Breaking Down WSU's Resume

[Rlh04d]

I can't complain too much about the seeding. If teams were properly seeded we wouldn't have as many upsets. Biased voters make the tournament more entertaining.

[Jamar Howard 4 President]

I'm disputing your entire notion that looking at RPI breakdowns of wins is a better way to evaluate resumes than the Massey Consensus ranking. Yes, there are a number of teams that have more top 50 RPI wins, but the differences in scheduling alone (much less the validity of RPI wins as a tool; the real difference between beating the #50 RPI team and the #51 RPI team; how to weigh good wins vs. bad losses; margin of victory/loss; etc.) make the entire approach suspect. Not only do I believe that the consensus score is a better evaluation tool, it has a better track record at predicting outcomes of NCAA games than the RPI, the AP and Coaches Polls, actual NCAA seeding, and Vegas Favorites.

You may be right about how the committee will look at the resumes (although they have the right to use the rankings in the Massey Consensus), but I am personally more interested in the discussion of how teams ought to be seeded than how they might actually be seeded. I know that your job is much more to evaluate what you think the committee will actually do, but I want to reinforce that there are other evaluation approaches possible than the ones you present. Therefore I can disagree with your assessment that any particular team has a better or worse resume than WSU without disagreeing on the likelihood that they will be seeded higher than WSU by the committee.

I completely agree that what SHOULD happen and what WILL happen can be 2 separate discussions. I also agree that the massey concensus is a better individual analysis than a single RPI rating. If I had easy access to find a team's record vs 1-25, 26-50, etc. using Massey ratings, I would use that instead. Would that type of analysis be good in your mind?

[shox1989]

I'm still and advocate of using the Massey Comparison Site as the first gauge of team rankings. By averaging 52 different ranking systems, it gives you the result of a lot of different approaches to measuring team strength while avoiding outliers of any particular system. The reality is that comparing different teams in different conferences is not easy and the more data and approaches utilized, the better.

http://masseyratings.com/cb/compare.htm

I'm not saying that the committee has to seed directly from this composite ranking (although I wouldn't mind that), but that it is a more statistically objective starting point for seeding than most other methods.

I would love it if the committee seeded by the massey ratings. That would make us a 3 seed right now with a chance to move up to a 2 seed.

[ShockerFever]

Alabama is #27 in the composite ranking by Massey. Many bracketologists have them as a 10 seed. 10's should be #s 37-40. That's 10-13 spots off.
Texas is #40 in the composite rankings. Many bracketologists have them as a 12 seed. 12's should be #s 45-48. That's 5-8 spots off.

WSU's 5-8 spots off isn't that big of a deal.

Uh, your justification is that the other brackets are off? All you did was prove more overrated BS. Texas is not an NCAA Tournament team. I have no clue why that team is in the field. They are about as watered down as they come. Again, same thing with Bama. "Most bracketologists" have them as a 10 seed. They should be a 12. They are teetering.

You basically just made an argument out of everything I was arguing against. My whole point was saying the "expert" bracketologists are overrating a lot of teams based on certain affiliations or names on jerseys. And your counterpoint was "that's because most bracketologists have them there".

What the hell was that? You just basically cemented my point. Thanks.

[Jamar Howard 4 President]

I would love it if the committee seeded by the massey ratings. That would make us a 3 seed right now with a chance to move up to a 2 seed.

My question is, would you like to see the committee use the vs top 25, vs 26-50, etc. style of looking at things if the massey ratings were used instead of the rpi to establish those categories? Or, would you like to just use the massey rating directly and let the computer select the field and seeds with no human input?

[ShockerFever]

The infatuation with the 1-25, 26-50 breakdowns is amazing. Things can be weighed a whole array of directions with these little BCS-skewed nuggets. A team could have one win over a 25 team and another could have wins over teams 26,27,28, and 29 and by breaking them into these arbitrary columns, the second team is adjusted to look "worse". Great logic and skewing.

[ShockerFever]

My question is, would you like to see the committee use the vs top 25, vs 26-50, etc. style of looking at things if the massey ratings were used instead of the rpi to establish those categories? Or, would you like to just use the massey rating directly and let the computer select the field and seeds with no human input?

Might as well let the computers decide. Then at least nobody could blame it on any type of bias.

[Jamar Howard 4 President]

Uh, your justification is that the other brackets are off? All you did was prove more overrated BS. Texas is not an NCAA Tournament team. I have no clue why that team is in the field. They are about as watered down as they come. Again, same thing with Bama. "Most bracketologists" have them as a 10 seed. They should be a 12. They are teetering.

You basically just made an argument out of everything I was arguing against. My whole point was saying the "expert" bracketologists are overrating a lot of teams based on certain affiliations or names on jerseys. And your counterpoint was "that's because most bracketologists have them there".

What the hell was that? You just basically cemented my point. Thanks.

You completely mis-understood my point. Texas is rated better in the massey computer ratings than the seed most bracketologists are giving them. Same with Alabama. When WSU gets less than what the computers say they should, you call it BCS bias. What do you call it when BCS teams get less than what the computers say they should?

[shox1989]

Might as well let the computers decide. Then at least nobody could blame it on any type of bias.

That is my understanding what they do for the NCAA division 1 College Hockey Championship. They use a computer ratings system to choose the 11 at large teams that get to compete to get to the frozen 4. No human bias, every school knows what goes into the computer ratings and they schedule to try to improve their chances.

[The Mad Hatter]

I completely agree that what SHOULD happen and what WILL happen can be 2 separate discussions. I also agree that the massey concensus is a better individual analysis than a single RPI rating. If I had easy access to find a team's record vs 1-25, 26-50, etc. using Massey ratings, I would use that instead. Would that type of analysis be good in your mind?

I think it would be an improvement, however I still have reservations about using a team record vs. 1-25, 26-50, etc. approach. The ranking systems used in Massey are built to adjust for the schedule you played. Focusing on top 25 or top 50 wins only works to push the system back in favor of teams with a lot of opportunities to play other top teams. That destroys the very reason for these systems to exist (i.e. to objectively compare teams with much different kinds of schedules). Furthermore, I believe that there is an inherent psychological bias towards rewarding good wins more than we penalize bad losses, even if we don't like to admit it. Once again, the statistical measures do a better job applying objectivity to those evaluations.

Record vs. 1-25, 25-50, etc. should not be viewed as more important than the team's own ranking, IMHO. I have no problem using it as a tie-break between similarly ranked teams (for instance - WSU, Marquette, and Baylor all have a median score of 10 and close mean scores with Missouri in front by a wide margin and Wisconsin behind by a wide margin. Record vs. establish the seeding for WSU, Marquette, and Baylor makes perfect sense, but not to move a team well behind one of these ahead of them).

I don't necessarily want Massey to be the final say with no human input. What I do want is for it to serve as the starting point so that the committee has to argue to change that ranking rather than have a different ranking and trying to argue in favor of Massey for change.

[Rlh04d]

Might as well let the computers decide. Then at least nobody could blame it on any type of bias.

We'd just complain about what computers to use.

I think the computers get more blame than anything else in the BCS football system.

[Jamar Howard 4 President]

I think it would be an improvement, however I still have reservations about using a team record vs. 1-25, 26-50, etc. approach. The ranking systems used in Massey are built to adjust for the schedule you played. Focusing on top 25 or top 50 wins only works to push the system back in favor of teams with a lot of opportunities to play other top teams. That destroys the very reason for these systems to exist (i.e. to objectively compare teams with much different kinds of schedules). Furthermore, I believe that there is an inherent psychological bias towards rewarding good wins more than we penalize bad losses, even if we don't like to admit it. Once again, the statistical measures do a better job applying objectivity to those evaluations.

Record vs. 1-25, 25-50, etc. should not be viewed as more important than the team's own ranking, IMHO. I have no problem using it as a tie-break between similarly ranked teams (for instance - WSU, Marquette, and Baylor all have a median score of 10 with Missouri in front by a wide margin and Wisconsin behind by a wide margin. Record vs. establish the seeding for WSU, Marquette, and Baylor makes perfect sense, but not to move a team well behind one of these ahead of them).

Fair enough.

[The Mad Hatter]

Fair enough.

And same to you. I get that there is no perfect approach to team evaluation and you and other bracketologists have the added problem of trying to predict how the committee will act, not how they should act (even if that could be objectively answered).

[ShockerFever]

You completely mis-understood my point. Texas is rated better in the massey computer ratings than the seed most bracketologists are giving them. Same with Alabama. When WSU gets less than what the computers say they should, you call it BCS bias. What do you call it when BCS teams get less than what the computers say they should?

You're right, I did. Forget everything I typed. If Massey says that's where they should be, then I agree with it. Although, other factors do need to be considered... like recent trends, conference records, injuries, suspensions. etc.

[ShockerFever]

We'd just complain about what computers to use.

I think the computers get more blame than anything else in the BCS football system.

Maybe. But the primary numbers used to generate the BCS Standings are... you guessed it... the human polls. You wouldn't use that human BS bias here.