That is the best way to increase your RPI. How exactly are they going to entice schools to play them and why should they be forced to travel? Why on earth would a team like Monmouth, High Point etc. want to have their student athletes take several long trips out west? When the Pac12 played the ACC, all the games are on the road. Why do you believe the West should travel instead of having the NCAA fix an obviously flawed system (remember the NCAA has made changes in basketball and baseball to make it more equitable for teams that travel)? I'm not sure where anyone said there were no weak teams. Instead the issue is there are no weak conferences. Also the number you provide are the skewed RPI numbers. The RPI is mostly a measure of a relative distribution. So if everyone in your region is above, those at the bottom will necessarily have lower RPI rankings since they are at the bottom of the regional distribution. I'm not sure where anyone has stated all the teams are strong. Instead the problem is that the RPI significantly under ranks teams from the West coast in Men's soccer because the large majority of the teams are above average, and contrary to your simplistic solution, it is not possible to schedule games in such a way that would make such a flawed system equitable. On the other hand, you could almost assuredly hire someone to design much more equitable system for the less than the cost of cross country trip for one team. Why does it make more sense to you to have men's teams from the west coast try to travel all over the country than it does to change a broken system? They already do. The problem is that teams like Davis, Northridge, St Mary's, Cal, Stanford don't benefit from it because they play a top 50 team with a record like a top 100 team such as East Tenn St, Detroit, Appalachian St, Radford, Monmouth etc. No the issue is that the NCAA should have a system that allows the most deserving teams to make the playoffs and it clearly doesn't which taints their playoffs. They also claim that RPI is just one of many factors that determines participants in the playoff system. However it should be pretty obvious to anyone with a reasonable understanding of statistics that the most deserving teams in terms of performance are not getting in and the reason for this is because contrary to what the NCAA claims, the selection committee has made their selection based on the flawed RPI calculations.
If you don't understand yet, I doubt there is much more I can tell you that will help. Most people intuitively would understand that beating a team like Fullerton should not make Davis' chances of making the NCAA playoffs worse while someone on the other side of the country can increase the chances of making those same playoff by beating a less capable team like Monmouth. Davis is a very good team and has really improved their speed of play over the past two years. If they played in the Big East, they would easily get an at-large bid, but since they play in the west their prospects are very dicey at best. I wish them good luck.
Conjecture is not proof. The same issue (a win against a weak opponent can lower RPI) can happen to any team in any conference. This is recognized and is one of the factors considered when selecting the teams for the tournament. To draw a historical parallel, Aritotle observed that the planets move wioth respect to the fixed stars, Copernicus observed the motion was centered on the sun, Kepler observed the paths were ellipses, and Newton proved they must be ellipses. In your analysis of RPI, you are somewhere between Copernicus and Kepler.
Nice! This is a first for the NCAA and a very positive development. With all due respect to GauchDan - and he's changed the way most of us look at college soccer in the regular season - having the official numbers the committee uses is awesome.
Those guys were all way smarter than I am. But more to the point, is there anything that would qualify as proof? 1) It is an indisputable fact that the RPI formulas provide will skew rankings its rankings lower for an isolated population of above average teams. 2) Now the next question is whether the Western conference teams are above average compared to the general population. Massey says the Pac12 is the best conference. Their record is 2-2-1 against the ACC this year. But since all the games were at ACC venues most computer systems would say the Pac12 outperformed the ACC something like 1.3 to .7. Of course you could counter that by saying 1 of the wins and one of the ties were by the top team in the Pac12 and even if the Pac12 won all 5 games that wouldn't prove anything.
That is being kind. Here are some comments from an article last year about the men's basketball RPI (keep in mind the RPI calculation in Men's basketball takes into account home field advantage and the number of teams is greater that women's soccer so you don't have any group of teams absurdly skewed rankings like you do with western teams like Men's soccer.) http://www.nytimes.com/2012/03/11/s...-outside-the-ncaa-tournament-bubble.html?_r=0 Yet people on this board are constantly quoting RPI numbers when claiming one team is stronger than another or teams from the West need to schedule tougher opponents. The reality is that even when it is improved to include things like home field advantage and is used on a larger mover balanced pool of teams it is a very poor system. When it is applied to a population of teams like Men's soccer where one region of the country has no below average teams it turns into a travesty.
The Massey rating system is focused on "prediction" since it was developed as a tool for gamblers and bookmakers. The RPI was developed as a publicly accessible method to compare the performance of temas based on game results. There is no gimmick or tweak in the formula that takes into account any geographical information about the teams being analyzed.
How much does seeding usually deviate from the NCAA RPI that was released yesterday? Northeastern is #21 for RPI. In College Soccer News they have them projected to play at Brown who has a 34 RPI. I cant imagine that is very likely.
Why not? Neither team is high enough to get a bye and the schools are close enough that it's cheap for the NCAA to have them play each other. A lot of early NCAA Tournament matchups are based on geography because of the costs savings.
Most of my point was that NU is playing AT Brown versus the other way around. And if the take geography into account that much I would think BC would play AT NU since they are 38
The CSN article was written a few weeks ago. Correct? Don't worry in 6 hours, we'll know (the good, the bad and the ugly).
Each conference gets an automatic bid for their champion. Most conferences have a tournament to determine their champion/automatic bid. Three conferences - the West Coast, Pac 12 and Ivy - do not have a conference tournament and instead award their automatic bid to the conference's regular season champion. There are 22 conferences that play college soccer, ergo there are 22 automatic bids. The remaining 26 bids are "at-large" bids. They are determined by a selection committee made up of mostly of ADs and associate ADs. They are tasked with selecting the 26 teams deemed most worthy of participating in the NCAA tournament that did not win an automatic bid. The primary criteria they use for this selection is the RPI rating. However, they do not use it exclusively so teams towards the bottom of the Top 26 list are on the proverbial bubble and a team outside the Top 26, as it were, might leapfrog a team that is towards the bottom of the 26. (Also, teams must not have losing records to be included for at-large bids.) Besides RPI, they look at things like big wins, poor losses, finishing strong and strength of schedule. It's not perfect and there are always teams that feel like they got snubbed. The committee is also tasked with deciding the 16 teams that get "seeded" and get the first round bye and the matchups and where they play within the brackets they pick. Hope that helps.
It starts with the 22 automatic qualifiers from the 22 conferences regardless of their RPI ,then you fill in the remaining 26 slots to get to a field of 48. The notable element here is that every automatic bid winner with an RPI greater than 48 means one of the top 48 teams is not getting it. It usually works out that the cut off falls to about an RPI of 42. This year there are nine teams with an RPI > 48 83 Air Force (MPSF) 80 Cleveland State (Horizon) 108 Farleigh Dickinson (Northeast) 65 Florida Gulf Coast (Atlantic Sun) 115 Lafayette (Patriot) 81 San Diego (West Coast) 51 UMBC (America East) 170 Western Illinois (Summit) 73 Winthrop (Big South) So the cut line looks to be around 39.Here are the four teams above and below that line in the sand. 35 SIU Edwardsville Missouri Valley 13-7-0 3-4-0 3- 2- 0 7-1-0 0-0-0 36 Xavier Atlantic 10 13-2-5 4-1-3 0- 1- 1 8-0-1 1-0-0 37 Coastal Carolina Big South 18-2-2 6-1-1 2- 1- 0 10-0-1 0-0-0 38 Boston College Atlantic Coast 8-5-5 4-2-3 0- 0- 0 4-3-2 0-0-0 40 Syracuse Big East 12-6-0 5-2-0 2- 1- 0 5-3-0 0-0-0 41 Northwestern Big Ten 11-5-4 2-1-3 3- 0- 0 6-4-1 0-0-0 42 UC Davis Big West 10-7-4 3-3-4 1- 0- 0 6-4-0 0-0-0 43 Villanova Big East 12-6-2 4-3-1 1- 1- 0 7-2-1 0-0-0 Perks along the way.... Top 16 seed get an opening round bye and open at home. Top 4 seeds play at home through quarterfinals The wrinkle in it all is filling in in the 26. Rarely do they follow RPI exactly but it is pretty darn close so those deviations from the RPI are the points that make it all interesting
Based on current NCAA RPI (Nov 11)...Top 48 by conference.................. Big East 10 (ND, GT, CT, MQ, LU, STJ, USF, RU, SU,VNova) ACC 6 (MD, UNC, VA, WF, BC, Duke) A10 4 (STL, VCU, CHAR, XAV) CUSA 4 (TUL, KU, SMU.,UAB) Big Ten 4 (IND, MU, MSU, NWest) CAA 3 (NEAST, ODU, DREX) Pac-12 3 (UCLA, WASH, CAL) Ivy 2 (CU, BU) MVC 2 (CU, SIUE) MPSF 2 (NM, DU) Big South 2 (CC, CU) Big West 2 (CSN, Davis) MAAC 2 (Loyola, Niag) Southern 1 (Elon) MAC 1 (Akron)
Here's what the NCAA needs to do to help fix the sport. 1) incorporate home field advantage into the system. That seems like a no brainer. Surprised it's not currently part of the formula since it is for basketball 2) Screw the Geographical aspect of the tournament. I cannot imagine travel costs are truly that prohibitive that you should screw up the seeding of the NCAA tournament just so teams travel less. The best teams should make the tournament regardless of Geography and the teams should be rewarded for good seasons by getting to play a weak opponent in the first round if they've earned that right. You shouldn't force two good teams to play each other in the first round just because they're close to each other. 3) They should set up a system like the bracket busters tournament in basketball. They set up good games between quality mid majors to help them improve their RPI. I don't see why they couldn't do the same thing, but instead of mid majors which is less of a distinction in soccer just use it to have west coast teams to get more games in the midwest and east. 4) What would really help things is to just increase the number of schools that play men's soccer. But that's not something that can just happen. 5) The RPI is not a great system for Basketball in terms of predicting the NCAA tournament, but that's not what it's designed to do. In addition, I don't have the facts in front of me. But someone mentioned in an RPI thread I think on the women's side that the RPI actually does a great job of predicting success in Soccer, so the fact that it doesn't work in basketball isn't really relevant. The large size of the basketball tournament means there really aren't teams who are getting left out who deserve to be in. This isn't true for men's soccer. 6) I really like the advanced statistics for basketball and follow them closely and think they're often better than the RPI but they're very flawed as well. There's never going to be a computer model that captures everything perfectly. There are human aspects to the games that can't be measured. I really don't think soccer lends itself to these advance stats the way basketball does where there are lots of stats to measure. I don't know how things like offensive efficiency and things like that factor in to soccer. I mean your FG% doesn't really matter nor make you more likely to win in a game of just a few goals. A team could shoot 95% and lose to a team that shot 20% but just took more shots. But a team could take twice as many shots as another team and still not win. Soccer is a game where much more often the best team doesn't win compared to a sport like basketball. The advanced statistics in basketball also work due to the large enough sample size of games played to connect all the teams. I'm not sure that happens in soccer. 7) I'm not saying the RPI isn't somewhat flawed, but I'm not positive there's a better model out there currently. Where you lose me is when you say things like there are no below average teams in the west or the best teams aren't getting in. I don't know how you can definitively say that. I don't know what it's based on. Is it based on Massey? I don't know what that system entails, but I'm sure there are flaws in it as well. I think the RPI actually is a good system overall for soccer because it focuses on the final results. The best teams are those that beat teams that beat teams. Soccer is a very stylistic game but in the end all that really counts is who scored more. A team can dominate another team but still lose 1-0. You can argue that the team that dominated shots and possession is a better team, but you are what your record says you are. Win games against teams that also win games that's the mark of a good team and that's what the RPI measures. 8) Outside of adjusting for home field advantage which I think is a good idea how do you change the RPI. From what I understand there are already bonuses for quality wins and negatives for bad loses which adjust for that fact that the RPI in a raw form cares more that you played a quality team rather than if you beat a quality team. So what changes do you make to the RPI to improve it.
quote="HoyaHooligan, post: 26765776, member: 133082"]Here's what the NCAA needs to do to help fix the sport. 1) incorporate home field advantage into the system. That seems like a no brainer. Surprised it's not currently part of the formula since it is for basketball[/quote] I'm pretty the bonus and penalty system used incorporates that a win on the road vs a Top 25 or Top 50 team gets you maximum bonus all the way down to a home loss vs a Bottom 25 or Bottom 50 is the maximum penalty. Its pretty complex from what I remember but indirectly, home field advantage is factored
I think they do that, do a degree, already. That's why the NCAA's RPI differs slightly from GauchoDan's. Doesn't mean the couldn't tweek it. $$$ You're talking about airline tix, hotel, food for 25 or so people. It adds up and doesn't bring much, if any, revenue in return. Again, $$$ Who is going to pay for it? If costs didn't matter, UCLA would play all their non-con games against Big East, Big Ten and ACC schools plus Akron. Bracket-busters are (besides a great idea) profitable for the NCAA and the schools involved. They'd just add to the money lost in men's soccer. And, still more $$$. If college soccer brought more to the schools in terms of money and recognition, schools would be tripping over themselves to add men's soccer.
How (mathematically how) would you incorporate home field advantage? And how much advantage is there? It's not like soccer games are held in an indoor location with 10,000 screaming home fans. Perhaps an additional analysis that accounted for referee errors in close games? NCAA hockey starts with RPI as part of the system, and then compares all of the teams with a reasonable chance for the tournament with each other, based on objective measures such as head-to-head results, common opponents, and records against teams with .500 or better RPI. RPI is used as entry for the comparison process (.500 or better RPI) and as the tie-breaker in some of the comparisons. The teams that win the most comparisons are usually selected. There are still some wrinkles having to do with conference autobids that would not otherwise be selected, and whether to include them in the comparisons, but all the factors are public. http://www.uscho.com/faq/pairwise-rankings-explanation/