Seeding Methodology

[FirstDownB]

Has anyone heard a concise explanation from the committee about seeding methodology, particularly the mess in the 3-5 range?
And here I thought worst-case scenario would be the committee over-valuing RPI. Turns out they didn't really value any of the metrics. Seems to me they just went with personal preference.

Purdue
RPI: 14
BPI: 10
KenPom: 10
Sagarin: 9
NCAAT: 18

Iowa St
RPI: 21
BPI: 16
KenPom: 17
Sagarin: 19
NCAAT: 16

 

[BBG]

The only safe conclusion is that the committee had been drinking.

 

[Purdue Grad in Texas]

Has anyone heard a concise explanation from the committee about seeding methodology, particularly the mess in the 3-5 range?
And here I thought worst-case scenario would be the committee over-valuing RPI. Turns out they didn't really value any of the metrics. Seems to me they just went with personal preference.

Purdue
RPI: 14
BPI: 10
KenPom: 10
Sagarin: 9
NCAAT: 18

Iowa St
RPI: 21
BPI: 16
KenPom: 17
Sagarin: 19
NCAAT: 16

Between those two teams.....obviously pretty close, but maybe just better top wins. Iowa State beat both Kansas and Oklahoma, slightly better than Purdue's wins over Michigan State and Maryland. Iowa State also beat Iowa and Texas while Purdue did beat Wisconsin twice.

Although I'm no big fan of Calipari, he had a good point saying the criteria is a moving target from year-to-year.

The better teams at handling the tournament are eventually going to win......should see some interesting results. Lace em up, play whoever is in front of you, and go down swinging.

 

[FirstDownB]

Between those two teams.....obviously pretty close, but maybe just better top wins. Iowa State beat both Kansas and Oklahoma, slightly better than Purdue's wins over Michigan State and Maryland. Iowa State also beat Iowa and Texas while Purdue did beat Wisconsin twice.

Although I'm no big fan of Calipari, he had a good point saying the criteria is a moving target from year-to-year.

The better teams at handling the tournament are eventually going to win......should see some interesting results. Lace em up, play whoever is in front of you, and go down swinging.

I agree with your last sentence and after upsets some things will change anyway. Happens every year.. But if you're the committee and going to go to all that trouble to order the teams, might was well explain your methodology so all can understand.

Side note: I still think we need to schedule fewer of the RPI 300+ type teams. Even beating those teams by 40 hurts you this time of year. You know years in advance that NC A&T and Howard are going to be dog crap.

 

[Dryfly88]

Between those two teams.....obviously pretty close, but maybe just better top wins. Iowa State beat both Kansas and Oklahoma, slightly better than Purdue's wins over Michigan State and Maryland. Iowa State also beat Iowa and Texas while Purdue did beat Wisconsin twice.

Although I'm no big fan of Calipari, he had a good point saying the criteria is a moving target from year-to-year.

The better teams at handling the tournament are eventually going to win......should see some interesting results. Lace em up, play whoever is in front of you, and go down swinging.

Tex, I agree with you about Calipari. Can't stand him, but have to agree with his rant on ESPN after the bracket was announced. There is no method to this madness (see what I did there)! Their locked in a room for all those hours the least they can do is give a logical explanation for what method they are using.

While I enjoy the fact that IU and KY are looking at playing each other in round of 32, it makes 0 sense. You have two power 5 schools that won/tied for their conference and one that took the conference tournament playing in the round of 32? That makes no sense at all.

It's like they have so many criteria they can explain away any bad decision. Needs to be streamlined and have some accountability for their decisions.

 

[Purdue Grad in Texas]

I agree with your last sentence and after upsets some things will change anyway. Happens every year.. But if you're the committee and going to go to all that trouble to order the teams, might was well explain your methodology so all can understand.

Side note: I still think we need to schedule fewer of the RPI 300+ type teams. Even beating those teams by 40 hurts you this time of year. You know years in advance that NC A&T and Howard are going to be dog crap.

yep....the "so-called" bottom feeders damage SOS, RPI, etc. But, you have to fill out a schedule. I don't know if a school can totally avoid them unless it's willing to go on the road more often in the non-conference. I would like to see Purdue try to get some tough non-conf home/home series the way Michigan State did a few years back and drop one or two of the really low-ranked opponents. Purdue didn't really have any bad losses other than maybe Illinois (and that's a conference game), but it appears that the committee felt Purdue didn't have enough "big" wins to justify better than a #5-seed.

 

[IamBob]

Has anyone heard a concise explanation from the committee about seeding methodology, particularly the mess in the 3-5 range?
And here I thought worst-case scenario would be the committee over-valuing RPI. Turns out they didn't really value any of the metrics. Seems to me they just went with personal preference.

Purdue
RPI: 14
BPI: 10
KenPom: 10
Sagarin: 9
NCAAT: 18

Iowa St
RPI: 21
BPI: 16
KenPom: 17
Sagarin: 19
NCAAT: 16

Don't worry about it and focus on beating Little Rock. Ain't nothin gunna change about it by critiquing the committee.

It's always going to be better than the 80's. Didn't we have to play at LSU one year and at Memphis another?

 

[TopSecretBoiler]

To quote one of my favorite movies: "I don't see any method at all."

 

[Purdue Grad in Texas]

Don't worry about it and focus on beating Little Rock. Ain't nothin gunna change about it by critiquing the committee.

It's always going to be better than the 80's. Didn't we have to play at LSU one year and at Memphis another?

Yessiree, Bob. Some of us posters can recall those days......sometimes it's a little foggy. 1986.....Purdue as a #6 seed opens up @ LSU (the 11th seed) and loses in 2OT. LSU went on to make the Final Four, beating Kentucky in the regional Final.

And for those thinking the Big Ten conference got shafted yesterday....consider 1984 (when the bracket rules were different and it wasn't yet a 64-team field but 52). Co-Big 10 Champ Purdue had a first-round bye (as a 3-seed) and waiting in the second round ended up being Memphis @ Memphis, who then advanced to the regional in St. Louis. The other Co-Big 10 Champ? Illinois......they did get a 2-seed but ended up playing the top-seed Kentucky in the Regional Final @ Rupp Arena. The Illini gave them a very good game but eventually lost.

As imperfect as it is today, at least there are not those types of advantages for any teams like in the earlier days.

 

[SIBoiler2]

The only rule that is written in stone is "thou shalt always seed Duke higher than they deserve". Once that is satisfied just throw darts at one board with teams and one with brackets until it's filled.

 

[Purdue Grad in Texas]

The only rule that is written in stone is "thou shalt always seed Duke higher than they deserve". Once that is satisfied just throw darts at one board with teams and one with brackets until it's filled.

and "thouest shall play in thy Greensboro, thy Charlotte or thy Raleigh unlessest thy Tarheels are such higher ranked.....thennest thou shall be kings of that county untillest the next succeeding bracket year."

So let it be written.....

 

[cprh9u]

I thought we should be a 4, but even if we "deserved" it, what does that matter? We got a very nice bracket, and even if we swapped with Iowa State, we would still meet them in the round of 32 if both teams do what they should. If we lose to Arkansas Little Rock, nobody should be blaming the committee. This is not CMP's best team (the team where Robbie went down at the Barn) but it's fairly close, and this tourney lines up for us to have our best chance at an Elite 8 or Final 4 over the past 15+ years.

 

[FirstDownB]

yep....the "so-called" bottom feeders damage SOS, RPI, etc. But, you have to fill out a schedule. I don't know if a school can totally avoid them unless it's willing to go on the road more often in the non-conference. I would like to see Purdue try to get some tough non-conf home/home series the way Michigan State did a few years back and drop one or two of the really low-ranked opponents. Purdue didn't really have any bad losses other than maybe Illinois (and that's a conference game), but it appears that the committee felt Purdue didn't have enough "big" wins to justify better than a #5-seed.

Sorry, but saying "you have to fill out a schedule" is a cop out. There are plenty of mid major and low major teams who want the exposure and experience of playing a B1G team like Purdue. Just avoid the absolute bottom feeder teams and conferences (MEAC, SWAC, Southland) and you'll see a bump in RPI and SOS.

 

[Dryfly88]

I thought we should be a 4, but even if we "deserved" it, what does that matter? We got a very nice bracket, and even if we swapped with Iowa State, we would still meet them in the round of 32 if both teams do what they should. If we lose to Arkansas Little Rock, nobody should be blaming the committee. This is not CMP's best team (the team where Robbie went down at the Barn) but it's fairly close, and this tourney lines up for us to have our best chance at an Elite 8 or Final 4 over the past 15+ years.

I thought by the metrics we should be a #3 and was convinced we would get that (shows what I know).

That said, I agree with your comment. We have a very winnable bracket set up for us and no excuses. Just play like we did yesterday (for the most part) and we will be fine.

 

[Purdue Grad in Texas]

I thought we should be a 4, but even if we "deserved" it, what does that matter? We got a very nice bracket, and even if we swapped with Iowa State, we would still meet them in the round of 32 if both teams do what they should. If we lose to Arkansas Little Rock, nobody should be blaming the committee. This is not CMP's best team (the team where Robbie went down at the Barn) but it's fairly close, and this tourney lines up for us to have our best chance at an Elite 8 or Final 4 over the past 15+ years.

agree here.....given the "ups" and "downs" of this season and the difficulties from a few years back, the Boilers have an excellent opportunity in front of them.....could have been a lot worse from a match-up stand-point. one thing, however, if the Boilers don't bring a focused effort, or have a poor shooting night......could be a really short tournament......no mystery there.

 

[Grover]

I am yet to see any metric that has Purdue this low

 

[FirstDownB]

Don't worry about it and focus on beating Little Rock. Ain't nothin gunna change about it by critiquing the committee.

It's always going to be better than the 80's. Didn't we have to play at LSU one year and at Memphis another?

Why do I need to focus on beating Little Rock? It isn't going to help the team's preparation. The thread was clearly titled. If you don't want to discuss seeding methodology, don't participate.

 

[Purdue Grad in Texas]

Sorry, but saying "you have to fill out a schedule" is a cop out. There are plenty of mid major and low major teams who want the exposure and experience of playing a B1G team like Purdue. Just avoid the absolute bottom feeder teams and conferences (MEAC, SWAC, Southland) and you'll see a bump in RPI and SOS.

Maybe so, FDB....I hear what you're saying, but I've heard that line a few times over the years.....maybe we'll start seeing better scheduling.

Candidly, I don't know the ins/outs from a timing perspective as to scheduling non-conference opponents....but I suppose you can accomplish this with planning well in advance.

 

[FirstDownB]

For the record I don't mind our draw. And I love that we're in the Chicago region.
The #18 slot is just puzzling to me.
And if we had won yesterday, many of us would be irate if we were given a 4 seed, which is likely what would have happened.

 

[PBoiler84]

The committee used Common Core math methodology to get to their otherwise unexplainable seeds.

 

[02Boilermaker]

FDB, in addition to those rankings in your OP, the selection committee back of the napkin comparisons also includes how teams fared against the RPI top 25/50/100 with demerits for "bad losses." The problem is they use RPI to justify one seed/selection and then point out good wins to justify another choice (when RPI doesn’t back that up)...if they were consistent on what metric was most important, I think people would be a lot happier with their brackets.

It's easy to see what they used to rank Purdue so low: Purdue’s 4 RPI top 50 wins was lacking compared to their peers...of teams that got a 5 seed or better, no one other than Kentucky (with 3) had fewer than 5 wins vs the top 50 (used CBS team comparison page for info).

The B1G has 0 teams in the RPI top 10 (Big12 has 3, ACC - 3, Pac12 - 2, Big East - 2). The B1G had 2 teams outside the RPI top 200 & 6 of 14 teams outside the RPI top 100. That meant B1G teams couldn’t compile the “good wins” that other conferences could...and that seems to have really hurt the top B1G teams when it came to seeding.

 

[rgarlitz]

On better non-conference scheduling:

I'd like to see Purdue go back to playing a regular home-and-home or 2-for-1 with Valpo like in the late 90s. Sure, the Crusaders won't always be as good as they are this year, but they're a quality mid-major opponent more often than not. Purdue should play at their place from time to time, too.

 

[FirstDownB]

FDB, in addition to those rankings in your OP, the selection committee back of the napkin comparisons also includes how teams fared against the RPI top 25/50/100 with demerits for "bad losses." The problem is they use RPI to justify one seed/selection and then point out good wins to justify another choice (when RPI doesn’t back that up)...if they were consistent on what metric was most important, I think people would be a lot happier with their brackets.

It's easy to see what they used to rank Purdue so low: Purdue’s 4 RPI top 50 wins was lacking compared to their peers...of teams that got a 5 seed or better, no one other than Kentucky (with 3) had fewer than 5 wins vs the top 50 (used CBS team comparison page for info).

The B1G has 0 teams in the RPI top 10 (Big12 has 3, ACC - 3, Pac12 - 2, Big East - 2). The B1G had 2 teams outside the RPI top 200 & 6 of 14 teams outside the RPI top 100. That meant B1G teams couldn’t compile the “good wins” that other conferences could...and that seems to have really hurt the top B1G teams when it came to seeding.

When you break it down like that it does help to align methodology with the results (though, as you said, not always consistent). Essentially, the committee valued RPI above all other metrics (RPI favors PAC12 this year). Then, on top of that they gave additional weight to RPI top 25/50 wins. And unfortunately, several of Purdue's big wins came against teams in the 51-60 range.

I won't go to the bother to dig the thread up, but there was some lengthy discussion earlier in the season about scheduling and effect on SOS and RPI. It was mentioned by some that RPI was outdated and the selection committee would give it minimal attention. That does not appear to be the case.

 

[SqueakyClean]

There was an interview with Brian Castiglione (SP?) the Committee chairman on the radio. From what he was saying, the committee seemed to put alot of emphasis on A) scheduling the tougher teams OOC and B) Top 50 wins.
Unfortunately, our "big" OOC opponents, Florida, Vanderbilt, Pittsburgh, and even Butler, all ended up on the wrong side of the Top 50 line (52, 60, 51, and 54 respectively). If Florida and Vanderbilt had each won one more SEC tournament games, we likely would have been the last 4 seed.

 

[Purdue Grad in Texas]

When you break it down like that it does help to align methodology with the results (though, as you said, not always consistent). Essentially, the committee valued RPI above all other metrics (RPI favors PAC12 this year). Then, on top of that they gave additional weight to RPI top 25/50 wins. And unfortunately, several of Purdue's big wins came against teams in the 51-60 range.

I won't go to the bother to dig the thread up, but there was some lengthy discussion earlier in the season about scheduling and effect on SOS and RPI. It was mentioned by some that RPI was outdated and the selection committee would give it minimal attention. That does not appear to be the case.

FDB.....that does not appear to be the case THIS YEAR. There was that discussion this year and even last year how RPI was becoming less critical (but still a metric tool). This is what some folks are criticizing.....how can you properly plan if RPI is less important one year (or with some decisions) and more important others? Plus, as some "experts" have pointed out, RPI can be manipulated and may not be the best measure. I think some of the committee metrics starts getting into paralysis of analysis.

It's a hard task, and I think the Selection Committee usually does a good job. In the end, the field is set.....let March Madness reign.....

The Boilers have had plenty of opportunities this year......and still has more in front of them.....think they just need to relish this latest opportunity.....go play.....have fun.......and let the chips fall.

Easy for me to write I suppose.

 

[BoilerDaddy]

There was an interview with Brian Castiglione (SP?) the Committee chairman on the radio. From what he was saying, the committee seemed to put alot of emphasis on A) scheduling the tougher teams OOC and B) Top 50 wins.
Unfortunately, our "big" OOC opponents, Florida, Vanderbilt, Pittsburgh, and even Butler, all ended up on the wrong side of the Top 50 line (52, 60, 51, and 54 respectively). If Florida and Vanderbilt had each won one more SEC tournament games, we likely would have been the last 4 seed.

50 is such an arbitrary line, especially with a metric as crude as RPI. Nevertheless, Purdue got a reasonable draw and its time to move forward with enthusiasm.

 

[BoilerBiker]

Re: scheduling
Izzo mentioned this Sunday (below).
No matter how much you try, it still takes some luck with other teams performance and having other teams to adopt a similar philosophy.
Non power five teams may not be what you expected. And While it's not nearly as far in advance as football, power five teams that do agree to play may not be as expected few years later either (say we schedule series with Duke, coach k retires and they are not nearly the same). Iirc bloom mentioned few years back how difficult it is especially for series like we had with West Virginia.

Re seeding:
I like when they look at rankings such as RPI, but dislike when coupled with seemingly arbitrary comparisons/cutoffs such as records vs top 50, 100.
(If two teams have similar total rpi's, but say team A could be 4-1 vs top 50 with wins over teams 30+ rank, and team B just 1-1 vs top 50 who are also top 10 plus maybe 3-0 vs teams ranked 51-60 . point being - Not always Apple to Apple comparisons).


Izzo yesterday:
“That’s kind of been the MO,” Izzo said. “I don’t look at it as disrespect. There’s got to be a reason for it. I’m most disappointed with this year with what our strength of schedule fell to. The reason I’m disappointed is we played teams like Northeastern that we thought would be really good and win their league. … Those are hard to schedule and you try to put yourself in a good position. We played so many good teams but then played like three 300 (RPI) teams because we couldn’t schedule anybody else."
http://www.detroitnews.com/story/sp...opens-ncaa-play-vs-middle-tennessee/81742198/

 

[purdu1378]

On better non-conference scheduling:

I'd like to see Purdue go back to playing a regular home-and-home or 2-for-1 with Valpo like in the late 90s. Sure, the Crusaders won't always be as good as they are this year, but they're a quality mid-major opponent more often than not. Purdue should play at their place from time to time, too.

Totally agree. I would love to see 2 for 1 or 3 for 1s against some of the other state schools that Have been ok lately. Valpo, IPFW, Indiana St, IUPUI. The trick is simply timing it when their teams will likely have a good set of upperclassman. Worse case you get a team likely better than 300+ RPI

 

[FirstDownB]

Totally agree. I would love to see 2 for 1 or 3 for 1s against some of the other state schools that Have been ok lately. Valpo, IPFW, Indiana St, IUPUI. The trick is simply timing it when their teams will likely have a good set of upperclassman. Worse case you get a team likely better than 300+ RPI

Yes to any of those teams, consistently in the 100's and low 200's. Or Evansville. Or Ball State. Or the most of the MAC.. And I would argue with schools like those you don't even worry about timing. On any given year they are better than teams like Howard and NC A&T.
And part of the problem this year was getting involved in that Hall of Fame Tip off Tourney, which had an awful field. Locked us into five opponents, only one of which was even NIT quality. Need to get back to Hawaii or even Puerto Rico or Orlando.

 

[Dryfly88]

Yes to any of those teams, consistently in the 100's and low 200's. Or Evansville. Or Ball State. Or the most of the MAC.. And I would argue with schools like those you don't even worry about timing. On any given year they are better than teams like Howard and NC A&T.
And part of the problem this year was getting involved in that Hall of Fame Tip off Tourney, which had an awful field. Locked us into five opponents, only one of which was even NIT quality. Need to get back to Hawaii or even Puerto Rico or Orlando.

Next year won't be any better with Mexico tournament. Auburn, Utah State, Texas Tech. Not terrible teams but certainly nothing to get excited about compared to some other Thanksgiving tournaments.

 

[Do Dah Day]

FDB, in addition to those rankings in your OP, the selection committee back of the napkin comparisons also includes how teams fared against the RPI top 25/50/100 with demerits for "bad losses." The problem is they use RPI to justify one seed/selection and then point out good wins to justify another choice (when RPI doesn’t back that up)...if they were consistent on what metric was most important, I think people would be a lot happier with their brackets.

It's easy to see what they used to rank Purdue so low: Purdue’s 4 RPI top 50 wins was lacking compared to their peers...of teams that got a 5 seed or better, no one other than Kentucky (with 3) had fewer than 5 wins vs the top 50 (used CBS team comparison page for info).

The B1G has 0 teams in the RPI top 10 (Big12 has 3, ACC - 3, Pac12 - 2, Big East - 2). The B1G had 2 teams outside the RPI top 200 & 6 of 14 teams outside the RPI top 100. That meant B1G teams couldn’t compile the “good wins” that other conferences could...and that seems to have really hurt the top B1G teams when it came to seeding.

Several talking heads from several different media outlets said that the two most important criteria are AWAY wins and wins against the TOP 50. If there is no clear distinction at that point, then others come into play, but if not, those are what they use (or used). I didn't do any research to see if this is validated, but what you say seems to partially support this.

 

[Do Dah Day]

Maybe so, FDB....I hear what you're saying, but I've heard that line a few times over the years.....maybe we'll start seeing better scheduling.

Candidly, I don't know the ins/outs from a timing perspective as to scheduling non-conference opponents....but I suppose you can accomplish this with planning well in advance.

I think we are now committed through 2020. Which means that we are guessing who is going to be good with players that aren't even there yet.

 

[honey_badger]

Has anyone heard a concise explanation from the committee about seeding methodology, particularly the mess in the 3-5 range?
And here I thought worst-case scenario would be the committee over-valuing RPI. Turns out they didn't really value any of the metrics. Seems to me they just went with personal preference.

Purdue
RPI: 14
BPI: 10
KenPom: 10
Sagarin: 9
NCAAT: 18

Iowa St
RPI: 21
BPI: 16
KenPom: 17
Sagarin: 19
NCAAT: 16

Don't worry about it and focus on beating Little Rock. Ain't nothin gunna change about it by critiquing the committee.

It's always going to be better than the 80's. Didn't we have to play at LSU one year and at Memphis another?

It's a farce of a seeding system. They need to prosecute the people that did this.