Tuesday, November 29, 2011

NCAA 2011 Women's Preview: Introducing the Conference-Adjusted Combined Offensive-Defensive Rating

With the NCAA Division I women's tournament starting Thursday and rampant displeasure at the tournament committee's seedings (believed to be based heavily on RPI ratings), the fans at VolleyTalk are awash in different alternative rating schemes for evaluating the teams.

The major known ranking systems -- the AVCA Coaches' Poll, RPI, Pablo Rankings, and Rich Kern Rankings -- all appear to take teams' win/loss records and strength of schedule into account. As anyone who has read my blog over the past five years knows, my focus has always been on hitting percentage. I think it's a great singular statistic for incorporating many aspects of the game.

If you hit well (not just keep the ball in play, but get kills), your (individual or team) hitting percentage goes up. An attack kept in play by the other team hurts, as does a hitting error (spiking the ball out of bounds or getting stuff-blocked for an opponent's point). In order to hit well, a team must pass and set well. If you block or dig your opponent's spike attempts, that drives down the opponent's hitting percentage.

What I've done, therefore, is create a national ranking metric based heavily on each team's ratio of its own overall season hitting percentage (offense) divided by the overall hitting percentage it has allowed the opposition (defense). A ratio is maximized when a large numerator is divided by a small denominator. For example, hitting .300 for the season and allowing one's opponents (in the aggregate) to hit .100 yields a ratio of 3. Hitting .250 and allowing one's opponents to hit .200 yields a ratio of 1.25.

But that's not all. Teams play in differentially tough conferences, so I wanted to adjust for that. I came up with a very simple adjustment system out of thin air. We'll see how well my rankings predict this year's tournament matches and I can modify my conference adjustments as needed. Here's my current adjustment system:
  • If a team plays in the Big 10, Big 12, or Pac 12, I multiplied its hitting percentage-to-opponent hitting percentage ratio by 1.25. This way, teams that faced what I (and others) consider the top opposition are rewarded for doing so.
  • Teams from the ACC, SEC, Big East, Atlantic 10, Mid America, Missouri Valley, West Coast, Big West, Mountain West, Western Athletic (WAC), or Conference USA had their ratios multiplied by 1.00 (i.e., leaving their ratios alone).
  • Teams from all remaining conferences, whose schools tend to have relatively low athletic budgets and little or no track record of national success in women's volleyball, had their ratios multiplied by 0.75. Teams dominating these smaller conferences could hit really well and keep their opponents' hitting low, so to account for this, I adjusted their ratios downward.
What follows now are my 1-through-64 rankings of the NCAA tournament teams, based on my Conference-Adjusted Combined Offensive/Defensive (CACOD) rankings. You may click on the chart (which is divided into three panels) for an enlarged view.



Nebraska coming out top-ranked seems to give my system a little "face validity." Further, I have USC ranked higher than does the NCAA tournament committee! And if Dayton or Colorado State makes a big run in the tourney, you heard it here first. Like all other ranking systems, mine will stand or fall on how well it predicts tournament games. For any given match, we would predict the higher-ranked team to beat the lower-ranked one. We'll see how it works.

A note on sources:  I obtained all teams' (offensive) hitting percentages from the NCAA statistics page (see link in right-hand column). To glean teams' opponent (defensive) hitting percentages, I looked at a variety of conference and team-specific pages. When looking at conference and team pages, I checked whether the listed offensive hitting percentages matched those on the NCAA site, to verify that the statistics were from the same time-frame. As it happened, a few tiny discrepancies appeared between the NCAA and conference/team pages regarding teams' offensive hitting percentages (e.g., the NCAA page said Yale had hit .253, whereas the Ivy League page said the figure was .254).

Saturday, November 26, 2011

Cal Women Defeat Stanford Despite Wopat's Near Error-Free Hitting

Cal came into Stanford's Maples Pavilion last night and upset the Cardinal in a mostly tight match, 24-26, 29-27, 25-23, 25-14 (box score). Cal led 24-22 in the opening game, but Stanford ran off four straight points. Stanford had a couple of opportunities to go up two sets to none, having set points in Game 2 at 25-24 and 26-25. Cal had set points at 27-26 and 28-27, the latter of which the Golden Bears capitalized on.

The Golden Bears outhit the Cardinal, .306-.233. Leading Cal were Kat Brown, .500 (7 kills, 0 errors, 14 attempts); Correy Johnson, .500 (16-3-26); and Shannon Hawari, .375 (12-3-24). (I always confuse Kat Brown with Texas’s Khat Bell.)

Carly Wopat, Stanford's 6-foot-2 sophomore middle blocker, continues to shine. She has committed only 13 hitting errors in her last 234 swings (over the team's last nine matches). With 118 kills during this time, she is hitting .449 during her hot stretch. Here are Wopat's hitting statistics for the second half of Pac 12 play...

Opponent
Kills
Errors
Attempts
Percentage
@Arizona
9
4
20
.250
@ASU
18
7
35
.314
USC
11
0
18
.611
UCLA
7
1
21
.286
@Colorado
14
4
26
.385
@Utah
10
1
18
.500
@Washington
11
1
33
.303
@WSU
13
1
22
.545
Oregon
18
0
26
.692
OSU
12
4
32
.250
Cal
22
1
38
.553

Elsewhere in Pac 12 play last night...

USC avenged its earlier defeat to UCLA, winning in Westwood 24-26, 26-24, 25-21, 25-20 (box score). Based on the teams' low hitting percentages -- USC's .208 and UCLA's .136 -- it looks like it was a grind-it-out kind of match, with long rallies and many hitting errors. The pattern I identified earlier this season of the Bruins' often experiencing a dip in hitting percentage after the post-Game 2 intermission was again evident: .196 in Game 1, .263 in Game 2, and -.086 in Game 3 (though rebounding back to .143 in Game 4).

Oregon avenged Tuesday night's loss to Oregon State, sweeping the Beavers (box score). OSU was very anemic offensively, illustrated by its .090 hitting percentage for the match (with many unforced errors, as UO had only 5 team blocks) and its never exceeding a 50% side-out percentage in any game.

Lastly, just as I was getting a little excited about my graduate-school alma mater Michigan's improved play of late, the Wolverines lost in four at home to Ohio State (box score). The Buckeyes, led by Kelli Barhorst at .444 (10-2-18), outhit the Wolverines 275-.222.

Friday, November 25, 2011

Post-Thanksgiving Round-Up (2011)

I hope everyone had a nice Thanksgiving. Here are a few tidbits from Tuesday and Wednesday nights' action to get caught up on recent developments.

No sooner did I write an entry about how well Oregon was doing than the Ducks got swept Tuesday night by Oregon State. The teams have a rematch tonight in Eugene.

A few weeks ago, the college-volleyball world marveled at the 34-32 game played by Penn State and Wisconsin. Wednesday night, Oklahoma took a 35-33 set over Texas Tech to close out a three-game sweep.

In another Wednesday match featuring overtime games, Tennessee held off Kentucky in Knoxville, 28-30, 25-20, 30-28, 25-18, to win the SEC title. The potent Lady Volunteer offense did not disappoint, racking up a .291 overall team hitting percentage (box score). Leading contributors were DeeDee Harrison, .444 (15 kills, 3 errors, 27 attempts); Shealyn Kolosky, .421 (8-0-19); and Leslie Cikra, .370 (14-4-27).  The Wildcats were led by Whitney Billings .367 (14-3-30); and Becky Pavan, .353 (13-1-34). The latter is the sister of former Nebraska great Sarah Pavan.

One additional Wednesday match I wanted to mention, as a Michigan alumnus, is the Wolverines' four-game win over Michigan State. UM's left-handed slugger Alex Hunt (shown in this video from last year) had her best conference hitting night of the season, registering a .361 percentage. Her previous bests had been .353 at Nebraska and .308 at home vs. Minnesota. She has missed (or had her action greatly limited in) five Big 10 matches due to injury, and in her remaining matches has had a lot of lackluster hitting percentages (e.g., -.021 at Northwestern, .000 at Purdue, .083 at Illinois, and .120 at home vs. Indiana).

I try to follow as many UM matches as possible via Internet video or audio broadcasts and I've heard a couple of schools' announcers make the point that Hunt hits the ball as hard as anyone, yet she is inconsistent in her hitting percentages. If she can continue the hot hitting she exhibited vs. MSU, then the Wolverines' journey in the NCAA tournament might be termed the "Hunt for Blue December." Michigan hosts Ohio State tonight to conclude the regular season.

Finally, national No. 4 UCLA hosts No. 1 USC tonight. The Trojans (19-2 in conference) have already wrapped up the Pac 12 championship, ahead of the second-place Bruins (17-4). However, UCLA routed USC in the teams' first match-up. The Trojans will be out for revenge, but perhaps there's something about the Bruins that matches them up well against 'SC.

Tuesday, November 22, 2011

Nebraska's Slow Starts

The Lincoln Journal Star notes in this piece that Nebraska lately "has gotten off to slow starts, losing the first set in seven of its last nine matches. While Nebraska is 7-2 in those matches, it's something the Huskers hope to get a better handle on before the NCAA Tournament begins next week." (Thanks to "Red in Colorado" for posting the link on VolleyTalk.)

To probe the matter further, I created the following table based on Nebraska's schedule/results page, showing the Cornhuskers' opponents and lost sets/games (in yellow) in the matches alluded to above.

OpponentGame 1Game 2Game 3Game 4Game 5
Illinois
24-26
25-18
25-19
25-11
---
@Ohio St.
19-25
25-22
25-19
25-17
---
@Penn St.
17-25
15-25
25-17
17-25
---
Michigan
17-25
26-24
25-20
25-21
---
Mich. St.
25-11
25-18
25-23
---
---
@Indiana
25-19
25-19
25-20
---
---
@Purdue
19-25
19-25
16-25
---
---
Minnesota
17-25
25-18
25-15
25-16
---
Wisconsin
12-25
25-17
25-13
25-22
---

Here are a few things to note:
  • Most of Nebraska's Game-1 losses during this stretch were pretty decisive; in six of them, Nebraska didn't even get 20 points.
  • After five of the Game-1 losses, Nebraska bounced back to sweep the next three sets and take the match.
  • The two times the Huskers failed to turn things around after losing the opening set were both on the road and against very tough opponents -- Penn State and Purdue.
UPDATE: Tonight the Huskers beat Iowa -- without losing the first game or any game for that matter -- to win the Big 10 championship.

Monday, November 21, 2011

Oregon's Impressive Play Continues

The Oregon Ducks, who opened the season in August with a shocking win at Penn State, have been making a lot of noise of late. This past weekend, Oregon went down to the Bay Area and knocked off both the Stanford and Cal women. With these latest wins in Pac 12 play, the Ducks have now won 8 of their last 10 matches (game-by-game log).

One feature of the Oregon team -- if not the primary feature -- that appears to give it strength is the number of players who can hit for high percentages on a given night. Below, I've selected four major Duck wins and, for each, the following table displays players' hitting percentages (with number of spike attempts in parentheses). Hitting percentages of .300 or higher, based on at least 10 attempts, are highlighted.

Player
@PSU (8/26)
vs. UCLA (11/11)
@Stan (11/18)
@Cal (11/19)
Bergsma (OH)
.239 (46)
.290 (62)
.234 (47)
.220 (50)
Brenner (OH)
.579 (19)
-.257 (15)
.651 (43)
.043 (47)
Fischer (OH)
.317 (41)
.250 (32)
.143 (28)
.412 (34)
Krstojevic (MB)
.400 (5)
.571 (7)
---
.500 (2)
Paffen (MB)
-.500 (2)
.429 (14)
.100 (10)
.400 (10)
Williams (MB)
.118 (17)
.333 (21)
.194 (31)
.429 (21)

Junior Alaina Bergsma has been getting the most sets and has recorded respectable, if not spectacular, hitting percentages. Frosh Liz Brenner seems to be getting nearly as many sets as Bergsma lately; Brenner seems to be more of a hot and cold hitter. Junior Katherine Fischer gets fewer sets, but is capable of some big nights. Ariana Williams (Soph.) and Savannah Paffen (Fr.) are the main figures in the middle; they've had some strong matches, but aren't totally consistent. Junior Milica Krstojevic seems to record high hitting percentages whenever she plays, but doesn't get that many attempts per match. Orchestrating the Ducks' young offensive attack is sophomore setter Lauren Plum.

The Ducks have a busy Thanksgiving week, closing out the regular season with two matches against Oregon State, Tuesday in Corvallis and Friday in Eugene.

Other statistical notes from this past weekend...

Stanford soph MB Carly Wopat led her team in its five-game loss to Oregon with an error-free hitting night. With her 18 kills in 26 attempts, she hit .692.

Cal lost not only to Oregon this past weekend, but also to Oregon State. One apparent reason is that the Beavers outblocked the Golden Bears 15-8.

UCLA lost for the second time this season to Arizona, this time via a three-game sweep in Los Angeles. Both teams sided out well, but the Wildcats' percentages for the three games (65, 70, 75) exceeded the Bruins' (60, 66, 60).

Michigan State, which fell to then-No. 1 Illinois in five games in Champaign on October 15, won the return match in East Lansing last Friday, 3-1. The Spartans held the normally powerful Illini to a .165 hitting percentage.

Wednesday, November 16, 2011

Texas Nearly Flawless In Win Over Baylor

In sweeping Baylor tonight, Texas committed only one hitting error total in the final two games (box score). In Game 2, the Longhorns had 15 kills and 0 errors on 24 attempts, for a .625 hitting percentage. In Game 3, UT had 14 kills and 1 error in 28 swings, yielding a percentage of .464.  

Sunday, November 13, 2011

Top Women's College Teams Fall in Weekend of Upsets

The nation's top-ranked teams in last week's AVCA poll took it on the chin this weekend.

After taking the two opening games (sets) Friday night on the Oregon Ducks' home court, No. 1 UCLA dropped the final three and rather decisively at that, 25-15, 25-18, and 15-9 (box score).

Whichever team compiles a higher side-out rate (winning points on the opponents' serve) wins the game. Thus, it is not which team had the better side-out rate that is noteworthy, but rather the margin by which the higher team did so. Here's a graph I made of the UCLA and Oregon side-out rates by game.

By Game 2, UCLA's siding-out proficiency had begun to slide, and slide it did for the rest of the match. Siding out in the percentage range of the 40s to the low-50s is not going to win a team many matches.

Also, in a previous posting, I noted a tendency for UCLA to suffer a decline in hitting percentage after the post-Game 2 intermission. That's what happened against the Ducks. After hitting.286 and .226 in the first two games, the Bruins fell all the way to .000 (11 kills with 11 errors, on 43 swings) in Game 3, as Oregon climbed back into the match.

Alaina Bergsma carried the hitting load for the Ducks, taking 62 attempts and achieving 25 kills with only 7 errors, for a .290 evening. Milica Krstojevic (.571), Savannah Paffen (.429), and Ariana Williams (.333) also had strong hitting nights for Oregon, albeit on far fewer attempts. 

No. 2 Nebraska's loss at No. 10 Purdue was quick and decisive, 25-19, 25-19, and 25-16. What stood out to me from the box score was Purdue's microscopically low rate of hitting errors: only 6 miscues in 105 attempts (with 49 kills) for a team hitting percentage of .410 for the match. Nebraska recorded only 1 total team block. Three Boilermakers hit at clips of .500 or higher: Catherine Rebarchak (.625), Kierra Jones (.625), and Tiffany Fisher (.500). Cornhusker setter Lauren Cook, who had been suspended for her traffic-related legal problems, played in this match.

No. 11 Washington, which came into the weekend with five losses in its last eight matches, rose to the occasion Friday night with a five-set win over No. 3 Stanford (box score). Stanford outhit (.215-.193) and outblocked (16-8.5) Washington. However, the Huskies had two games, the fifth (.333) and the first (.326), in which their hitting percentages exceeded any game-specific percentage by Stanford. Another area is which U-Dub had the edge was its fewer service errors than Stanford (6 vs. 15). An error-free hitting night by Morgan Boukather (10-0-22, .455) led the Cardinal, whereas a near-error-free night from Bianca Rowland  (11-1-22, .455)  paced the Huskies.

No. 4 USC was the highest-ranked team to go unscathed this weekend, sweeping both of the Oregon schools. The Trojans have now won 17 of their last 18 matches.

No. 5 Cal swept the two Washington schools. Last year's star outside hitter Tarah Murrey has struggled for the Bears in some of their big matches this season; against the Huskies last night, she registered another lackluster hitting line (17-13-65, .062). Shannon Hawari (12-2-18. .556) and Correy Johnson (11-2-22, .409) were able to provide Cal with some firepower, however. The aforementioned Bianca Rowland capped off an outstanding weekend for the Huskies with a .435 hitting performance against Cal (12-2-23). 
     
No. 7 Illinois, which came into the weekend with three losses in its last five matches, swept No. 9 Penn State in a tight match Friday night, 25-21, 25-23, 28-26 (box score).With the win, the Illini have taken both matches from the Nittany Lions this season, the first time that has ever happened. Erin Johnson (9-0-14, .643) and Liz McMahon (10-1-16, .562) turned in monster hitting nights for Illinois, which finished with a .340 attack percentage as a team (on 42 kills and only 8 hitting errors, on 100 swings). Penn State's Deja McClendon had a big hitting night against Illinois (17-3-40, .350) and then added a .439 performance the next night in a three-game win over Northwestern.
  
Lastly, I wanted to discuss No. 24 Michigan, a team I follow closely as an alumnus (Ph.D., 1989) of the school. The Wolverines have continued to be ranked nationally, even though they lacked an impressive Big 10 conference win (with a 5-10 league mark) coming into this afternoon's contest at No. 16 Minnesota. Further, Michigan has exhibited an unfortunate penchant for failing to capitalize on game and match points during conference play (here, here, and here).

Somehow, today, the Wolverines managed to keep their focus and sweep the Gophers, 25-21, 25-21, 25-22 (box score). With only one hitting error between them, Claire McElheny (7-0-18, .389), Jennifer Cross (7-1-16, .375), and Molly Toon (9-0-27, .333) powered the Maize and Blue. Whenever the Gophers seemed to be digging themselves out of a hole, either Michigan came up with a timely block (8 total) or Minnesota committed a service error (11 total).

All was not perfect for the Wolverines, however. One problem was serve-receipt, with the Gophers recording 8 aces. Minnesota received strong hitting performances from two of its stalwarts, Tori Dixon (11-1-24, .417) and Ashley Wittman (14-3-28, .393). Michigan's Alex Hunt reportedly suffered an injury in warm-ups and was used very sparingly.

Wednesday, November 9, 2011

Texas Sweeps Texas Tech; Eckerman's Spike Attempts Increase Without Bell in Lineup


No. 8 Texas came here to Lubbock and made short work of Texas Tech, 25-16, 25-15, 25-18. Above is a shot I took of the Longhorns walking toward the net to shake hands with the Red Raiders at the conclusion of the match. The contest was the first for Texas without middle blocker Khat Bell, who injured her knee in a weekend match in the midst of an amazing three-match hitting stretch.

How did the Longhorns reallocate their spike attempts without Bell? The most direct comparison would be Texas's statistics in its two matches this season against Texas Tech, with Bell (Oct. 12 in Austin) and without her (tonight). Here are links to the box scores of the first and second matches. Looking at only two matches is not ideal, but it's a start. Additional matches can be examined as the Longhorns' season moves toward a close. Here are graphs of the Horns' hit allocations in their two matches against the Red Raiders (the first column doesn't add to 100%, due to rounding).

The Longhorns seemed to compensate for Bell's absence tonight by increasing their outside sets to Haley Eckerman. The move worked well, as Eckerman hit .595 on 23 kills with only 1 attack error, on 37 spike attempts (UT as a team took 106 swings, hitting .396). Rachael Adams also exceeded .500 tonight (.556), albeit on far fewer attempts (6-1-9). One reason Texas was able to hit so well tonight was that Tech had only 1 total team block for the match.

The Red Raiders hit only .126 tonight. A bright spot for them, however, was junior right-side hitter Miara Cave, who hit .471 for the evening (10-2-17). 

Monday, November 7, 2011

Khat Bell (Texas) Out for Season; Had Been on Spectacular Hitting Run

Texas frosh MB Khat Bell is out for the season with a knee injury, it was announced today. I was actually getting ready to do a posting on her recent error-free hitting. Bell's statistics lines (Kills, Errors, Total Attempts, Percentage) for hitting in her last three matches were:
I'm sure all VolleyMetrics readers will join me in wishing Ms. Bell a speedy and healthy recovery. I had been looking forward to seeing her (and her Longhorn teammates) play Wednesday night here at Texas Tech, but we'll just have to wait till next year.

Sunday, November 6, 2011

Penn State-Wisconsin Marathon (2011)

For drama in a regular-season match, it would be hard to top today's Penn State-Wisconsin battle in Madison. In the end, it was the unranked Badgers pulling a five-game upset over the No. 6 Nittany Lions, 26-24, 25-19, 32-34, 14-25, 15-12.

The set score that probably jumps out at you is that for Game 3, won by Penn State 34-32 in preventing a Wisconsin sweep. As gleaned from this statistics sheet (which includes the box score and play-by-play) and shown in the following table, the Badgers had five game (and thus match) points in the third set, and the Lions also had five game points.

Held Game PointScoreHeld Game Point
Wisconsin
24-23
25-24
Penn State
26-25
Penn State
27-26
Penn State
Wisconsin
28-27
Wisconsin
29-28
Wisconsin
30-29
Wisconsin
31-30
32-31
Penn State
33-32
Penn State*
*Won game 34-32.

Penn State outhit Wisconsin for the match, .212-.173. One game about which the Nittany Lions are probably kicking themselves is the opener. Despite outhitting Wisconsin .250-.182, Penn State dropped Game 1 to the Badgers, 26-24, thanks in part to five Lion service errors.

For the match, total team blocks were essentially even, 15 for PSU to 14 for Wisconsin.

Side-out rates (i.e., winning points on the opponent's serve) are instructive. In any particular game (set), whichever team records the higher side-out percentage will win. As seen in the graph below, Wisconsin was the steadier team at siding out, registering in the .60s in four of the five games, whereas Penn State jumped around a bit more.


In the closing stages of the match, Wisconsin built a 14-10 lead in Game 5 (which is played only to 15). Fighting till the end, as always, Penn State held off two more match points, cutting the Badgers' lead to 14-12. At that point, Wisconsin finally got the game-winning kill from Bailey Reshel.