>
> Betting lines have their own flaws, they just tend to average
> out.

True. Of course, don't forgot that the people betting can read Sagarin
as well. That's part of the advantage. Sagarin helps aggregate data,
just as the betting markets do.

> There's actually an interesting analogy to things like the stock market
> and algorithmic investing, but I doubt either of us know enough to be
> useful.

I know a little bit, but the last time we discussed this in rsbc, about,
what, 7-8 years ago?, I got accused of being a liar because someone
couldn't view the academic papers I linked to, not having a subscription
nor being part of a university network with a subscription. Suffice it to
say that any algorithm that gets publicized for betting doesn't work after
people know about it. There's quite a few interesting papers with actual
data using the sports lines looking for strategies. In one case they
found a strategy that worked, but then a followup paper two years later
found that it no longer worked with the new data.

John


05 Dec 2004 03:53:44
Tom Adams
Re: A good laugh, was Re: I'm seeing Orange

"Edward M. Kennedy" <nospam@baconburger.com > wrote in message news:<copt7g$vqe$1@gargoyle.oit.duke.edu>...
> "Tom Adams" <tadamsmar@yahoo.com> wrote:
>
> > > > > 10. Eastern Kentucky
>
> > > > > You'd be real proud of this if it weren't for the other 4 mid
> > > > > majors ahead of you.
> > > >
> > > > EKU? Someone is delusional!
> > >
> > > I find Sagarin's ratings are wacky before ~January,
> > > when the anomolies start to get overwhelmed by
> > > normal performance. I'm sure the RPI would have
> > > equal if not greater bozoticy.
> >
> > They stay wacky. As the season proceeds, these computer rating system
> > become a mishmash of the performance teams with injury-pocked season
> > histories.
>
> "Wacky" is not accurate. Sagarin beats humans over the long
> haul consistently -- ask James Armstrong if you want. That said,
> his average picks also do very well.

I'm not sure what you mean by "beats humans". Sagarin does not beat
the betting markets.

Average over the long haul is not a good test. This year's average
temperature would pass that test if we used every day next year to
predict the next day's temperature. The optimal predictor will
provide more than good average results over the long haul.

>
> Betting lines have their own flaws, they just tend to average
> out.

As far as I know, there are no flaws in the betting lines when judged
as a predictor. I have never seen such a claim that was backed up
with evidence.

>
> > By far, the best rankings come from the betting markets, like
> > www.tradesports.com.
> >
> > This is one of the reasons that the current BCS is a joke. Don't get
> > me started.
>
> In terms of predictive power you may be right but sports
> isn't, in general, about expected performance. Past perfomance
> with injuries or not is what you are judged by (in most
> sports -- some start with a fresh slate if you make a cut).
>
> As far as an evaluation of a team's entire season goes, the
> Sagarin ratings would be better.

Sagarin calls his ratings "Predictor". You are correct that Sagarin
ratings represent a kind of season summary, sort of like the standing
in a conference. But since Sagarin calls them a predictor, I can call
them wacky.

In recent memory, the wackiness of the Sagarin ratings peaked just
after Kenyon Martin broke his leg. But a weaker wack factor is at
work all the time, IMO.

> When seeding brackets,
> betting lines would be useful (but by definition, they are
> not sufficient).
>
> There's actually an interesting analogy to things like the
> stock market and algorythmic investing, but I doubt
> either of us know enough to be useful.
>
> --Tedward


06 Dec 2004 01:12:31
Edward M. Kennedy
Re: A good laugh, was Re: I'm seeing Orange

"Tom Adams" <tadamsmar@yahoo.com > wrote

> > > > > > 10. Eastern Kentucky
> >
> > > > > > You'd be real proud of this if it weren't for the other 4 mid
> > > > > > majors ahead of you.
> > > > >
> > > > > EKU? Someone is delusional!
> > > >
> > > > I find Sagarin's ratings are wacky before ~January,
> > > > when the anomolies start to get overwhelmed by
> > > > normal performance. I'm sure the RPI would have
> > > > equal if not greater bozoticy.
> > >
> > > They stay wacky. As the season proceeds, these computer rating system
> > > become a mishmash of the performance teams with injury-pocked season
> > > histories.
> >
> > "Wacky" is not accurate. Sagarin beats humans over the long
> > haul consistently -- ask James Armstrong if you want. That said,
> > his average picks also do very well.
>
> I'm not sure what you mean by "beats humans". Sagarin does not beat
> the betting markets.

Individual human beings. Even the best humans over the course
of a season lose to Sagarin.

> Average over the long haul is not a good test. This year's average
> temperature would pass that test if we used every day next year to
> predict the next day's temperature. The optimal predictor will
> provide more than good average results over the long haul.
>
> > Betting lines have their own flaws, they just tend to average
> > out.
>
> As far as I know, there are no flaws in the betting lines when judged
> as a predictor. I have never seen such a claim that was backed up
> with evidence.

Depends on what you are measuring. Any individual *game*
could have a line that is way out of wack for whatever reason,
like lots of homer bets, or hardly anyone really knows about
team X. Sagarin does not have that problem.

> > > By far, the best rankings come from the betting markets, like
> > > www.tradesports.com.
> > >
> > > This is one of the reasons that the current BCS is a joke. Don't get
> > > me started.
> >
> > In terms of predictive power you may be right but sports
> > isn't, in general, about expected performance. Past perfomance
> > with injuries or not is what you are judged by (in most
> > sports -- some start with a fresh slate if you make a cut).
> >
> > As far as an evaluation of a team's entire season goes, the
> > Sagarin ratings would be better.
>
> Sagarin calls his ratings "Predictor". You are correct that Sagarin
> ratings represent a kind of season summary, sort of like the standing
> in a conference. But since Sagarin calls them a predictor, I can call
> them wacky.

In that context, you have a point, but they are less wacky than
a human, so therefore I conclude you're statement that Sagarin
is wacky is more wacky than Sagarin.

> In recent memory, the wackiness of the Sagarin ratings peaked just
> after Kenyon Martin broke his leg.

Of *course* they did, by definition. He makes no attempt
to account for such obviously important information and is
instead trying to predict on past game results alone. He is
trying to be algorhtmic, not subjective, even if one could
attempt to take player ratings and injuries into account.

--Tedward






08 Dec 2004 06:52:23
tadamsmar
Re: A good laugh, was Re: I'm seeing Orange


Edward M. Kennedy wrote:
> "Tom Adams" <tadamsmar@yahoo.com> wrote
> > I'm not sure what you mean by "beats humans". Sagarin does not > >
> beat the betting markets.
>
> Individual human beings. Even the best humans over the course
> of a season lose to Sagarin.

This is not true. Your human is a strawman. Any human can use the
betting market spreads to beat or, at least, match Sagarin. Such a
human will beat Sagarin when injuries are an issue.

> > In recent memory, the wackiness of the Sagarin ratings peaked
> > just after Kenyon Martin broke his leg.
>
> Of *course* they did, by definition. He makes no attempt
> to account for such obviously important information and is
> instead trying to predict on past game results alone. He is
> trying to be algorhtmic, not subjective, even if one could
> attempt to take player ratings and injuries into account.
>

We agree on Sagarin. I am saying the glass is half empty.
But we don't agree on what a human can do.



08 Dec 2004 11:34:58
Edward M. Kennedy
Re: A good laugh, was Re: I'm seeing Orange


"tadamsmar" <tadamsmar@yahoo.com > wrote in message news:1102517543.003103.58240@z14g2000cwz.googlegroups.com...
>
> Edward M. Kennedy wrote:
> > "Tom Adams" <tadamsmar@yahoo.com> wrote
> > > I'm not sure what you mean by "beats humans". Sagarin does not > >
> > beat the betting markets.
> >
> > Individual human beings. Even the best humans over the course
> > of a season lose to Sagarin.
>
> This is not true. Your human is a strawman. Any human can use the
> betting market spreads

And if everyone in the market used the betting market to make
their decision, the system collapses. You have to have the part
where you gather opinions.

> to beat or, at least, match Sagarin. Such a
> human will beat Sagarin when injuries are an issue.

I meant when humans pick on their opinion (i.e. a market of sorts
is created) which should have been obvious because I pointed out
that the average picks did really well too.

Granted a few idiots might enter a picking *contest* and use a
market (or Sagarin) to try and look all basketball schmarty, but
that is a pointless and trivial thing to do. Yeah, you could
expressly test some market by entering its market output in a
contest, but that's got nothing to do with the point of a picking
contest.

Using Sagarin and/or markets to *help* make a decision is
just doing your homework.

> > > In recent memory, the wackiness of the Sagarin ratings peaked
> > > just after Kenyon Martin broke his leg.
> >
> > Of *course* they did, by definition. He makes no attempt
> > to account for such obviously important information and is
> > instead trying to predict on past game results alone. He is
> > trying to be algorhtmic, not subjective, even if one could
> > attempt to take player ratings and injuries into account.
>
> We agree on Sagarin. I am saying the glass is half empty.
> But we don't agree on what a human can do.

You don't agree on what I was comparing Sagarin to.

--Tedward




09 Dec 2004 20:28:00
Milt Epstein
Re: A good laugh, was Re: I'm seeing Orange

"Edward M. Kennedy" <nospam@baconburger.com > writes:

[ ... ]
>I find Sagarin's ratings are wacky before ~January, when the
>anomolies start to get overwhelmed by normal performance. I'm sure
>the RPI would have equal if not greater bozoticy.

Let me take advantage of this lead-in to gauge interest in doing the
picking contest I've done off and on the last few years. I had called
it "Pick The Top 25" (PTT25), and although it may be forgotten, the
original intent was to see how well people could pick as compared to a
system like Sagarin's.

In the past, I typically didn't start until the new year, which about
coincided with the start of conference seasons -- that was mostly a
matter of my not getting my act in gear until then, with consideration
to the fact that Sagarin rankings would have settled down by then,
with enough games in the books to have a basis for comparison. But I
could start it sooner this year.

The format in the past was that people would pick all the games
involving Top 25 teams each week (winner and margin of victory). That
format was mostly a matter of convenience, and it worked pretty well,
with perhaps the biggest problem being that the Top 25 didn't come out
until Monday afternoon, so there wasn't much time for people to get
their picks in for Monday evening games. IIRC, people also expressed
interest in varying the format in terms of what teams were involved,
for example, using games involving "major"/top 6 conferences teams.

So, let me know, are you interested in participating, and what format
would you prefer. My main consideration on the format is that it has
to be conducive to automation. I have code in place to handle things
using the format I've used in the past (i.e., weekly Top 25), but I'm
willing to make minor modifications to handle a different format.

--
Milt Epstein
mepstein@uiuc.edu


09 Dec 2004 15:22:13
Drew
Re: A good laugh, was Re: I'm seeing Orange

"Milt Epstein" <mepstein@uiuc.edu > wrote in message news:cpacgg$tto$1@news.ks.uiuc.edu...
>
> Let me take advantage of this lead-in to gauge interest in doing the
> picking contest I've done off and on the last few years. I had called
> it "Pick The Top 25" (PTT25), and although it may be forgotten, the
> original intent was to see how well people could pick as compared to a
> system like Sagarin's.

Color me interested...

> In the past, I typically didn't start until the new year, which about
> coincided with the start of conference seasons -- that was mostly a
> matter of my not getting my act in gear until then, with consideration
> to the fact that Sagarin rankings would have settled down by then,
> with enough games in the books to have a basis for comparison. But I
> could start it sooner this year.

Eh, why bother? The holidays will just make it harder for many to
get their picks in.

> So, let me know, are you interested in participating, and what format
> would you prefer. My main consideration on the format is that it has
> to be conducive to automation. I have code in place to handle things
> using the format I've used in the past (i.e., weekly Top 25), but I'm
> willing to make minor modifications to handle a different format.

I'm interested, and not so picky about the format.

Drew






09 Dec 2004 22:23:01
Edward M. Kennedy
Re: A good laugh, was Re: I'm seeing Orange

"Milt Epstein" <mepstein@uiuc.edu > wrote

> Let me take advantage of this lead-in to gauge interest in doing the
> picking contest I've done off and on the last few years. I had called
> it "Pick The Top 25" (PTT25), and although it may be forgotten, the
> original intent was to see how well people could pick as compared to a
> system like Sagarin's.
>
> In the past, I typically didn't start until the new year, which about
> coincided with the start of conference seasons -- that was mostly a
> matter of my not getting my act in gear until then, with consideration
> to the fact that Sagarin rankings would have settled down by then,
> with enough games in the books to have a basis for comparison. But I
> could start it sooner this year.

I'm gin.

--Ted