|
|
Post by notwvb on Nov 9, 2021 22:27:46 GMT -5
Our historical season has come to an end. Amusing. |
|
|
|
Post by coahc21 on Nov 9, 2021 22:30:18 GMT -5
The NCAA will tell you that it is not as simple as some equation. It is standardized, in a way that the Region X softball RAC should be able to look at the Region 3 Football data and come up with a similar ranking, but if it were as simple as a mathematical formula they wouldn't need coaches and administrators to serve on RACs. The process is as much art as science. And if a team has a significantly higher Win % and SOS the committee is told NOT to "drill down" to the rest of the data, so it should not be surprising at all that those two columns give you 19 of the 20. The part we couldn't know is what order they were working from. Despite Springfield beating Babson again, the new ranking still has Babson ahead of Springfield which I don't think any of us expected. For all of the speculation about Muskingum, we now know they were blocked by Denison and likely were not considered at all. Also, some of Stevenson's opponents got upgraded. Desales got back into the rankings in Region 4, and York (x2) in Region 5 which put Stevenson's record vs ranked at 11-5, and if I'm remembering this right, I think their SOS went up about 10 points. At that point I don't think they were the bubble team people expected them to be. And if they were, just by circumstance of their order in Region 5, and they ended up on the board with Denison who had .070 lower win % and a record vs ranked of 4-5, everybody would take Stevenson there. Hey NE, I love talking about this stuff. Not sure the difference between equation and standardization. But I agree it should be an art when it needs to be a art in the last 5-7 rounds, that should not be too hard. To use Win % and SOS and stop there is a formula. All the criteria should be considered. Stevenson's win % is higher than Denison since they played 6 more matches. Win % and SOS can be qualifying values, but once you get to a qualifying level, the difference of high(731) or higher(813) is not the key thing. And there is a certain amount of luck to a high SOS. Ranked results then become an important comparison. Stevenson was 11-5 versus ranked teams. Denison was 4-5 versus ranked. OK, that means Stevenson right? Nope, not if you look at it. Denison's 4-5 record is better than Stevenson's 11-5. Denison's wins: Ohio Northern(top 25), Wittenberg(top 25), Mount Union(top 50),Muskingum(top 50) Stevenson's wins: Haverford(top 100), Coast Guard(top 100), Messiah(top 100), VA Wesleyan(top 100), and 7 schools outside, some well outside, the top 100. Those 7 wins should count for nothing. The committee is supposed to notice this and select accordingly, the art part. I am less concerned with losses, but Denison's 5 were to better teams in total. They had no common opponents so can not go there. To me Denison is the easy choice. Are you using Massey for top 50, 100 etc? |
|
|
|
Post by d3follower on Nov 10, 2021 4:11:56 GMT -5
|
|
|
|
Post by ned3vball on Nov 10, 2021 6:25:13 GMT -5
Unranked losses are not necessarily as bad a thing as implied. Only 69 of the top 100 teams were ranked. With the regional imbalances this is going to happen. Finishing the math means 16 schools outside the top 100 were ranked. Oshkosh's 5 'unranked' losses were to Carthage(top 50), La Crosse(top 50), Cal Lu(top 100), Wheaton(IL)(top 100), and St. Catherines(over 100). Everybody gets 1 bad day. Those schools would have been ranked in a lot of regions further east. As with the Denison/Stevenson ranked compare, Oshkosh beats Ithaca and Springfield in the same way. As you can guess by now, I think just counting ranked wins is not the way to go. How are you defining top 100? I use Pablo(RichKern), Massey, and HuskerGeek. I take an average of those 3 ratings systems. In 2019 I just used Pablo but I thought the NE schools were getting a little too much support in that system so I wanted to smooth it out by adding in a few other methodologies. Yes, I know. The NCAA manual says they do not use outside ratings systems, but I have heard the terms 'top 50' and 'top100' used in discussions in prior years so they are probably using RPI or something similar to get to these numbers. If they are not using data like this, they should. And a fine line is not needed, you do not need to worry who was 56th versus 48th. just use top 100 as a guideline. Only 30 pool A teams are in the top 100. That leaves 70 schools to find 20 more. |
|
|
|
Post by The Bofa on the Sofa on Nov 10, 2021 9:28:46 GMT -5
How are you defining top 100? I use Pablo(RichKern), Massey, and HuskerGeek. I take an average of those 3 ratings systems. In 2019 I just used Pablo but I thought the NE schools were getting a little too much support in that system so I wanted to smooth it out by adding in a few other methodologies. Was that before or after everyone but Pablo was dismissing Johns Hopkins? Although this year I am more concerned - the schedules have been much more localized. At the beginning of the season, I had to work very hard to get NE teams in a place that was realistic. I had to set convergence criteria exceedingly high and it was taking forever. |
|
|
|
Post by bigfan on Nov 10, 2021 10:08:13 GMT -5
You VTer’s who analyze this DIII selection info 💻 so thoroughly amaze me. 🤯 Thanks! 👍 and noreaster are the best analyst's on Volleytalk |
|
|
|
Post by ned3vball on Nov 10, 2021 10:21:30 GMT -5
I use Pablo(RichKern), Massey, and HuskerGeek. I take an average of those 3 ratings systems. In 2019 I just used Pablo but I thought the NE schools were getting a little too much support in that system so I wanted to smooth it out by adding in a few other methodologies. Was that before or after everyone but Pablo was dismissing Johns Hopkins? Although this year I am more concerned - the schedules have been much more localized. At the beginning of the season, I had to work very hard to get NE teams in a place that was realistic. I had to set convergence criteria exceedingly high and it was taking forever. Don't worry, you are still my favorite(don't tell Massey or Hgeek). It is just a general inferiority complex for our region versus the Midwest. So I add in my own pessimism factor. Although I will say I never bought into the Johnson & Wales(RI) rankings in 2018. Admittedly, that was a tough circumstance as they won 12 leagues matches 25-10, 25-10, 25-10. Using an average of 3 makes it harder for people to complain (I hope). I agree on the scheduling, hopefully next year people get out there more. I know some coaches were told to schedule local to avoid travel problems. |
|
|
|
Post by The Bofa on the Sofa on Nov 10, 2021 13:34:24 GMT -5
Was that before or after everyone but Pablo was dismissing Johns Hopkins? Although this year I am more concerned - the schedules have been much more localized. At the beginning of the season, I had to work very hard to get NE teams in a place that was realistic. I had to set convergence criteria exceedingly high and it was taking forever. Don't worry, you are still my favorite(don't tell Massey or Hgeek). It is just a general inferiority complex for our region versus the Midwest. So I add in my own pessimism factor. Although I will say I never bought into the Johnson & Wales(RI) rankings in 2018. Admittedly, that was a tough circumstance as they won 12 leagues matches 25-10, 25-10, 25-10. Let me tell you a story about this. A couple of years ago, I tried to do JC rankings. You talk about regionalization - that is much more of a problem. Well, there was a team in the NE that was beating up everyone they played 3x25-10, and they did get ranked high in Pablo, probably too high, to be honest. And I had discussion with coaches about it, on how they were ranked way too high because this happens every year where that team beats up weak opponents in their region, but when they get to nationals, they get blown out. So they didn't take it seriously. OK, we get to the end of the season and that team is, yeah, probably overrated at #1. And the coaches seed them as #6. Tournament starts and they lose a close match in the first round to the #3 seed. That was a place where the committee was right and Pablo was wrong (nominally). However, they go into the consolation and win the next 2 (including beating a higher seeded team), winning the consolation bracket. Meanwhile, they team they lost to in the first ronud goes on to win the championship. Thus, that first round loss might mean they weren't #1, but they still have an argument for being as high as #2. So, yeah, Pablo probably had them overrated at #1, but it might not be as far off as everybody thinks. In the end, in that tournament, Pablo outperformed the seeding committee in terms of predicting the winners, so the lesson is that while it's not perfect, there is a lot of insight there and don't be quick to dismiss it because it goes against conventional wisdom. So yeah, extensive regionalation of schedules can lead to some distortions in the rankings, but if you take that into account and don't put your expectations too high on the accuracy, there is a lot to be learned from Pablo rankings even in extreme circumstances. |
|
|
|
Post by Brutus Buckeye on Nov 10, 2021 13:47:18 GMT -5
|
|
|
|
Post by ned3vball on Nov 11, 2021 6:05:47 GMT -5
Emory, Calvin, and Eau Claire regionals start today. D3Follower prepared excellent guides in post just above. Why is Emory starting on a Thursday?
At Eau Claire, #19 Gustavus Adolphus v #20 Chicago, and #15 Northwestern(MN) v Illinois Wesleyan are good ones.
Don't get me wrong on Ithaca, yes I disagree on their pick this year, but I love the way they schedule and get after it, I wish more schools did. Let's see if they can give Whitewater a battle.
|
|
|
|
Post by noreaster on Nov 11, 2021 7:27:30 GMT -5
Emory, Calvin, and Eau Claire regionals start today. D3Follower prepared excellent guides in post just above. Why is Emory starting on a Thursday? At Eau Claire, #19 Gustavus Adolphus v #20 Chicago, and #15 Northwestern(MN) v Illinois Wesleyan are good ones. Don't get me wrong on Ithaca, yes I disagree on their pick this year, but I love the way they schedule and get after it, I wish more schools did. Let's see if they can give Whitewater a battle. I wondered that about Emory too. So for the benefit of those who don't know, there are certain schools that won't play on Sundays, so the NCAA moves up the regionals to Thursday-Saturday instead of Friday-Sunday to honor those schools beliefs. However, once that school plays on Sunday even once, in any sport, the school loses the exemption forever. Calvin, Hope, Greenville, and Northwestern are regular tournament teams that have this issue. Hope, Northwestern and Greenville were all sent to the same regional rather than spread them out because they want as many regionals as possible to play Friday-Sunday to avoid missed class time at the other schools. So why Emory? I believe it is for Southern Virginia. This was originally known as the BYU rule and Southern Va is an LDS institution like BYU so it would make sense they follow the same rule. I looked through a couple of schedules on the website and I don't see where anyone has Sundays on their schedule. And of course Christopher Newport and UT Dallas are public, and just looking through schedules, Emory, Berry, Southwestern, Transy and W&L have all played on Sunday this year so now I'm confident this is Southern Va. |
|
|
|
Post by ned3vball on Nov 11, 2021 8:02:44 GMT -5
Here are some stats for each region using each school's average of their Pablo/Massey/HuskerGeek rankings. | Regional | ave rank | median rank | # top 25 | # top 50 | # top 100 | # over 100 | # AVCA top 25 |
|---|
| MIT | 135 | 119 | 0 | 2 | 2 | 4 | 2 | | RIT | 110 | 73 | 1 | 1 | 4 | 2 | 1 | | johns hopkins | 121 | 80 | 1 | 1 | 3 | 3 | 1 | | Emory | 45 | 50 | 3 | 2 | 2 | 1 | 3 | | calvin | 52 | 45 | 4 | 0 | 2 | 2 | 3 | | Eau Claire | 25 | 21 | 6 | 1 | 1 | 0 | 5 | | wartburg | 22 | 20 | 6 | 1 | 1 | 0 | 6 | | Stevens | 105 | 70 | 2 | 1 | 3 | 2 | 2 |
note: the top 25 columns foot to 23 since MHB and Santa Cruz did not get in. These numbers are typical. just highlights the uneven distribution of teams that makes the selection process as tough as it is. |
|
|
|
Post by The Bofa on the Sofa on Nov 11, 2021 9:53:16 GMT -5
Just to let you all know, I am working on an updated D3 ranking using the most up-to-date data from RichKern.com. Unfortunately, with non-D1 divisions, some match results trickle in late, either because they are played later on Sunday or their entry is just delinquent. There are about 8 matches that are there now that were not there when I last downloaded on Sunday. Because it's the end of the regular season, I am also hitting it very hard to get the best convergence. Short answer, I will post them here when they are done, but it might not be until later this afternoon after some of the matches today. However, it will not include any results from today.
|
|
|
|
Post by bigfan on Nov 11, 2021 10:29:22 GMT -5
The sooner Emory is eliminated the better |
|
|
|
Post by bigfan on Nov 11, 2021 10:33:58 GMT -5
Thanks for posting this. CMS with a #2 seed in the Wartburg regional makes me wonder why no other SCIAC team made the tournament, Pomona and Chapman are as good as some teams that got in. |
|
