AP and Coaches Poll Watch Thread (2015-16 Edition)

[WUpigsooie]

To be honest, @jdshock: brought it up. I was just fixing his data for him based on his comment about needing to exclude tournament games. I'm not sure what the goal of that comparison is, relative to this thread's discussions so far.

It might be interesting to look at for its own sake. I'll see if I have time later to pull it together.

If I understand correctly, the point of that comparison would be to take a metric that try's to identify how good a team is (KenPom) against a metric more based on what a team has accomplished (RPI). If there is consistently a descrepency between the two (kenpom over RPI) then you could argue that the schedule is holding a team back from accomplishing as much as they could.

I'm not sure how solid that logic is, but it might be interesting to look at anyway.

[jdshock]

I have no problem with that statement regarding RPI bias toward BCS conferences. If you think WSU would earn an RPI of 24 in the MVC and an RPI of 22 in a BCS conference, same year, just due to different SOS, I can go along with that. Unfortunately, I regularly see people claiming much more than a "slight" bias. It's the claims that the bias is so great that a BCS team at #40 is not as good as an MVC team at #90 that I find wildly off base.

If I understand correctly, the point of that comparison would be to take a metric that try's to identify how good a team is (KenPom) against a metric more based on what a team has accomplished (RPI). If there is consistently a descrepency between the two (kenpom over RPI) then you could argue that the schedule is holding a team back from accomplishing as much as they could.

I'm not sure how solid that logic is, but it might be interesting to look at anyway.

That was definitely the logic, but I regret posting before I thought about the selection sunday data. I also am not sure that any bias that exists in RPI wouldn't exist in kenpom. I just don't know enough about the intricacies of the math.

As an interesting aside, I tried to look into the math to see if I could figure out how a bias might exist. One of the first papers I came across suggests that the bias is actually the reverse of what has been suggested (and the opposite of what I intuitively believed to be the case). The paper suggested that RPI has a bias against mid-level teams in elite conferences and a bias in favor of high-level teams in bad conferences. (at least that's how I read it. Maybe someone smarter than I am would understand it better!) http://www.accessecon.com/pubs/EB/2007/Volume4/EB-07D80023A.pdf

I also found an article stating the seeding bias in favor of SEC teams over ACC teams during a certain set of years, which I thought was funny. http://www.accessecon.com/pubs/EB/20...-07L80004A.pdf

[Kung Wu]

I have no problem with that statement regarding RPI bias toward BCS conferences. If you think WSU would earn an RPI of 24 in the MVC and an RPI of 22 in a BCS conference, same year, just due to different SOS, I can go along with that. Unfortunately, I regularly see people claiming much more than a "slight" bias. It's the claims that the bias is so great that a BCS team at #40 is not as good as an MVC team at #90 that I find wildly off base.

From a team perspective, the bias is slight. Let's say it's 5% (I am not saying it's 5%, I'm just using a number). If you expected a 100 RPI spot, you would say it might end up being 105. That doesn't _seem_ too bad at first glance -- just like you said, "Hey look RPI is close to KenPom". But it skewed low in 3 of four cases, with the one notable exception being when we ran the table.

However, if you look at it across how it affects all the non-BCS teams, you really have a major problem. If all non-BCS teams are _each_ skewed by 5%, then there are always a handful of teams that should have been "protected" or "first four in" or higher seeded and playing easier matches that didn't get those perks. And by corollary, there are always several BCS teams getting slightly better matches, or being protected that really shouldn't be protected, or in the tournament when they shouldn't have been (and that can cost non-BCS tremendous amounts of money!).

Of course, if you look at the MVC's RPI versus KenPom over some reasonable timeframe and that skewed error isn't there (there is no negative bias), then this is all poppy-cock.

[Jamar Howard 4 President]

It might be interesting to look at for its own sake. I'll see if I have time later to pull it together.

I'm going to start a new thread with this data. I'm seeing some interesting things.

Hopefully that will allow us to get his back on the topic of AP/Coaches polls in this thread.

[Dave Stalwart]

Bump. I'll go with #18

This was the game people needed to see and the Shocks delivered. Take care of next week and we will be back around or below #15

[ROOT ANGRY!]

Today's performance moves us to #12 in Kenpom.
We're #27 in offensive efficiency, #10 in def.

(Flagship is #15, 15 in off, 39 in def)

[another shocker]

Today's performance moves us to #12 in Kenpom.
We're #27 in offensive efficiency, #10 in def.

(Flagship is #15, 15 in off, 39 in def)

man, i would like to see grady assert himself offensively, consistently. i think he has it in him, but he started so late in this program.

[T.VuOP]

Teams ranked higher than WSU lost this week:
#3 Iowa - to #8 Maryland
#4 KU - to #14 Iowa St
#5 Texas A&M - to Arkansas
#9 West Virginia - to Florida
#10 Providence - to #7 Xavier
#14 Iowa St - to #5 Texas A&M
#15 Miami - NC St
#16 Louisville - to #11 Virginia
#18 Arizona - to #23 Oregon
#19 Indiana - to Wisconsin
#20 Kentucky - to #4 KU
Two other ranked teams #24 Duke and #25 Notre Dame also lost this week.
A brutal week for the ranked, 13 out of 25

[Dave Stalwart]

Teams ranked higher than WSU lost this week:
#3 Iowa - to #8 Maryland
#4 KU - to #14 Iowa St
#5 Texas A&M - to Arkansas
#9 West Virginia - to Florida
#10 Providence - to #7 Xavier
#14 Iowa St - to #5 Texas A&M
#15 Miami - NC St
#16 Louisville - to #11 Virginia
#18 Arizona - to #23 Oregon
#19 Indiana - to Wisconsin
#20 Kentucky - to #4 KU
Two other ranked teams #24 Duke and #25 Notre Dame also lost this week.
A brutal week for the ranked, 13 out of 25

Sounds good for the good guys

[Stickboy46]

17 in Seth Davis' ballot for tomorrow

[Steeleshocker]

17 in Seth Davis' ballot for tomorrow

And a couple fun snarky tweets regarding the shockers.

Are you suggesting I'm cheating on Wichita State? https://t.co/bG05qwohg1

Edit: i suck at trying to copy things from Twitter. Someone else do it.

[WuDrWu]

There's not a couple of fun snarky tweets, there's about 40. That's good stuff.

https://twitter.com/SethDavisHoops

[WuShock Reaper]

BOZICH | My weekly AP college basketball Top 25 ballot

First ballot published and has the Shockers at...?

http://www.wdrb.com/story/31103576/bozich-my-weekly-ap-college-basketball-top-25-ballot

[ShockerFever]

BOZICH | My weekly AP college basketball Top 25 ballot

First ballot published and has the Shockers at...?

http://www.wdrb.com/story/31103576/b...-top-25-ballot

I'm not gonna rip a guy that's being very positive about WSU.

But I told ya people would forget about KU's loss on Monday. No drop.

[WSUwatcher]

Speaking of Seth Davis, I just remember him picking UE to beat the Shocks 78-77. At least he got the winner's score right, but apparently he has no idea how hard it is for a Valley team to score 78 against the Shocks these days. This year's high was today's 65; last year's was 70 by UNI in WSU's loss; the undefeated year before, Loyola got 74 once in a high scoring game in Chicago and UNI got 73 at their place.

So in the last three seasons, Valley foes have scored 70+ three times against the Shocks, with a max of 74. I'm thinking that a prediction of 78 by anybody in this league is not a winning proposition for the mighty Seth.