THINKING CRITICALLY ABOUT MOMENTUM
by: Paul Arrington, M.D.
Having found that calling time out at the elite level of the sport is not effective in stopping the opponent’s “momentum”, it became apparent that it was time to try to more clearly define the concept of “momentum” and the roles emotions and random events might play in the development of this concept.
“Momentum” is thought of by most coaches as: if a teams scores one or more consecutive points then that team is more likely to score on the next attempt than if it had lost the preceding point. It is similar to the “hot hand” concept in many sports, and would correspond to positive momentum. Conversely, negative momentum would be when loss of a point would indicate that the next point would probably also be lost.
Although where there are two opposing forces (teams) positive momentum and negative momentum would both have to occur simultaneously, coaches tend to be much more consciously aware of positive momentum than they are of negative momentum. In physics, momentum is a measure of energy derived from mass times velocity. To stop an object’s momentum then requires a level of energy equal to or greater than the kinetic energy the object possesses as a result of its momentum.
In discussing “momentum” with three top level Women’s volleyball coaches (several NCAA DI championships among them), they all had a somewhat different idea about what momentum meant. However, they all agreed that “momentum” is not what TV announcers and most coaches commonly mean when they use the term. The three coaches thought that having a run of successes perhaps would relate to an emotional component that would:
- Result in an improvement in team defense in later plays
- Lead to successes later in the match
- Result only from a major come-from-behind win that would carry over into the next game of the match.
Everyone watching volleyball can see that emotions are intimately involved with “momentum.” It has always been thought that the positive emotions are a cause of the positive momentum and that negative emotions result in negative momentum. From a different perspective, the other possibility that is not usually considered, is that the emotions are results of “momentum” and not the cause.
People in general, and coaches in particular, are very uncomfortable with the concept of randomness. We like to have explanations for everything that happens in our lives, and we will accept some explanation of a random occurrence whether the explanation is based in fact or not. We think that “we believe what we see,” when in fact most psychologists understand that “we see what we already believe.”
To best envision the role of randomness in athletics let us take a child who has never shot a basketball. The goal for this child is to make free throws. During his initial attempts his lack of skill and strength will result in perhaps a 1 out of 10 success rate (about equal to Shaquille O’Neal). During this phase it is fairly obvious that his attempts are fairly randomly distributed in the general area of the rim. As years pass and he improves with age and practice, he may begin to hit more and more until he is able to make 9 out of 10 attempts. At this time the random nature is not as apparent, but, if we were able to track an imaginary marble in the center of the basketball as it is shot toward an imaginary small (1 inch) circle in the middle of the hoop, we would still see a very random distribution around the circle. All the athlete has done by practice is to decrease his area of randomness. Continued practice will continue to decrease the area of randomness until none is apparent to the eye – the randomness still, however, persists and would be seen if the ball and target were both reduced to dot size.
This concept of practice decreasing the area of randomization is represented on the next page (Figure 1) using random numbers generated by a random number generator and sequentially decreasing the area of randomization from 1 to 256 to 1 to 16 to represent the impact of practice. Although “practice” increases the numbers within the circle, even then, the red dots are still randomly distributed around a central point inside the circle.
|
|
|
|
|
IMPACT OF PRACTICE ON RANDOM DISTRIBUTION
Figure 1
To help clarify the role of emotions and random events in causing “momentum,” the following games were exhaustively examined by recording the number and length of all scoring runs from a single point to eight in a row, and by determining the percentage of points scored by side outs and by service:
- Seven matches from the 2004 Women’s’ NCAA Division I Championship Tournament (1195 points)
- Six matches from NCAA Division I Men’s early season 2004 (1424 points)
- Eight matches from the 18 and under 2006 Volleyball Festival Championship Division (865 points)
- Eight matches from the 13 and under 2007 Volleyball Festival Championship Division (882 points)
All matches examined were between equally matched teams as determined by minimal total point differences over all matches.
If “momentum,” as it is commonly conceived, is a real phenomenon:
- Then scoring a point or a run of points will increase the chances of scoring on the next opportunity, thus one will expect to see larger and more runs than would be expected on the basis of random distribution.
- Improvement in team defense in later plays will occur, this will also increase the size and number of runs higher than expected those based on random distribution
- Winning the first game of the match will increase the chances of winning the match more frequently than seen on the basis of random distribution.
- A major come-from-behind win carries over and will increase the chances of winning the next game of the match at a rate higher than expected from random distribution.
Examination of the NCAA matches (combined men’s and women’s) showed point runs as follows:
- Single isolated points 43%
- Runs of 2 points 31%
- Runs of 3 points 14%
- Runs of 4 points 8%
- Runs of 5 points 3%
- Runs of 6 points or more 1%
Based on the NCAA teams’ observed percentage of scoring on serve reception and on service, using a random number generator, “volleyball games” were played. The point runs in these “games” are as follows:
- Single isolated points 42%
- Runs of 2 points 31%
- Runs of 3 points 15%
- Runs of 4 points 8%
- Runs of 5 points 3%
- Runs of 6 points or more 1%
Based on the percentage of scoring on serve reception and on service from the actual matches, a mathematical formula was used to predict the point runs based on random distribution. The predicted point runs are as follows:
- Single isolated points 42%
- Runs of 2 points 30%
- Runs of 3 points 16%
- Runs of 4 points 8%
- Runs of 5 points 3%
- Runs of 6 points or more 1%
A graphic comparison of these three sets of data is seen in Graph 1.
Comparison of Scoring Runs In Real and Random Number Games With Mathematical Predictions
Evaluation of the 18 and under as well as the 13 and under matches revealed results similar to those found from the collegiate matches. The point runs in the real matches are virtually identical with those predicted from mathematics and from random number generators.
Now let’s look at some other factors that are considered by some as “momentum” builders and see if they impact outcomes. Some of these factors that will be studied are: 1) the impact of winning the first game of a match; 2) the impact of being the first team to reach 10 points and the first team to reach 20 points; and 3) the impact of coming from behind to win the game.
From the 13 NCAA matches, the team that won the first game won the match 10 times (77%), whereas from the twenty random number games, the team that won the first game won the match 15 times (75%).
For the first team to reach 10 points, the NCAA teams won 34 of 47 (72%), whereas the random number game “teams” won 35 of 50 (70%). For the first team to reach 20 points the NCAA teams won 41 of 47 (87%) and the random teams won 43 of 50 (86%).
There were only 4 NCAA matches to evaluate in which a team came from behind late in the game (after 20 points) to win. Of the four teams to come from behind late in the game to win (from scores of 19 –21 to 30-24; 21-22 to 30-27; 26-28 to 32-30; and 26-27 to 36-34), only 1 of the four teams won their next game. It was not felt necessary to evaluate the random number games in this regard and with the small sample size the results would have lacked statistical significance.
All of the above data indicate very strongly that “momentum” is not a real event, but is simply our perception of random events and is a result of our desire to see patterns and causes even where none exist. Additionally the observed emotional changes during the game are more likely the result of a run of points rather than a cause of the run.
As mentioned earlier, people are uncomfortable with random events and have a strong need to explain events that happen in our lives. In addition to the need to explain events in our lives, the other difficulties in our accepting random events are:
- Persisting belief in superstitions and streaks
- Persisting belief that chance should not be “lumpy,” i.e., that bunches should not appear in random sequences
- Persisting belief in determinism
- Persisting in seeing “patterns” in random events where none exist.
The findings presented here are consistent with much research, which has been done on the so-called “hot hand” phenomenon–it has never been found. Tom Gilovich of Cornell University presented material in this regard in the 1980’s carefully evaluating the “streak” shooting in professional basketball. Despite the complete absence of findings to support the concept of “momentum,” one of the early researchers in this field reported that he had been in thousands of arguments on this topic, won them all, but had convinced no one. We do not give up our beliefs easily, as demonstrated from the following exerpt from Dr. Gilovich’s book, “How We Know What Isn’t So”:
Red Auerbach, the brains behind what is arguably the most successful franchise in American sport history, the Boston Celtics, had this to say upon hearing about our results: “Who is this guy? So he makes a study. I couldn’t care less.” Another prominent coach, Bobby Knight of the 1987 NCAA champion Indiana Hoosiers, responded by saying “….. there are so many variables involved in shooting the basketball that a paper like this really doesn’t mean anything.” These comments are not terribly surprising. Because a truly random arrangement of hits and misses contains a number of streaks of various lengths, the belief in the hot hand should be held most strongly by those closest to the game. Furthermore, simply hearing that the hot hand does not exist, or merely taking another look at the game is not sufficient to disabuse oneself of this belief. It is only through the kind of objective assessment we performed that the illusion can be overcome.
In discussing the findings of this research with other volleyball coaches, their responses fall into one of two categories. One group persist in their beliefs, despite solid evidence to the contrary. The others ask, “So what, what does it mean for coaching?”
It is difficult to overstate the importance of accepting the fact that the outcome of a match is a result of talent, preparation, and match up of random events between the two contestants and not a lack of effort, desire, focus or some other nebulous concept. With small numbers, 25 or 30 for example, the weaker team will sometimes win due to the impact of random events. However, if the games were played to 1000, the better team will win nearly every time.
By not taking into account the impact of random distribution, athletes are being abused when coaches tell them: “You just didn’t want it bad enough!” “You didn’t stay focused.” “You let them get away from you.” Or even worse like Pat Knight SOB (son of Bob) at Texas Tech, who ran his team to the point of exhaustion after a lopsided loss. He attributed their huge win the next night to the positive impact of his methods. He failed to report, however, the reasons for the teams nearly 50 point loss in the subsequent game.
If coaches can began to accept that a major determinant of outcome is the impact of random events, not only will they be less likely to abuse their athletes, but they also may be less hard on themselves when the team is not as successful as they would have liked.












