Selection Sunday Discussion

[stever20]

one thing- based on what the selection committee did- I really think if Marquette had beaten Xavier and Seton Hall- we could have seen them in the tourney. 5 top 50 wins at that point and given how huge of a criteria that was- I think there would have been a great chance.....

[DudeAnon]

Here is an interview with Castiglione
http://www.si.com/college-basketball/20 ... h-carolina

He pretty much sounds like stever, just cherry picking stats to include/exclude teams. Its horrible.

[Jet915]

Here is an interview with Castiglione
http://www.si.com/college-basketball/20 ... h-carolina

He pretty much sounds like stever, just cherry picking stats to include/exclude teams. Its horrible.

Yeah, I saw his interview w/Katz. Pretty much cherry picked a stat to exclude one team, the next team, he chose a different metric to exclude a team.

[MUWarrior1090]

and frankly regarding the RPI, I don't recall many folks here last year complaining about the RPI when Temple got jobbed, despite having the #34 RPI.

Also, I guess then if we go straight RPI, you're fine in Butler missing the tourney with their #56 RPI this year.

Didn't Temple lose to a division 2 school last year? If that happens, you don't deserve to get in. Period.

no Temple didn't, Tulsa did.

Iowa lost to a D2 school this year. They shouldn't be in?

[stever20]

that was a preseason game, not a counting regular season game. Just like St John's one was.

[JPSchmack]

To the implication that St. Bonaventure could get OOC games, they just don’t want to…

- the Bonnies had to sign a home & home contract with BINGHAMTON last year to get their 11th OOC game.
- the Bonnies didn’t even PLAY a 12th OOC game last season, only 11.
- In 2013-14, they had to play two transitioning schools to fill their schedule.

Bona fans aren’t blowing smoke. NO ONE wants to play them.

Even if you think "Bona would rather sell $100,000 worth of tickets and win against a non-Division I school than going on the road for $50,000 and losing to a BCS school and having no chance at an NCAA bid when they can't beat Missouri" (which is crazy since we've competed for an at-large twice in 30 years)...

You have to admit the Bonnies would rather play ANYONE and make money than play no one and get no money.

[JPSchmack]

Isn’t OOC SOS a really stupid metric if your OOC schedule included wins over Virgnia & Seton Hall? I mean, who in their right mind would say “GW didn’t do anything OOC” when they beat two ridiculously good teams like that? EVERYONE beats bad teams. We’re counting decimal places for teams 120 vs 220 vs 320… when if you’re in the conversation, you’re SUPPOSED TO WIN 90% of those games.

And that’s the biggest misconception out there. We all point to bad losses to say “that other team that isn’t ours shouldn’t get a bid because.” However, absolutely zero people except maybe Will Wade and myself ever put NATIONAL CONTEXT into the discussion.

Wade (VCU head coach) did 20 minutes in his presser Tuesday on the Bonnies getting the shaft and the perils of OOC scheduling.

Statistically speaking, winning on the road at #100 is just as hard as beat #20 at home.
Statistically speaking, Top 50 RPI teams win 90% of their games against teams 125+ of the RPI. And the percentage on the road is insanely lower than that.

For example, the big bad BCS which dominates everyone… this season, OOC on the road at non-BCS teams: .657 total. UNC lost at Northern Iowa. Virginia at GW, etc, etc.

So a team like Monmouth plays 20 MAAC conference games, 10 on the road and 7 against bad teams. They go 6-2. They get HAMMERED for it.
While Kansas only played two games on the road vs 125+ of the RPI… and couldn’t even with them both.


Baseball had a stat revolution and stopped giving the MVP to the guy with the most RBI, because counting stats are stupid. But here’s the 1981 RPI criteria in use in 2016 counting Top 50 wins and counting bad losses as if everyone plays the same schedule and we have an ounce of context for it.


You wanna fix this in the easiest way possible?

Go back to the old RPI formula (no home/road/neutral win percentage nonsense). Then run a second set of numbers.

Take everyone’s breakdown vs Top 20, 21-60, 61-100, 101-140, 141-220, 221-350 and divides out there home / road / neutral splits.

Whatever the average win percentages are against each of those 18 categories by teams 21-60 is the litmus test: The “BUBBLE AVERAGE.”

Have the computer program compare everyone’s win percentages to the Bubble Average, and spit out a number for what each school would be against “The Ideal NCAA Schedule For an At-Large.” Obviously, it would build in a regression to the average (for each RPI range) for a lack of games played in a category.

>> Example, Northern Iowa wouldn’t be credited with going 1.000 vs the top group just because they beat UNC and Utah in their only two games. They’d get between that number and what’s expected for a 61-100 team vs the Top 20 in 2-3 more games, so like, .650 or something.
>> Monmouth’s 11-2 on the road at bad teams would be about .846 because that’s about the average.
>> Kansas’ 1-1 on the road at bad teams would regress to the mean and be about .650.

Weight each category how we see fit: 21-60 should be the most important group; you’d get a bonus for beating teams in 1-20, and a penalty for losing to teams in the bottom groups.

You don’t need to bother with silly crap like OOC SOS (Which, by the way is ALREADY INCLUDED IN RPI) and you’ve got how strong everyone’s resume actually is. In one number.


And you tell everyone exactly how to game the system, like they did with RPI. Because the ONLY WAY to game that formula is to have a good record in a few games against each category.

Kansas would know that a 1-1 vs bad teams on the road would hurt them. And Oklahoma State is a tough “bad team.” So the only way for Kansas to fight that is to call up a couple bad teams and say ““We need to come beat you at your place.”

If they go lose to the 8th place MAAC school on the road, they can say “Wow, sorry for screwing you in 2016, Monmouth. That’s harder than we thought.”
If they go whoop the 8th place MAAC school on the road… nothing changes. Kansas gets a bid, the 8th place MAAC team is still in the 200s, but they got to sell some tickets and make some dough.

It would encourage every kind of team in the country to play games against every kind of team in the country. And that’s good.

[stever20]

To the implication that St. Bonaventure could get OOC games, they just don’t want to…

- the Bonnies had to sign a home & home contract with BINGHAMTON last year to get their 11th OOC game.
- the Bonnies didn’t even PLAY a 12th OOC game last season, only 11.
- In 2013-14, they had to play two transitioning schools to fill their schedule.

Bona fans aren’t blowing smoke. NO ONE wants to play them.

Even if you think "Bona would rather sell $100,000 worth of tickets and win against a non-Division I school than going on the road for $50,000 and losing to a BCS school and having no chance at an NCAA bid when they can't beat Missouri" (which is crazy since we've competed for an at-large twice in 30 years)...

You have to admit the Bonnies would rather play ANYONE and make money than play no one and get no money.

so St Bonnie's didn't try to get road games against tougher team?

What you are saying is No one wants to play them..... at their place....

I'm sorry, but that's a you problem. You can't force the big boys to play at your place. But then you can't say, the big boys won't play us. Because they would. At their place. So either you have to take a risk and play at their place, or deal with what happened on Sunday.

[stever20]

Isn’t OOC SOS a really stupid metric if your OOC schedule included wins over Virgnia & Seton Hall? I mean, who in their right mind would say “GW didn’t do anything OOC” when they beat two ridiculously good teams like that? EVERYONE beats bad teams. We’re counting decimal places for teams 120 vs 220 vs 320… when if you’re in the conversation, you’re SUPPOSED TO WIN 90% of those games.

And that’s the biggest misconception out there. We all point to bad losses to say “that other team that isn’t ours shouldn’t get a bid because.” However, absolutely zero people except maybe Will Wade and myself ever put NATIONAL CONTEXT into the discussion.

Wade (VCU head coach) did 20 minutes in his presser Tuesday on the Bonnies getting the shaft and the perils of OOC scheduling.

Statistically speaking, winning on the road at #100 is just as hard as beat #20 at home.
Statistically speaking, Top 50 RPI teams win 90% of their games against teams 125+ of the RPI. And the percentage on the road is insanely lower than that.

For example, the big bad BCS which dominates everyone… this season, OOC on the road at non-BCS teams: .657 total. UNC lost at Northern Iowa. Virginia at GW, etc, etc.

So a team like Monmouth plays 20 MAAC conference games, 10 on the road and 7 against bad teams. They go 6-2. They get HAMMERED for it.
While Kansas only played two games on the road vs 125+ of the RPI… and couldn’t even with them both.


Baseball had a stat revolution and stopped giving the MVP to the guy with the most RBI, because counting stats are stupid. But here’s the 1981 RPI criteria in use in 2016 counting Top 50 wins and counting bad losses as if everyone plays the same schedule and we have an ounce of context for it.


You wanna fix this in the easiest way possible?

Go back to the old RPI formula (no home/road/neutral win percentage nonsense). Then run a second set of numbers.

Take everyone’s breakdown vs Top 20, 21-60, 61-100, 101-140, 141-220, 221-350 and divides out there home / road / neutral splits.

Whatever the average win percentages are against each of those 18 categories by teams 21-60 is the litmus test: The “BUBBLE AVERAGE.”

Have the computer program compare everyone’s win percentages to the Bubble Average, and spit out a number for what each school would be against “The Ideal NCAA Schedule For an At-Large.” Obviously, it would build in a regression to the average (for each RPI range) for a lack of games played in a category.

>> Example, Northern Iowa wouldn’t be credited with going 1.000 vs the top group just because they beat UNC and Utah in their only two games. They’d get between that number and what’s expected for a 61-100 team vs the Top 20 in 2-3 more games, so like, .650 or something.
>> Monmouth’s 11-2 on the road at bad teams would be about .846 because that’s about the average.
>> Kansas’ 1-1 on the road at bad teams would regress to the mean and be about .650.

Weight each category how we see fit: 21-60 should be the most important group; you’d get a bonus for beating teams in 1-20, and a penalty for losing to teams in the bottom groups.

You don’t need to bother with silly crap like OOC SOS (Which, by the way is ALREADY INCLUDED IN RPI) and you’ve got how strong everyone’s resume actually is. In one number.


And you tell everyone exactly how to game the system, like they did with RPI. Because the ONLY WAY to game that formula is to have a good record in a few games against each category.

Kansas would know that a 1-1 vs bad teams on the road would hurt them. And Oklahoma State is a tough “bad team.” So the only way for Kansas to fight that is to call up a couple bad teams and say ““We need to come beat you at your place.”

If they go lose to the 8th place MAAC school on the road, they can say “Wow, sorry for screwing you in 2016, Monmouth. That’s harder than we thought.”
If they go whoop the 8th place MAAC school on the road… nothing changes. Kansas gets a bid, the 8th place MAAC team is still in the 200s, but they got to sell some tickets and make some dough.

It would encourage every kind of team in the country to play games against every kind of team in the country. And that’s good.

GW lost to DePaul for gosh sakes OOC. And the UVA/Seton Hall wins- at home...

Monmouth got hammered because they lost 3 sub 200 games. The rest of the top 100- had only 2 other teams with 3 or more losses. Top 150 had I think 11 teams with at least that many losses. Other teams play that many road games against bad teams, and they don't lose...

Also you say go back to the old RPI...
St Bonnies new RPI 30, old RPI 37
Monmouth new RPI 52, old RPI 65
Valpo 49 to 60
Akron 34 to 51
South Dakota St 28 to 49
Pittsburgh 53 to 39
Butler 56 to 47
Syracuse 72 to 63
Temple 60 to 53
Tulsa 58 to 55
Ohio St 74 to 56

so I think the big guys would love to go back to the old RPI. be careful for what you wish for- you might just get it....

[JPSchmack]

so St Bonnie's didn't try to get road games against tougher team?

No. You're missing the point. No one will play, period.

You know this to be true because there is zero reason, financial or competitive, to ever have an open slot on your schedule and not play a game. We’re not trying to play RPI games… This was the first time we’ve competed for an at-large in 17 seasons. And we were picked Preseason EIGHTH in the Atlantic 10.

As a bottom half A-10 team, we can't be relied on to have a winning record. You’d expect a 16-15 record at best from the 8th place A-10 team. We don’t help the RPI of anyone above us.

But we’re still .767 in the last 20 years against teams outside the BCS. So anyone in the OTHER 24 conferences don’t want to play us either.

Our lat 14 games against BCS teams have been decided by an average of 4 points.
3 wins
8 losses decided by 5 points or less (or in OT)
2 losses by 3-5 possessions (9-13 points)
>> 7-22 Bona at #16 Syracuse, we lost 70-82. That’s not the kind of battle you want in a guarantee game. Cuse can pay Cornell and win by 30.

The ONLY time we’ve been blown out by a BCS team in the last 20 years was the season we were 2-26. And no one wants to play a 2-26 team, either.


There is zero incentive for any school in the country to play us. Period. And that’s the only reason a school would ever only schedule 11 OOC games when they’e allowed to play 12.