“Not Everything That Can Be Counted, Counts”

As the club season is in full swing, coaches evaluate the skill development of their players, establish successful systems that take advantage of player strengths, and then use the information to organize their practice routines. The ability of the coach to accurately observe on-court happenings is the first evaluation component. Combining observations with statistical data will assist in supporting or refuting your observations and identifying the strengths and weaknesses of players and systems.

W. Edwards Deming

“If you can’t measure it, you can’t manage it” is an often-quoted admonition attributed to the late W. Edwards Deming. So, club coaches dutifully submit match videos to analysts or spend hours applying analytics to every contact by their players. Sabermetrics is not my thing. However, I’m sufficiently familiar with the process to know that valuable information can be gleaned from the statistics attached to practices, matches, or a season. Equally, misjudgments can be made if critical thought isn’t a component of the evaluation process.

Albert Einstein

Not to diminish the importance of statistical analysis of performance, but since I’m tossing out quotes, a quote from Albert Einstein illustrates that statistical data will not answer all questions; “Not everything that can be counted counts, and not everything that counts can be counted.” What type of statistical information is relevant to team success? By relevance, are the statistics measuring the items of most importance, and can you effectively communicate the results to your team and staff?

With statistical information, much of the available software that club or high school coaches use for monitoring performance uses some form of the Coleman System of Volleyball Statistics. The Coleman system generally evaluates the serve and pass with a point system (3,2,1,0). The hitting efficiency would be attack kills minus the errors divided by the attempts (10 kills-5 errors/20 attempts=25% efficiency).

Dr. Jim Coleman

When reviewing the data from any statistical system, the overriding question is, “What do I do with these numbers?” Statistical information that doesn’t impact behavior (team systems, practice planning, technical training) wastes your valuable time.

I tend to use the numbers in the Coleman System in a manner that might be considered outside the norm. Regardless, I’ll present a few thoughts about evaluating player and team performance that I deem valuable.

Sample Size of the Information
It is always better to have a large sample size when measuring performance. Larger sample sizes of information will lead to greater data accuracy, a better predictor of performance, and, ultimately, the coaching decisions made. I caution against making significant changes based on the results of a match or a single tournament. Those scenarios are a snapshot of the performance data. I encourage using statistics throughout the season, in both practice and competition. The best predictor of future performance is past performance.

Recency Bias
Like most humans, coaches tend to focus on the most recent on-court occurrences. The focus on recent on-court events could lead to rash decisions not aligned with a player’s historical performance. For example, a player gets aced twice by a tough server. There is an urge for the coach to change the reception pattern or take the player off the court. Even though there were two errors, historically, the player has been a quality passer. Players will return to the statistical mean over time. With this concept in mind, I avoid making substitutions after mistakes. Besides the potential damage to confidence, if they were sufficiently skilled to be in your reception pattern, don’t let a mistake or two change their role. Player errors are an unavoidable component of the game. It is a better use of time to work with players on positively handling mistakes rather than taking them off the court or making lineup changes.

Purpose of Statistics
Statistics have a two-fold purpose. One is to provide the coach with meaningful information to evaluate players and systems, be a predictor of execution during competition, and plan future practices.

The second and equally important purpose is to provide the player with a snapshot of current performance or trendlines of ongoing improvement. Often, the player will not easily understand the statistics that might be meaningful to a coach. For example, telling a player that 70% of their receptions are in-system and allowing the setter to run a great offense might be more meaningful to players than informing a player they are passing at a 2.49 in the Coleman system. The coach should provide straightforward feedback that promotes the player’s comprehension of the information. I use the most visually appealing way (video, pictures, drawings, etc.) to communicate information to the player. Numbers on a spreadsheet might be meaningful to a coach, but better ways exist to communicate the same information to a player.

Relative Statistics
Relative statistics will measure your team’s performance in specific areas relative to the opponent’s in those same areas. One of the most critical relative statistic that correlates directly to winning is the sideout percentage by your team relative to the sideout percentage of the opponent. Generally (always), the team with the highest sideout percentage will win a set. Your team may sideout at 51% (not very good), but you’ll most likely win the set relative to your opponent’s side out percentage of 48%. Not all statistics equate to match success. For example, your team may dig more balls than your opponent but can still lose the match.

Measuring What is Important
I focus on the goal of each skill and try to incorporate statistics based on the goal. My skill goals might be different than yours. For example, the goal of the server is to put the ball in bounds in a manner that leads to point scoring. If the serve is too easy, the opponent will sideout effectively. If the serve is too risky, the error rate is high. In both cases, servers that are high error or serve a ball allowing the sideout percentage by the opponent to be high will not have many serve attempts. Therefore, I evaluate the best server by the points scored during their serve. I don’t care how the points are scored (opponent error, a transition kill by our side, etc.) I understand that a stuff block by my teammate may or may not have been impacted by my serve. I value that the serve went in-bounds, and we scored a point. A player not scoring points when serving might be a candidate for a serving substitution. By default, your best server is the player serving the most points in a match. The server scoring the most points should probably be one of your first servers in your rotation. Again, this is a more straightforward statistic for players to understand. I prefer to say, “We scored 12 points from your serve” versus “You served at a 2.33.” on the Coleman scale. Or, if measuring the effectiveness of a server, a points per serve metric indicates a server’s effectiveness.

Similarly, the goal of a passer is not necessarily to pass every ball perfectly. The primary goal of the passer is to avoid getting aced. Of course, we like in-system passes because we score at a higher rate. We work hard to make that happen as often as possible. But we can still score if the pass is out of system. Obviously, a reception error is an immediate loss of a point. One of the statistical items I value highly is the number of reception errors in a set. Generally, depending on the competitive level, the goal for my teams is one reception error per set.

Volleyball is a Team Activity; The Statistics You Keep Can Reflect This
Data that evaluates single-skill and single-player performance is essential. However, I also measure some skills collectively. For example, along with individual passing scores, I approach serve reception as a team activity. I focus on the percentage of in-system passes as a collective group of passers. I value how they work as a collective unit (communication, movement, covering seams, etc). I set goals for the passers as a collective unit.

Connecting The Skills
I was discussing serving in tennis with Ty Tucker, the highly successful tennis coach at The Ohio State University. As with volleyball coaches, Ty places value on serving aces. However, he also places high value on a serve that will result in the server being able to hit a forehand on the return, which means that the location and speed of the serve impacted the return. Tennis, similar to volleyball, has correlated skills. The skills evaluation should not be isolated but correlated to ensuing contacts.

For example, I value digs resulting in kills rather than just the number of digs. A dig that results in a free ball does have value, but not as much as a dig resulting in a kill. So, they should be recorded differently.

I use the same logic in serve receive. A pass that results in a first-ball sideout has a high value. The better the pass, the better the chances for a first-ball sideout. I record each passer’s sideout percentage to provide passer(s) feedback and expand court coverage responsibilities for the passers with the highest sideout percentage. A passer with a high sideout percentage should cover more court in the serve-receive pattern. I want that passer to receive as many serves as possible.

Check the Accuracy of the Information
One of the first criteria for having in-match statistics is the accuracy of the information. Is the coach taking the data, or are the players? Both players and coaches can get distracted during a match and miss plays or enter things inaccurately. As a head coach, I was the worst possible statistician. Equally, I never had players take statistics during a match. I wanted them to focus on court events and ensure they were mentally and physically ready to go on the court. The adage “garbage in, garbage out” certainly applies to statistics.

Noticing Trends Is Really Important!
Coaches should pay special attention to trends in the data. Of course, observing a trend requires a significant sample size. As the season progresses, statistical trends should be noted. If a team or player is not trending up, a review of the training might be in order. An upward trend is also great feedback for the hard-working player.


As I indicated earlier, statistics is not my thing. However, I know that observations, backed by statistics, should impact coaching decisions, future planning and strategies, and player feedback. Coaches should not feel trapped in a statistical format that software kicks out. A coach can review each skill, determine what’s essential at their level with their team, and value the numbers that correlate to success. These components are needed to plan the training sessions for continued improvement and winning results.