Since this is a discussion that has come up in several different threads, and seems to derail all of them, I'll try to isolate further discussion to this thread.

As Tiger fans, we are all painfully aware of how Tiger related this topic is. The Committee seems to use the RPI as a crutch for Big-6 teams whose performances have otherwise been mediocre, while arguing for the "eye test" as an excuse to discredit a non-Big 6 team with a high RPI. See: The Memphis Tigers of 2012.

For those Tiger fans who may wish to enlighten themselves both on the selection committee's process as well as a better way to evaluate teams, here is some reading:

RPI: Why it's a lousy way to pick teams for the NCAA Tournament

Ridiculous Performance Indicator

Just give me the d*** field

For those Tiger fans (we know who you are), who aren't interested in enlightenment and would prefer to keep their minds securely clamped closed, feel free to use this thread for wailing and gnashing of teeth.

One reason why the RPI is important is that an index like the Pomeroy rankings could actually send teams in over other teams even if they were not deserving. For instance, say team A and team B had the exact same schedules and team A went undefeated and team B lost one game. Pomeroy could actually say team B should have a higher seed or get in the tournament before team B. Here's how: Even though team B lost a game, Pomeroy may send them in ahead of team A because team B beat their opponents so bad that it overrode their one loss so much that it still gave them a higher rating than team A who was undefeated. This is NOT how teams should get in over other teams. Even though team A didn't run through their schedule as convincingly as team B did theirs, they should still get in over team B because the didn't lose a game. It should still matter who you beat, not how well you did per possession.


Another reason - the RPI is still good to keep people honest. For instance, two years ago a crappy Alabama team would have made the field if it were not for the RPI. The media were jumping on the bandwagon with them because they had won a few games in the SEC against name opponents, but Alabama’s problem that year was that their SOS was so bad because they played a terrible non conference slat, and the RPI wasn't having it. These days anlysts/media/selection committee want to forget about how well you did against crap teams. It should still matter, and the RPI does this. The RPI keeps allot of the BCS teams the media wants to jump on that are not deserving, and keeps non BCS teams in that should be in that have a top 30 RPI or so, but may not have beaten big name teams. The RPI is still important.

Don't get me wrong, we should not get rid of the Pomeroy index, but we certainly should not get rid of the RPI either.

(02-28-2013 11:40 AM)memphis mania Wrote:  One reason why the RPI is important is that an index like the Pomeroy rankings could actually send teams in over other teams even if they were not deserving. For instance, say team A and team B had the exact same schedules and team A went undefeated and team B lost one game. Pomeroy could actually say team B should have a higher seed or get in the tournament before team B. Here's how: Even though team B lost a game, Pomeroy may send them in ahead of team A because team B beat their opponents so bad that it overrode their one loss so much that it still gave them a higher rating than team A who was undefeated. This is NOT how teams should get in over other teams. Even though team A didn't run through their schedule as convincingly as team B did theirs, they should still get in over team B. It should still matter who you beat, not how well you did per possession.

While this is technically true, if team A beat every team by 1 point, while team B beat every team by 20 points, and only lost their game by one point (say, when their star player got hurt), which team is actually the better team? Which team would most likely win if A and B played? Team B.

Selection should be about picking the better team. Always.

(02-28-2013 11:40 AM)memphis mania Wrote:  Another reason - the RPI is still good to keep people honest. For instance, two years ago a crappy Alabama team would have made the field if it were not for the RPI. The media were jumping on the bandwagon with them because they had won a few games in the SEC against name opponents, but Alabama’s problem that year was that their SOS was so bad because they played a terrible non conference slat, and the RPI wasn't having it, and for good reason too. The RPI keeps allot of the BCS teams the media wants to jump on that are not deserving, and keeps non BCS teams in that should be in that have a top 30 RPI or so, but may not have beaten big name teams. The RPI is still important.

Don't get me wrong, we should not get rid of the Pomeroy index, but we certainly should not get rid of the RPI either.

Kenpom, Sagarin and the like all have SOS as well.

(02-28-2013 11:17 AM)MemphisCanes Wrote:  Since this is a discussion that has come up in several different threads, and seems to derail all of them, I'll try to isolate further discussion to this thread.

As Tiger fans, we are all painfully aware of how Tiger related this topic is. The Committee seems to use the RPI as a crutch for Big-6 teams whose performances have otherwise been mediocre, while arguing for the "eye test" as an excuse to discredit a non-Big 6 team with a high RPI. See: The Memphis Tigers of 2012.

For those Tiger fans who may wish to enlighten themselves both on the selection committee's process as well as a better way to evaluate teams, here is some reading:

RPI: Why it's a lousy way to pick teams for the NCAA Tournament

Ridiculous Performance Indicator

Just give me the d*** field

For those Tiger fans (we know who you are), who aren't interested in enlightenment and would prefer to keep their minds securely clamped closed, feel free to use this thread for wailing and gnashing of teeth.

I'll sum up what you said to save everyone from wasting time reading it. Nobody cares about any of the other measures you listed, or any of the links you provided. Only the RPI is referenced by the selection committee.

(02-28-2013 11:45 AM)MemphisCanes Wrote:  

(02-28-2013 11:40 AM)memphis mania Wrote:  One reason why the RPI is important is that an index like the Pomeroy rankings could actually send teams in over other teams even if they were not deserving. For instance, say team A and team B had the exact same schedules and team A went undefeated and team B lost one game. Pomeroy could actually say team B should have a higher seed or get in the tournament before team B. Here's how: Even though team B lost a game, Pomeroy may send them in ahead of team A because team B beat their opponents so bad that it overrode their one loss so much that it still gave them a higher rating than team A who was undefeated. This is NOT how teams should get in over other teams. Even though team A didn't run through their schedule as convincingly as team B did theirs, they should still get in over team B. It should still matter who you beat, not how well you did per possession.

While this is technically true, if team A beat every team by 1 point, while team B beat every team by 20 points, and only lost their game by one point (say, when their star player got hurt), which team is actually the better team? Which team would most likely win if A and B played? Team B.

Selection should be about picking the better team. Always.

(02-28-2013 11:40 AM)memphis mania Wrote:  Another reason - the RPI is still good to keep people honest. For instance, two years ago a crappy Alabama team would have made the field if it were not for the RPI. The media were jumping on the bandwagon with them because they had won a few games in the SEC against name opponents, but Alabama’s problem that year was that their SOS was so bad because they played a terrible non conference slat, and the RPI wasn't having it, and for good reason too. The RPI keeps allot of the BCS teams the media wants to jump on that are not deserving, and keeps non BCS teams in that should be in that have a top 30 RPI or so, but may not have beaten big name teams. The RPI is still important.

Don't get me wrong, we should not get rid of the Pomeroy index, but we certainly should not get rid of the RPI either.

Kenpom, Sagarin and the like all have SOS as well.

SOS based on both teams scoring garbage points at the end of games and mailing it in against crap teams. In other words; meaningless stuff that doesn't matter.

(02-28-2013 11:47 AM)Stammers Wrote:  

(02-28-2013 11:17 AM)MemphisCanes Wrote:  Since this is a discussion that has come up in several different threads, and seems to derail all of them, I'll try to isolate further discussion to this thread.

As Tiger fans, we are all painfully aware of how Tiger related this topic is. The Committee seems to use the RPI as a crutch for Big-6 teams whose performances have otherwise been mediocre, while arguing for the "eye test" as an excuse to discredit a non-Big 6 team with a high RPI. See: The Memphis Tigers of 2012.

For those Tiger fans who may wish to enlighten themselves both on the selection committee's process as well as a better way to evaluate teams, here is some reading:

RPI: Why it's a lousy way to pick teams for the NCAA Tournament

Ridiculous Performance Indicator

Just give me the d*** field

For those Tiger fans (we know who you are), who aren't interested in enlightenment and would prefer to keep their minds securely clamped closed, feel free to use this thread for wailing and gnashing of teeth.

I'll sum up what you said to save everyone from wasting time reading it. Nobody cares about any of the other measures you listed, or any of the links you provided. Only the RPI is referenced by the selection committee.

Nice of you to join us, Stams.

(02-28-2013 11:45 AM)MemphisCanes Wrote:  

(02-28-2013 11:40 AM)memphis mania Wrote:  One reason why the RPI is important is that an index like the Pomeroy rankings could actually send teams in over other teams even if they were not deserving. For instance, say team A and team B had the exact same schedules and team A went undefeated and team B lost one game. Pomeroy could actually say team B should have a higher seed or get in the tournament before team B. Here's how: Even though team B lost a game, Pomeroy may send them in ahead of team A because team B beat their opponents so bad that it overrode their one loss so much that it still gave them a higher rating than team A who was undefeated. This is NOT how teams should get in over other teams. Even though team A didn't run through their schedule as convincingly as team B did theirs, they should still get in over team B. It should still matter who you beat, not how well you did per possession.

While this is technically true, if team A beat every team by 1 point, while team B beat every team by 20 points, and only lost their game by one point (say, when their star player got hurt), which team is actually the better team? Which team would most likely win if A and B played? Team B.

Selection should be about picking the better team. Always.

(02-28-2013 11:40 AM)memphis mania Wrote:  Another reason - the RPI is still good to keep people honest. For instance, two years ago a crappy Alabama team would have made the field if it were not for the RPI. The media were jumping on the bandwagon with them because they had won a few games in the SEC against name opponents, but Alabama’s problem that year was that their SOS was so bad because they played a terrible non conference slat, and the RPI wasn't having it, and for good reason too. The RPI keeps allot of the BCS teams the media wants to jump on that are not deserving, and keeps non BCS teams in that should be in that have a top 30 RPI or so, but may not have beaten big name teams. The RPI is still important.

Don't get me wrong, we should not get rid of the Pomeroy index, but we certainly should not get rid of the RPI either.

Kenpom, Sagarin and the like all have SOS as well.

Nope, sorry. The team that had more wins against the same schedule should get in. It's the way the playoffs work. The NFL doesn't say such and such team doesn't deserve to be in over the other because the other team lost their QB. It should at least be a mix. I like it how they have it now, but I wish they would put more emphasis on taking care of business, but they don't care.

For instance:
You could have team A lose to the 300th ranked team in the country but beat the 25th team in the country. You could have team B beat the 300th team in the country but lose to the 25th. They will take team A because the beat the 25th even though they lost to the 300th. It should cancel out, but it doesn't. This is the main thing that should change imo.

(02-28-2013 11:49 AM)Stammers Wrote:  

(02-28-2013 11:45 AM)MemphisCanes Wrote:  

(02-28-2013 11:40 AM)memphis mania Wrote:  One reason why the RPI is important is that an index like the Pomeroy rankings could actually send teams in over other teams even if they were not deserving. For instance, say team A and team B had the exact same schedules and team A went undefeated and team B lost one game. Pomeroy could actually say team B should have a higher seed or get in the tournament before team B. Here's how: Even though team B lost a game, Pomeroy may send them in ahead of team A because team B beat their opponents so bad that it overrode their one loss so much that it still gave them a higher rating than team A who was undefeated. This is NOT how teams should get in over other teams. Even though team A didn't run through their schedule as convincingly as team B did theirs, they should still get in over team B. It should still matter who you beat, not how well you did per possession.

While this is technically true, if team A beat every team by 1 point, while team B beat every team by 20 points, and only lost their game by one point (say, when their star player got hurt), which team is actually the better team? Which team would most likely win if A and B played? Team B.

Selection should be about picking the better team. Always.

(02-28-2013 11:40 AM)memphis mania Wrote:  Another reason - the RPI is still good to keep people honest. For instance, two years ago a crappy Alabama team would have made the field if it were not for the RPI. The media were jumping on the bandwagon with them because they had won a few games in the SEC against name opponents, but Alabama’s problem that year was that their SOS was so bad because they played a terrible non conference slat, and the RPI wasn't having it, and for good reason too. The RPI keeps allot of the BCS teams the media wants to jump on that are not deserving, and keeps non BCS teams in that should be in that have a top 30 RPI or so, but may not have beaten big name teams. The RPI is still important.

Don't get me wrong, we should not get rid of the Pomeroy index, but we certainly should not get rid of the RPI either.

Kenpom, Sagarin and the like all have SOS as well.

SOS based on both teams scoring garbage points at the end of games and mailing it in against crap teams. In other words; meaningless stuff that doesn't matter.

Just to see if I understand your point, Team A should get the same credit for beating Team C by 1 point in triple OT, as Team B gets for beating Team C by 20?

(02-28-2013 11:51 AM)MemphisCanes Wrote:  

(02-28-2013 11:47 AM)Stammers Wrote:  

(02-28-2013 11:17 AM)MemphisCanes Wrote:  Since this is a discussion that has come up in several different threads, and seems to derail all of them, I'll try to isolate further discussion to this thread.

As Tiger fans, we are all painfully aware of how Tiger related this topic is. The Committee seems to use the RPI as a crutch for Big-6 teams whose performances have otherwise been mediocre, while arguing for the "eye test" as an excuse to discredit a non-Big 6 team with a high RPI. See: The Memphis Tigers of 2012.

For those Tiger fans who may wish to enlighten themselves both on the selection committee's process as well as a better way to evaluate teams, here is some reading:

RPI: Why it's a lousy way to pick teams for the NCAA Tournament

Ridiculous Performance Indicator

Just give me the d*** field

For those Tiger fans (we know who you are), who aren't interested in enlightenment and would prefer to keep their minds securely clamped closed, feel free to use this thread for wailing and gnashing of teeth.

I'll sum up what you said to save everyone from wasting time reading it. Nobody cares about any of the other measures you listed, or any of the links you provided. Only the RPI is referenced by the selection committee.

Nice of you to join us, Stams.

My pleasure. Just making sure that if there are a few stray non agendites reading this, that don't know better; that they don't buy into the garbage.

(02-28-2013 11:52 AM)memphis mania Wrote:  

(02-28-2013 11:45 AM)MemphisCanes Wrote:  

(02-28-2013 11:40 AM)memphis mania Wrote:  One reason why the RPI is important is that an index like the Pomeroy rankings could actually send teams in over other teams even if they were not deserving. For instance, say team A and team B had the exact same schedules and team A went undefeated and team B lost one game. Pomeroy could actually say team B should have a higher seed or get in the tournament before team B. Here's how: Even though team B lost a game, Pomeroy may send them in ahead of team A because team B beat their opponents so bad that it overrode their one loss so much that it still gave them a higher rating than team A who was undefeated. This is NOT how teams should get in over other teams. Even though team A didn't run through their schedule as convincingly as team B did theirs, they should still get in over team B. It should still matter who you beat, not how well you did per possession.

While this is technically true, if team A beat every team by 1 point, while team B beat every team by 20 points, and only lost their game by one point (say, when their star player got hurt), which team is actually the better team? Which team would most likely win if A and B played? Team B.

Selection should be about picking the better team. Always.

(02-28-2013 11:40 AM)memphis mania Wrote:  Another reason - the RPI is still good to keep people honest. For instance, two years ago a crappy Alabama team would have made the field if it were not for the RPI. The media were jumping on the bandwagon with them because they had won a few games in the SEC against name opponents, but Alabama’s problem that year was that their SOS was so bad because they played a terrible non conference slat, and the RPI wasn't having it, and for good reason too. The RPI keeps allot of the BCS teams the media wants to jump on that are not deserving, and keeps non BCS teams in that should be in that have a top 30 RPI or so, but may not have beaten big name teams. The RPI is still important.

Don't get me wrong, we should not get rid of the Pomeroy index, but we certainly should not get rid of the RPI either.

Kenpom, Sagarin and the like all have SOS as well.

Nope, sorry. The team that had more wins against the same schedule should get in. It's the way the playoffs work. The NFL doesn't say such and such team doesn't deserve to be in over the other because the other team lost their QB. It should at least be a mix. I like it how they have it now, but I wish they would put more emphasis on taking care of business, but they don't care.

For instance:
You could have team A lose to the 300th ranked team in the country but beat the 25th team in the country. You could have team B beat the 300th team in the country but lose to the 25th. They will take team A because the beat the 25th even though they lost to the 300th. It should cancel out, but it doesn't. This is the main thing that should change imo.

I understand your point, but you're effectively advocating putting an inferior team into the dance over a superior team.

The NFL analogy doesn't work, as the NFL doesn't seed teams and doesn't take SOS into account, which we both concede is important.

In your "for instance", both of those teams would have the same RPI. Kenpom or Sagarin would be able to tell you more about each team in that scenario, but RPI would just have them tied.

(02-28-2013 11:56 AM)MemphisCanes Wrote:  

(02-28-2013 11:52 AM)memphis mania Wrote:  

(02-28-2013 11:45 AM)MemphisCanes Wrote:  

(02-28-2013 11:40 AM)memphis mania Wrote:  One reason why the RPI is important is that an index like the Pomeroy rankings could actually send teams in over other teams even if they were not deserving. For instance, say team A and team B had the exact same schedules and team A went undefeated and team B lost one game. Pomeroy could actually say team B should have a higher seed or get in the tournament before team B. Here's how: Even though team B lost a game, Pomeroy may send them in ahead of team A because team B beat their opponents so bad that it overrode their one loss so much that it still gave them a higher rating than team A who was undefeated. This is NOT how teams should get in over other teams. Even though team A didn't run through their schedule as convincingly as team B did theirs, they should still get in over team B. It should still matter who you beat, not how well you did per possession.

While this is technically true, if team A beat every team by 1 point, while team B beat every team by 20 points, and only lost their game by one point (say, when their star player got hurt), which team is actually the better team? Which team would most likely win if A and B played? Team B.

Selection should be about picking the better team. Always.

(02-28-2013 11:40 AM)memphis mania Wrote:  Another reason - the RPI is still good to keep people honest. For instance, two years ago a crappy Alabama team would have made the field if it were not for the RPI. The media were jumping on the bandwagon with them because they had won a few games in the SEC against name opponents, but Alabama’s problem that year was that their SOS was so bad because they played a terrible non conference slat, and the RPI wasn't having it, and for good reason too. The RPI keeps allot of the BCS teams the media wants to jump on that are not deserving, and keeps non BCS teams in that should be in that have a top 30 RPI or so, but may not have beaten big name teams. The RPI is still important.

Don't get me wrong, we should not get rid of the Pomeroy index, but we certainly should not get rid of the RPI either.

Kenpom, Sagarin and the like all have SOS as well.

Nope, sorry. The team that had more wins against the same schedule should get in. It's the way the playoffs work. The NFL doesn't say such and such team doesn't deserve to be in over the other because the other team lost their QB. It should at least be a mix. I like it how they have it now, but I wish they would put more emphasis on taking care of business, but they don't care.

For instance:
You could have team A lose to the 300th ranked team in the country but beat the 25th team in the country. You could have team B beat the 300th team in the country but lose to the 25th. They will take team A because the beat the 25th even though they lost to the 300th. It should cancel out, but it doesn't. This is the main thing that should change imo.

I understand your point, but you're effectively advocating putting an inferior team into the dance over a superior team.

The NFL analogy doesn't work, as the NFL doesn't seed teams and doesn't take SOS into account, which we both concede is important.

In your "for instance", both of those teams would have the same RPI. Kenpom or Sagarin would be able to tell you more about each team in that scenario, but RPI would just have them tied.

Why keep tab on who won the games then? If it's just by per possesion, who cares who won?

(02-28-2013 12:03 PM)memphis mania Wrote:  Why keep tab on who won the games then? If it's just by per possesion, who cares who won?

"Margin of Victory" inherently implies that victories are accounted for in the metric. Strength of Schedule is accounted as well.

But you seem to be focusing directly on Pomeroy's system (efficiency), what about Sagarin, Massey, BMI?

(02-28-2013 12:07 PM)MemphisCanes Wrote:  

(02-28-2013 12:03 PM)memphis mania Wrote:  Why keep tab on who won the games then? If it's just by per possesion, who cares who won?

"Margin of Victory" inherently implies that victories are accounted for in the metric. Strength of Schedule is accounted as well.

But you seem to be focusing directly on Pomeroy's system (efficiency), what about Sagarin, Massey, BMI?

The only things that matter are wins and losses. By being measured against your opponents' wins and losses and their opponents' wins and losses, you get a very accurate and very fair measure of who the best teams are.

Adding all the other stuff doesn't matter and that's why the selection committee doesn't give a crap about it. It's meaningless.

(02-28-2013 12:18 PM)Stammers Wrote:  

(02-28-2013 12:07 PM)MemphisCanes Wrote:  

(02-28-2013 12:03 PM)memphis mania Wrote:  Why keep tab on who won the games then? If it's just by per possesion, who cares who won?

"Margin of Victory" inherently implies that victories are accounted for in the metric. Strength of Schedule is accounted as well.

But you seem to be focusing directly on Pomeroy's system (efficiency), what about Sagarin, Massey, BMI?

The only things that matter are wins and losses. By being measured against your opponents' wins and losses and their opponents' wins and losses, you get a very accurate and very fair measure of who the best teams are.

Adding all the other stuff doesn't matter and that's why the selection committee doesn't give a crap about it. It's meaningless.

But it's not accurate.



The RPI is not the Real Predictive Indicator

The RPI is the worst indicator of "which team is better" in this study, clearly worse than Kenpom in all categories.

(02-28-2013 11:47 AM)Stammers Wrote:  Only the RPI is referenced by the selection committee.

Stammers, I actually thought pretty much the same thing until reading this:

http://sportsillustrated.cnn.com/college...-thoughts/

"Debunking Selection Sunday's biggest myths" - Seth Davis (Mock NCAA Selection Participant"

Quote:I won't get into the tiresome debate about the RPI, but I do want to point out that much of the complaining is based on a myth -- namely, that the committee ignores the other valid metrics that are readily available. In fact, every member of the committee is not only free to look at other rankings, they are encouraged to do so.

On the NCAA's own internal website, there is a link that reads "Links to Other Sites." Those sites are: NCAA RPI, Polls, Sagarin Ratings, CollegeRPI.com, RPI Ratings, LRMC Rankings, Pomeroy Rankings, Composite Ranking, Conference Monitoring, ESPN BPI.

Jeanie Boyd, a veteran NCAA staffer who works closely with the committee from inside the room, told us that she has seen members arrive at selection week with their own rankings that aren't among those listed. Those other metrics are definitely a part of the discussion. If there is a big discrepancy between the rankings, the committee will go into a discussion to probe further into the numbers.

caveat, this was a MOCK process and for all we know, the committee throws darts at boards for 6 days and drinks beer.

I think the point is, the committee uses a vast range of inputs to come to their conclusions. However, RPI is certainly 'the industry standard' and a place to start the discussion. And a great RPI doesn't mean automatic acceptance (the most glaring case I can remember is the Missouri State - 20 RPI snub).

So, I'd say:

1. RPI is very important and having a good one is a major plus, while having a bad one is often fatal.
2. It is not the sole metric used, but it is the 'industry standard'.
3. It is irrelevant our own feelings on the RPI (I think you can game it), it is still a factor.
4. However, it's only one factor.

I'm curious about threads / comments that reference us as a bubble team or needing to win out or auto bid. Seems pretty rough to me. I think we're pretty much in, but it's 1 6-12 seed at stake.

As always, all of the above is just my very humble opinion.

(02-28-2013 12:22 PM)MemphisCanes Wrote:  

(02-28-2013 12:18 PM)Stammers Wrote:  

(02-28-2013 12:07 PM)MemphisCanes Wrote:  

(02-28-2013 12:03 PM)memphis mania Wrote:  Why keep tab on who won the games then? If it's just by per possesion, who cares who won?

"Margin of Victory" inherently implies that victories are accounted for in the metric. Strength of Schedule is accounted as well.

But you seem to be focusing directly on Pomeroy's system (efficiency), what about Sagarin, Massey, BMI?

The only things that matter are wins and losses. By being measured against your opponents' wins and losses and their opponents' wins and losses, you get a very accurate and very fair measure of who the best teams are.

Adding all the other stuff doesn't matter and that's why the selection committee doesn't give a crap about it. It's meaningless.

But it's not accurate.



The RPI is not the Real Predictive Indicator

The RPI is the worst indicator of "which team is better" in this study, clearly worse than Kenpom in all categories.

Whether or not it is better or worse as a predictive metric is fairly irrelevant as the NCAA clearly uses it. It may be imperfect, but it is a tool the powers that be use in selection and rankings.

However, it's just one tool. From the Seth Davis article, they seem to reference A LOT of other metrics to determine where to place teams...

Again, I'm not in the room. I think they just pick the minimum number of non-big 6 teams and then fill the rest with as many of the BCS 19-13 77 RPI types as they think we will swallow...

(02-28-2013 12:35 PM)Sundanceuiuc Wrote:  

(02-28-2013 11:47 AM)Stammers Wrote:  Only the RPI is referenced by the selection committee.

Stammers, I actually thought pretty much the same thing until reading this:

http://sportsillustrated.cnn.com/college...-thoughts/

"Debunking Selection Sunday's biggest myths" - Seth Davis (Mock NCAA Selection Participant"

Quote:I won't get into the tiresome debate about the RPI, but I do want to point out that much of the complaining is based on a myth -- namely, that the committee ignores the other valid metrics that are readily available. In fact, every member of the committee is not only free to look at other rankings, they are encouraged to do so.

On the NCAA's own internal website, there is a link that reads "Links to Other Sites." Those sites are: NCAA RPI, Polls, Sagarin Ratings, CollegeRPI.com, RPI Ratings, LRMC Rankings, Pomeroy Rankings, Composite Ranking, Conference Monitoring, ESPN BPI.

Jeanie Boyd, a veteran NCAA staffer who works closely with the committee from inside the room, told us that she has seen members arrive at selection week with their own rankings that aren't among those listed. Those other metrics are definitely a part of the discussion. If there is a big discrepancy between the rankings, the committee will go into a discussion to probe further into the numbers.

caveat, this was a MOCK process and for all we know, the committee throws darts at boards for 6 days and drinks beer.

I think the point is, the committee uses a vast range of inputs to come to their conclusions. However, RPI is certainly 'the industry standard' and a place to start the discussion. And a great RPI doesn't mean automatic acceptance (the most glaring case I can remember is the Missouri State - 20 RPI snub).

So, I'd say:

1. RPI is very important and having a good one is a major plus, while having a bad one is often fatal.
2. It is not the sole metric used, but it is the 'industry standard'.
3. It is irrelevant our own feelings on the RPI (I think you can game it), it is still a factor.
4. However, it's only one factor.

I'm curious about threads / comments that reference us as a bubble team or needing to win out or auto bid. Seems pretty rough to me. I think we're pretty much in, but it's 1 6-12 seed at stake.

As always, all of the above is just my very humble opinion.

People have shown him this in multiple threads. He'll ignore it, then probably insult you or make a comment about an agenda.

(02-28-2013 12:38 PM)Sundanceuiuc Wrote:  

(02-28-2013 12:22 PM)MemphisCanes Wrote:  

(02-28-2013 12:18 PM)Stammers Wrote:  

(02-28-2013 12:07 PM)MemphisCanes Wrote:  

(02-28-2013 12:03 PM)memphis mania Wrote:  Why keep tab on who won the games then? If it's just by per possesion, who cares who won?

"Margin of Victory" inherently implies that victories are accounted for in the metric. Strength of Schedule is accounted as well.

But you seem to be focusing directly on Pomeroy's system (efficiency), what about Sagarin, Massey, BMI?

The only things that matter are wins and losses. By being measured against your opponents' wins and losses and their opponents' wins and losses, you get a very accurate and very fair measure of who the best teams are.

Adding all the other stuff doesn't matter and that's why the selection committee doesn't give a crap about it. It's meaningless.

But it's not accurate.



The RPI is not the Real Predictive Indicator

The RPI is the worst indicator of "which team is better" in this study, clearly worse than Kenpom in all categories.

Whether or not it is better or worse as a predictive metric is fairly irrelevant as the NCAA clearly uses it. It may be imperfect, but it is a tool the powers that be use in selection and rankings.

However, it's just one tool. From the Seth Davis article, they seem to reference A LOT of other metrics to determine where to place teams...

Again, I'm not in the room. I think they just pick the minimum number of non-big 6 teams and then fill the rest with as many of the BCS 19-13 77 RPI types as they think we will swallow...

RPI is always quoted. They might look at other stuff, but every single time they are asked to justify their decisions; out comes the RPI.

(02-28-2013 12:38 PM)Sundanceuiuc Wrote:  Whether or not it is better or worse as a predictive metric is fairly irrelevant as the NCAA clearly uses it. It may be imperfect, but it is a tool the powers that be use in selection and rankings.

However, it's just one tool. From the Seth Davis article, they seem to reference A LOT of other metrics to determine where to place teams...

Again, I'm not in the room. I think they just pick the minimum number of non-big 6 teams and then fill the rest with as many of the BCS 19-13 77 RPI types as they think we will swallow...

We all know the NCAA uses RPI. They use it as "the main metric" when looking at a team's record. They can use others, but RPI is the "industry standard".

The argument is that it's subpar to most of the other measurements out there at determining the better team, and should be discarded.

(02-28-2013 12:38 PM)MemphisCanes Wrote:  

(02-28-2013 12:35 PM)Sundanceuiuc Wrote:  

(02-28-2013 11:47 AM)Stammers Wrote:  Only the RPI is referenced by the selection committee.

Stammers, I actually thought pretty much the same thing until reading this:

http://sportsillustrated.cnn.com/college...-thoughts/

"Debunking Selection Sunday's biggest myths" - Seth Davis (Mock NCAA Selection Participant"

Quote:I won't get into the tiresome debate about the RPI, but I do want to point out that much of the complaining is based on a myth -- namely, that the committee ignores the other valid metrics that are readily available. In fact, every member of the committee is not only free to look at other rankings, they are encouraged to do so.

On the NCAA's own internal website, there is a link that reads "Links to Other Sites." Those sites are: NCAA RPI, Polls, Sagarin Ratings, CollegeRPI.com, RPI Ratings, LRMC Rankings, Pomeroy Rankings, Composite Ranking, Conference Monitoring, ESPN BPI.

Jeanie Boyd, a veteran NCAA staffer who works closely with the committee from inside the room, told us that she has seen members arrive at selection week with their own rankings that aren't among those listed. Those other metrics are definitely a part of the discussion. If there is a big discrepancy between the rankings, the committee will go into a discussion to probe further into the numbers.

caveat, this was a MOCK process and for all we know, the committee throws darts at boards for 6 days and drinks beer.

I think the point is, the committee uses a vast range of inputs to come to their conclusions. However, RPI is certainly 'the industry standard' and a place to start the discussion. And a great RPI doesn't mean automatic acceptance (the most glaring case I can remember is the Missouri State - 20 RPI snub).

So, I'd say:

1. RPI is very important and having a good one is a major plus, while having a bad one is often fatal.
2. It is not the sole metric used, but it is the 'industry standard'.
3. It is irrelevant our own feelings on the RPI (I think you can game it), it is still a factor.
4. However, it's only one factor.

I'm curious about threads / comments that reference us as a bubble team or needing to win out or auto bid. Seems pretty rough to me. I think we're pretty much in, but it's 1 6-12 seed at stake.

As always, all of the above is just my very humble opinion.

People have shown him this in multiple threads. He'll ignore it, then probably insult you or make a comment about an agenda.

I don't think the specific article had been quoted, and as I said, I actually thought the same until recently (I can't process everything in my own life, I generally don't concern myself with the NCAA selection details, I just make educated guesses on watching lots of Selection Sundays)

I don't see an agenda in either of you actually:

You: The RPI is not a good metric (I'm inclined to agree)
Stammers: The RPI is the only metric used (Not true, but it is the first one they go to) and a high RPI should protect our bid and hopefully help our seed (both true I think).

The rest is really just noise, correct?