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 Dec 1, 2014 14:36:21 GMT -5
Trojan, Good work as always. I think that committee missed one as well with MSU but oh well. Was wondering if you might try a slightly different format for listing profiles next time. Just the 6 items that are primary criteria: ● Overall record ● SOS ● RPI ● Head-to-head competition. ● Results versus common opponents. ((with teams of similar criteria above) ● Significant wins and losses. The last 10 games and home/away are secondary criteria and only looked at if primary are effectively tied. I think including the secondary stuff just creates confusion. Trojan and Blue, With primary criteria being basically 4 items (with RPI being the tiebreaker for record & SOS), you could almost use check boxes to determine who meets primary criteria better. With so little Head-to-Head (and much of that being split), it usually comes down to who wins RPI and either of the last two categories. There is lots of analyzing of "significant wins and losses". But have noticed that there is almost no comparisons were made by anyone involving "common opponents" this year or in past years. Seems like it this could be a very important and vastly overlooked part of selection that we are missing. Thoughts on this and an easy way to track it? I kind of think we overthink this? My take - the committee is made up of some pretty smart people who try to put together the best bracket possible and limit the # of 400+ mile travel. They probably spend many hours looking at the top 16 seeds and review just about everything there is to determine a consensus that UCLA needs to jump Iowa State. Or Penn State is the #5 seed and not #8 or #9. I just don't see how they could do that kind of detail when trying to figure out the last 4 or 5 teams in the tournament, so they end up mostly using RPI. How do quantify a team like Pittsburgh, that lost a couple tough matches towards the end of the season w/o their best player. Pacific or Seton Hall that swept a team that is in the tournament? They could spend weeks going over this - and never come to a consensus. They tend to bypass smaller conference teams on the bubble (Ohio) in favor of larger conferences, but for the most part just follow RPI for the final bids. Maybe there is more to this...
|
|
|
Post by c4ndlelight on Dec 1, 2014 15:11:22 GMT -5
I'm actually starting to understand UCLA at #12. Once you've decided that you're going to seed the Bruins, it doesn't make sense to put them below these guys -
Victories over RPI Top 25/Victories over seeded teams UCLA: 4/2 Kentucky: 2/0 Nebraska: 2/1 Colorado St.: 2/0 Kansas: 3/0
Also, 2 wins over USC gave them a common opponent edge over Kentucky, and the ILL win might have helped vs Nebraska (though the PSU loss, hmmm). I bet Nebraska would have been the real #13 seed, but the Committee adjusted to put Kentucky in Louisville and it makes sense to avoid adding the Huskers to the Wisky/PSU group of death.
The outlier is Kansas v Colorado St and Kentucky. It's worthwhile to note that Kansas had the worst loss of the group and also had an unfavorable common opponent results vs Kentucky (Lipscomb). This may also show that the Committee was a bit skeptical of the Big 12's RPI (coupled with the decision not to seed Iowa St.)
|
|
|
Post by FOBRA on Dec 1, 2014 15:36:48 GMT -5
Arizona also moved from #16 to RPI to an 11 seed for probably the same reasons. They had less Top 25 wins than UCLA but beat them head to head and placed higher in the conference.
|
|
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 Dec 1, 2014 15:39:37 GMT -5
I'm actually starting to understand UCLA at #12. Once you've decided that you're going to seed the Bruins, it doesn't make sense to put them below these guys - Victories over RPI Top 25/Victories over seeded teams UCLA: 4/2 Kentucky: 2/0 Nebraska: 2/1 Colorado St.: 2/0 Kansas: 3/0 Also, 2 wins over USC gave them a common opponent edge over Kentucky, and the ILL win might have helped vs Nebraska (though the PSU loss, hmmm). I bet Nebraska would have been the real #13 seed, but the Committee adjusted to put Kentucky in Louisville and it makes sense to avoid adding the Huskers to the Wisky/PSU group of death. The outlier is Kansas v Colorado St and Kentucky. It's worthwhile to note that Kansas had the worst loss of the group and also had an unfavorable common opponent results vs Kentucky (Lipscomb). This may also show that the Committee was a bit skeptical of the Big 12's RPI (coupled with the decision not to seed Iowa St.) I agree about UCLA. There just wasn't anyone else clearly worthy of the #12. I think the last 4 was determined by 'other' factors - geography, conference (although I don't know why I still hang on to the idea that conference matters at all to the committee after the 2nd round). Agree that Nebraska was next (#13), but was bumped to get Kentucky in Louisville. #15 and 16 doesn't matter, but at least they didn't put Kansas in with Texas. I think Iowa State being left out was more of believing in the Pac 12 being the dominate conferfence this year, then Big 12 RPI skepticism.
|
|
|
Post by c4ndlelight on Dec 1, 2014 15:43:09 GMT -5
I'm actually starting to understand UCLA at #12. Once you've decided that you're going to seed the Bruins, it doesn't make sense to put them below these guys - Victories over RPI Top 25/Victories over seeded teams UCLA: 4/2 Kentucky: 2/0 Nebraska: 2/1 Colorado St.: 2/0 Kansas: 3/0 Also, 2 wins over USC gave them a common opponent edge over Kentucky, and the ILL win might have helped vs Nebraska (though the PSU loss, hmmm). I bet Nebraska would have been the real #13 seed, but the Committee adjusted to put Kentucky in Louisville and it makes sense to avoid adding the Huskers to the Wisky/PSU group of death. The outlier is Kansas v Colorado St and Kentucky. It's worthwhile to note that Kansas had the worst loss of the group and also had an unfavorable common opponent results vs Kentucky (Lipscomb). This may also show that the Committee was a bit skeptical of the Big 12's RPI (coupled with the decision not to seed Iowa St.) I agree about UCLA. There just wasn't anyone else clearly worthy of the #12. I think the last 4 was determined by 'other' factors - geography, conference (although I don't know why I still hang on to the idea that conference matters at all to the committee after the 2nd round). Agree that Nebraska was next (#13), but was bumped to get Kentucky in Louisville. #15 and 16 doesn't matter, but at least they didn't put Kansas in with Texas. I think Iowa State being left out was more of believing in the Pac 12 being the dominate conferfence this year, then Big 12 RPI skepticism. You're probably right. It was easier to not seed Iowa St. as they are the Big 12 school that can drive to Madison or Champaign. The multiple bad losses provide ample cover as well.
|
|
|
Post by trollhunter on Dec 1, 2014 15:48:58 GMT -5
Arizona also moved from #16 to RPI to an 11 seed for probably the same reasons. They had less Top 25 wins than UCLA but beat them head to head and placed higher in the conference. FOBRA, Thanks for pointing that out, most people just assume RPI and top50 wins matters (even this is really "significant wins"). The two primary criteria often overlooked on VT are head to head and common opponents. The head to head should be an important factor when it applies. Also, placing higher in conference is not part of the selection criteria, but it does imply that they had a better common opponents record, at least in conference. Although unbalanced schedule warrants a closer look at this.
|
|
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: 28,163
|
Post by trojansc on Dec 1, 2014 17:03:50 GMT -5
Trojan, Good work as always. I think that committee missed one as well with MSU but oh well. Was wondering if you might try a slightly different format for listing profiles next time. Just the 6 items that are primary criteria: ● Overall record ● SOS ● RPI ● Head-to-head competition. ● Results versus common opponents. ((with teams of similar criteria above) ● Significant wins and losses. The last 10 games and home/away are secondary criteria and only looked at if primary are effectively tied. I think including the secondary stuff just creates confusion. Good note. I will change the format to something closer to what you put, I have more data on the teams than I Post on the thread obviously. I pay attention to results of common opponents, but to an extent, everybody always beats everybody. Pacific beats UALR who beats Miami who beats Virginia Tech who beats Illinois who beats Penn State! That has to mean Pacific is better than Penn State, right? But I get the point that sometimes, the results are very important. Last year, I think the H2H wasn't a huge factor: 2013 amongst all the bubble teams: Colorado beat UCLA UCLA beat CSUN Tulsa beat Miami Tulane beat Tulsa (TWICE) Xavier beat Butler (TWICE) Arizona State split with UCLA and Colorado Tulane beat Tulsa twice (Tulane boasted a 28-4 record) but Tulsa didn't get in. Tulsa beat Miami. Miami gets in NCAA tournament! Tulsa & Tulane, OUT. UCLA beat CSUN but CSUN made it in, UCLA OUT. Colorado, ASU both in. While last 10 is a secondary criteria, I actually think performance down the stretch (however you choose to measure it) is important to the committee. They are notorious for leaving out a team who looked like they were in until a last-second sketchy loss, and the committee puts them out. The exception is Lipscomb, because they hung in with a just good enough RPI. Thanks for the appreciation, and I appreciate your comments. I think I'm getting better at doing this. My main goal is to figure out who the committee thinks are the best-at large teams every year according to their published criteria (& selection history) but I can be honest and say I wasn't comfortable with the last bid of Virginia Tech, so I actually do think that was a miss on my part and not too surprised. I do wonder what the committees next choice would have been, if I saw Virginia Tech over Pacific (and that is not an opinion, only based on committee criteria) maybe the committee would have too.
|
|
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: 28,163
|
Post by trojansc on Dec 1, 2014 17:12:45 GMT -5
I think your second sentence, says it all blue. When I made the projection, I knew this but I thought if the committee would be willing to overlook three teams with a higher RPI (using RKPI projections) I thought they could look past Michigan State. Your statement about MSU vs. 50-100 level teams is very true. I was just not impressed with a 2-10 record vs. top-50 teams, and a 6-4 finish granted the opposition that they played. So what, they beat Northwestern(twice), Indiana, Rutgers, Purdue, and Iowa. None of those teams are NCAA-teams! Their losses however, were not bad. Illinois, Wisconsin, Nebraska. Minnesota not a tournament team but by far not a bad loss. Michigan State just hadn't done anything that impressive in the B1G. They swept Purdue, which may have done it. But that is only how I see it. I cannot blame the committee for (in my opinion, for whatever that's worth) missing on the last bid. Truthfully, if Western Kentucky fell in its conference tournament, there wouldn't even be a discussion between Michigan State, Pacific, Virginia Tech for that last bid. You can't be mad when you submit a resume that's sketchy, you just have to hope for the best. That's why I'm surprised by Pacific's release. They are clearly upset, but even if you were to get in, it may have only been by a 55/45 vote. The margins are so thin, it's easy to make a case for anyone including CSUN, SMU, Ohio.. either of those three could have been in and I don't know if you can really say the committee missed that bad.
|
|
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: 28,163
|
Post by trojansc on Dec 1, 2014 17:31:13 GMT -5
They beat Marquette twice, I am not considering Creighton Top-25. With top-25 bonus win over Florida, I think Marquette finishes top-25. Another concern is does Pittsburgh finish top-50? That may key VTEch in. 4 hours to selections Marquette was 26 and went 0-1, while teams around them won. Not sure they're moving up. c4ndlelight, I remember discussing this with you and you were skeptical about Seton Hall. I just looked at the RPI finals. Creighton and Marquette both finished top-25.. I think that may have been crucial to Seton Hall getting in... but you had right to doubt them. Maybe both two spots down, and I think you give the bid to Pacific. Margins so thin.
|
|
|
Post by trollhunter on Dec 1, 2014 18:47:28 GMT -5
Trojan, for common opponents I am suggesting tracking them as they do in D2. Teams A & B each play some of the exact same teams C, D, E and F during the season. Team A goes 6-2 versus those teams, while team B goes 2-3. Clearly team A win that criteria.
You do have to be careful with certain situations with this comparison. For instance, if team A splits 1-1 with team E, while team B just plays them once and goes 1-0 (or 0-1). In D2 that is counted as a wash (tie) since it is unclear if they are truly better.
NCAA does have a mechanism for tallying these that it uses in D2, but just in small groups so the data is meaningful. I'm guessing it would tell something useful for last 8 bubble teams.
|
|
|
Post by c4ndlelight on Dec 1, 2014 19:12:14 GMT -5
Marquette was 26 and went 0-1, while teams around them won. Not sure they're moving up. c4ndlelight, I remember discussing this with you and you were skeptical about Seton Hall. I just looked at the RPI finals. Creighton and Marquette both finished top-25.. I think that may have been crucial to Seton Hall getting in... but you had right to doubt them. Maybe both two spots down, and I think you give the bid to Pacific. Margins so thin. You're probably right there. I also wonder if, with the BEast final played so lately, they just made a contingency plan and put SHU in so they wouldn't have to drastically reorder the brackets if they did get in?
|
|