Thursday, December 20, 2012

NCAA Women's Final Four Wrap-Up: Powerful Texas Overcomes Grind-it-Out Michigan and Oregon

Looking back on this past weekend's NCAA women's Final Four, what stands out to me is the contrast in teams' strong suits. The champion Texas Longhorns set the ball high, hit over the block, and delivered surgical strikes to win points quickly. Michigan and Oregon, UT's opponents in the semifinals and finals, respectively, got as far as they did with more of a grind-it-out approach. Defensively, Michigan and Oregon persistently dug opponents' spike attempts, and when the Wolverines and Ducks went on offense, they sometimes would need multiple hitting attempts on a single rally to finally put the ball away.

The following graph conveys the Longhorns' devastating offensive prowess in their final three matches of the tournament (vs. USC in the Elite Eight, and the Michigan and Oregon matches). In addition to the conventional hitting-percentage statistic ([Kills - Errors]/Total Attempts), I have presented a slight variation, namely kill percentage (Kills/Total Attempts). You may click on the graphics to enlarge them.


Against USC, the Longhorns achieve a kill on half the swings they took (.505, 50/99), whereas the other three Final Four teams all had kill percentages in the high .30s in their Elite Eight matches. Then, in the final, a 25-11, 26-24, 25-19 Longhorn win over the Ducks (box score), UT nearly reached .500 in kill percentage again, clocking in at .483 (43/89).

If a team commits no hitting errors, then kill percentage and hitting percentage are equal to each other. That the black and orange bars for Texas against Oregon are nearly equally tall tells us that the Longhorns made very few attack errors in that match -- 4 to be exact, compared to 43 kills on 89 swings. As a further sign of how unstoppable Texas was offensively, Oregon came up with only 1 team block in the final, compared to the Longhorns' 15.

In the semifinal round, Michigan extended Texas to five games, with the Longhorns prevailing 25-11, 21-25, 23-25, 25-12, 15-11. For the match as a whole,Texas's hitting percentage of .316 was literally double Michigan's .158. However, when game-specific hitting percentages are examined, we can see that in the two sets the Wolverines won, they somehow managed to slow down the Horns' offense, while raising their own hitting percentages into the mid-high .200s.


Above, I characterized Michigan as having a "grind-it-out" style, for which being able to dig opposing spikes to keep setting up the offense is essential. As one possible measure of a team's grind-it-out propensity, I've simply taken that team's total number of attack attempts (TA) in a given match and divided it by the total number of points in the same match (to control for match length). Looking at all matches in the Sweet 16 and beyond (a total of 15 matches and thus 30 team-specific values), Michigan took first and second in grind-it-out propensity:

Mich (vs. Tex): 193 total points, Mich 196 TA (1.02)
Mich (vs. Stan): 180 total points, Mich 173 TA (.961)
IowaSt (vs Stan): 127 total points, ISU 121 TA (.953)
Stan (vs. IowaSt): 127 total points, Stan 118 TA (.929)
Ore (vs. PSU): 195 total points, Ore 176 TA (.903)
Min (vs. Pur): 192 total points, Minn 173 TA (.901)
Pur (vs. Min): 192 total points, Pur 173 TA (.901)
Neb (vs. Ore): 172 total points, Neb 155 (.901)
Ore (vs. BYU): 178 total points, Ore 159 TA (.893)
Ore (vs. Neb): 172 total points, Ore 151 TA (.878)
Ore (vs. Tex): 130 total points, Ore 114 TA (.877)

Oregon had four of the top 11 grind-it-out values. Grind-it-out teams will not always out-dig their opponents, as a lot of their own spikes are being dug. Indeed, Texas out-dug Michigan 76-71. However, Oregon did out-dig Penn State 80-67 in their national semifinal match (box score).

With the exception of the Michigan match, Texas didn't take that many swings in its final three matches, yielding relatively low grind-it-out ratios for the Longhorns: (vs. USC, 99 swings/130 points =.762; vs. Michigan 158/193 = .819; and vs. Oregon, 89/130 = .685).

***

My Conference-Adjusted Combined Offensive-Defensive (CACOD) ranking, which is based almost entirely on teams' regular-season hitting percentages and the hitting percentages they allowed their opponents, correctly predicted 47 of the 63 tournament matches this year. Thus, the CACOD placed right up there with other, more established ranking systems (screen capture from here):



For each of the 63 tournament matches (shown below), the team with the higher CACOD ranking (shown in parentheses) would be predicted to win. Matches in which the team with the higher CACOD lost are shown in red.

1st Round (26-6)

PSU (1) d Bing (63) 3-0
BGSU (51) d Yale (47) 3-2
Ohio St (28) d ND (34) 3-0
Ky (33) d ETS (59) 3-0
FSU (6) d Hof (61) 3-0
Pur (23) d ColSt (20) 3-0
Crei (14) d Marq (27) 3-0
Minn (13) d Lib (60) 3-0

Ore (9) d NCol (55) 3-0
Day (17) d Pepp (48) 3-2
Ok (37) d ASU (39) 3-2
BYU (4) d NMSU (38) 3-0
Wash (8) d CArk (49) 3-0
Haw (18) d SClara (54) 3-0
UNI (31) d KSt (21) 3-0
Neb (5) d MdES (53) 3-0

Tex (2) d Colg (64) 3-0
TAMU (29) d NCSU (41) 3-1
CofC (57) d Mia (22) 3-2
Fla (7) d Tulsa (43) 3-0
KU (10) d CleveSt (42) 3-1
WichSt (46) d Ark (36) 3-2
StM (52) d SDSU (35) 3-2
USC (12) d Fair (58) 3-0

UCLA (11) d LIU (45) 3-0
MichSt (15) d USD (40) 3-2
Mich (26) d Tenn (32) 3-2
Lou (16) d Bel (62) 3-1
IaSt (24) d IPFW (56) 3-2
NCar (25) d Cal (30) 3-1
WKy (19) d LMU (44) 3-0
Stan (3) d JxSt (50) 3-0

2nd Round (11-5)

PSU (1) d BGSU (51) 3-0
Ky (33) d OhioSt (28) 3-1
Pur (23) d FSU (6) 3-2
Minn (13) d Crei (14) 3-1

Ore (9) d Day (17) 3-0
BYU (4) d Ok (37) 3-0
Wash (8) d Haw (18) 3-2
Neb (5) d UNI (31) 3-0

Tex (2) d TAMU (29) 3-1
Fla (7) d CofC (57) 3-0
WichSt (46) d KU (10) 3-1
USC (12) d StM (52) 3-0

Mich St (15) d UCLA (11) 3-1
Mich (26) d Lou (16) 3-1
IaSt (24) d NCar (25) 3-2
Stan (3) d WKy (19) 3-0

3rd Round (6-2)

PSU (1) d Ky (33) 3-0
Minn (13) d Pur (23) 3-1

Ore (9) d BYU (4) 3-1
Neb (5) d Wash (8) 3-0

Tex (2) d Fla (7) 3-0
USC (12) d Wich St (46) 3-0

Mich (26) d MSU (15) 3-0
Stan (3) d Iowa St (24) 3-0

4th Round (2-2)

PSU (1) d Minn (13) 3-1
Ore (9) d Neb (5) 3-1
Tex (2) d USC (12) 3-0
Mich (26) d Stan (3) 3-1

Final Four (1-1)

Tex (2) d Mich (26) 3-2
Ore (9) d PSU (1) 3-1

Championship (1-0)

Tex (2) d Ore (9) 3-0

The CACOD also did well in predicting the 2011 tournament.

Wednesday, December 12, 2012

Preview of the NCAA Women's Final Four

The NCAA women's Final Four begins Thursday night, with Michigan facing Texas, followed by Penn State taking on Oregon. To preview the Final Four, I've decided to look back on each team's victory in the Elite Eight. To begin, for each of the victorious teams, let's look at the result of every spike attempted in its respective Elite Eight match (you may click on the following graphics to enlarge them). Box scores for the matches are available at the following links: Penn State-Minnesota, Oregon-Nebraska, Texas-USC, and Michigan-Stanford.

Let's walk through the first case, to illustrate the format. Penn State attempted 156 spikes vs. Minnesota. Sixty-two of these attempts (39.7%) were successful, resulting in kills. Outcomes depicted in black (or dark blue, later) are good. Twenty-five Nittany Lion attacks resulted in hitting errors, 14 (9%) because they were blocked right back onto the Penn State side of the floor for immediate Gopher points and 11 (7%) because Penn State hit the ball out-of-bound or never cleared the net. These bad outcomes for Penn State are depicted in red and purple, respectively. Minnesota dug up 61 (39.1%) of Penn State's spike attempts, and the remaining 8 (5%) are designated as "Other" (e.g., Minnesota blocking the ball back, but Penn State keeping it in play).


Looking at the four horizontal bars, it is clear that Texas (who played USC in its regional final) did the best of any Final Four team at putting the ball away for kills (50.5% of the time, with no other team reaching 40%). There is no way to know for sure whether the four losing teams in the Elite Eight (Minnesota, Nebraska, USC, and Stanford) were equally good defensively, so comparing the hitting statistics of Penn State, Oregon, Texas, and Michigan must be done with caution. Still, it must be noted that USC features Natalie Hagglund, one of the top diggers in the the country (here and here), so it's not like the Longhorns faced a weak opponent in the backcourt.

Oregon had nearly half (46.4%) of its attack attempts dug up by Nebraska, the highest such percent among the Final Four teams. Penn State, Oregon's Thursday opponent, is probably a little better digging team than Nebraska (see Big 10 rankings in statistical categories), so the Ducks may struggle somewhat to put balls away. Penn State (9%) and Michigan (8.7%) exhibited the greatest proneness to getting their spikes blocked.

That Michigan took 173 swings against Stanford shows what a grind-it-out team the Wolverines are. To control for match length, let's note that there were 180 total points in the match (20-25, 25-20, 25-20, 25-20). The Wolverines thus took nearly one hitting attempt for each point played (173/180 = .96). Because many rallies don't have any spike attempts (e.g., aces, free balls), Michigan would thus have taken multiple swings on many plays. The swings/total points ratios for the other teams are as follows:
  • Oregon, 151 swings/172 total points (15-25, 25-22, 25-18, 25-17), ratio = .88
  • Penn State, 156 swings/181 points (25-19, 19-25, 26-24, 25-18), ratio = .86.
  • Texas, 99 swings/130 points (25-19, 25-22, 25-14), ratio = .76.
As seen, none of the other Final Four teams played as swing-intensive a match in the Elite Eight as Michigan. By comparison, Texas hardly tired out its arms at all.

Next, let's look at how the Final Four teams did defensively, focusing on how all of the opponents' spike attempts turned out.


Penn State was best at minimizing the percent of opponent spike attempts that yielded kills, as only 49 (31%) of Minnesota's swings bore immediate fruit (remember, red is bad, so the bigger the red bands, the worse the defense). Texas allowed 38.3% of USC's swings to go for kills, in part because the Longhorns dug a relatively small percent (42.1) of Trojan attempts.

Whether tomorrow night's matches will follow the scripts suggested by the above analyses -- Michigan trying to keep the ball alive on Texas's big swings off the high outside sets and the Wolverines scrapping for points on offense, and Penn State likewise digging up Oregon's attacks -- remains to be seen.  If nothing else, I hope these analyses provide viewers with something to think about during the matches.

Saturday, December 8, 2012

Favorites March On in Friday's NCAA Women's Sweet 16

A tournament in which favored teams overwhelmingly prevail is known, via an expression from horse racing, as following the "chalk." This year's NCAA women's competition is definitely looking like a chalk tournament, as all but one of the eight remaining teams after last night's Sweet 16 round is a top-eight seed. Today's match-ups thus include the following (see bracket for starting times):

No. 1 Penn State vs. No. 8 Minnesota

This match does not look to be all that competitive, at least on paper, as the Nittany Lions took both matches from the Gophers during Big 10 conference play. The first match, in Minneapolis, was a 25-23, 25-8, 25-20 Penn State rout, as the Lions outhit the Gophers .404-.098. The second match, in State College, was closer, but still decisive for Penn State, 25-21, 25-19, 23-25, 25-21.

If one delves further into the statistics of the second match, however, there are some bright spots for the Gophers. Minnesota actually outhit (.207-.204) and outblocked (16-9) Penn State. The Nittany Lions' advantages seemed to come from two other areas. Penn State dug 72 of Minnesota's 121 non-error spike attempts (59.5%), whereas the Gophers dug only 49 of the Lions' 125 non-error attacks (39.2%). Also, Penn State's balance of service aces (11) and errors (12), though negative, was nevertheless more favorable than Minnesota's (5 and 10).

Tonight's rematch features a great coaching match-up, with Penn State's veteran mentor Russ Rose going against new Minnesota coach Hugh McCutcheon, of Olympic coaching fame. McCutcheon definitely has Minnesota on the right track after a tough stretch in early November. Since losing both matches November 9-10 on a swing through Michigan State and Michigan, the Gophers have won seven in a row. The most notable wins during this streak include a home conference match against Nebraska and last night's tight four-game win vs. Big 10 rival Purdue on the Boilermakers' home court (game-by-game log). As shown in the following chart (on which you can click to enlarge), some hot hitting appears to be fueling the Gophers' resurgence.


Yesterday's Minneapolis Star Tribune had a feature article on  middle blocker Dana Knudsen and setter Alexandra Palmer, who both transferred from Santa Clara to join the Gophers this season. Knudsen's hitting, as shown above, has consistently been around .400 of late, with the obvious exception of last night's Purdue match. Fellow middle Tori Dixon has also been pummeling the ball.

The Penn State-Minnesota contest is the only one of today's four matches that can really be said to be taking place on a neutral court, as Purdue is hosting this regional.

No. 2 Stanford vs. unseeded Michigan

Tonight represents the third NCAA-tournament meeting between the Cardinal and Wolverines in four years, with Michigan winning on Stanford's home court in 2011 and 2009. The location for tonight's match is slightly different, Berkeley instead of Palo Alto. The way Stanford is currently playing, Michigan will obviously have a tall order this time.

The Cardinal got a big hitting performance from frosh outside hitter Jordan Burgess (16 kills and only 3 errors on 29 attempts, for a .448 percentage) in sweeping Iowa State last night. Fellow frosh outside Brittany Howard didn't do too badly either (8-0-27, .296). The strong hitting of Burgess and Howard will likely prevent opponents from focusing their block too heavily on Stanford's middle hitters, Carly Wopat (.250 last night) and Inky Ajanaku (yet another frosh; .312 last night). Hitting diagrams I compiled on Stanford earlier this season are available here and here.

Up until last night's win over conference rival Michigan State in three, Michigan had recently been riding the improved hitting of OH Molly Toon. Last night, however, Toon was off-key with a -.048 hitting percentage. MB Jennifer Cross, who has arguably been the Wolverines' steadiest hitter this season, came up big last night, ripping the Spartans for a .500 percentage (12-1-22). Earlier this week, Michigan coach Mark Rosen appeared on Internet radio's "The Net Live" (see link in right-hand column for archives of the show). He characterized Lexi Erwin as a workhorse type of hitter, who is not a kill machine, but produces well over the long haul. Erwin fit exactly that profile last night, hitting .234 on a match-leading 47 attempts.

No. 3 Texas vs. No. 6 USC

The conventional wisdom on this match between the host Longhorns and visiting Trojans, as discussed on The Net Live and elsewhere, is two-fold:
  • Can USC's blockers slow down UT's outside hitting duo of Bailey Webster and Haley Eckerman?
  • Can the Trojans' huge "A-O" middles, Alexis Olgard (6-5) and Alicia Ogoms (6-4), hit well enough to keep the Longhorn block from keying exclusively on 'SC frosh OH Samantha Bricio?
Bricio has led the Trojans in hitting attempts this season with 1,294, outdistancing the next closest player, right-side hitter Katie Fuller, by more than 300 swings. Bricio's season-long .244 hitting percentage may not look spectacular, but is pretty good considering how often she hits. My hitting diagrams on USC are available here and here. The digging of junior libero Natalie Hagglund has also been a major part of the Trojans' success this season.

In addition to Eckerman, recently named Big 12 Player of the Year, and Webster, another Texas stalwart is MB Khat Bell, who suffered a season-ending knee injury last year. Longhorn coach Jerritt Elliott has been resting Bell periodically this season, presumably to avoid re-aggravating the knee, so she should be in good shape for tonight.

No. 4 Nebraska vs. No. 5 Oregon

Finally, we have Oregon taking on Nebraska in Omaha. Needless to say, the Cornhuskers appear to have recovered from a mid-season slump, last night sweeping a Washington squad that was ranked No. 2 in the nation earlier this season and owned wins over USC, UCLA, and Oregon. Scores were 25-14, 25-21, 25-23.

Morgan Broekhuis (9-1-18, .444) and Hannah Werth (9-3-20, .300) led the Huskers' hitting.  Washington had nothing going on offensively, hitting no better than .178 in any of the games. OH Krista Vansant, the Huskies' go-to gal, registered a .242 hitting night (11-3-33). Washington's 9 service errors (to only a pair of aces) didn't help, either. Blocking was a strong suit for the Huskies all season and, indeed, UW outdid Nebraska in this area, 11-4.

Oregon stopped BYU in what again might be called a close, but decisive match, 25-23, 25-21, 22-25, 25-12. According to the linked game article, the Ducks' Liz Brenner "hit .417 - her 11th +.400 total of the season - and added 13 digs for her seventh double-double." Ariana Williams (13-3-23, .435) and Alaina Bergsma (17-4-43, .302) also paced the Quack Attack, which generated a .333 team hitting percentage.

BYU had come out fourth (behind only Penn State, Texas, and Stanford) on my Conference-Adjusted Combined Offensive-Defensive (CACOD) ranking at the end of the season, so I thought an upset over Oregon might be possible. However, the Ducks were too strong.

Sunday, December 2, 2012

NCAA Women's Tourney -- the First Weekend

The Big 10 (which actually has 12 schools) and Pac 12 conferences each had seven teams in this year's NCAA women's tournament. At this stage, with two rounds complete after the first weekend, the Big 10 is looking a bit stronger than the Pac 12. The former has lost only one team (Ohio State), whereas the latter has lost three (No. 7-seed and defending national champion UCLA, Cal, and Arizona State).

Many of the top teams from the two conferences  -- No. 1-seed Penn State and No. 4 Nebraska of the Big 10; and No. 2 Stanford, No. 5 Oregon, and No. 6 USC of the Pac 12 -- made it through to next weekend's Sweet 16 without having lost as much as a single game (set). Other teams faced tougher competition and thus struggled to varying degrees to advance. From the Big 10:
  • No. 8 Minnesota dropped the first set of its second-round match against Creighton, before taking the next three (20-25, 25-17, 25-23, 25-17). The Gophers hit .314 as a team against the Blue Jays, with five Minnesota players hitting .250 or higher on at least 18 attempts each (Dana Knudsen, .500; Tori Dixon, .360; Ashley Wittman, .286; Daly Santana, .280; and Katherine Harms, .258).
  • Next up for Minnesota will be Purdue, which upset No. 9 Florida State in five on the Seminoles' home court. The Boilermakers, who hit .266 as a team vs. FSU, featured three players who exceeded .400 on at least 15 swings each (Rachel Davis, .467; Anna Drewry, .421; and Kierra Jones, .421). Purdue also heavily outblocked FSU, 17-7 in total team blocks. The Minnesota-Purdue match will be on the Boilermakers' home court, a predetermined regional site. Minnesota beat Purdue 3-1 in Minneapolis in the teams' only regular-season meeting.
  • Michigan State is the team who beat UCLA, on the Bruins' home court, no less. UCLA's Tabi Love hit .344 against the Spartans (14 kills and 3 errors on 32 attempts), which according to a previous analysis, should have put the Bruins in a strong position to win. However, Love took only 22.5% of UCLA's total swings (32/142). Looking back at roughly the past month, Love regularly took at least 30-35% of the Bruins' hitting attempts, and sometimes as many as 40-45% of them (in UCLA's October 28 win over Washington, Love took 71 of the Bruins' 157 attempts, which is 45%). With three Spartans reaching attack percentages around .300 (Lauren Wicinski, .294 with 51 attempts; Alexis Mathews, .312; and Taylor Galloway, .310), MSU outhit UCLA, .293-.261. (Wicinski is a transfer from Northern Illinois, whose hitting statistics I analyzed last year.)
  • In another all-Big 10 Sweet 16 match-up, Michigan State will face Michigan in Berkeley, California. The Wolverines escaped with a five-set win over Tennessee in the first round after leading 2-0 in games, then upset No. 10 Louisville in four, on the Cardinals' home court. The hitting star for Michigan has been outside hitter Molly Toon. You might say that most of the regular season was a "tune-up" for Toon, as she (and some of her teammates) didn't start hitting well until fairly late in the conference schedule. Toon's hitting percentages in her last four matches have been .519 at Michigan State and .542 at Ohio State to close Big 10 play, and then .370 and .306 against Tennessee and Louisville, respectively. Michigan and MSU split their two conference matches, with each team winning on the road (box score for match at Ann Arbor).
Besides Stanford, Oregon, and USC, the additional remaining Pac 12 team is No. 13 Washington. The Huskies received a tough second-round challenge from Hawai'i, eking out a 27-25 Game-4 win to stay alive (after the Rainbow Wahine had held match point at 25-24) and then prevailing 15-11 in the fifth. Washington faces a tall task in its next match, playing No. 4 Nebraska in Omaha, the Cornhuskers' home away from their Lincoln home. Against Hawai'i, star Husky hitter Krista Vansant (.297) received support from teammates Cassie Strickland (.312), Kaleigh Nelson (.310), and Kylin Munoz (.289), with all four of these players taking at least 29 swings each.

The only top-10 seed not discussed thus far is, of course, No. 3 Texas from the Big 12. In the second round, the Longhorns faced former conference rival Texas A&M (now of the SEC) and won in four. In vanquishing the Aggies, UT hit a spectacular .389 as a team, led by Bailey Webster (.556), Khat Bell (.421), and Haley Eckerman (.342), with each taking at least 19 swings.

An updated bracket is available here.