|
Post by n00b on Nov 25, 2018 16:55:07 GMT -5
Now that figstats updated with all matches included, here are the discrepancies between fig and blue
6/7 Wisconsin/Pittsburgh 12/13 Creighton/Penn State 21/22 Texas State/Michigan 31/32/33 ETSU/Missouri/South Carolina 39/40 Arizona/Denver 41/42 SFA/Yale 46/47 FGCU/Hawaii 48/49/50 Kansas/NM State/Colorado 56/57/58 Hofstra/LMU/Maryland
Then finally, Iowa State is the biggest difference. 65 from blue, 60 from figgie.
|
|
trojansc
Legend
All-VolleyTalk 1st Team (2022, 2021, 2020, 2019, 2018, 2017), All-VolleyTalk 2nd Team (2016), 2021, 2019 Fantasy League Champion, 2020 Fantasy League Runner Up, 2022 2nd Runner Up
Posts: 28,419
|
Post by trojansc on Nov 25, 2018 17:25:40 GMT -5
Now that figstats updated with all matches included, here are the discrepancies between fig and blue 6/7 Wisconsin/Pittsburgh 12/13 Creighton/Penn State 21/22 Texas State/Michigan 31/32/33 ETSU/Missouri/South Carolina 39/40 Arizona/Denver 41/42 SFA/Yale 46/47 FGCU/Hawaii 48/49/50 Kansas/NM State/Colorado 56/57/58 Hofstra/LMU/Maryland Then finally, Iowa State is the biggest difference. 65 from blue, 60 from figgie. Could K-State be Top 50 and not Colorado ? They are tied on figstats
|
|
bluepenquin
Hall of Fame
4-Time VolleyTalk Poster of the Year (2019, 2018, 2017, 2016), All-VolleyTalk 1st Team (2021, 2020, 2019, 2018, 2017, 2016)
Posts: 12,447
|
Post by bluepenquin on Nov 25, 2018 18:20:03 GMT -5
The difference between #48 and 53 is so small that just the small improvement in opponent record from Illinois winning could/should mean several places. So, are the differences in RPI at the cutoff less pronounced than they have been in past years? If so, might that lead other factors to be more important than RPI when determining who gets in? If they are going to use RPI (or something like Pablo) - one should be looking at the 'score/rating' and not just the ranking. But I bet that the committee - like most people - will opt for the ease of just looking at the ranking. Since RPI is already so flawed as a measure...
|
|
|
Post by ay2013 on Nov 25, 2018 18:31:31 GMT -5
So, are the differences in RPI at the cutoff less pronounced than they have been in past years? If so, might that lead other factors to be more important than RPI when determining who gets in? If they are going to use RPI (or something like Pablo) - one should be looking at the 'score/rating' and not just the ranking. But I bet that the committee - like most people - will opt for the ease of just looking at the ranking. Since RPI is already so flawed as a measure... it's so true. Some of these differences are SOOOOOOOO small. There is almost the same difference between #1 and #2 in the RPI and #50 and #25......It's almost as if they should just generally junk teams together: Regional hosts (RPI 1-10), Seeds (RPI 1-25), Tournament bids (RPI 1-60), and then COMPLETELY DISREGARD THE AGGREGATE RPI RANK. At that point, use all the other criteria OTHER THAN the actual RPI rank.
|
|
|
Post by tallguy86 on Nov 25, 2018 18:40:45 GMT -5
If they are going to use RPI (or something like Pablo) - one should be looking at the 'score/rating' and not just the ranking. But I bet that the committee - like most people - will opt for the ease of just looking at the ranking. Since RPI is already so flawed as a measure... it's so true. Some of these differences are SOOOOOOOO small. There is almost the same difference between #1 and #2 in the RPI and #50 and #25......It's almost as if they should just generally junk teams together: Regional hosts (RPI 1-10), Seeds (RPI 1-25), Tournament bids (RPI 1-60), and then COMPLETELY DISREGARD THE AGGREGATE RPI RANK. At that point, use all the other criteria OTHER THAN the actual RPI rank. I see your point, but at the end of the day, RPI is the only real objective tool. If they were to put into groupings based on close enough ratings, then use the other factors, it becomes more subjective and chaos ensues if the committee picks/seeds the wrong teams (which will happen every time). I get that RPI isn’t perfect, but when the RPI generates rankings pretty damn close to the other subjective rankings, you’ve got to go with the objective quantitative approach rather than the committee (or VT) who they think should be where. It’s like a jury during a trial. A jury doesn’t decide if someone is guilty or innocent based on their guts, they have decision trees to use based on the laws.
|
|
|
Post by ay2013 on Nov 25, 2018 19:45:22 GMT -5
it's so true. Some of these differences are SOOOOOOOO small. There is almost the same difference between #1 and #2 in the RPI and #50 and #25......It's almost as if they should just generally junk teams together: Regional hosts (RPI 1-10), Seeds (RPI 1-25), Tournament bids (RPI 1-60), and then COMPLETELY DISREGARD THE AGGREGATE RPI RANK. At that point, use all the other criteria OTHER THAN the actual RPI rank. I see your point, but at the end of the day, RPI is the only real objective tool. If they were to put into groupings based on close enough ratings, then use the other factors, it becomes more subjective and chaos ensues if the committee picks/seeds the wrong teams (which will happen every time). I get that RPI isn’t perfect, but when the RPI generates rankings pretty damn close to the other subjective rankings, you’ve got to go with the objective quantitative approach rather than the committee (or VT) who they think should be where. It’s like a jury during a trial. A jury doesn’t decide if someone is guilty or innocent based on their guts, they have decision trees to use based on the laws. Well if that is the case, why have a seeding committee at all? just go strictly by RPI. They already deviate from the ACTUAL rankings when it comes to seeding placement as shown in the RPI/Seed differential chart in the other thread. IMO, the RPI prognosticators on volley talk are talking about a very real issue with how we evaluate teams in regards to RPI. The actual rank should mean something in general context NOT specific context, because with specific context, not looking at the actual RATING is where the issues with RPI and it's calculation continues to compound. Lets use a real world scenario here: (based on figstats) Oregon #19 RPI Marquette #15 RPI Lets compare these two teams....The difference between these two teams are 0.0065. To put that in real context, #19 Oregon is closer to #15 Marquette in the RPI than #15 Marquette is to #14 WSU (just one spot up). I think it would be better served to just say that Oregon and Marquette are in a general group and we will compare the teams based on OTHER criteria rather than the completely small and highly irrelevant difference in the aggregate RPI rating. I'd rather know that Oregon has 4 top 25 wins on its resume and Marquette only has 3. Or that Marquette has a LOSING RECORD (5-6) against RPI top 50 (tournament quality teams) compared to Oregon that has an 11-9 record against tournament quality teams. Meaning that not only has Oregon played far more tournament quality teams, they've WON the majority of them. The problem in relying on the actual RPI rank rather than grouping by general ranking and using other criteria is that the RPI is just a computer number completely unrelated to the quality of the team it purports to represent. Marquette gets a huge BONUS for having a team like St. John with its 23-11 record counted THREE times on Marquette's SOS, but this is despite the fact that St. John isn't even an RPI top 100 team. Meanwhile Oregon plays teams like ASU and Cal, which have LOSING records but significantly higher RPI than St. John. In terms of the aggregate RPI rank, ASU and CAL are much better than St. Johns, and yet even though all three of these teams are NOT going to the tournament, Marquette actually benefits MORE in SOS because of St. John's record multiplied 3 times over than Oregon does with ASU and Cal even though the RPI rank actually has ASU and Cal as definitely better top 100 teams. I'm sure you can see a flaw like this. So when looking at how close the aggregate rating between Marquette and Oregon is, NO I don't want the committee to look JUST at where they are ranked in the RPI, because REALLY small differences like this can be 100% explained away in completely fair terms that shouldn't reward Marquette/Punish Oregon.
|
|
|
Post by hammer on Nov 25, 2018 20:21:15 GMT -5
So, are the differences in RPI at the cutoff less pronounced than they have been in past years? If so, might that lead other factors to be more important than RPI when determining who gets in? If they are going to use RPI (or something like Pablo) - one should be looking at the 'score/rating' and not just the ranking. But I bet that the committee - like most people - will opt for the ease of just looking at the ranking. Since RPI is already so flawed as a measure... According to People Magazine, "A New Study Found That Lazy People Are Smarter Than Those Who Keep Busy"
|
|
|
Post by FreeBall on Nov 26, 2018 9:52:06 GMT -5
bluepenquin - Is the updated official RPI available anywhere online at this time?
|
|
bluepenquin
Hall of Fame
4-Time VolleyTalk Poster of the Year (2019, 2018, 2017, 2016), All-VolleyTalk 1st Team (2021, 2020, 2019, 2018, 2017, 2016)
Posts: 12,447
|
Post by bluepenquin on Nov 26, 2018 10:28:20 GMT -5
bluepenquin - Is the updated official RPI available anywhere online at this time? I would expect it to come out at its normal Monday time ~ 1-2 PM central time.
|
|