Emotion and arousal in archery events: Medeiros 2008

Three Brazilian archers were followed through competitions for an entire season, marking on a grid after each end how activated they felt and how good they felt, while a heart rate monitor recorded their beats. They were 24, 27 and 57 years old, with a mean of 13.7 years of practice, trained 12 hours a week for 44 weeks a year, and all three were among the three best Brazilian archers in their category.

For each archer, and for each shooting distance, the authors split the ends into thirds by score and then used ordinal logistic regression to derive the zones of arousal, pleasure and heart rate associated with the best third. The zones differ from one archer to another and from one distance to another. At 70 metres, for instance, the optimal arousal zone ran from 4.7 to 5.9 for the first archer, from 5.0 to 6.6 for the second and from 6.1 to 7.15 for the third, on a 1 to 9 scale. One thing, though, held for all of them: with reported arousal very low (1, 2, 3) or very high (8, 9) none of them landed in the best third, and performance was already falling below 4 and above 7.

The same pattern returns for pleasure and for heart rate. One interesting exception is the first archer, whose pleasure zone was the same at four outdoor distances (30, 50, 70 and 90 metres) and different and wider in the 70-metre elimination round and the 18-metre indoor round. The authors also note that pleasure does not run in one direction only: unpleasant zones could go with the best ends.

The third finding is about sequence. An end in the best third did not make the next one more likely to be: in 67.6%, 74.3% and 67.3% of cases no other followed, and runs of three or more optimal ends covered between 9 and 17% of the season's ends. All three archers moved in and out of their own zones over the season and within a single competition day too.

The data come from a whole season, from March to December 2006, in a state in south-eastern Brazil, with at least forty readings of arousal, pleasure, heart rate and score for each archer at each distance. All three shot the state outdoor championship; two also competed at the Pre Pan-American and one at the South American championship, so the number of data points collected is not the same for the three.

Affect is recorded with the Affect Grid, a nine-by-nine grid on which a single mark gives pleasure (from 1, displeasure, to 9) and arousal (from 1, sleepiness, to 9). Performance is the total score for the end, written by the archer on a sheet immediately after shooting, at the same moment as filling in the grid. Heart rate comes from a Polar S610i monitor with a chest belt, which averages the rate every 5 seconds; the values are then averaged over each end, so over 4-minute series at 90 and 70 metres. To express heart rate in relative terms, the authors estimated each archer's maximal rate in a laboratory using the Bruce protocol, an incremental test to exhaustion.

The analysis follows a probabilistic method published earlier. For each distance the ends are split into poor (first to thirty-third percentile), moderate (thirty-fourth to sixty-sixth) and optimal (above the sixty-sixth); the poor and moderate ends are then separated according to whether they fall below or above the mean of the values recorded in the optimal ones, giving five categories. Ordinal logistic regression is applied across these five categories, returning for each level of arousal, pleasure and heart rate the probability of falling into each. Arousal, pleasure and heart rate are treated separately, with no search for interactions between them.

The sample is three archers, all men, and the design is openly idiographic: the authors write that they do not intend to generalise the results to the archery population. The zones are built around each individual and do not transfer to anyone else.

Optimal is not an absolute level, it is the best third of that archer's ends at that distance. The performance table shows this clearly: at 90 metres the first archer's best third sits between 78 and 83% of the maximum possible score, while at 18 metres it runs from 94 to 100%. This definition also matters for the finding about sequences: if by construction about a third of the ends are optimal, then an optimal end failing to bring another two thirds of the time is what independent events would produce, and independence is exactly what the authors conclude.

The authors themselves flag three limitations: the grid was filled in immediately after the end, so with the score already known, which may have coloured the judgement of arousal and pleasure; interactions between affective and physiological states were not examined; and stability is described rather than quantified. To these can be added that the scores are written down by the archers themselves, and that heart rate is the average over a four-minute end, so it says nothing about what the heart was doing at the moment of release. In the analysis section the list of distances leaves out 50 metres, which nevertheless appears in the performance table.

The study describes a way of looking at your own state in competition, and for someone who shoots the interesting point is that the way is personal. The three archers had arousal zones that differed from one another, and differed by distance: the number that went with one archer's best ends was not the other's. The same holds for heart rate, which in the same archers shifted zone from one distance to the next.

Beyond the individual differences one common boundary remains: none of the three shot their best ends when reporting themselves very flat or very charged. That is an observation on three people, not a threshold, and it does not show that getting into the right zone produces better scores: the direction of the link was not tested, and the state was marked down after the score had been seen.

Two things matter for anyone keeping a competition diary. The first is that a good end does not announce the next one: in two cases out of three, in these three archers, an end in the best third was followed by something else, and runs of three or more covered a small share of the season. The second is that the state does not hold still, not even within a single competition day, from which the authors conclude that one measurement before or after a competition is not enough to describe what happens during it. None of the three, finally, was given any intervention: the authors leave to future work the attempt to use these zones to change anything.