EEG microstates before the release: Gu 2022 study

The study recorded the EEG of high-level Chinese archers at rest with their eyes closed, and in the 3 seconds before release during a series of 35 shots at 70 metres. Of the 31 archers recruited, 26 were analysed: 13 from the Chinese national team, the "elite" group, and 13 from the Beijing team, the "expert" group. The elite archers had more years of practice (8.1 against 4.3) and a higher mean score per arrow (8.7 against 6.8). Microstate analysis divides the brain's electrical activity into brief stable configurations, four in this case, and measures their duration, frequency, share of total time and transitions from one to another, within each archer's alpha band.

At rest the elite archers spent more time in microstate D, the one the literature associates with voluntary attention: a share of 26.01% against 19.24%, and 2.68 occurrences per second against 2.07. During aiming the maps differed from the classic resting ones, and the authors interpret them mainly through source localisation: in both groups the most frequent transitions were towards microstate A, located in the left temporal region. In the elite group the duration and share of microstate C during aiming were negatively related to score (r -0.798 and -0.726): less C, better score. In the expert group no measure was related to score.

The EEG had 32 channels, recorded with a portable TMSi SAGA amplifier at 500 Hz on the national team's outdoor range. At rest the archers were recorded seated, 3 minutes with eyes closed and 3 with eyes open; they then shot 35 arrows at their own pace with their own bows, at a 52 x 52 cm target at 70 metres with a 10 ring 10 cm across and each following ring 5 cm wider, being told the result after every arrow. An infrared sensor marked the moment of the shot. The signal was filtered to each archer's individual alpha band, the peak alpha frequency with eyes closed plus or minus 2 Hz, cleaned of eye artefacts with independent component analysis and cut into 2-second windows at rest, 160 seconds per archer on average, and 3-second windows before each shot, 30 per archer on average.

The four microstate classes were obtained by clustering the maps, with their number chosen by a cross-validation criterion, and their sources were estimated with sLORETA. Groups were compared with the Wilcoxon test and FDR correction, and the relationship with score with Spearman correlation. Four archers were excluded for excessive artefacts and one because, with an average of 2.4 points, the authors thought he had not aimed in earnest.

The statistics suggest that the group comparisons were run on individual signal windows rather than on archers: with 13 people per group a Wilcoxon test cannot produce p values of the order of 10 to the minus 12, yet such values appear repeatedly. If so, the differences are much less robust than they look, because windows from the same archer are not independent observations; this is our inference, and the text does not allow it to be confirmed. The correlation with score also comes in two versions: p of 0.001 and 0.005 in the table, 1.1 times 10 to the minus 9 in the text.

There are other inconsistencies: the archer excluded for a low score is called expert13, but in the table the 2.4 average belongs to expert11, and the target described is not the regulation face for 70 metres. The interpretations (dorsal attention network, neural efficiency, suppression of language areas) rest on links between microstates and brain networks established in other studies, and the authors admit that the aiming maps do not match the classic ones. The introduction repeats, without discussion, the claim that 80% of a shooter's performance is psychological. The comparison is cross-sectional and cannot say whether the differences come from training or from something else.

For archers the study offers nothing practical: it measures the brain's electrical activity with a method that needs 32 electrodes and complex analysis, and the results are interpreted through brain networks identified in other contexts. The underlying idea, that stronger archers have a different attentional profile even at rest, is plausible and in line with other EEG work on archers, but here it rests on statistics that probably treat many windows from the same archer as independent.

The finding closest to shooting is the negative correlation between microstate C during aiming and score in the national team archers: an association in 13 people, with no clear account of what that microstate represents while aiming. It does not point to anything to train and does not correspond to a sensation an archer could recognise or change.