Not taking offense, but curious in your ranking of FSU. We've been to the elite 8 or farther 12 of the last 14 years. More CC appearances than other top programs. Played #1 SOS last year with nearly twice as many ranked opponents. I would think our trend line would be steady or even up. What gives, especially in light of WVU and GU ahead of us in your rankings.
I don't know anything about the UVA players, but would suggest more broadly that evaluating players based on the recent U17 WNT experience is problematic. The team that was selected was beaten by Korea 3-0 I believe, and by Germany 4-0, and beat Cameroon playing up 11 vs. 9 and did not progress through the group stage, so it would be difficult to draw any lines between team selection and player quality in college.
Tom81, here's how the data chart out for Florida State. The lines to look at are Massey's blue and the ARPI's red. As you can see, the ARPI says FSU is trending slightly better year-to-year and Massey says they're trending slightly poorer year-to-year. From my perspective, these are very similar, given that they're dealing with statistical ranking systems' rankings of teams over time. When you're looking at statistical system rankings like these, try not to get perturbed about the exact placement of teams. Rather, think about whether where your team is is "in the ball park," which is about as good as statistical systems can do in terms of predictions. (And, since they're better on average than most humans, I suggest not expecting your own predictions to be precise either.)
FSU for sure. I think i told you about Patten well before anyone else. South Carolina, UNC, PSU, Duke , UCLA, USC to name a few or the higher profile ones
Im confused as to why its problematic? It depends whether you actually have watched the players and its one of many data points OR its just based on the result. I think some people do the latter, which I agree is problematic. Im doing the former. My comments above are based on that. UVA are not going to replace McLernon or Morse with a freshman unless forced by injury. Its possible they go 3 back and then maybe Staude gets a shot, but there are other more experienced options. Her time will come but i dont think she will have a big role in 2019 if it plays out how it should on paper. At least not at CB
i think you have to look at all the games. The road to the CC is not the same for everyone. I would guess that a statistical system simply evaluates every game and a CC game has no more statistical significance than a regular season game.
Look forward to the season. Should be another good one. There is no odds on favorite unlike last year, and we see how that worked out. New players and transfers can make huge impacts. In 17 we lost to Stanford 1-0 in the quarterfinals. While the score was close, the game was not. They clobbered us and dominated possession and every key statistic. Their midfield was particularly dominant. One season later with the addition of Yujie Zhao, Jaelin Howell and others, we totally turned the tables on Stanford. We dominated (not as much as them the year before) the midfield and consequently held them to 1 SOG. My point being, we don't know which team(s) is going to get that key piece or pieces that puts them over the top. I'm looking forward to seeing it. Quite frankly, if I take FSU out of the equation, I wouldn't mind seeing either Anson getting back on top or UVA and Swanson getting the CC trophy. College women's soccer rivalries tend to being much more friendly!
To be honest, Stanford were the favorite and deserved to be. I think a healthy Stanford were teh best team in College soccer last year, but we rarely saw that. They lost a lot thru injury last year. Other than them, I think all the teams were pretty close last year. I just look at each game and dont really look at the CC as much more than another set of games. Teams travelling across the country, playing 2 games etc are all factors that detract from optimum performance, but are obstacles to overcome. i rarely get involved in the whos going to win the CC stuff. at the end of the day there are 10 or so teams that to me have pretty equal chances and the breaks will determine who gets there. For me, we kinda do know who the pieces are but we dont know what role injuries or good fortune will play.
This would be true if I included the NCAA Tournament games in my system, but as I indicated way back when, I don't even include those games in my analyses. That's a limitation on what I do, but it's with a good reason: My field of work is the ratings and other data that the Women's Soccer Committee uses when it makes its NCAA Tournament decisions and how teams and conferences can schedule so that their ratings and other data will put them in the best light when the Committee is considering their profiles. Teams that do "better" in the NCAA Tournament would come out better if I included tournament games (last year, Florida State moved from #5 to #1 in Massey's rankings), but on average I don't think the difference would be huge. If a team has a history of "outperforming" its end-of-regular-season ranking, in the NCAA Tournament, that's where there could be a noticeable difference. But, if you think of the question as whether teams are "in the ball park" of where they need to be to be a College Cup participant, and even a winner, using the end of regular season data seems to work just fine.
Excellent points. But what's your definition of "long term"? Although you have used UNC as your poster child, surely the "long term" factor doesn't have to be anywhere near 39 years with the same head and assistant coaches? And can this stability factor transfer from school to school if a head coach brings the assistants along to the new school?
Good questions. For "long term," it's got to be long enough for the team to have reached a stable point under the coaching staff. I haven't done a detailed project on it, but I think it takes about 8 years for a coaching staff to reach that point. It would take more years after that for the data to show that that point indeed is stable. I use a 12-year period for my trending, because that's the number of years in my data base. But, for coaches who are new during that period but with 4 or more years as their team's head coach, I also look at the team's trend during that coach's tenure, as compared to its 12-year trend, to see which seems to better indicate where the team is going. And, sometimes neither seems to give a good picture of where the team is going, in which case I default to last year's rank. Interestingly, for the top 12 teams my system identified based on trends going into the 2019 season, here is the head coach longevity: Stanford 16 years UCLA 6 So California 5 North Carolina 37 Georgetown 15 West Virginia 23 Florida State 14 Duke 18 Virginia 19 South Carolina 18 Penn State 12 Tennessee 7 Plus, of the three relatively shorter term coaches, UCLA's and Tennessee's coaches previously had established themselves as top coaches over a significant period of time. So California's had over a much shorter period of time. For these three, however, one might look at their programs as potentially still in flux, whether for good or for ill. On the above list, the one that was most tricky to assign a position to was Georgetown. I look at straight line trends and also at order 2 polynomial trends to see which seems best based on a team's history. For Georgetown I used an order 2 polynomial trend, which gave it a significantly better rank than a straight line. If I made the wrong pick, then Georgetown will be down on the list and maybe several positions off it. Picking which trends to use for a particular team is where the potential for human error comes in. Mostly I use straight line trends. But, the system seems to work pretty well: One might argue about where the teams should fit on the list, but I think it's hard to argue that the list overall has the wrong teams.
OK, here's me handicapping seeding for 2019 I see three teams at the top tier, and as solid 1-seeds - FSU, North Carolina, and UCLA. FSU - the ‘18 champions come back with a roster that is mostly intact from the previous championship season. Solid bet to play through the College Cup. North Carolina are always a solid bet to earn a 1-seed. Possess the experience and firepower to play through the College Cup. UCLA have the most experienced team with championship aspirations. Last season, I had FSU as the most experienced team heading into the season, and they sealed it as national champions. UCLA have the most experienced team this season and also the firepower to win it all. I see two teams on the 2nd tier - Stanford, and Virginia - and will compete for the 4th 1-seed. Stanford graduated a lot of experienced players, but also return a lot of talented players. Virginia have the experience and firepower to earn themselves a 1-seed. The odd team out between Stanford and Virginia will join USC, TAMU, and Santa Clara as 2-seeds. USC lost some to graduation but have the firepower to challenge. TAMU have the experience and firepower to earn themselves at least a 2-seed. Can they find answers to their post-season struggles and play to their seed? Much like TAMU, Santa Clara have the experience and firepower to earn themselves at least a 2-seed. They score a lot, but give up a ton of goals as well. Get good results against College Cup opponents North Carolina and Stanford, but RPI takes a hit playing against mediocre conference opponents. Solid 2-seed if playing in a P5 conference, but the bottom of their conference drags their whole conference down. 8 teams, 8 seeds. BC will make a lot of noise this season, and make a case for a high seed
Solid logic. Hard to argue. LIke I mentioned above, not all new players or transfers are created equal. What team has a new player or transfer that will be that team's Yujie Zhao, or Macario, or Ashley?.... That team will take a leap forward. Yujie and Jaelin Howell did that for us last year.
I think that I am agreeing with you here. An earlier post had an incoming freshman with U17 WNT experience starting for UVA, and you countered that assertion, and I believe I am supporting your argument by noting that playing for a U17 WNT that performed very poorly was not an endorsement for starting for a top 10 college team like UVA.
I think a more interesting question.... In the last 10 years nine programs have made the final four for the first time: Boston College Ohio State Wake Forest Virginia Tech Texas A&M Rutgers South Carolina West Virginia Georgetown So if we're averaging about one new final four participant a year, what program will be making their first appearance in a final four this year?
Interesting to see who the new comers to the final four were in the last 10 years. Ignoring, for the moment, the surprising absence of PAC-12 teams as newcomers, the list of 9 more or less seems to speak to the strength of the conferences. ACC - 3 newcomers Big 10 - 2 newcomers SEC - 2 newcomers Big 12 - 1 newcomer Big East - 1 newcomer So to answer the question of what school might make the last 4 standing in 2019, the PAC 12 would have to have several of its second tier teams get better seeding for a chance to end up in the final four. If two or more of Colorado, Arizona, Washington State, AZ State, or Cal step up a notch and are not in the same bracket as Stanford, USC, of UCLA then they could possibly join the College Cup.
I have a complex statistical model. . The model predicts that unc goes to the college cup every year. My model has a decent record over the past roughly 38 years.