RPI Element 1 is a team's winning percentage Element 2 is the average of its opponents' winning percentages (against teams other than the team itself) Element 3 is the average of its opponents' opponents' winning percentages So Elements 2 and 3 are strength of schedule. The effective weights of the three elements, at the end of the season, are roughly 50% Element 1, 40% Element 2, 10% Element 3. So overall it's roughly 50% winning percentage and 50% strength of schedule. Earlier in the season, strength of schedule has a greater effective weight and winning percentage a lesser weight.
If you go to the RPI: This Year's Reports page at the RPI for Division I Women's Soccer website, at the bottom of the page there are links to detailed weekly reports that include, towards the end, conference average ratings. For this week's report, the one you want is 2018 RPI Report 9.24.18. For conferences, scroll down to page 34. If you really want to peruse the information and have Excel, you can download the report. For teams, it's pretty difficult to use on-line. For conferences it's easier.
The NCAA RPI is out... with some glaring mistakes. Most obvious is they don't have the BC vs. WFU result logged for some reason. https://www.ncaa.com/rankings/soccer-women/d1/ncaa-womens-soccer-rpi
Great catch, I just finished vetting my numbers for games through yesterday (September 30) and saw that. The most likely reason they don't have that game would be that the schools for some reason didn't report it on time. It looks to me like its in the NCAA's statistics system now, but it may have come in after the release of the RPI. That's the only glaring error, but it has a pretty big effect. There is one other data error, which is that the have the result of 9/30 Alabama State 10 v Alcorn State 0 backwards. They also have a minor programming error, which is that although the number of teams competing at Division 1 increased from 333 to 335 this year, which means they should have adjusted upward by 2 positions the tiers for penalty adjustments for poor wins and ties, they forgot to do it. As I do weekly after vetting my numbers, I've informed the NCAA of all of these issues and I'm confident they'll make corrections. When I issue my RPI materials later today, they'll be exactly correct. Also, the All White Kit numbers are exactly correct.
I just have posted my weekly actual ARPI rankings and simulation updates at the Blogspace: 2018 RPI REPORT 10.1.2018 2018 SIMULATED RPI RANKS 10.1.2018 2018 SIMULATED CONFERENCE STANDINGS AND CONFERENCE TOURNAMENTS 10.1.2018 2018 SIMULATED NCAA TOURNAMENT BRACKET 10.1.2018 In addition, at the RPI for Division I Women's Soccer website, on the NCAA Tournament: Predicting the Bracket, Track Your Team page, I've put an attachment at the bottom of the page that will help you see the details of how each team fares in terms of the NCAA Tournament, within the simulation. The title of the attachment for this week is 2018 Website Factor Workbook 10.1.2018. If you're a coach or a fan, you can use this to see how your team's simulated record matches up with a whole series of the Women's Soccer Committee's seeding and at large selection decision-making patterns over the last 11 years. The webpage itself includes an explanation of the attachment.
I just noticed that TDS seems to still be publishing "Power Rankings" different from their national rankings. Anybody have any insight into what goes into that? Is it supposed to be their version of an RPI? Not even sure if its kept up to date, they don't really publicize its existence.
I've discovered and corrected a programming error that, in very limited circumstances, affected my conference tournament simulations. One of those circumstances occurred this past week. I've therefore replaced the tables in the posts covered by the above links (except for the 2018 RPI Report 10.2.2018, which wasn't affected by the error) with corrected tables.
RPI DATA ERROR ALERT: The NCAA's RPI system has some data errors this week and a continuing programming error: 1. Most significantly, their RPI rankings and reports for games through yesterday, October 7, do not include the South Carolina (home) 1 v Florida 0 game. In the NCAA Statistics system, this game has a reported game date of today, October 8. Apparently because of that, the computation of teams' ratings and rankings through yesterday, October 7, do not include the game. This obviously is a significant game to have not be in the system. 2. The NCAA's data have the 10/7 game between Bryant and Fairleigh Dickinson (home) as having been won by Bryant. It wasn't, it rather was won by Fairleigh Dickinson 1-0. 3. The NCAA's data do not include the 10/4 game of Delaware State (home) 1 v South Carolina State 1. It appears to me that these schools failed to enter this game into their schedules filed with the NCAA. 4. The NCAA, notwithstanding notice of the problem last week, still has not corrected the tiers opponents for which poor results against them cause the imposition of penalties. As a result of this, for this week poor results (ties or losses) against Vermont and Lafayette have resulted in penalties where there shouldn't be penalties. And, poor results against Chicago State and Alcorn State are not resulting in penalties when they should. Ordinarily, this would be inconsequential. Villanova, however, ranked in the 60s, had a poor result against Lafayette and being penalized when it shouldn't be. The NCAA is aware of all these issues and hopefully will fix them.
I just have posted my weekly actual ARPI rankings and simulation updates, for games through Sunday, October 7, at the Blogspace: 2018 RPI REPORT 10.8.2018 2018 SIMULATED RPI RANKS 10.8.2018 2018 SIMULATED CONFERENCE STANDINGS AND CONFERENCE TOURNAMENTS 10.8.2018 2018 SIMULATED NCAA TOURNAMENT BRACKET 10.8.2018 In addition, at the RPI for Division I Women's Soccer website, on the NCAA Tournament: Predicting the Bracket, Track Your Team page, I've put an attachment at the bottom of the page that will help you see the details of how each team fares in terms of the NCAA Tournament, within the simulation. The title of the attachment for this week is 2018 Website Factor Workbook 10.8.2018. If you're a coach or a fan, you can use this to see how your team's simulated record matches up with a whole series of the Women's Soccer Committee's seeding and at large selection decision-making patterns over the last 11 years. The webpage itself includes an explanation of the attachment.
I just have posted my weekly actual ARPI rankings and simulation updates, for games through Sunday, October 14, at the Blogspace: 2018 RPI REPORT 10.15.2018 2018 SIMULATED RPI RANKS 10.15.2018 2018 SIMULATED CONFERENCE STANDINGS AND CONFERENCE TOURNAMENTS 10.15.2018 2018 SIMULATED NCAA TOURNAMENT BRACKET 10.15.2018 In addition, at the RPI for Division I Women's Soccer website, on the NCAA Tournament: Predicting the Bracket, Track Your Team page, I've put an attachment at the bottom of the page that will help you see the details of how each team fares in terms of the NCAA Tournament, within the simulation. The title of the attachment for this week is 2018 Website Factor Workbook 10.15.2018. If you're a coach or a fan, you can use this to see how your team's simulated record matches up with a whole series of the Women's Soccer Committee's seeding and at large selection decision-making patterns over the last 11 years. The webpage itself includes an explanation of the attachment.
One place you can look, that has Average of Opponents' Winning Percentages (Element 2) -- which the NCAA calls Strength of Schedule; and Average of Opponents' Opponents' Winning Percentages (Element 3) -- which the NCAA calls Opponents' Strength of Schedule is the Team Sheets that the NCAA publishes weekly. Here's a link for the Team Sheets for games played through 10/14. For the RPI formula's combination of these two numbers, starting tomorrow, in the RPI Report I publish, I'll include the combination of those two numbers per the NCAA formula, which I call Strength of Schedule. And, I'll include the SoS Rank. The formula is (2* Opponents' Average Winning Percentage + Opponents' Opponents' Average Winning Percentage)/4. If you divide your team's Winning Percentage by 4 and add it to that number, you'll get your team's unadjusted RPI. I haven't checked to see, but one thing I just saw, in answering the question above, is that Baylor's Strength of Schedule Rank, through the 10/14 games, was #2. Combined with their 12-4-0 record, that gave them a #5 ARPI rank. That rank put them at the high end of the potential #1 seeds as of last week's games, which are teams #1 through #19. So, it doesn't surprise me to see them as a possible #1 seed at this stage of the season. They're also currently at the top of the Big Twelve regular season competition. For games through 10/14, the Big Twelve was the #2 ranked conference, behind the SEC and ahead of the ACC.
Thanks cpthomas! Regarding Baylor and the Big 12.... that's shocking. Are the bottom teams in the Pac 12 and ACC really dragging the conferences down that much? I'm shocked Georgia's RPI is as high as it is (among other SEC teams as well like Miss St.).
I just have posted my weekly actual ARPI rankings and simulation updates, for games through Sunday, October 21, at the Blogspace: 2018 RPI REPORT 10.22.2018 2018 SIMULATED RPI RANKS 10.22.2018 2018 SIMULATED CONFERENCE STANDINGS AND CONFERENCE TOURNAMENTS 10.22.2018 2018 SIMULATED NCAA TOURNAMENT BRACKET 10.22.2018 In addition, at the RPI for Division I Women's Soccer website, on the NCAA Tournament: Predicting the Bracket, Track Your Team page, I've put an attachment at the bottom of the page that will help you see the details of how each team fares in terms of the NCAA Tournament, within the simulation. The title of the attachment for this week is 2018 Website Factor Workbook 10.22.2018. If you're a coach or a fan, you can use this to see how your team's simulated record matches up with a whole series of the Women's Soccer Committee's seeding and at large selection decision-making patterns over the last 11 years. The webpage itself includes an explanation of the attachment.
My simulation likewise has the Big 12 receiving a #1 (Baylor), #2 (Texas), and #3 (West Virginia) seed. It has Baylor finishing with a rank of #2. Also, the Big Twelve, SEC, and ACC are bunched extremely close together in the current conference actual average ARPI standings, in that order, with the Pac 12 only slightly behind in 4th. Baylor's current actual strength of schedule rank is #1. And, on another matter, my simulation has Wake Forest ending the season ranked #31 but with an 8-9-1 record (current record of 8-7-1, with a loss to North Carolina in the last game of the season and again to North Carolina in the first round of the ACC Tournament). If that happens, they will be barred from getting an NCAA Tournament at large selection. This would be the first time in the 11 years I've been tracking this that a team that otherwise would have gotten an at large selection almost for sure will not get one due to a below 0.500 record.
Under the current version of the RPI, Duke in 2014 would have been #61 in the ARPI rankings. If their ranking had been better, they might have been considered for an at large selection, except that their below 0.500 record would have DQ'd them. Even if they had been considered, the Committee's historic patterns indicate they likely would not have gotten a selection. Wake Forest this year, on the other hand, is a whole different situation. They're likely to end up #31 and meet the profile of teams that always have gotten selections over the last 11 years, except for their likely record below 0.500.
Only 8 teams will make the ACC tournament correct? If that’s the case, it looks like they will finish at 9th in the standing if they were to lose to UNC. Will they make it in then? Some of these teams in ACC and SEC that aren’t good enough to make their conference tournament shouldn’t make it in to the NCAA tournament imo. Why have a conference tournament if you don’t need to play? Do it like WCC or Pac-12 then
My understanding (as a Wake fan): the Wake women are guaranteed a .500 finish. Currently 8-7-1, with 1 regular season game left, and are currently in a 3-way tie for the 8th spot in the ACC Tourney... but if my math is correct will lose any tiebreaker with any combination of VT, ND, and NC State at 12 points by virtue of having at best a 1-1 cumulative record against the other three (loss against ND, win against NC State). So the three outcomes of the UNC game: - Wake wins, moves to 9-7-1, clinches above .500. - Wake draws, moves to 8-7-2, clinches .500 even with loss in ACC Tourney - Wake loses, falls to 8-8-1, clinches .500 due to not qualifying for ACC Tourney
Ah, you may be right. If Wake doesn't get into the ACC tournament, they'll have at least an 0.500 record and should get into the NCAA tournament.
CP, Thanks for running this.. qq for you.. first round of NCAA, top 32 in RPI host a game? or how is home scheduling determined for first round games?
The 16 seeded teams host first round games, so that's half of the games. I believe, with confidence but not certainty, that after the Committee does the 16 seeds, it builds out from there with the next group of teams, at some point arriving at the teams they think are 17 to 32. Ordinarily, those teams get to host the remaining games. For travel expense reasons it may not work exactly that way, but it's pretty close. Those teams are not necessarily 17 to 32 in the RPI rankings, although they may be close.