The AAC's OOC lightweights are killing us...

[partyATkemba]

The committee absolutely does judge "what have you done for me lately" on selection Sunday for the bubble teams. SMU lost their last 3 going into tournament time last year, which was one of the primary reasons they got shafted. They're not looking at these resumes blind. If UConn wins 5 of their last 6 which would be 11 for 14 we are IN (assuming that loss comes to SMU in the finals and it's a close game).

 

[pepband99]

You do realize that the net impact isn't that great if they lose the games right?

Sure if they played better teams and had the same record. They wouldn't.

If you're going to flip your smarmy button on (again), at least be right...

e.g. Houston (http://www.rpiforecast.com/teams/Houston.html)
Their top 100 loss @ Harvard has a net positive impact on their RPI, as does their home win against us (ugh)
Their crappy 300+ home wins are net negative.

Even in context for an individual team, you're wrong. Conceptually it makes sense, too, as teams in the 300+ range have a disproportionately big impact on 2/3 of the RPI measure. You simply have to upgrade to "somewhat awful" from "we played roadkill" to make a significant impact. To put it simply - when you play a 300+, they likely have both a) a bad record and b) bad opponents, so you're nuking the majority of your RPI measure.

In a wider context, the conference RPI is a measure in and of itself, that impacts the collective perception of individual teams. These teams torpedoing their OOC SOS hurts us, and it REALLY hurts us when we're on the bubble.

 

[Ts5034]

The committee absolutely does judge "what have you done for me lately" on selection Sunday for the bubble teams. SMU lost their last 3 going into tournament time last year, which was one of the primary reasons they got shafted. They're not looking at these resumes blind. If UConn wins 5 of their last 6 which would be 11 for 14 we are IN (assuming that loss comes to SMU in the finals and it's a close game).

Over the last 5 years or so, I thought the committee has been stressing the "entire body of work" compared to putting emphasis on how teams finish. A good win in a November tournament means just as much as a good win in February. Im pretty sure thats why you dont see the media talking and comparing the bubble teams record over the last 10 games of the season like they used to because teams are not judged on that anymore.

 

[whaler11]

If you're going to flip your smarmy button on (again), at least be right...

e.g. Houston (http://www.rpiforecast.com/teams/Houston.html)
Their top 100 loss @ Harvard has a net positive impact on their RPI, as does their home win against us (ugh)
Their crappy 300+ home wins are net negative.

Even in context for an individual team, you're wrong. Conceptually it makes sense, too, as teams in the 300+ range have a disproportionately big impact on 2/3 of the RPI measure. You simply have to upgrade to "somewhat awful" from "we played roadkill" to make a significant impact. To put it simply - when you play a 300+, they likely have both a) a bad record and b) bad opponents, so you're nuking the majority of your RPI measure.

In a wider context, the conference RPI is a measure in and of itself, that impacts the collective perception of individual teams. These teams torpedoing their OOC SOS hurts us, and it REALLY hurts us when we're on the bubble.

'Net impact isn't that great.' Tell me I'm wrong when I'm wrong. thx.

BTW you said 150-200 in your post, not 100.

 

[pepband99]

'Net impact isn't that great.' Tell me I'm wrong when I'm wrong. thx.

BTW you said 150-200 in your post, not 100.

Yes, because finding a 100s OOC game in the sea of 300-rated-teams is a bit hard. See how that works?

So - since everyone is pointing at the Houston L - let's see how this impacts vs OOC SOS:

UConn now - 69 RPI / 70 SOS
w/ Houston flipped to W: 54 / 63

Now, I'll replace Houston completely with a 2nd instance of South Florida (similar RPI, 186 OOC SOS)
UConn now (ie swap @Houston L for an @USF L) - 68/63
w/ @USF flipped to W: 54/63

Our SOS jumps 10 points based on 1 away game for teams that are both essentially the same, minus OOC SOS. The RPI stays flat because it impacts the bigger measure of RPI (opponent W/L is 50%, and we would play them twice, vs once for Houston) Extrapolate this across the whole conference, and you can see where I'm going.
It's not a perfect calculation (knock-on impact of flipping the game), but i would call this a statistically significant "net impact"

 

[7774]

Yes, because finding a 100s OOC game in the sea of 300-rated-teams is a bit hard. See how that works?

So - since everyone is pointing at the Houston L - let's see how this impacts vs OOC SOS:

UConn now - 69 RPI / 70 SOS
w/ Houston flipped to W: 54 / 63

Now, I'll replace Houston completely with a 2nd instance of South Florida (similar RPI, 186 OOC SOS)
UConn now (ie swap @Houston L for an @USF L) - 68/63
w/ @USF flipped to W: 54/63

Our SOS jumps 10 points based on 1 away game for teams that are both essentially the same, minus OOC SOS. The RPI stays flat because it impacts the bigger measure of RPI (opponent W/L is 50%, and we would play them twice, vs once for Houston) Extrapolate this across the whole conference, and you can see where I'm going.
It's not a perfect calculation (knock-on impact of flipping the game), but i would call this a statistically significant "net impact"

A lot of this doesn't make sense to me. Why would our strength of schedule improve based upon changing a Houston loss to a win?

 

[tzznandrew]

A lot of this doesn't make sense to me. Why would our strength of schedule improve based upon changing a Houston loss to a win?

Because, I presume, when he ran the scenario it calculated the next two games--Memphis and @Temple. You can't just flip it in RPIforecast--it will forecast all the way to the end regardless.

 

[7774]

Because, I presume, when he ran the scenario it calculated the next two games--Memphis and @Temple. You can't just flip it in RPIforecast--it will forecast all the way to the end regardless.

OK, thanks.

 

[pepband99]

Correct, but kept it same for all calcs.

Aaaaand - you just made me realize that I'm not going to able to do this with one game. Sigh. I might be completely wrong, too. Let me see if i can come up with something...

 

[whaler11]

Yes, because finding a 100s OOC game in the sea of 300-rated-teams is a bit hard. See how that works?

So - since everyone is pointing at the Houston L - let's see how this impacts vs OOC SOS:

UConn now - 69 RPI / 70 SOS
w/ Houston flipped to W: 54 / 63

Now, I'll replace Houston completely with a 2nd instance of South Florida (similar RPI, 186 OOC SOS)
UConn now (ie swap @Houston L for an @USF L) - 68/63
w/ @USF flipped to W: 54/63

Our SOS jumps 10 points based on 1 away game for teams that are both essentially the same, minus OOC SOS. The RPI stays flat because it impacts the bigger measure of RPI (opponent W/L is 50%, and we would play them twice, vs once for Houston) Extrapolate this across the whole conference, and you can see where I'm going.
It's not a perfect calculation (knock-on impact of flipping the game), but i would call this a statistically significant "net impact"

Literally none of that makes any sense but whatevs.

 

[whaler11]

1/2 of RPI is your opponents winning percentage. 1/4 is their opponents winning percentage.

It's not weighted for home/road.

So if you flip Tulane and Houston wins to losses against better teams... those teams are going to have to be a hell of a lot better because their record is weighed half as much as Tulane and Houston's.

Teams ranked 150-200 are not going to make up that gap at all, nevermind make it a meaningful net positive.

 

[Fishy]

The OP's point was dead on before it got sidetracked by some nitwit yelling "BUT YALE!"

The bottom half of this conference needs to be told by Aresco that they can't do business like a mid-major.

 

[pepband99]

Sorry I deleted once i realized i misread, and just before you posted. I can admit when I'm wrong.

 

[MattMang23]

The answer for us is to win our games. Yes our conference mates should schedule better but if we won any of our games vs Stanford, Duke or West Virginia and beat Yale and Houston as we should have, we would be solidly in the field.