(01-13-2019 04:34 PM)sfink16 Wrote:  

(01-13-2019 04:20 PM)WhoseHouse? Wrote:  

(01-13-2019 04:01 PM)sfink16 Wrote:  Of course Temple goes down to 43 from 41 in the newly released NET, I expected that with the offensive efficiency calculations, whatever they are, but how does the same game get played with USF and they go up from 100 to 94.

So winning ugly is worse then losing ugly, got it. I understand other games are played but it still makes no sense.

I'm not actually sure what you're railing against. Are you actually saying you're unhappy that Temple and USF moved up in the NET? Just accept that nice things are happening for you and move on. Sheesh.

You didn't read that correctly, Temple went down to 43 from 41, Temple didn't go up.

Well Temple played a team ranked twice their own, so an ugly win was not expected and lightly punished. USF played a much better team by rank and stayed in a game they shouldn't have and could have won, they were mildly rewarded for that. Then it also depends on what all the teams they played already did, a very good day for USF opponents will also help then while a bad day for temple's could make them sink a little. Finally you must take into account this is a ranking by score. It may be that Temple's score mildly improved but the people immediately behind them had better games, so did their opponents and so their score improved moving them in front of Temple who has the same score as before.

(01-13-2019 04:48 PM)stever20 Wrote:  

(01-13-2019 04:36 PM)ShockerFever Wrote:  

(01-13-2019 03:23 PM)stever20 Wrote:  

(01-13-2019 12:07 PM)sfink16 Wrote:  

(01-13-2019 11:51 AM)Foreverandever Wrote:  RPI is exclusively about wins and losses, that's why he goes back to the RPI.

Want a metric that counts W/L's it's the RPI, so when you say things like wins and losses should be important, he is literally using the metric you want W/Ls to compare to the NET.

Using the method you prefer (W/Ls) Purdue is ranked higher in the RPI than the NET even though you think 10-6 isn't good enough. They are also ranked higher than Temple even though Temple has more Ws. In the RPI it's because the teams they have played have a much better record than the teams Temple has played.

That's always been my argument against SOS. If Purdue wins 3 mores games to make them 13 and 3 and Temple loses 3 more games to make them 10 and 6, than the opponents records change with those wins and losses. Purdue opponents than wouldn't have as good a record while Temples opponents record would be better. That's all I'm saying and all I've been saying. Disagree or not, I'm consistent with what I'm saying.

By definition NET is supposed to reward road wins but I question if it really matters much. Temple has a 4 and 2 road record while Purdue is 1 and 4 on the road. Secondary to W/L, I think home/away should be the next more important factor. Scheduling mostly home games doesn't indicate how good a team is overall IMO. We've seen recently in our conference games how hard it is to win on the road.

For the record, I'm not comparing these two teams, it's all teams records I'm saying this for.

sorry but it's impossible for W/L to be the #1 factor. The difference in schedules makes it like that. Otherwise you'd have teams scheduling like ECU and USF and how they did all around..... No thank you.

Agreed. It's not just about wins and losses. If that was the case, then we'd still be using the RPI.

Style IMO does matter to a certain extent when differentiating teams from different types of wins and losses. I think it's pretty much agreed upon that a team winning by 30 looks better than when somebody else playing the same opponent is trailing throughout and only wins by 1. Yes, both result in wins, which should make up the majority portion of the formula, but also the way you do so should matter when rating them.

As a real life example, Louisville's 21 point road win over UNC looks (and is) much better than Texas' 3 point neutral win over them.

RPI has a lot of SOS in there....

It does? I thought SOS was a separate measure.

I thought RPI was your winning percentage, your opponent's winning percentage, and your opponent's opponent winning percentage.

(01-13-2019 05:02 PM)ShockerFever Wrote:  

(01-13-2019 04:48 PM)stever20 Wrote:  

(01-13-2019 04:36 PM)ShockerFever Wrote:  

(01-13-2019 03:23 PM)stever20 Wrote:  

(01-13-2019 12:07 PM)sfink16 Wrote:  That's always been my argument against SOS. If Purdue wins 3 mores games to make them 13 and 3 and Temple loses 3 more games to make them 10 and 6, than the opponents records change with those wins and losses. Purdue opponents than wouldn't have as good a record while Temples opponents record would be better. That's all I'm saying and all I've been saying. Disagree or not, I'm consistent with what I'm saying.

By definition NET is supposed to reward road wins but I question if it really matters much. Temple has a 4 and 2 road record while Purdue is 1 and 4 on the road. Secondary to W/L, I think home/away should be the next more important factor. Scheduling mostly home games doesn't indicate how good a team is overall IMO. We've seen recently in our conference games how hard it is to win on the road.

For the record, I'm not comparing these two teams, it's all teams records I'm saying this for.

sorry but it's impossible for W/L to be the #1 factor. The difference in schedules makes it like that. Otherwise you'd have teams scheduling like ECU and USF and how they did all around..... No thank you.

Agreed. It's not just about wins and losses. If that was the case, then we'd still be using the RPI.

Style IMO does matter to a certain extent when differentiating teams from different types of wins and losses. I think it's pretty much agreed upon that a team winning by 30 looks better than when somebody else playing the same opponent is trailing throughout and only wins by 1. Yes, both result in wins, which should make up the majority portion of the formula, but also the way you do so should matter when rating them.

As a real life example, Louisville's 21 point road win over UNC looks (and is) much better than Texas' 3 point neutral win over them.

RPI has a lot of SOS in there....

It does? I thought SOS was a separate measure.

I thought RPI was your winning percentage, your opponent's winning percentage, and your opponent's opponent winning percentage.

What do you think your opponent's winning percentage and your opponents opponents winning percentage is?

(01-13-2019 05:09 PM)stever20 Wrote:  

(01-13-2019 05:02 PM)ShockerFever Wrote:  

(01-13-2019 04:48 PM)stever20 Wrote:  

(01-13-2019 04:36 PM)ShockerFever Wrote:  

(01-13-2019 03:23 PM)stever20 Wrote:  sorry but it's impossible for W/L to be the #1 factor. The difference in schedules makes it like that. Otherwise you'd have teams scheduling like ECU and USF and how they did all around..... No thank you.

Agreed. It's not just about wins and losses. If that was the case, then we'd still be using the RPI.

Style IMO does matter to a certain extent when differentiating teams from different types of wins and losses. I think it's pretty much agreed upon that a team winning by 30 looks better than when somebody else playing the same opponent is trailing throughout and only wins by 1. Yes, both result in wins, which should make up the majority portion of the formula, but also the way you do so should matter when rating them.

As a real life example, Louisville's 21 point road win over UNC looks (and is) much better than Texas' 3 point neutral win over them.

RPI has a lot of SOS in there....

It does? I thought SOS was a separate measure.

I thought RPI was your winning percentage, your opponent's winning percentage, and your opponent's opponent winning percentage.

What do you think your opponent's winning percentage and your opponents opponents winning percentage is?

Ok, then that's my point. RPI is based solely on winning and losing.

(01-13-2019 05:25 PM)ShockerFever Wrote:  

(01-13-2019 05:09 PM)stever20 Wrote:  

(01-13-2019 05:02 PM)ShockerFever Wrote:  

(01-13-2019 04:48 PM)stever20 Wrote:  

(01-13-2019 04:36 PM)ShockerFever Wrote:  Agreed. It's not just about wins and losses. If that was the case, then we'd still be using the RPI.

Style IMO does matter to a certain extent when differentiating teams from different types of wins and losses. I think it's pretty much agreed upon that a team winning by 30 looks better than when somebody else playing the same opponent is trailing throughout and only wins by 1. Yes, both result in wins, which should make up the majority portion of the formula, but also the way you do so should matter when rating them.

As a real life example, Louisville's 21 point road win over UNC looks (and is) much better than Texas' 3 point neutral win over them.

RPI has a lot of SOS in there....

It does? I thought SOS was a separate measure.

I thought RPI was your winning percentage, your opponent's winning percentage, and your opponent's opponent winning percentage.

What do you think your opponent's winning percentage and your opponents opponents winning percentage is?

Ok, then that's my point. RPI is based solely on winning and losing.

no it's not. Opponents winning percentage and opponent's opponents winning percentage- that's all SOS.....

RPI is 50% opponents winning percentage..

(01-13-2019 05:27 PM)stever20 Wrote:  

(01-13-2019 05:25 PM)ShockerFever Wrote:  

(01-13-2019 05:09 PM)stever20 Wrote:  

(01-13-2019 05:02 PM)ShockerFever Wrote:  

(01-13-2019 04:48 PM)stever20 Wrote:  RPI has a lot of SOS in there....

It does? I thought SOS was a separate measure.

I thought RPI was your winning percentage, your opponent's winning percentage, and your opponent's opponent winning percentage.

What do you think your opponent's winning percentage and your opponents opponents winning percentage is?

Ok, then that's my point. RPI is based solely on winning and losing.

no it's not. Opponents winning percentage and opponent's opponents winning percentage- that's all SOS.....

RPI is 50% opponents winning percentage..

Lol dont do that.

You're complicating it and confusing terms. You're correct. But you have to break down the parts of it. You skip around and people can't follow how it works. You're kind of underwear grooming people, step 1 this, step 2, that, step 3 we dont know, step 4 NET.

(01-13-2019 05:27 PM)stever20 Wrote:  

(01-13-2019 05:25 PM)ShockerFever Wrote:  

(01-13-2019 05:09 PM)stever20 Wrote:  

(01-13-2019 05:02 PM)ShockerFever Wrote:  

(01-13-2019 04:48 PM)stever20 Wrote:  RPI has a lot of SOS in there....

It does? I thought SOS was a separate measure.

I thought RPI was your winning percentage, your opponent's winning percentage, and your opponent's opponent winning percentage.

What do you think your opponent's winning percentage and your opponents opponents winning percentage is?

Ok, then that's my point. RPI is based solely on winning and losing.

no it's not. Opponents winning percentage and opponent's opponents winning percentage- that's all SOS.....

RPI is 50% opponents winning percentage..

Ok, I get what it is. I guess you're not following what I'm saying. RPI is strictly based on winning and losing, whether it's the primary team, the team you're playing, or the teams they played, whereas the NET uses much more analytics, efficiency metrics, and margin of victory (aka style points). RPI is comprised all about winning and losing in all its forms. NET is adding other aspects to the formula besides winning and losing, but how you're winning and losing.

(01-13-2019 05:49 PM)ShockerFever Wrote:  

(01-13-2019 05:27 PM)stever20 Wrote:  

(01-13-2019 05:25 PM)ShockerFever Wrote:  

(01-13-2019 05:09 PM)stever20 Wrote:  

(01-13-2019 05:02 PM)ShockerFever Wrote:  It does? I thought SOS was a separate measure.

I thought RPI was your winning percentage, your opponent's winning percentage, and your opponent's opponent winning percentage.

What do you think your opponent's winning percentage and your opponents opponents winning percentage is?

Ok, then that's my point. RPI is based solely on winning and losing.

no it's not. Opponents winning percentage and opponent's opponents winning percentage- that's all SOS.....

RPI is 50% opponents winning percentage..

Ok, I get what it is. I guess you're not following what I'm saying. RPI is strictly based on winning and losing, whether it's the primary team, the team you're playing, or the teams they played, whereas the NET uses much more analytics, efficiency metrics, and margin of victory (aka style points). RPI is comprised all about winning and losing in all its forms. NET is adding other aspects to the formula besides winning and losing, but how you're winning and losing.

yep, totally agree.... Never had the RPI described quite like that- but you're spot on.....

Lunardi updated today....
UCF 9 E vs Ole Miss, Virginia in Columbia
Houston 5 S vs Lipscomb, Auburn in Salt Lake City
Temple 11S FF vs Kansas St, Purdue, Va Tech in Hartford
Cincy 9 MW vs Iowa, Tennessee in Columbus
FF-First Four
Temple last team in

(01-15-2019 09:42 AM)stever20 Wrote:  Lunardi updated today....
UCF 9 E vs Ole Miss, Virginia in Columbia
Houston 5 S vs Lipscomb, Auburn in Salt Lake City
Temple 11S FF vs Kansas St, Purdue, Va Tech in Hartford
Cincy 9 MW vs Iowa, Tennessee in Columbus
FF-First Four
Temple last team in

So 4 in, but 3 of the teams are dangerous low on seed.

Really need UCF/Houston/Temple/Cincy to win the vast majority of their games and only lose to each other. Then maybe have a 5th team get hot and win the AAC tourney. That would be 5 bids in our "down" year

Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

Because they now use NET, not RPI. Temple is #44 in the NET and the NCAA at-large cut line is gonna land somewhere in the mid-40s.

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

NET of 44. RPI isn't used.

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

The effect of NET - based on data to date - is higher rankings for the top 6 conferences (P5 + BE) than under the defunct RPI method.

(01-15-2019 10:08 AM)Tiger1983 Wrote:  

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

The effect of NET - based on data to date - is higher rankings for the top 6 conferences (P5 + BE) than under the defunct RPI method.

actually this is wrong.... AAC in that group as well...
AAC- avg NET rating 103.08
AAC- avg RPI rating 114.83

(01-15-2019 09:57 AM)Stickboy46 Wrote:  

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

NET of 44. RPI isn't used.

Then not sure why Houston is a 5. We have a net of 8.

(01-15-2019 09:57 AM)HuskyU Wrote:  

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

Because they now use NET, not RPI. Temple is #44 in the NET and the NCAA at-large cut line is gonna land somewhere in the mid-40s.

Hopefully this NET thing works itself out over the next few weeks because looking at Temple I agree with the other poster that they look like a team that should be firmly in but this net formula is dinging them for some reason.

(01-15-2019 10:18 AM)Joprior23 Wrote:  

(01-15-2019 09:57 AM)Stickboy46 Wrote:  

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

NET of 44. RPI isn't used.

Then not sure why Houston is a 5. We have a net of 8.

It is always about who you play, where you play them, and who you beat. RPI, NET, Kenpom, and like metics help sort teams but do not tell the entire picture. Playing all but two games in Houston during the non-conference hurts. Kenpom has Houston at 33. I think Houston can be anywhere from a three to a six based on thier current resume. I am guessing they end up a four or five if they win the league. I like Temple’s resume better that Cincinnati at this point. I also like both far better than UCF because the Knights didn’t do anything outside thier home gym in the non-conference.

(01-15-2019 10:18 AM)Joprior23 Wrote:  

(01-15-2019 09:57 AM)Stickboy46 Wrote:  

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

NET of 44. RPI isn't used.

Then not sure why Houston is a 5. We have a net of 8.

I am sure they will say eye test or something else stupid. Talking heads going to talk but offer no real data to back up what they have done.

(01-15-2019 10:32 AM)bearcatlawjd2 Wrote:  

(01-15-2019 10:18 AM)Joprior23 Wrote:  

(01-15-2019 09:57 AM)Stickboy46 Wrote:  

(01-15-2019 09:53 AM)Joprior23 Wrote:  Temple is 13-3 with an RPI of 17. Not really sure why they are even considered on the bubble. They should be solidly in.

NET of 44. RPI isn't used.

Then not sure why Houston is a 5. We have a net of 8.

It is always about who you play, where you play them, and who you beat. RPI, NET, Kenpom, and like metics help sort teams but do not tell the entire picture. Playing all but two games in Houston during the non-conference hurts. Kenpom has Houston at 33. I think Houston can be anywhere from a three to a six based on thier current resume. I am guessing they end up a four or five if they win the league. I like Temple’s resume better that Cincinnati at this point. I also like both far better than UCF because the Knights didn’t do anything outside thier home gym in the non-conference.

I’m glad you said it. Temple has a better resume than Cincy and UCF. Temple should be nowhere near the bubble.