Coach’s Clipboard: Revolutionizing the Way We Look at a Player’s Hitting Percentage
Brent Lewis Anyone who knows anything about me knows that I am in love with statistics- and I have been as long as I can remember. My earliest memory with stats was way back in 1994, when, at the age…
Access all of Prep Dig
Continue reading this article and more.
Continue ReadingAnyone who knows anything about me knows that I am in love with statistics- and I have been as long as I can remember. My earliest memory with stats was way back in 1994, when, at the age of seven, I kept a baseball score book of the MLB All-Star Game from my living room couch. This is my favorite All-Star Game of all time, featuring nine Hall of Famers (one of which was Tony Gwynn, who scored the winning run in the bottom of the 10th) and many more that will join them once they become eligible for induction. This was an historic game for a multitude of reasons: it was the last of four All-Star Games to be played at Three Rivers Stadium in Pittsburgh, Willie Stargell threw out the first pitch, it was the first win for the National League since 1987, and Leonard Coleman became the first African-American to present the Most Valuable Player award. Also, the legend Bob Costas did play-by-play next to Mr. Baseball Bob Uecker (juuuust a bit outside!) and Hall of Famer Joe Morgan. Twenty years later, I’ve transmuted my love of statistics into my career: volleyball statistical analysis. Most people find this kind of thing boring, however I enjoy it so much I’ve written papers about it. I do a number of different things with statistics throughout the year, and I’d like to share a little bit about it.
My largest, most detailed statistical study was my Master’s thesis. I took the seven common box score statistics (hitting %, opponent hitting %, assists, kills, blocks, aces, digs) and compared them with team standings in order to determine which of the statistics is the greatest indicator of a higher team ranking. What I found were that the two most significant factors for winning were hitting % and opponent hitting % (more specifically, the difference between the two). Blocking was the least significant statistic. (Parallel studies have reported similar results across different levels of play.) Based on this data, my teams spend a lot of time finding ways to put the ball on the floor and rarely ever work on blocking. My teams are never good blocking teams, and I’m perfectly OK with that. As long as we aren’t getting “tooled” on the block, I am satisfied.
This past year at Iowa Western, I performed a serving study during the season. I categorized all (4,003) of our serves based on the quality of the opponent’s pass (among other factors) and placed the data into a spreadsheet. I had two primary foci for this: first, to find out which rotations were our strongest defensively, and second, to determine the relationship between the strength of serve and result of the side-out attempt. I’ll spare you all the data, but I’ll share some of my results. Our strongest rotations (based on the criteria in the spreadsheet) were, in order: 5, 6, 1, 2, 4, 3. This data allowed us to see that by starting in rotation five, we would give ourselves the greatest chance to score more points than if we started in a different rotation (we scored 757 points in 5 as opposed to 561 in 3). The second part of my study showed that the more “perfect” passes we allowed resulted in a higher side-out rate for our opponents. I took this one step farther to combine service error rate and side-out rate (either way we lose the point), and compared that to serve quality. What I found was that weaker serves (including low service errors that accompanied them) led to a higher side-out rate, while stronger serves, although resulting in more errors, led to a significantly (up to 6%) lower side-out rate. So what did I do with this data? It encouraged me to promote tougher serving and not penalize service errors, as someone who is scared of making a service error will not serve tough. I am always OK with service errors if they are committed on a tough serve. Of course this can create an issue, as four consecutive service errors will not win you a game. This is why I am also an advocate of teaching different types of serves to everyone. This way when the team commits consecutive service errors, the next server knows we need to score some points in a row so she hits a lesser, still effective serve. If she usually hits a high, hard topspin serve, maybe now she stays down and hits a low-driven float serve. Still an effective serve, not as potent as the hard top, but still useful towards getting the opponent out-of-system.
At my previous school (Hardin-Simmons University) I created a setting spreadsheet which analyzed each set by our setters. It was broken down into several different categories: serve-receive, transition, by rotation, by hitter, etc. What this showed us was the success (or lack of) each setter had setting each hitter in each rotation. This was very useful in serve-receive, where we could see if a hitter had little success hitting a particular set so we could adjust to find something that worked better for her. Likewise, it showed us our strongest set location/hitter combination which we encouraged our setters to use as often as possible. Our theory was if Player X kills the quick set at 80% success in rotation three, set her every chance you can on a quick set in rotation three. Contrarily, if Player Y had a 10% success rate killing a high outside set in rotation five, we wouldn’t set that anymore. Instead, we would set her a quicker outside ball or move her to the inside where she had a higher conversion rate. The whole purpose of this was to determine the best opportunities for our hitters to get a kill (especially in serve-receive). We were the best offensive team in the conference both of my years on staff there as we led the conference in kills and assists both seasons. A lot of that had to do with the natural talent on the team, but the spreadsheet (and similar data) helped maximize it.
Most recently, I created a new statistic and wrote a paper about it. I’ve been called Moneyball before, and I take that with pride as I’m sure if Billy Beane knows anything about volleyball he would appreciate my thinking behind this stat. I first envisioned a need for a new offensive statistic when I would see a hitter blast a ball out of bounds, get yelled at by their coach, so then on their next attempt when they would get a difficult ball to kill they would hit a roll shot, tip, or other “useful” shot under the circumstances. My issue with this was that their hitting percentage would drop (and possibly rather significantly) every time this occurred. The hitter is making a good play, following coach’s advice, and yet their offensive numbers would decline. That didn’t seem fair to me, so I started playing with some numbers and came up with what I think is a viable solution: (kills/errors) + hitting %. I call it KEHP, and I won’t go into mass detail here about how I arrived at this formula (I cover it in full in the paper), but I will tell you why I like it. First of all, as long as a hitter keeps the ball in play, you are guaranteed that your KEHP stat will not significantly drop (it will drop slightly, but I won’t explain that here- it is explained in the paper). This is the reward a hitter gets for keeping the ball in play and not trying to kill what I refer to as an “unkillable” ball. Instead of penalizing a hitter for maintaining control of the ball, they are now rewarded. Additionally, there is no maximum score; the better a hitter performs, the higher the statistic goes. Hitting percentage caps at 1.000 regardless if that is two kills on two attempts or 100 kills on 100 attempts. Per KEHP, these same numbers would be represented as 4.000 and 102.00, respectively. That more equally represents the kill numbers, in my opinion. (If you calculate those numbers using the formula stated above I know that they don’t equate; there is a reason for that that I won’t cover here, because, again, it is explained in the paper). I have debated the usefulness of this statistic with some high quality volleyball minds and have garnered mixed results. Billy Beane also received mixed results when he first implemented the theories behind Moneyball. I’m not saying this one statistic will have the impact on volleyball that Moneyball has had on baseball, but it’s something I’d encourage everyone to consider. I use it wholeheartedly and beat it into the heads of my outside hitters. I’m sure they’re sick of hearing me talk about it, but I’m sure they are also pleased to see their numbers go up.
That is just a glimpse into some of my bigger projects. I’m a proponent of teaching the game and expanding the knowledge of the game, so if any reader wants to see a copy of anything mentioned above, email me (blewis@iwcc.edu) and I will send you one. I also enjoy fun-spirited debates so if anyone wants to haggle with me bring it on!
Brent Lewis serves as the head coach for Top Ten Volleyball Club and is the assistant coach for NJCAA D1 Iowa Western Community College. He arrived at Iowa Western in May 2013 after spending the previous two seasons as an assistant coach at NCAA D3 Hardin-Simmons University in Abilene, Texas where he received his Master’s degree in Kinesiology.
The 2013 Iowa Western team went 44-5 and earned a seventh place finish at the National Tournament after being ranked as high as #2 during the season. The team won the Region XI and District D titles en route to the National Tournament appearance. With his assistance the Reivers led the nation in kills and assists as well as finishing seventh in total digs. In his two years at HSU he helped the team to a 58-10 overall record, 38-3 record in the American Southwest Conference, two conference titles, and two NCAA regional tournament appearances. The Cowgirls led the conference in kills and assists both years he was on staff. Following the 2012 season HSU was named American Southwest Conference Coaching Staff of the Year.
In three years as an assistant coach at the collegiate level he has amassed a record of 102-15. He has been a part of two school record winning streaks, first with the 2012 HSU squad that won 25 straight games and then again in 2013 with an IWCC team that tied the school record with 33 consecutive victories. He has coached a total of eight All-Conference players, one conference MVP, five All-Region players, and three All-Americans. A bizarre fact he is not real fond of is that three times his team has blown a 13-8 lead in the fifth set (it happened a fourth time but the team managed to hold on for the win).
A majority of his work lies in the statistical and analytical realm. While a graduate student he wrote a Master’s thesis entitled “Discovering Success in Volleyball: An Integral Match Analysis of the Relationship between Statistics and Team Rankings” which compared general statistics to team records in order to determine the most significant numbers for predicting the outcome of a win/loss record. He most recently finished a paper where he proposed a new statistic to measure a hitter’s effectiveness.
A native of Marshall, Texas, Brent was a hitter on the men’s club team at East Texas Baptist University where he received his Bachelor’s degree in Mass Communication. He lives in Council Bluffs, Iowa.
You may email Brent at blewis@iwcc.edu with any questions or items you would like to see him write about.