Introduction
Throughout the 1980s and 1990s, I used real-time hand notation to record performance in international rugby union games. In the 1980s I used live BBC television broadcasts, in the 1990s I had the good fortune to attend many of the games as a notational analyst.
One of my interests in that period (after playing rugby at Loughborough College, London Welsh and Rosslyn Park) was to provide some evidence about game content and game time, particularly ball in play time, in order to support the coaching of expansive rugby. I was interested specifically in kicking, passing, set piece play and stoppages for injury as indicators of the flow of activity in games.
The most comprehensive set of data I have from that period is the competition between England and Wales 1987 to 1992 inclusive.
Game Content
In these games, the totals for the activities I was monitoring were:
Year |
Kicks |
Passes |
Lineouts |
Scrums |
Pens/FKs Conceded |
Stoppages for Injury |
1987 |
108 |
68 |
53 |
28 |
39 |
10 |
1988 |
94 |
183 |
64 |
27 |
25 |
15 |
1989 |
122 |
102 |
46 |
51 |
20 |
7 |
1990 |
84 |
196 |
40 |
27 |
24 |
10 |
1991 |
114 |
131 |
54 |
34 |
28 |
3 |
1992 |
77 |
182 |
35 |
39 |
36 |
12 |
The odd year games were played in Cardiff and the even year games at Twickenham. Wales won the games in 1987, 1988 and 1989. England won in 1990, 1991 and 1992. Three of the games were affected by rain (1987, 1989, 1992) and there were strong winds in the 1990 game.
Ball in Play Time
For the ball in play time I started my stopwatch at the kick off and stopped it on the referree’s whistle or when the ball had clearly left the field of play. I recorded time in minutes (m) and seconds (s).
Year | First Half | Second Half | Game Total |
1987 |
7m 32s |
9m 24s |
16m 57s |
1988 |
13m 43s |
12m 28s |
26m 11s |
1989 |
11m 08s |
11m 25s |
22m 33s |
1990 |
10m 38s |
14m 11s |
24m 49s |
1991 |
11m 36s |
10m 59s |
22m 35s |
1992 |
11m 14s |
14m 51s |
26m 05s |
These total ball in play times as a percentage of total game time were:
Year |
Ball in Play (BiP) |
Elapsed Time (ET) |
Percentage of Game BiP |
1987 |
16m 57s |
83m 45s |
20.24 |
1988 |
26m 11s |
88m 35s |
29.56 |
1989 |
22m 33s |
85m 40s |
26.32 |
1990 |
24m 49s |
82m 29s |
30.09 |
1991 |
22m 35s |
83m 30s |
27.05 |
1992 |
26m 05s |
89m 12s |
29.24 |
Activity Cycles
In the 1992 game at Twickenham there were 120 activity cycles (defined as play between the referee’s whistles or the start of action without a whistle when the ball was introduced into play, for example, a scrum). Their durations of these cycles were:
Duration of Cycle in Seconds |
First Half Activity Cycles |
Second Half Activity Cycles |
Game Total Activity Cycles |
As a Percentage of all Activity Cycles in the Game (n=120) |
0-4 |
23 |
16 |
39 |
32.5 |
5-8 |
9 |
9 |
18 |
15 |
9-14 |
12 |
10 |
22 |
18.33 |
15-19 |
6 |
6 |
12 |
10 |
20-24 |
2 |
7 |
9 |
7.5 |
25-29 |
4 |
4 |
8 |
6.67 |
30-39 |
4 |
5 |
9 |
7.5 |
40+ |
1 |
2 |
3 |
2.5 |
My hand notations for the 1992 game are:
A PDF copy of this hand notation Actions
and Activity Cycles:
A Pdf copy of the notation ACEW.
Methodological Note
The data presented here are from real-time hand notation. I had been doing hand notation in rugby union since 1980 and by the mid-1980s had established a stable set of game events to notate on a single sheet of landscape A4 paper.I used one sheet per half of the game. The example shared here from 1992 is a transcription of two halves onto one summary sheet. I made a separate record of activity cycles and they are transcribed onto a single sheet here too. In all games I had two stop watches running, one for total game time and the second for ball in play.
During this period I was keen to profile teams in terms of two ratios:
- kicks: passes
- lineouts: scrums
I was very keen to identify those teams that played rugby handball more than rugby football.
I was mindful of the literature on systematic observation in educational studies and physical education and sport. I was a trained observer and was confident that my observations were valid and reliable. My aim was to provide real-time information to coaches if required.
I did not undertake any intra- or inter observer reliability studies of the data presented here. I was aware of Paul Croll’s (1986: 154) argument that:
Stability of observations depends primarily not on characteristics of the observer or observation system but on the naturally occurring patterns of whatever is being observed … It seems unsatisfactory that an observation procedure that provides a highly accurate description of classroom events should be described as unreliable. In some cases the extent to which a characteristic is a stable feature of individuals and the extent to which it varies for different people may be of interest to the researcher and is itself a focus of analysis rather than a constraint upon whether the data are sufficiently reliable. (Systematic Classroom Observation. Lewes: The Falmer Press.)
In 1987, the ball was in play for less than 21% of the total game time. Given the relatively small number of game events I was monitoring there was ample time to record events. I was very aware that using a temporal measure required close attention to accuracy in starting and stopping the stopwatch. I was aware of the potential of mathematical error of using stopwatches. I used new batteries for each game recorded and prior to all games compared the performance of watches.
Throughout my real-time notation I was conscious of observer drift. No game lasted longer than ninety minutes and no half of a game was longer than 48 minutes. I used breaks in play and half time to refocus my attention. Real-time hand notation does require concentration and as a trained observer I felt comfortable with the cognitive load of the activity. I had clear operational definitions for all items to be notated and had made a very conscious decision not to try to capture granular details in real-time. With improvements in video technology in the late 1980s and 1990s I was able to undertake detailed lapsed-time analysis of performance for research purposes.
Conclusion
This post presents some of the data I captured in real-time twenty years ago. My overall aim was to develop a notation system that might offer decision support to coaches within games and subsequent opportunities to reflect on performance. During the period discussed here I immersed myself in the literature on observation. I was fortunate that at that time I was researching and writing Using Video in Sport (1988. Huddersfield: National Coaching Foundation) and completing a PhD that used ethnographic methods to record observations.
There is a story within a story here. By coincidence the data presented chart the rise of the England rugby team. In 1987 Wales defeated England in the 3rd/4th play off game at the inaugural World Cup. In 1991, England won in Cardiff for the first time since 1963 (there had been a draw in 1983) and at the end of the year contested the World Cup final against Australia at Twickenham.
Photo Credit
The National Stadium Cardiff (Walt Jebsco, CC BY-NC-ND 2.0)
[…] A transcript of the hand notation is available for download here 1991 RWC Final. The methods used are described in this post. […]