This is absolutely correct. Scheduling the right group of opponents, for teams "in play" for NCAA Tournament seeds and at large selections, is very important. There are tools available to look at what teams have done for their opponents' strengths of schedule over time, what their trends might be, what the standard deviations are for their ranks as contributors to strengths of schedule, and where they fit on the reliability list. But in addition to those tools, a coach scheduling opponents ideally will try to know current information about a potential opponent, that won't show up in the team's prior performance. And, there always are unforeseeable events that can occur that will change a team's performance. And, as has occurred this year, weather events can mess up the best of scheduling plans. Those interested in this can check this webpage, which is specifically intended as a resource for the scheduling process and can give casual fans a sense of what goes into scheduling for high level teams: Scheduling Resource.
What is going on with BennettRank? That thing looks a mess... https://herosports.com/rankings/college-d1-womens-soccer
Yea, seems a bit wackers. Are there 13 teams better than UCLA? I doubt it. Is the Ivy League the best conference? I seriously doubt it. Rankings are all over the map.
It appears BennettRank has "lost" a bunch of teams. For example, the ACC only has 4 teams: Wake, NC State, Clemson, Notre Dame. The Pac-12 only has 3 teams: Washington State, UCLA, and Cal. The Ivy only has Princeton.
This is the week I start publishing actual RPI reports covering teams' current actual ratings. Hopefully, the NCAA will start publishing its reports too, at the RPI Archive. If you want to see teams' current RPI ranks, you can go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the most recent post titled: 2017 RPI Ranks 9.18.2017. If you want more detail on teams' ratings and ranks, as well as on conferences' and regional playing pools' average ratings and ranks, follow the instructions in that post to get to a detailed Excel workbook that you can download. Also, I now have substituted into my weekly simulations the actual results of all games through Sunday, September 17, and have produced an updated simulation for the entire season that includes end-of-season ratings, ranks, and other data and a simulation for the #1, #2, #3, and #4 seeds, the Automatic Qualifiers, and the unseeded at large selections for the NCAA Tournament. If you're interested, go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the two most recent posts titled: 2017 Simulated RPI Ranks 9.18.2017 and 2017 NCAA Tournament Bracket Simulation 9.18.2017. PLEASE be sure to read the introductory comments to the 2017 RPI Ranks 9.18.2017 and the 2017 Simulated RPI Ranks 9.18.2017 posts so that you don't put too much emphasis on the actual rankings you see in those reports. And, If you are really serious about understanding the details of how your team is doing or how the overall bracket simulation system works, go to NCAA Tournament: Predicting the Bracket, Track Your Team at the RPI for Division I Women's Soccer website, read that page, and then follow the instructions for using the 2017 Website Factor Workbook 9.18.2017, which is attached at the bottom of the page.
If you're interested in the best data sources for Division I women's soccer, I've provided a description of, and links to, the three best systems at Good Data Sources.
This is the second week of my publishing RPI reports covering teams' current actual ratings. We're still waiting for the NCAA to start publishing its reports too, at the RPI Archive. If you want to see teams' current RPI ranks, you can go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the most recent post titled: 2017 RPI Ranks 9.25.2017. If you want more RPI detail, including information about teams' Non-Conference RPIs and the conferences' average RPIs, you can go to the RPI website's This Year's Reports page. At the bottom of that page, there's a link to the 2017 RPI Report 9.25.2017 that is an excel workbook with a lot more detail. Also, I now have substituted into my weekly simulations the actual results of all games through Sunday, September 24, and have produced an updated simulation for the entire season that includes end-of-season ratings, ranks, and other data and a simulation for the #1, #2, #3, and #4 seeds, the Automatic Qualifiers, and the unseeded at large selections for the NCAA Tournament. If you're interested, go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the two most recent posts titled: 2017 Simulated RPI Ranks 9.25.2017 and 2017 NCAA Tournament Bracket Simulation 9.25.2017. And, If you are really serious about understanding the details of how your team is doing or how the overall bracket simulation system works, go to NCAA Tournament: Predicting the Bracket, Track Your Team at the RPI for Division I Women's Soccer website, read that page, and then follow the instructions for using the 2017 Website Factor Workbook 9.25.2017, which is attached at the bottom of the page.
CP, can you tell me how Penn State would have been affected in your rankings and RPI simulator if the match against UVA ending up in a tie? It seemed to me that a UVA player intereferred with PSU's goalkeeper on their 3rd goal, which should've been called off. Thanks.
I just have finished publishing my reports for this week: An RPI report covering teams' current actual ratings. If you want to see teams' current RPI ranks, you can go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the most recent post titled: 2017 RPI Ranks 10.2.2017. If you want more RPI detail, including information about teams' Non-Conference RPIs and the conferences' average RPIs, you can go to the RPI website's This Year's Reports page. At the bottom of that page, there's a link to the 2017 RPI Report 10.2.2017 that is an excel workbook with a lot more detail. So far, although the NCAA has published teams' ARPI ranks each of the last few weeks, it has not published the details of teams' ratings. Over the last few years, by this stage of the season, the NCAA has published detailed rating (as distinguished from ranking) reports at the RPI Archive. I'm hoping that this does not mean the NCAA no longer is going to publish these detailed reports during the course of the season, but I have a bad feeling about it. In any event, however, I've compared the rankings in my reports to the NCAA's current rankings and they match exactly. Also, I now have substituted into my weekly simulations the actual results of all games through Sunday, October 1, and have produced an updated simulation for the entire season that includes end-of-season ratings, ranks, and other data and a simulation for the #1, #2, #3, and #4 seeds, the Automatic Qualifiers, and the unseeded at large selections for the NCAA Tournament. If you're interested, go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the posts titled: 2017 Simulated RPI Ranks 10.2.2017 and 2017 NCAA Tournament Bracket Simulation 10.2.2017. And, If you are really serious about understanding the details of how your team is doing or how the overall bracket simulation system works, go to NCAA Tournament: Predicting the Bracket, Track Your Team at the RPI for Division I Women's Soccer website, read that page, and then follow the instructions for using the 2017 Website Factor Workbook 10.2.2017, which is attached at the bottom of the page.
I just have finished publishing my reports for this week: An RPI report covering teams' current actual ratings, covering games through Sunday, October 8. If you want to see teams' current RPI ranks, you can go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the most recent post titled: 2017 RPI Ranks 10.9.2017. If you want more RPI detail, including information about teams' Non-Conference RPIs and the conferences' average RPIs, you can go to the RPI website's This Year's Reports page. At the bottom of that page, there's a link to the 2017 RPI Report 10.9.2017 that is an excel workbook with a lot more detail. The detail includes conferences' average RPIs and ranks. The NCAA continues to not publish its detailed RPI reports at the RPI Archive. This appears to be true for all Fall sports. This is very unfortunate. Also, I now have substituted into my weekly simulations the actual results of all games through Sunday, October 8, and have produced an updated simulation for the entire season that includes end-of-season ratings, ranks, and other data and a simulation for the #1, #2, #3, and #4 seeds, the Automatic Qualifiers, and the unseeded at large selections for the NCAA Tournament. If you're interested, go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the posts titled: 2017 Simulated RPI Ranks 10.9.2017 and 2017 NCAA Tournament Bracket Simulation 10.2.2017. And, If you are really serious about understanding the details of how your team is doing or how the overall bracket simulation system works, go to NCAA Tournament: Predicting the Bracket, Track Your Team at the RPI for Division I Women's Soccer website, read that page, and then follow the instructions for using the 2017 Website Factor Workbook 10.9.2017, which is attached at the bottom of the page.
Yes, the Women's Soccer Committee does consider conferences in the at large selection process. It considers conference average ratings and conference ranks, together with how well a team has done in both its conference regular season and conference tournament. Here's a list of the conferences' current average ARPIs and their ranks. If you want to track this weekly, if you'll go to the preceding post, in the second paragraph you'll see a link to This Week's Reports. If you use that link, it will take you to a webpage that's part of the RPI for Division I Women's Soccer website. Each week, at the bottom of that page, I add an attachment with the current RPI data. The attachment is in the form of an Excel workbook. Unless you want to download the workbook, which you're free to do, you simply can click on the link for the newest workbook, today's being 2017 RPI Report 10.9.2017. Once you've opened the workbook with the link, if you scroll way down to about page 23 on your screen, you'll see the conference RPI reports. Current Conference Average Adjusted RPIs and Ranks: SEC 0.6114 1 ACC 0.6087 2 PacTwelve 0.5997 3 BigTen 0.5842 4 BigTwelve 0.5679 5 American 0.5611 6 Ivy 0.5402 7 WestCoast 0.5353 8 Colonial 0.5338 9 BigEast 0.5268 10 BigWest 0.5022 11 ConferenceUSA 0.5015 12 MountainWest 0.4975 13 OhioValley 0.4938 14 AtlanticTen 0.4937 15 AtlanticSun 0.4913 16 MissouriValley 0.4912 17 SunBelt 0.4787 18 MidAmerican 0.4683 19 Patriot 0.4670 20 BigSky 0.4602 21 AmericaEast 0.4595 22 BigSouth 0.4529 23 Southland 0.4503 24 Southern 0.4503 25 Horizon 0.4499 26 MetroAtlantic 0.4496 27 Northeast 0.4382 28 WAC 0.4303 29 Summit 0.4092 30 Southwestern 0.3574 31
I just have finished publishing my reports for this week: An RPI report covering teams' current actual ratings, covering games through Sunday, October 16. If you want to see teams' current RPI ranks, you can go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the most recent post titled: 2017 RPI Ranks 10.16.2017. If you want more RPI detail, including information about teams' Non-Conference RPIs and the conferences' average RPIs, you can go to the RPI website's This Year's Reports page. At the bottom of that page, there's a link to the 2017 RPI Report 10.16.2017 that is an excel workbook with a lot more detail. The detail includes conferences' average RPIs and ranks. The NCAA continues to not publish its detailed RPI reports at the RPI Archive. This appears to be true for all Fall sports. This is a departure from what it's done the last few years and is very unfortunate. Also, I now have substituted into my weekly simulations the actual results of all games through Sunday, October 16, and have produced an updated simulation for the entire season that includes end-of-season ratings, ranks, and other data and a simulation for the #1, #2, #3, and #4 seeds, the Automatic Qualifiers, and the unseeded at large selections for the NCAA Tournament. If you're interested, go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the posts titled: 2017 Simulated RPI Ranks 10.16.2017 and 2017 NCAA Tournament Bracket Simulation 10.16.2017. And, If you are really serious about understanding the details of how your team is doing or how the overall bracket simulation system works, go to NCAA Tournament: Predicting the Bracket, Track Your Team at the RPI for Division I Women's Soccer website, read that page, and then follow the instructions for using the 2017 Website Factor Workbook 10.16.2017, which is attached at the bottom of the page. And, here's an editorial comment: We're now about 80% through the season. The simulations at this point are starting to point to where we are likely to end up. There's still room for some significant changes, but the likelihood of big changes is getting smaller.
Good stuff, thanks--always look forward to the reports, especially now that we're late in the season and approaching tournament time.
Good stuff...Question for you...Southern California's schedule is heavily back loaded with 5 very strong teams. (the 2 Washingtons, Cal, Stanford, and UCLA) should they go 4-1 do you see the Comittee not giving a seed to the defending champs with a record of 15-2-1? Was the non-conference scheduling their kiss of death?
My simulation currently has Southern California going 1-4 over its remaining games. I ran the simulation sequentially change the results of games, one at a time. First, I changed Washington to a win, then Cal, then UCLA. With the Washington win, the simulation brings So Cal in at #33, which would mean no seed. Also adding the Cal win, they come in at #25, which would make them a #4 seed candidate. Also adding the UCLA win they come in at #18, which would make them a #3 seed candidate. It's a little too much work to run my bracket simulation when I've manually altered the ranking simulation, but I'm guessing that if they were to go 4-1, they'd be a very good shot for a #3 seed. I don't think it matters to the Committee that they're the defending champions. Rather, I think they focus on this year's results.
Wow...thanks so much for doing that and for god sakes I hope your simulations are wrong! Been a fun season all the way around. Cheers
I finally have finished publishing my reports for this week. They're later than usual for two reasons. First, there are discrepancies between my RPI rankings and the NCAA's. It took me several hours to track them down. They appear all to be due to three NCAA data errors, of which I've notified the NCAA. Based on past experience, the NCAA will check to see if I'm right about the errors and then will make corrections. I'll know next week. Second, now that we have some actual conference tournament brackets, my process involves more manual work on my part, which takes longer. Here are this week's publications: An RPI report covering teams' current actual ratings, covering games through Sunday, October 23. If you want to see teams' current RPI ranks, you can go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the most recent post titled: 2017 RPI Ranks 10.23.2017. If you want more RPI detail, including information about teams' Non-Conference RPIs and the conferences' average RPIs, you can go to the RPI website's This Year's Reports page. At the bottom of that page, there's a link to the 2017 RPI Report 10.23.2017 that is an excel workbook with a lot more detail. The detail includes conferences' average RPIs and ranks plus other information. The NCAA continues to not publish its detailed RPI reports at the RPI Archive. This appears to be true for all Fall sports. This is a departure from what it's done the last few years, is very unfortunate, and is one of the reasons it took me so long today to identify the NCAA's data errors for this week. Also, I now have substituted into my weekly simulations the actual results of all games through Sunday, October 23, have included the actual conference tournament brackets for those conferences that have completed conference regular season play, and have produced an updated simulation for the entire season that includes end-of-season ratings, ranks, and other data and a simulation for the #1, #2, #3, and #4 seeds, the Automatic Qualifiers, and the unseeded at large selections for the NCAA Tournament. If you're interested, go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the posts titled: 2017 Simulated RPI Ranks 10.23.2017 and 2017 NCAA Tournament Bracket Simulation 10.23.2017. And, If you are really serious about understanding the details of how your team is doing or how the overall bracket simulation system works, go to NCAA Tournament: Predicting the Bracket, Track Your Team at the RPI for Division I Women's Soccer website, read that page, and then follow the instructions for using the 2017 Website Factor Workbook 10.23.2017, which is attached at the bottom of the page.
Here are this week's publications: An RPI report covering teams' current actual ratings, covering games through Sunday, October 30. If you want to see teams' current RPI ranks, you can go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the most recent post titled: 2017 RPI Ranks 10.30.2017. If you want more RPI detail, including information about teams' Non-Conference RPIs and the conferences' average RPIs, you can go to the RPI website's This Year's Reports page. At the bottom of that page, there's a link to the 2017 RPI Report 10.30.2017 that is an excel workbook with a lot more detail. The detail includes conferences' average RPIs and ranks plus other information. Also, I now have substituted into my weekly simulations the actual results of all games through Sunday, October 30, have included the actual conference tournament brackets for all conferences, and have produced an updated simulation for the entire season that includes end-of-season ratings, ranks, and other data and a simulation for the #1, #2, #3, and #4 seeds, the Automatic Qualifiers, and the unseeded at large selections for the NCAA Tournament. If you're interested, go to the RPI and Bracketology for D1 Women's Soccer Blogspace and check out the posts titled: 2017 Simulated RPI Ranks 10.30.2017 and 2017 NCAA Tournament Bracket Simulation 10.23.2017. And, If you are really serious about understanding the details of how your team is doing or how the overall bracket simulation system works, go to NCAA Tournament: Predicting the Bracket, Track Your Team at the RPI for Division I Women's Soccer website, read that page, and then follow the instructions for using the 2017 Website Factor Workbook 10.30.2017, which is attached at the bottom of the page.
I must admit I find the RPI calculations confounding at times but math was never my best subject. Here is an example: Duke beats Virginia (#21 Adjusted RPI) and UNC beats NC State (#28 Adjusted RPI) but UNC's Adjusted RPI increases to .6851 from .6810 and Duke's RPI only increases to .6817 from .6806 meaning UNC is now further ahead of Duke on an adjusted RPI basis. This all assumes the woso-stats.tk RPI calculations are accurate this morning. I know that the other games being played have an impact but it seems that beating a higher RPI team should help more? What am I missing?
Here's a simple explanation -- the total explanation might be a little more complicated, but this should suffice. 1. NC State's record, apart from the UNC game, is 14 W, 4 L, 1 T. Virginia's, apart from the Duke game, is 11 W, 4 L, 4 T. In computing what those two teams add to UNC's and Duke's strength of schedule element of the RPI, those are their winning percentages to use. (Team B's contribution to Team A's strength of schedule is Team B's record against other teams.) 2. Team A's RPI consists of three elements: Element 1: Team A's winning percentage Element 2: The average of Team A's opponents' winning percentages against other teams. Element 3: The average of Team A's opponents' opponents' winning percentages. 3. In terms of effective weights, Element 1 is 50%, Element 2 40%, and Element 3 10% (all "roughly") of a team's RPI. 4. As you can see from NC State's and Virginia's records, NC State has a significantly better winning percentage than Virginia. Thus the strength of schedule contribution that UNC got from NC State was greater than the contribution that Duke got from Virginia. This is true even though Virginia has a better rank than NC State. So, that's why UNC benefited more even though playing a more poorly ranked team. As you can see, the RPI has an issue: A team's RPI rank can be different than its rank in terms of what it contributes to its opponents' strength of schedule. In fact, the differences can be quite large. I have an alternative version of the RPI, which I call the 5 Iteration RPI, that does not have this problem. And, it doesn't have another RPI problem, which is that the RPI discriminates against stronger conferences and regions and in favor of weaker ones. My 5 Iteration version has a little discrimination regionally, but not much. And, the 5 Iteration version's ratings match better than the RPI with the Women's Soccer Committee's actual NCAA Tournament decisions. The NCAA knows about my 5 Iteration version, but so far hasn't indicated any interest in trying it out.