Phases of match-play in professional Australian Football: Descriptive analysis and reliability assessment

To examine the frequency and time spent in the phases of Australian Football (AF) match-play and to assess the intra-assessor reliability of coding these phases of match-play.

Observational, intra-reliability assessment.

Video footage of 10 random quarters of AF match-play were coded by a single researcher. Phases of offence, defence, contested play, umpire stoppage, set shot and goal reset were coded using a set of operational definitions. Descriptive statistics were provided for all phases of match-play. Following a 6-month washout period, intra-coder reliability was assessed using typical error of measurement (TEM) and intra-class correlation coefficients (ICC).

A quarter of AF match-play involved 128±20 different phases of match-play. The highest proportion of match-play involved contested play (25%), followed by offence (18%), defence (18%) and umpire stoppages (18%). The mean duration of offence, defence, contested play, umpire stoppage, set shot and goal reset were 14, 14, 10, 11, 28 and 47s, respectively. No differences were found between the two coding assessments (p>0.05). ICCs for coding the phases of play demonstrated very high reliability (r=0.902-0.992). TEM of the total time spent in each phase of play represented moderate to good reliability (TEM=1.8-9.3%). Coding of offence, defence and contested play tended to display slightly poorer TEMs than umpire stoppages, set shots and goal resets (TEM=8.1 vs 4.5%).

Researchers can reliably code the phases of AF match-play which may permit the analysis of specific elements of competition.