2017 Colorado Buffaloes

Discussion in 'Women's College' started by unkiemark, Dec 17, 2016.

  1. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
    Wow - Buffs have taking it out on the Pioneers. Now 6-1.
     
  2. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    Not tonight. Two goals 14 seconds apart. Katie Joella got the first of her career, I think more to come. Substitutes played abundantly.
     
  3. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    Elsewhere on Friday ...

    UCLA 1, Pepperdine 1

    USC 2, Loyola Marymount 1 both USC goals came in the final five minutes

    Arizona 2, BYU 1, OT BYU must have done something to anger the soccer gods.
     
  4. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    BigSoccer won't let me add to the previous post, so ...

    UCLA 1, Pepperdine 1, at UCLA. Short bus ride for the Waves, the campus is in Malibu.
    All UCLA in the first half (shots 7-0), much less so the rest of the game.

    USC 2, at Loyola Marymount 1. USC overcame the effects of the short bus ride, both USC goals came in the final five minutes. Shots were 25-7, first half 10-0, but LMU scored on a free kick from midfield that bounced in.

    Arizona 2, BYU 1, OT BYU must have done something to anger the soccer gods. They are 1-4-3, but as we saw, they are a very good team. Although I wish it otherwise (go Broncos !!), I'll be surprised if they don't win the WCC. They end their season at Santa Clara (go ... oh, wait, I already said that).
     
  5. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    Another elsewhere, today ...

    California 3, #14 Oklahoma State 0.

    Shots on goal for the Cowgirls: 0


    Did anybody attend the Denver game?
     
  6. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    The first 2017 RPI is out:

    upload_2017-9-18_19-48-19.png

    Comments to follow. USC #64?
     
  7. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    #257 cpthomas, Sep 18, 2017
    Last edited: Sep 18, 2017
    derbarkasmann, see my comments introductory to my RPI blog post here. They'll give you a perspective on what the current actual RPI rankings mean and don't mean. Simply put, don't take them too seriously, they aren't intended to be correct at this stage of the season. They may signal something -- such as that a team's strength of schedule, at this stage of the season, looks strong or weak -- but beyond that, I wouldn't put too much stock in them.
     
  8. Gilmoy

    Gilmoy Member+

    Jun 14, 2005
    Pullman, Washington
    Nat'l Team:
    United States
    See the raw data at All White Kit. Sort by "Unadjusted RPI" for the same as above, or by "Conference". Click on any school's name to pop open a very nice summary of all of results and opponents' records. (Note: there are some discrepancies between this and NCAA D1 RPI for 09/18, e.g. Southern California is #73 at AWK, and Monmouth is +5=2-2 at NCAA but +4=2-2 at AWK. I think NCAA has some bad data bugs.)

    We already saw inklings of USC's weirdly low rating in cpthomas's 2017 Simulation thread. Basically, USC's opponents are doing poorly. It's partly the Santa Clara effect (they bravely played 5 Pac-12 + Michigan, and won none of them), plus other regional foes just being not good this preseason.

    The formula is URPI = [ Ele1 + (2 * Ele2) + Ele3 ] / 4. There can be enormous gaps in Element 2s, e.g. +/-0.25. Double that (and then divide by 4) and you get a whopping +0.125 to ARPI, which totally swamps anything else. For example, the gap between #1 South Carolina and USC is only +0.191, so almost all of that is just from having a better Element 2, doubled.

    +W=D-L ele1 ele2 ele3 URPI
    +7=1-0 .875 .779 .616 .762 #01 South Carolina
    +7=1-0 .875 .636 .588 .684 #09 Princeton -- mostly from -.14 in Ele2
    +5=2-1 .750 .536 .583 .601 #48 Washington State -- slightly better Ele2 and Ele3
    +6=2-1 .778 .517 .550 .591 #54 Colorado -- your Ele2 beats USC's
    +6=1-0 .857 .402 .623 .571 #73 Southern California

    If USC's opponents had just managed .500 W/L, that would boost USC by +0.049 = 0.620, which slots them between .623 #35 Alabama and .617 #36 Butler. And if their opponents were at .600, then USC would get another +0.049 = 0.669, just above .668 #13 North Carolina State. (+0.05 might seem like small points, but it's a huuuge jump in ranks because so many teams are tightly packed into a range that's only 0.500 wide from best to worst.)

    All teams are kind of bunched up still, so large swings in Element 2s cause gigantic swings in early RPI. cpthomas's 2017 projections compound this effect by estimating future results using these ARPIs, so USC is projected to lose even more, because they're rated lower :) Their projected end-of-season result has actually improved, though, from #153 and 10th using last week's data, to #107 and 9th in this week's.

    All data-based systems are slaves to the data. USC might secretly be #25 in actual playing strength, but we don't measure margin of victory, and so we can't deduce how much better a team is, just from a win. FIFA woso, and FIDE Elo, can't do it either. Life is hard :D
     
  9. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    Thoughts on RPIs:

    That puts the loss at Texas in a bit better perspective. We don't have a win over a team rated above us because we haven't played anybody else above us. The tie with CSUN didn't hurt as much as I thought it would, and the wins over BYU and TCU didn't help as much as I thought they would. We have three wins and a tie against teams in the 200's (there are 333 teams, same as last year). Conference play will give us a chance to pad the RPI, or not. We have five home games and six on the road, and three of the five are against rather strong teams.

    USC #64? They have a loss at #35 Kansas. Does home or away make any difference in RPI calculations? There's a home win over #236 UC Davis, no help there. Then a win at #136 Missouri, not much help there. They rebounded from the loss at Kansas with home wins over #43 Santa Clara (go Broncos !!) and #178 Iowa State. OK, that's three wins against teams not highly rated, but we have four and one tie. Then came a win over #23 Loyola Marymount, and at #220 San Diego. That's four not powerful teams, I sort of see the problem, their pre-conference schedule wasn't very brutal. But they have two wins and one road loss against teams rated quite a bit above them.

    I was a math whiz in high school, then majored in English at Santa Clara (I already said it), a decision I regret (the major, not the university). I'm sure there is a logical mathematical reason why USC is #64, but I have no clue what it is.

    UCLA at #19 surprised me also. The tie with Pepperdine was unexpected, but the Waves have been pretty good that last few years and their RPI is 29. The win at #295 Weber State didn't help, and San Diego State (go Aztecs !!) isn't as good as they usually are. But the Bruins have a win at #18 Virginia and over #43 Santa Clara (Jon, don't repeat yourself). I guess the moral of the story is, don't schedule Weber State.

    I'm guessing that Stanford (RPI 2) will be ranked #1. They got by Santa Clara (it would be VERY repetitive) 2-1 Sunday night. It will be interesting to see if UCLA or South Carolina (RPI 1) will be second, but that USC has a 3-2 loss at #15 Wake Forest.
     
  10. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    The strength of schedule does seem important. BYU is 1-4-3, but their RPI is not catastrophically low because they have played good or very good teams (except Oregon State, their one win).

    They did defeat Washington State 2-1 and tied Michigan, but lost to all four California Pac-12 teams, three of them by 2-1. Their worst loss this year was 4-1 at Notre Dame, it really got away from them in the last few minutes. They are usually competitive with Pac-12 teams, and did knock Stanford out of the playoffs last year. Their first four games last year were Pac-12, and Stanford came a little later, they won 2 at home and lost 3 on the road, all five games were one goal apart. It's good preparation for conference play, as the WCC usually has a few pretty good teams.

    Gilmoy, when I was in high school I might have understood your last post. 55 years later, I trust that you and cpthomas have a good handle on the RPI and I am just enjoying the ride.
     
  11. Gilmoy

    Gilmoy Member+

    Jun 14, 2005
    Pullman, Washington
    Nat'l Team:
    United States
    URPI is insensitive to many things. It doesn't care about home/road/neutral, nor margin of victory, nor player availability due to injuries/int'l duty/team flu. The only data it ingests is W/L records. (ARPI adds small bonuses/penalties for top wins and bottom losses.)

    I think URPI doesn't even care about "good" wins (upsets) or "bad" losses (being upset). Obviously, for any 1 team considered in isolation, an upset is strictly better than the alternatives. But if two teams play similar schedules, and one upsets Stanford but loses to Cal, and the other does the opposite, URPI essentially doesn't distinguish them! (Somebody double-check that -- my intuition still doesn't fully accept it.)

    The real benefit of playing high-RPI ranked teams is that their ranks already imply (encode) their W/L records -- and so just playing against a high-rank team means your Element 2 gets a prorated share of a fraction that's usually larger than it already was, so it gets dragged upward. The converse is true, too: low-rank teams will hurt your Element 2, even if you take the expected win to keep your Element 1 afloat.
     
  12. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    Gilmoy, the NCAA has a muck up on Monmouth because they're counting a win against Principia, which they shouldn't be doing since it isn't a D1 school.

    For Southern Cal, AWK's adjusted RPI rank for Southern California is 69, which is closer to what the NCAA has but still a number of positions off.

    Hopefully, the NCAA tomorrow will publish the rating details in the RPI Archive. Once they do that, I match my ratings with theirs and identify any discrepancies, which are due either to my muck ups or theirs. They're very good at correcting muck ups if pointed out to them -- and so am I.

    I checked my ratings against AWK's this morning, and they match exactly (after I corrected a data error I had).
     
  13. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    You're absolutely right.

    One way to think about this is: For RPI rating purposes, set aside all your pre-existing evaluations of teams. The NCAA says that for its purposes, they don't matter. For rating purposes, all the RPI cares about is your winning percentage, the average of your opponents' winning percentages, and the average of your opponents' opponents' winning percentages. It doesn't care about which of your opponents you beat or which you lost to (except for the very minor bonuses and penalties for good and poor results), it doesn't even care which opponents you played (except to the extent its reflected in their winning percentages). And, it doesn't care about game locations (this is good), except for the very minor differences in the bonuses and penalties related to game locations. The RPI is completely "blind" to anything other than the numbers.

    The Women's Soccer Committee, however, is not blind in that way. The RPI is one of their considerations, and I believe they use it primarily to identify groups of teams they will consider for seeds and for at large selections. Once they get to having to make actual decisions, however, they consider, among other things, (1) head to head results between teams, (2) results of teams against common opponents, (3) how teams did in their conferences, both regular season and tournaments, (4) the strength of teams' conferences, (5) how teams did against highly ranked teams, and (6) how teams did over their last eight games (which I take to be looking at poor results). When I simulate tournament brackets, my system considers all of these factors and various combinations of them, in relation to the Committee's decision-making patterns as to the factors over the last 10 years. Gilmoy is exactly right, however, that if the current actual RPI ratings are skewed, my simulation uses those skewed ratings and assumes they are the right ratings for purposes of simulating future game results. Week by week, as the season progresses, as teams' results change, any improper skewing gets reduced. But still, given that the teams play only about 19 games each and that it takes about 28 to 30 games to produce a highly integrated system, there always are going to be some issues from the perspective of someone who thinks he or she knows which teams really are the strongest.

    One other thought: the RPI is concerned only with how a team has performed. It has nothing to do with predicting how a team will perform in the future ... retrodictive rather than predictive.
     
  14. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I stand partially corrected: After a thorough vetting of my rankings against the NCAA's, although the team records indicated with their current rankings include results against schools that are not Division 1, their RPI calculations do not include those games.

    There are two reasons why the AWK rankings and mine differ from the NCAA's:

    1. The AWK rankings (and mine currently) include the games played yesterday, September 18, whereas the NCAA's do not include those games.

    2. Two data issues within the NCAA's system:​

    a. The NCAA's system is treating the result of the 9/14 game @ Stony Brook v Hofstra as a tie, whereas actually Hofstra won the game 1-0. This appears to have happened because the game is double entered into the NCAA's system and the way it is double entered is causing the NCAA's system to read it as a tie even though the teams entered it as a Hofstra win.

    b. The 9/15 game @ Incarnate Word 0 v Texas RGV 2 is entered into the NCAA's system for that date and the identical game also is entered for 9/14, so the NCAA's system is counting that game twice.

    My standard practice after my weekly vetting of my numbers against theirs is to let them know if they have data errors. They then check them out and, if they agree there are errors (which I'm sure they will in these cases), they get the errors corrected.​
    When I back the 9/18 games out of my system and introduce these two NCAA data issues into my system, my rankings (and thus AWK's too, if they were to do the same thing) exactly match the NCAA's.

    Put differently, the AWK rankings (and mine) are exactly correct. The NCAA's also are correct for games through 9/17 except for the issues caused by those two data issues.
     
  15. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
  16. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
  17. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    The rankings that came out on Tuesday. I didn't think UCLA would still be #1, but

    1. UCLA
    2. Stanford
    10. Southern California
    13. California
    17. Texas (included because we played them)
    25. Utah
    33. Colorado
    40. Washington
     
  18. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
  19. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
  20. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    UCLA 2, Buffs 0

    And it wasn't really that close. A couple of situations didn't go our way that maybe could have, but for the result to be different would have been a massive injustice to the Bruins. Shots on goal at halftime were 1-1, but the run of play had been maybe 85% UCLA, and the second half was way more one-sided. Shots wound up 18-3, on goal 9-1. We got four yellow certificates of improper play, two to Alex Vidger who will miss the Washington State game.

    That said, UCLA cheated. The game is supposed to be played by humans. They brought precision-engineered machines, because those are the only things that could pass the ball the way they did. They were very realistic in appearance, but had to have been machines.

    Starters were

    upload_2017-9-22_20-26-46.jpeg


    We got the first shot of the game, sure didn't think we'd only get two more. It took UCLA 19 minutes to get their only shot on goal of the half, a harmless one right at JJ. But the Buffs were defending a LOT. UCLA disrupted everything we were doing, and when we did get the ball, they would instantly drop back to defend. In the 33rd minute they got way too many people into our box, but an open shot went over the goal. A player was massively alone in the left side of our box with the ball, this appeared to be doom, and Hannah Cardenas rocketed in from nowhere and stripped the ball, but it was off-sides anyway. Late in the half we strung a few short passes together and Camilla Shymka found herself with an open shot that went straight to the goalkeeper, our only shot on goal of the day. About a minute later, Camilla got back into their box with the ball but a defender took it from her.

    In the 50th minute, a Bruin fired the ball into Joss at short range, and it hit her arm. I've seen that called, and not called. It was called. I couldn't tell if she was just inside the box or just out. Referee said in. PK, a crappy way to give up a goal, especially when the defenders had played as heroically as they had. JJ to the rescue, she dove to her right and blocked the shot. A Bruin was tripped and fell hard, grabbing her ankle with one hand and pounding the ground in agony with the other hand. As soon as the yellow card was issued to Cassie Phillips, she sprang up and trotted away with no trace of discomfort. A few minutes later, Tatum Barton got into the box on a breakaway sort of except there was a defender with her stride for stride and the defender disrupted the run before Tatum could shoot. Two minutes after that, a Bruin beat two defenders at the left side of the box and had a choice of teammates to pass to, she chose correctly, JJ had no chance, 1-0.

    Then came the two corner kicks that I would like to see a replay of. The word "mugging" does not come close to describing what befell Taylor Kornieck on the first one, no call of course, I wonder if there would be a call if someone knifed her in the back on a corner kick. But a Bruin had knocked the ball over the end line, and on the ensuing corner kick a very large number of people saw the handball, which was not called. In the middle of the fourth quarter, JJ kept us in the game with a spectacular save, but a few minutes later a Bruin got by her defender and found an open teammate about at the penalty spot, 2-0. Then Taylor hit the crossbar. Alex Vidger committed a frustration foul which was an obvious grab of a player running past her, this was her second yellow card so she will miss next Thursday at Washington State. Oops, I already said that.

    Shots in the second half were 14-1, on goal 8-0. We committed no fouls in the first half, eight in the second. There was a good crowd, 1,261 on a not very warm or sunny afternoon.

    I don't think the Buffs could have given any more than they did. UCLA was just a lot better than us.
     
    cpthomas repped this.
  21. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    This week;s rankings:

    1. UCLA sure looked like it against us
    2. Stanford
    7. Southern California
    8. Texas 10-0-0
    11. California
    24. Utah
    37. Colorado still receiving four votes

    And the RPI:

    upload_2017-9-26_15-57-26.png
     
  22. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
  23. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
    BuffsPios repped this.
  24. derbarkasmann

    derbarkasmann Member

    1.FC Koeln (Cologne, Germany)
    United States
    Oct 27, 2008
    Grand Junction, Colorado
    Club:
    FC Köln
    Nat'l Team:
    United States
    Buffs 2, at Washington State 0

    Huge. I was very worried about this one. WSU played Stanford tough last week, losing 2-1, and they had owned us for several years until we stole one from them on their field last year and I expected a definite revenge factor. Shots were 18-13 Buffs, on goal even at 5, but corner kicks were 9-2 Cougars. They outplayed us most of the first half but Marty Puketapu ran down a long rebound and shot past the defender on the end line. Becca Rasmussen set it up after taking on an abundance of Cougars and launching a short-range shot that the GK blocked but didn't corral.

    The Buffs came out very strong for the second half and just a few minutes in, Becca again dispatched three defenders and her lob over everybody including the GK was just off goal. A Cougar would have cleared it or tapped it over the end line, except any possible space for her to do anything was occupied by Sarah Kinzner and the ball bounced off the Cougar and into the net for an own goal.

    JJ Tompkins contributed a couple of highlight video saves to preserve the shutout.
     
  25. unkiemark

    unkiemark Member+

    Dec 23, 2003
    Boulder
    Becca Rasmussen was brilliant tonight and the difference. Her one v one play set up both goals. Tonight she played up top rather than in the midfield - good move by Sanchez.

    This was an important game for the Buffs coming off the loss to UCLA and they have six away matches and five at home.

    They need to get another one on Sunday in Seattle.
     

Share This Page