Saturday, December 26, 2015

Huskers Sweep Horns to Win NCAA Women's Title

Playing in Omaha, not far from the school's Lincoln campus, the University of Nebraska swept the University of Texas, 25-23, 25-23, 25-21, to win the Cornhusker program's fourth NCAA title (the third under current coach John Cook) on Saturday, December 19. Here's a link to the box score and play-by-play sheet.

As shown in the following graph, Nebraska outhit Texas by a substantial margin in each game. Yet the games were tight (especially the first two). The Huskers were called for three "assist errors" (perhaps the setter backing into the net or contacting the ball above the net when starting out from the back row) and one ball-handling error for the match, whereas the Longhorns were called for none of either, thus giving UT a four-point advantage in that department and negating some of NU's hitting edge.


Pacing the Husker hitting attack were outside-hitter Mikaela Foecke (.385 percentage, based on 19 kills and 4 errors, on 39 swings) and middles Amber Rolfzen (.625, 10-0-16) and Cecilia Hall  (.500, 7-2-10). Hall's take on the situation, as reported in this Lincoln Journal-Star article, was that, “Mikaela just pulled their block away, and that made it easy to get kills,”

Nebraska called on its middles slightly less than did Texas, though. Husker middles attempted 26 out of the team's total 119 hitting attempts (21.8%), whereas Longhorn middles took 29 of the team's total 121 swings (24.0%).

The following ESPN graphic (of which I took a screenshot) shows the Longhorns' serve-receipt woes, which forced setter Chloe Collins to frequently deliver her sets from non-advantageous spots on the floor (shown as the dots, within the area within the imaginary blue arc being optimal). As television analyst Karch Kiraly put it, Collins often had to "put on her track shoes." For whatever reason, ESPN couldn't get Nebraska's spatial statistics to show up.


Of UT's two leading middles, Molly McCage (.429, 7-1-14) acquitted herself well, but Chiaka Ogbogu had as many hitting errors as kills (4 each) on 15 attempts, yielding a .000 percentage. One Longhorn who was not fazed by the team's erratic passing was frosh outside-hitter Yaasmeen Bedart-Ghani. In the championship match, she hit .500 (11-1-20), following up on a .583 (15-1-24) performance in the semifinal against Minnesota.

***

My Conference-Adjusted Combined Offensive-Defensive (CACOD) forecasting tool once again did well, even though none of the top-three-ranked teams on the CACOD (Washington, USC, and Penn State) made the Final Four. For each of the tournament's 63 matches, the team with the higher CACOD won 51 times and lost 12.

As shown in this summary of the VolleyTalk discussion board's Pick the Winner contest, someone correctly picked the winners of 54 matches. This is particularly impressive, in light of the following procedural aspect:

The PTW contest results... are based on the teams chosen before any of the games have been played and are not updated to show the actual teams playing in rounds subsequent to the first one. So if a high seed and consensus final four team is eliminated in the first round, anyone who picked that team for the following rounds automatically loses. On the other hand, RichKern... calculate[s] results a different way. RK looks at each succeeding game after the first round on a fresh basis to determine a win or loss based upon the last poll rankings for the two actual teams in each such game... 

Like RichKern, I use the CACOD to make "fresh" picks at the beginning of each new round (e.g., round of 64, round of 32, round of 16, etc.), using the actual teams in a given match.

Thursday, December 3, 2015

NCAA Women's Preview 2015

I didn't write much during the regular season, but I'm ready to go for the NCAA women's tournament. I have once again produced my Conference-Adjusted Combined Offensive-Defensive (CACOD) metric, the methodology of which is explained here. The heart of the CACOD is the ratio of a team's own hitting percentage to the hitting percentage allowed to opponents. The 64 NCAA teams are ranked below on the CACOD (unless you have amazing eyesight, you'll want to click on the images to enlarge them).

The principle is simple: Whenever two teams play, the one with the higher CACOD would be favored to win. The CACOD's success rate in predicting winners of NCAA women's tournament matches has been right up there with other forecasting tools such as Pablo, Massey/Elo, and RK (see here and here). Last year, the CACOD foresaw BYU's run of upsets in the NCAA tourney.




The Washington Huskies (28-2) come out No. 1 on the CACOD, with a clear margin over the next-closest teams. U-Dub also finished the season atop the AVCA coaches' poll, but received only a No. 5 seed from the NCAA. The Huskies' only losses were at USC (3-1) and at Stanford (3-2).

The No. 1 tournament seed is USC, which got off to a surprisingly torrid start this season, before losing to Washington and UCLA in the second half of conference play. A key to the Trojans' success this season was their mettle in road five-game matches. Four times (at Stanford, at Colorado, at Washington State, and at Arizona State), 'SC prevailed in five. However, in the final days of the season, UCLA took a five-setter from USC on the Trojans' home floor.

Looking over which teams have had the highest CACOD each year (since its introduction in 2011), I find that Washington has the highest score for a yearly leader of any school not named Penn State:

Penn State (2014) -- 3.09
Penn State (2013) -- 2.91
Penn State (2012) -- 2.85
Washington (2015) -- 2.67
Nebraska (2011) -- 2.29

That's all for now. I'll continue to comment on tournament results and the CACOD as "December Madness" progresses.

Sunday, October 4, 2015

USC Continues Undefeated; Penn State in Uncharacteristic Collapse vs. Nebraska

USC, having displayed an amazing ability to prevail in tight games and matches thus far, will almost certainly be ranked No. 1 in the next national poll. Now 16-0, the Trojans have staked their claim to the top ranking with wins at Stanford last Sunday (26-24, 19-25, 20-25, 25-17, 15-13) and at home vs. Washington today (25-16, 22-25, 25-21, 33-31).

This wasn't supposed to happen, at least according to the preseason AVCA poll, which had USC at No. 22. In fairness to the voters, the Trojans had were coming off a weak 16-16 record last season and were losing two transfer players, Ebony Nwanebu (to Texas) and Lauryn Gillis (to Wisconsin). It should be noted, however, that Nwanebu and Gillis did not put up great hitting percentages for the Trojans a year ago (.226 and .174, respectively).

Thus far in 2015, USC is playing lights-out both offensively and defensively, in terms of hitting percentage (you may click on the following graph to enlarge it).


The Trojan offense is led by a trio of upper-year players. As shown in the following table, senior outside-hitter Samantha Bricio is taking a huge number of attack attempts and hitting at an excellent .323 clip, with middle-blockers Alicia Ogoms and Elise Ruddins hitting at .419 and .412, respectively. (OH is considered a more difficult position than MB for attaining a high hitting percentage, as an OH tends to get a higher number of "desperation" sets when a team is out of system.)

Player (Position)
Year
Hitting Pct.
Spike Att.
Alicia Ogoms (MB)
Sr.
.419
198
Elise Ruddins (MB)
Jr.
.412
148
Samantha Bricio (OH)
Sr.
.323
609
Alyse Ford (OH)
Fr.
.285
316
B. Abercrombie (OPP)
So.
.247
304
Baylee Johnson (S)
Fr.
.190
147

***

After crushing Nebraska in the first game, 25-12 and eking out a 26-24 win in Game 2, top-ranked and previously undefeated Penn State not only lost the next three, but never really threatened to close out the match. The Cornhuskers' winning margins in these final three games were 25-14, 25-20, and 15-11 (box score).

While Nebraska hit pretty well in the third and fifth games, the big story in my view was the decline in Penn State's hitting. In the final three games, the Nittany Lions hit .059, .083, and .100. I don't know the last time Penn State hit .100 or below in three games of the same match, but I would imagine it was a long time ago!


Nebraska couldn't maintain the momentum from its win at Penn State, however, losing in five the next night at Ohio State. The Huskers are now 12-2.

Tuesday, September 8, 2015

'15 Women's World Cup Ends, College Play Begins

Needing a top-two finish at the FIVB World Cup in Japan to qualify for next year's Olympic Games in Rio de Janiero, the U.S. women took third. China (first) and Serbia (second) advance to Rio. (Brazil is automatically in the Olympics as host country, so per FIVB policy, did not participate in the World Cup.)

The World Cup uses a round-robin format exclusively, with no elimination rounds (i.e., seminfinals, finals). Under this format, the U.S. finished 9-2, with losses to Serbia (in five games) and Russia (in three). Final standings are shown here, with the U.S. match log shown here.

The U.S. still has another opportunity to qualify for the Olympics through competition at the North, Central America and Caribbean Volleyball Confederation (NORCECA) level.

Below, I've graphed how U.S. attackers performed against top competition (China, Serbia, fourth-place Russia, and fifth-place Japan). For each player, only matches in which she took at least 10 swings are shown. (Unless you have amazing vision, you'll want to click on the graphic to enlarge it.)


The three most consistent offensive threats for the U.S. (in terms of hitting percentage, at least) appeared to be Foluke Akinradewo (middle-blocker), Nicole Fawcett (opposite), and to a somewhat lesser extent, Jordan Larson-Burbach (outside-hitter). These are three of the more experienced players on the U.S. roster.

***

As the World Cup was winding down, the U.S. women's collegiate season was getting underway. This past weekend saw a match-up of No. 1 Penn State and No. 2 Stanford (won by the Nittany Lions), and between No. 3 Texas and No. 4 Florida (won by the Gators).

In Austin, three Florida players -- Ziva Recek (9-0-16, .562), Mackenzie Dagostino (5-0-10, .500), and Rhamat Alhassan (13-1-25, .480) -- committed only one hitting error amongst themselves.

In State College, Penn State featured an impressive trio of its own: Aiyana Whitney (13-2-25, .440), Haleigh Washington (12-2-19, .526), and Megan Courtney (11-1-25, .400).

Tuesday, August 11, 2015

"Volleyball Tracking" Site Provides Micro-Level Data

Over on a discussion thread at VolleyTalk, I learned of a service called Volleyball Tracking. One can find online a sample report of a match, providing information on such statistics as serve speeds, total number of meters run by each player during a match, and set-to-spike times (in fractions of a second). In addition, the report includes heat-maps and vector-type serve-trajectory diagrams. A separate Volleyball Tracking webpage shows animation of player movements on points during the match.

Sunday, July 5, 2015

Loyola-Chicago Retains Men's Title in Epic Final


I'm two months late in writing about the NCAA men's final, but an unintended benefit of the delay is that I was able to take the picture to the right, on the Loyola campus during a late-June trip to Chicago. Notice at the bottom of the three-story banner (displayed on a parking structure) that it lists both the Ramblers' 2014 and 2015 titles.

This year's championship was a lot harder to come by than last year's, though, as Loyola captured a marathon fifth game over in-state rival Lewis for the 2015 crown, as opposed to a relatively easy four-game victory over Stanford in 2014.

As shown in the chart below (which I created from the play-by-play sheet), Loyola didn't win until its eighth match point of Game 5 (and the Ramblers also had two match points in Game 4). Lewis had three match points of its own in Game 5.

Held Match PointScoreHeld Match Point
Loyola
14-11
Loyola14-12
Loyola14-13
15-14Lewis
Loyola
16-15
Loyola
17-16
Loyola
18-17

19-18Lewis
Loyola20-19
21-20Lewis
Loyola*22-21
*Won 23-21.

The NCAA final was the fourth Loyola-Lewis meeting of the season, with the Flyers taking both regular-season match-ups by 3-1 scores, and the Ramblers capturing the MIVA conference final, also by 3-1. The following graph shows the two teams' hitting percentages in the four matches.


















The nature of the four matches appeared to vary considerably, with offensive firepower having the upper hand in the MIVA tournament final and the NCAA final featuring a defensive battle.

In both of these matches, though, the Ramblers outhit the Flyers by slight margins. Loyola outblocked Lewis 20.5-13.5 in the NCAA final.

Monday, April 13, 2015

Determining My Vote for Men's OH of the Year

Once again, I have been invited by Off the Block blogger Vinnie Lopes to cast a ballot for men's college players of the year at the various volleyball positions. Because my votes involve a fair amount of statistical analysis, I typically only vote in one category. This year, I have chosen to vote for Outside Hitter of the Year.

My starting point was to look at the top ten outside hitters nationally in hitting percentage (this, and all other, rankings and statistics reported here are from roughly the end of March). These players were: Thomas Jaeschke (Loyola-Chicago, .383); Aaron Russell (Penn State, .379); Tamir Hershko (UC Irvine, .376); Josh Taylor (Pepperdine, .366); Cody Caldwell (Loyola, .325); Nicolas Szerszen (Ohio State, .322); Jonathan Martinez (Pfeiffer, .318); Jon Schaefer (Grand Canyon, .315); Alex Harthaller (IPFW, .305); and Enzo Mackenzie (Sacred Heart, .303).

Spiking may be the primary skill expected of outside hitters, but not the only one. They would also be expected to block, pass (on serve-receipt and in the middle of rallies), and serve. In the far right column of most box scores, one sees the heading "Points," which, for each player, shows the total number of points directly scored via kills, stuff blocks, and service aces. Though it encompasses multiple skills, I have never made much of the "points" statistic, as it does not reflect the number of points a player has cost his or her team through five possible types of errors: in attacking (hitting balls out of bounds or getting stuffed), serving, receiving serve, ball-handling, and blocking (i.e., touching the net).

To determine my vote for Outside Hitter of the Year, I am introducing a variation on the usual points statistic, which represents points earned minus points lost. I am calling my new statistic "Points Profit." (I could have called it "Net Points," reflecting "net" in the accounting sense, but such a term could easily be confused with points won at the net on a volleyball court.)

Using the top ten outside hitters nationally in hitting percentage (listed above), my plan was to compute Profit Points for each OH from his five toughest matches of the season. Match difficulty was based on opponent's rankings and match locations. The top ten teams (RPI) as of when I began the analyses were:  1. Loyola-Chicago; 2. UC Irvine; 3. Lewis; 4. Hawai'i; 5. Ohio State; 6. Penn State; 7. USC; 8. Pepperdine; 9. Indiana-Purdue Fort Wayne; and No. 10 BYU.

As an example, the five toughest matches for Loyola's Thomas Jaeschke were at Lewis, hosting Lewis, at Ohio State, at Penn State, and hosting Penn State. At Lewis, Jaeschke earned 21 points and cost the Ramblers 14, for a "profit" of 7. Each contending OH's five toughest matches and how he performed in them are shown in the following chart, which you may click to enlarge.


Although the candidate pool was drawn from the top ten national leaders among outside hitters in hitting percentage, only eight OH's were analyzed. Martinez was excluded because Pfeiffer played only one match against a top ten opponent (Ohio State). In addition, Grand Canyon's Schaefer missed substantial action in his team's most difficult matches, so he was likewise excluded. Sacred Heart's Mackenzie played in four of his team's five toughest matches, so I prorated his four-match total by multiplying it by (5/4) to estimate what it would have been had he played all five matches.

As shown in the above chart, the top three finishers in total Profit Points -- and for whom I cast my votes -- were Jaeschke (46 Profit Points, 1st); Hershko (43.5, 2nd); and Russell (42.5, 3rd).

Interestingly, even though hitting percentages (just one skill, from all matches up until late March) and Profit Points (multiple skills, in just five matches) were defined very differently, they had a near-perfect correlation (r = .93), where 1.00 is the maximum possible (see plot below). I'll have to conduct further studies to decide whether the extra work of calculating Profit Points is worth it, when hitting percentage is readily available. However, incorporating multiple skills allowed Hershko (third in hitting percentage) to leapfrog Russell (second in hitting percentage).


Click here for correlation plotter.

Saturday, March 7, 2015

Karch Kiraly Visit to Texas Tech: His Coaching Tips and My First Chat with Him in 33 Years!

Let's start with a little "Then and Now." Back in 1982, Karch Kiraly was a senior setter for UCLA, en route to leading the Bruins to three NCAA championships in his four years, and I was the men's volleyball writer for the Daily Bruin. Here's a photographic image of one of my articles (which you can click to enlarge)...


That was then. This is now...



Yesterday, Kiraly came to Texas Tech (where I'm a professor) to spend the day with Coach Don Flora and the Red Raider women's volleyball squad. After a day of private meetings, Kiraly gave a public forum in the evening -- part lecture, part clinic. Scheduled in conjunction with a girls' club tournament, the evening event drew hundreds (if not more) of young players and their parents. As shown in the previous picture, I brought along the old UCLA article and a page printed from my blog to show him. "You're going pretty far back," Karch said, upon seeing the 1982 Daily Bruin article.*

Kiraly's accomplishments after college are well-known: three Olympic gold medals as a player (indoor in 1984 and '88; beach in '96); co-recipient of the FIVB's greatest male player of the 20th century designation; head coach of the first USA national women's volleyball team to win a major international title (2014 world championships); and lead color analyst for ESPN's NCAA tournament broadcasts.



Flora (left) opened the evening, thanking audience members for their support of  Texas Tech volleyball and then introducing his fellow Californian.


Kiraly then alternated several times during the course of an hour between addressing the audience and leading a split Red Raider squad in some drills (equipped with a wireless mic). There was also a Q-and-A session at the end.


Kiraly began the drills with detail-oriented stuff; footwork, where to focus one's eyes, and calling out information for one's teammates (e.g., the opponent's play that appears to be unfolding). At one point, Kiraly said, "I'm not hearing that much noise. It's possible that my hearing is going, but that might not be it." In addition, he would often stop play during the drills and ask a player what her options were at a given point. He always followed up this question with "Why?"

He described volleyball as often involving "chaos," thus requiring teams to develop solutions. For example, if a team seemed to be getting out of control, digging balls higher would give the setter added time to function. Better yet, though, teams would have contingency plans: "If this happens, Player X hits the ball; if this happens, Player Y does."

A substantial time drilling, in fact, was done with the players assigned to positions other than their normal ones. Making a middle-blocker set, for example, or a libero (back-row defensive specialist) block at the net had two purposes: taking players out of their comfort zone so they could cope with having to improvise; and fostering "empathy" for the unique difficulties of playing a given position.

Much to your correspondent's delight, Kiraly made periodic references to empirical analysis of national-team play. One revelation was that, contrary to the convention of playing a libero in the left-back position, on the national team, the center-back location resulted in the libero touching more balls (see this "heat map" approach, put together by women's national team statistical analyst Joe Trinsey and associates).

Kiraly also reviewed the optimal location for sets to the outside-hitter (in the left-front area). In response to Kiraly's quizzing, the Texas Tech players identified the "5-by-5" principle: The set should be 5 feet before the net and 5 feet inside the left sideline. Kiraly noted that sets too close to the net are vulnerable to the opposing block, and sets beyond the sideline "limit your options" (presumably taking away the option of going down-the-line because the spike could hit the antenna, a violation). He noted that, "Our hitting percentage plummets" when sets deviate from the 5-by-5 location.

In conclusion, Kiraly has a very cerebral approach to volleyball and is passionate about building interest in the sport and developing young players. Here's hoping it doesn't take another 33 years before we again cross paths... 

---
*All quotes are to the best of my recollection.

Saturday, February 7, 2015

Belated Summary of Last December's NCAA Women's Final Four

It's been quite a while since Penn State swept BYU, 25-21, 26-24, 25-14, last December to give the Nittany Lion program its second straight NCAA women's title, sixth in the last eight years, and seventh overall. Greater drama was to be found in the two semifinal matches, in which Penn State defeated No. 1-seeded Stanford, and BYU held off No. 2-seeded Texas, both in four games. Accordingly, my statistical review will concentrate on the two semis.

PENN STATE VS. STANFORD

The Cardinal came into this match with a 33-1 record, including a five-game win over Penn State on September 5. Stanford's only regular-season loss had come at Washington on November 26.

In the Stanford-PSU rematch in the NCAA semifinals, however, the Nittany Lions had taken Games 1 (25-16) and 3 (25-22), to place the Cardinal one game away from elimination. Stanford may have been feeling that, if it could pull out Game 4, it would be in good shape, as the Cardinal had hit exceptionally well in Game-5 situations all year.

Date
Opponent
Game-5 
Hitting %
Sept. 5
Penn State
.412
Sept. 7
Illinois
.350
Oct. 17
at Colorado
.500
Nov. 5
Arizona State
.381
Nov. 7
Arizona
.429
All matches played at Stanford, unless noted otherwise.

However, needing a big performance in Game 4 of the NCAA semifinal match vs. Penn State to keep its season alive, Stanford laid an egg, hitting .159. In the Cardinal's 128 total games this season, this .159 tied for Stanford's sixth-worst game-specific hitting percentage of the season (you may click on the following histogram to enlarge it; graphing software).

















How well Stanford would have hit in a potential Game 5 of the national semifinal against Penn State will never be known, the Nittany Lions wrapping up Game 4, 25-21. For the match, PSU outhit Stanford, .279-.207.

BYU VS. TEXAS

In the last hurrah for Longhorn seniors Khat Bell and Haley Eckerman, Texas fell behind two games to none (25-23, 25-16), blew out BYU in Game 3 (25-17), but fell to the Cougars 26-24 in Game 4. Texas is typically bigger and stronger than its opponents, but it was BYU that benefited from height.

A .421 hitting performance by 6-2 outside-hitter Alexa Gray (19 kills and 3 errors on 38 attempts) and 17 total team blocks paced BYU. Six-foot-four middle Amy Boswell recorded 1 solo block and 8 assists (5 total, as an assist is credited as half a block), whereas 6-7 opposite-side hitter Jennifer Hamson and  Whitney Young each registered 3.5 total blocks based on 7 assists each. Young, at 6-0, was the short one of the bunch!

On the December 15 installment of Internet radio's The Net Live, Penn State men's assistant Jay Hosack argued that blocking should be evaluated more broadly than via direct stuff-blocks for points, because even the best blocking teams will generate only about three points per game this way(click here for archived broadcast; this discussion begins at roughly the 40:00 minute mark). As Hosack and fellow panelist Katie Charles agreed, beyond stuff blocks, a good blocking team can engage in "control blocking" (i.e., slowing the ball down so that the diggers behind the blockers can transition the team to offense) or cause an opposing hitter to alter his or her spike so that the ball is hit out of bounds.

I like Hosack's ideas and plan to pursue them. I would also say, however, that stuff-blocks in particular were key to BYU's victory over Texas. There was a sequence in Game 2 in which the Cougars expanded their lead from 12-9 to 18-11. During this stretch, BYU earned 5 of its 6 points through stuff blocks. These stuff blocks may have shaken Texas or possibly just revealed a pre-existing flaw in the Longhorns' offensive strategy.

Texas was able to turn things around after the break, avoiding any BYU stuff-blocks all through Game 3 and well into Game 4. However, with Game 4 tied 16-16, the Cougars stuffed three straight Longhorn spike attempts for a 19-16 lead. Texas rallied back, producing an extremely tight finish, but it is likely the match would have gone to a fifth game absent BYU's blocking resurgence.

Against the Longhorns, Hamson hit a solid, if unspectacular, .241 (22-9-54). For Texas, Eckerman (.033) and Bell (.120) were neutralized, with middle-blockers Molly McCage (.357) and Chiaka Ogbogu (.500), a junior and sophomore respectively, giving the Burnt Orange some spark.

CONCLUSION

BYU was a shadow of its semifinal performance in the final, with Gray (.214) and Hamson (.071) hitting much less effectively against Penn State than against Texas. As a team, the Cougars recorded 7.0 blocks against PSU (2.33 per game), compared to the 4.25 per game against the Horns. Senior Nittany Lion setter Micha Hancock went out on a high note, setting middle-blocker and fellow senior Nia Grant to a .500 hitting percentage (9-1-16).

Penn State dominated two metrics I've tracked this season, making its NCAA title unsurprising. One is my Conference-Adjusted Combined Offensive-Defensive measure. The other is the number of times scoring fewer than 20 points in a game (excluding fifth games, which are played only to 15). That happened to the Nittany Lions exactly zero times this season! PSU lost only two games total in the NCAA tourney, 25-22 in the opening game against Wisconsin in the Elite Eight and 25-23 in Game 2 vs. Stanford.