Friday, December 7, 2018

Prognostic Value of Seeds in NCAA Women's Tourney

The NCAA website has an article on how women's volleyball teams' seeds predict how they do in the tourney. Data are from 2000-onward, which is as long as teams have been seeded 1-16. Two findings I found interesting were as follows (quoting from the article):

Since 2000, the tournament’s No. 1 seed has won the national championship six out of 18 times.

Stanford [No. 11] in 2004 and No. 9 UCLA in 2011 are the only teams seeded lower than No. 6 to win a national championship.

The article also contains a chart at the bottom, listing the final four teams every year from 2000-onward, their seeds, and their ultimate national finish.

Thursday, November 29, 2018

2018 NCAA Women's Tourney Preview

With this year's NCAA women's tournament getting underway tonight, it's time for my annual Conference-Adjusted Combined Offensive-Defensive (CACOD) ratings of the leading teams (explanation). In a nutshell, the CACOD takes each team's season-long hitting percentage, divides it by the aggregate hitting percentage it has allowed its opponents, and then multiplies the result by an adjustment factor for difficulty of conference. Relevant numbers for this year's seeded teams are shown as follows (with the actual CACOD ratio listed under "adjratio"). You can click on the graphic to enlarge it.


The first thing that jumps out at me is that Nebraska is grossly under-seeded (No. 7), relative to its CACOD rating (best in the nation at 2.49). The Cornhuskers' defense, allowing a microscopic .130 opposition hitting percentage, is key to the high CACOD rating, along with the 1.25 adjustment factor for playing in the Big 10. And what a Big 10 season it was, with five of the top eight national seeds coming from the conference! Nebraska went 15-5 in Big 10 play, losing twice to Minnesota (No. 2 national seed), splitting with Illinois (No. 3 seed) and Penn State (No. 8 seed), and losing its only match to No. 6 seed Wisconsin (in five in Madison).

BYU, which charged out of the gate with an early win over Stanford and was ranked No. 1 for most of the season, allowed a comparably low .139 opposition hitting percentage, while the Cougars hit a very high .318. BYU had the highest unadjusted ratio of own to opponents' hitting percentage (2.29), but received no boost to its CACOD value by playing in the West Coast Conference.

The Cougars nearly went undefeated this year, losing only in their regular-season finale at Loyola Marymount. A likely contributor to BYU's late-season difficulties was outside-hitter McKenna Miller's season-ending ACL injury, suffered on November 8 vs. Santa Clara.

BYU's offense relied heavily during the season on two hitters: OH Roni Jones-Perry, who took a team-leading 882 hitting attempts on the season (30.3% of the Cougars' 2,910 attempts as a team), recording a very strong .346 hitting percentage; and Miller, who was second on the team in attack attempts with 602 swings (20.7% of the team total) and had a .307 hitting percentage. (Click here for BYU's season stats sheet.)

With someone who takes one-fifth of your team's swings -- Miller -- out of the lineup, how would the Cougars make up for this absence? Potentially, others among BYU's leading hitters, including middle-blocker Heather Gneiting (.393 hitting percentage on 349 attempts), MB Kennedy Eschenberg (.356 on 374 attempts), and OH Madelyn Robinson (.246 on 281 swings) would receive more sets in Miller's absence. However, as the following graph shows, this is not what has happened. 


Comparing individual players' hitting attempts (as a percent of the team's total) in the four matches before Miller's injury (left of grey bar) to hitting attempts in the four matches after, we see that the only hitter who has shown an increase in hitting attempts post-Miller injury is Jones-Perry (green line in graph), who already was the team's iron-woman. Jones-Perry is also among the team leaders in digs (1.93 per game) and blocks (0.67 per game).

On a concluding note (for now), we have seven years of data on the CACOD, so there's a pretty decent track record. One finding is that no team with a CACOD lower than 1.91 has ever won the NCAA title. As they say in television commercials for investment companies, "Past performance is not a guarantee of future performance." Still, I would say that this year's champion is likely to come from the pool of teams with CACOD values of 1.9 or higher.

Monday, September 10, 2018

Strange Start to 2018 NCAA Women's Season

The women's college volleyball season is only three weeks old, so the Top 10 teams in the AVCA pre-season poll shouldn't have too many losses at this point, should they?

Well, as shown in the following chart, the number of Top 10 teams who are still undefeated is... ONE. The number of these teams with three or more losses is... THREE.

Team Good Wins Over... Losses To...
1. Stanford Penn St.,
Minnesota
BYU
2. Nebraska Oregon (18),
Creighton (14)
Florida
3. Texas Oregon, Florida,
Kentucky
Wisconsin
4. Minnesota ? Oregon, Stanford
5. Kentucky ? Creighton, USC,
N. Iowa, Texas
6. Penn State Oregon Stanford
7. Florida Nebraska Texas, USC,
N. Arizona
8. BYU Stanford, USC ---
9. Wisconsin Texas Baylor
10. USC Kentucky, Creighton,
Florida
C. Florida, BYU,
Marquette

And it's not merely the number of losses by Top 10 teams. It's to whom. Florida losing to Northern Arizona? Wisconsin losing to Baylor? USC losing to Central Florida and Marquette?

I cannot say with certainly that this is the most upset-laden first three weeks ever in women's college volleyball, but I doubt there have been many seasons like this one.

The lone unbeaten team among those in the AVCA pre-season Top 10 is BYU, who is doing it with defense. Even excluding a season-opening exhibition match in which the Cougars held Utah Valley to an .086 hitting percentage, BYU's opposition hitting percentage is currently an extremely low .177. As shown in the following graph, the Cougars have been allowing some truly microscopic hitting percentages by opponents (BYU played two matches on a trip to Duke).


Another team that has been extremely consistent so far is the University of Michigan. The Wolverines are 9-0 and haven't even given up a game. The competition hasn't been great, but among Michigan's sweeps are ones over respectable teams such as LSU and Colorado State.

Sunday, July 1, 2018

Long Beach State's DeFalco Double-Trouble (Hitting and Digging) For UCLA in NCAA Men's Final

In reflecting on Long Beach State's five-game win (25-19, 23-25, 20-25, 26-24, 15-12) over UCLA back in May for the NCAA men's championship, there seemed to be a few possible angles to pursue.

One was the re-emergence of California schools -- which had been completely absent from the last three NCAA title matches and had last hoisted the trophy in 2013 (UC Irvine) -- as championship combatants.   

A second possible angle was UCLA's aggressive serving, which led to four straight service errors at a crucial point in Game 4 (at 20-20, 21-21, 22-22, and 23-23). I wrote about the Bruins' serving dilemma back in 2016 and it evidently is not something they have yet solved.

To me, however, it was the two-way performance of 49ers' outside-hitter TJ DeFalco, whose hitting and digging were indispensable to his team's success. DeFalco hit .419 (18 kills and 5 errors on 31 hitting attempts) and led his team with 12 digs.

A method for putting a player's dig total into perspective, which I previously discussed here, is to examine digs as a percentage of the opposing team's dig-able hitting attempts. UCLA delivered 97 attacks that were potentially dig-able: 62 that were not dug and ended up as kills, and 35 that were dug by LBSU (this formula excludes UCLA's hitting errors and balls Long Beach State immediately blocked back to the Bruins' side of the court and which were kept in play, neither of which the 49ers could be expected to dig). Thus, of all the dig-able spike attempts made by UCLA, DeFalco personally dug 12.4 percent of them (12 digs/97 dig-able spikes).

DeFalco's digging performance seems pretty good (especially for someone who also hit over .400 in the same match), but we must ask the eternal question: Compared to what? To address this question, I looked at box scores from all NCAA men's championship matches for the past 10 years. I then identified all players who hit .400 or higher (with a minimum of 15 attempts in a three-game match, 20 attempts in a four-game match, and 25 in a five-gamer). I then looked at the digging performances of the selected players, which are shown in the following table (ODA = Opponent's Dig-able Attacks).

Player (Team)
Year
Team's 
Result
Hitting
%
Hitting 
Att.
Digs
Digs/
ODA
Nicolas Szerszen (Ohio St.)
2017
W 3-0
.480
25
8
.151
TJ DeFalco (LBSU)
2018
W 3-2
.419
31
12
.124
Cody Caldwell (Loyola-Chicago)
2014
W 3-1
.562
32
10
.116
Carson Clark (UC Irvine)
2012
W 3-0
.465
43
8
.089
Tanner Jansen (USC)
2012
L 3-0
.500
18
7
.073
Brad Lawson (Stanford)
2010
W 3-0
.821
28
5
.062
Zackia Cavera (UC Irvine)
2013
W 3-0
.476
21
4
.059
Brenden Sander (BYU)
2016
L 0-3
.500
24
1
.014
Max Lipsitz (Penn St.)
2010
L 0-3
.450
20
1
.012
Jeff Jendryk (Loyola-Chicago)
2015
W 3-2
.500
28
0
.000

I also have an honorable mention for a player who fell slightly short of the .400 hitting percentage needed for inclusion in this analysis. In Ohio State's 2011 five-game win over UC Santa Barbara, the Buckeyes' Shawn Sangrey hit .389 (30-9-54) and recorded 7 digs. The Gauchos delivered 85 dig-able attacks (45 kills and 40 balls dug by the Buckeyes) and Sangrey dug a healthy 8.2% of these.

Based on the above statistics, it looks like Ohio State's Nicolas Szerszen had a better championship match in 2017 than DeFalco did in 2018, holding advantages in both hitting and digging percentage. However, one could argue on DeFalco's behalf that his performance came in a much tougher match (a five-gamer vs. UCLA on the Bruins' home court) than Szerszen's (a three-game win on his home floor at Ohio State).

Sunday, February 4, 2018

Looking Back at the 2017 College Women's Season: Nebraska, Penn State, and Mick Haley

Three stories stand out to me, in looking back at the 2017 college women's season:
  • Nebraska's turnaround from a slow start (coinciding with the loss of three senior starters from the 2016 squad) to an eventual NCAA championship.
  • The Cornhuskers' recent domination of Penn State, in the form of a seven-match winning streak over the Nittany Lions from 2015-2017.
  • The messy "divorce" taking place being USC and its now-former coach Mick Haley.
When I visited the University of Nebraska-Lincoln in mid-October on academic business, I stopped by Husker Headquarters near the end of my trip to pick up a souvenir. What I ended up purchasing was a Nebraska "School of Volleyball" shirt. Little did I know at the time that the numbers of NCAA championships and final appearances listed on the bottom line of the shirt (click on photo to enlarge) would have to be updated so quickly.

The Huskers started off the season with a 6-3 record, the losses coming to Oregon, Florida, and Northern Iowa (game-by-game log). However, by midseason, Nebraska had things rolling again, going 19-1 in B1G conference play (losing only at Wisconsin), en route to a 32-4 final record.

The Huskers' national championship, accomplished with a 3-1 (25-22, 25-17, 18-25, 25-16) win over Florida in the title match, raised Nebraska's number of women's NCAA volleyball titles to five (including two of the last three) and finals appearances to eight in school history. The Gators faced a difficulty in the finals that I identified in connection with their win over USC in the Elite Eight, namely allowing extremely high side-out rates to their opponents, meaning that Florida could not string together points on its own serve. As shown in the Nebraska-Florida box score, the Huskers sided-out at 77% in Game 2 and 76% in Game 4.

Florida has much to be proud of from this season, including a three-match winning streak in the tournament over Pac-12 schools -- UCLA, USC, and Stanford -- to reach the NCAA championship match (Gators' season log).

In the other semifinal (besides Florida-Stanford), Nebraska overcame a Penn State match-point in Game 4 to prevail in five, 25-18, 23-25, 24-26, 28-26, 15-11.

With NCAA titles in 2007, 2008, 2009, 2010, 2013, and 2014, Penn State won the championship six times in eight years. Perhaps even more remarkable than the fact the Nittany Lions have not won a national title in the past three seasons is that, as documented in the following chart, Penn State has lost seven straight matches to Nebraska during the same time span.

Date (Box 
Score Link)
Location
Games
Neb H%
PSU H%
12/14/17 KC, Mo.*
3-2
.249
.217
9/22/17 PSU
3-0
.347
.227
12/9/16 Neb*
3-2
.256
.234
11/16/16 Neb
3-0
.290
.065
11/4/16 PSU
3-2
.238
.166
11/28/15 Neb
3-0
.241
.126
10/2/15 PSU
3-2
.173
.164
*NCAA tournament.

Whereas the Huskers' hitting percentages have varied widely (from .173-.347) in these seven matches vs. Penn State, the Nittany Lions' hitting percentages have hit a wall in the low .230's, never exceeding .234 vs. Nebraska. How unusual is it for Penn State to hit at such a low clip? I counted a total of 13 matches the Nittany Lions played in 2015, 2016, and 2017 against Top 10 opponents other than Nebraska (four each against Stanford and Minnesota, two each against Wisconsin and Michigan State, and one against Hawaii). In a majority (seven) of these matches against top-shelf opposition, Penn State's hitting exceeded .250 (four times from .250-.299 and three times exceeding .300).

Our final topic is the controversial end to Mick Haley's 17-year run as coach at USC. Back in 2007, I looked at several top programs' NCAA tournament performance from 2003-2007, relative to what would have been expected from their seedings. For example, the No. 1 national seed in the NCAA tourney would be expected to win all six matches necessary for the title, the No. 2 seed would be expected to win five matches (before losing in the final), the No. 3 and 4 seeds would each be expected to win four matches, etc. I decided to repeat this analysis specifically for Haley over the past 10 seasons (2008-2017), the results of which appear in the following chart.

Year
USC Seed
Expected Wins
Actual Wins
Difference
2008
---
0.5*
1
+0.5
2009
---
0.5
1
+0.5
2010
6
3
4
+1
2011
7
3
4
+1
2012
6
3
3
0
2013
6
3
3
0
2014
---
0.5
1
+0.5
2015
1
6
3
-3
2016
---
0.5
0
-0.5
2017
10
2
3
+1
*In match-up of two unseeded teams, it is assumed each team has 50% chance of winning.

Totaling up the final column, Haley's USC teams won one more NCAA-tournament match over the 10-year period from 2008-2017 than would have been predicted from their seedings. In other words, Haley led his teams in the NCAA tourney about as far as they were expected to go, neither greatly exceeding nor greatly falling short of these benchmarks (one exception being 2015, in which the Trojans were the national No. 1 seed, but advanced only as far as the Elite Eight).

In Haley's favor, he led the Trojans to two NCAA titles (2002 and 2003) and, as recently as 2010 and 2011, his teams were making the Final Four. Also, his final (2017) 'SC squad was one point away from making the Final Four as a No. 10 seed.

On the negative side of the ledger, only once did a Haley-led Trojan squad attain better than a No. 6 NCAA-tournament seed in the last 10 years (2015). Further, USC's failure to make even a single Final Four from 2012-2015 with superstar Samantha Bricio on the team has to sting for Trojan fans.

2023 NCAA Women's Preview

Sixth-four teams are alive at the moment, but it sure looks like Nebraska (28-1) and Wisconsin (26-3) will meet for a third time this season...