# The First Goal – How Much of an Advantage?

Introduction

Sometime in the mid-1980s, an interesting football theory was put forward by a family member: the first goal curse. Distilled to a more workable format, the theory was that the team that kicks the first goal of a AVFL/AFL match will tend to lose. I immediately rejected this concept, arguing that kicking the first goal of a match can only help that team’s chances of winning and that, in the long-run, the team with the first goal on the board should win more often than not.

I have long entertained the idea of examining data on this aspect to assess the impact that a team kicking the first goal of a match has on their chances of winning. Now is my chance. In the time available to me, I have been unable to explore the first goal aspect for all VFL/AFL matches since 1897. Instead, I have examined all matches from the 2008 season onwards. The 2008 season was used as my starting point because the website afltables.com, a veritable statistical goldmine, sets out the scoring progression for each match from round 1, 2008[1]. The results make for some interesting reading.

The article begins by setting out the first goalscoring results for each season. The focus here is on general trends in first goalscoring data, including how the first goalscoring data has varied over time and the first goalscoring statistics for finals matches. The next section examines the relationships between first goalscoring data and overall performance. The article will then proceed to concentrate on the data for particular clubs and key trends. Finally, we will look to draw inferences from this data that could be of practical use.

1. FIRST GOALSCORING RESULTS FOR EACH SEASON

Table 1, below, sets out the number of times each club scored the first goal in each season. After this entry for each season, the number in brackets denotes the number of games won by the team when they scored the first goal. To reduce clutter in the tables, I have ignored draws in the individual season-by-season boxes for each club, although draws are set out in each club’s overall record for 2007-2014.

To get a feel for how each club performed in a season, I have shaded the season statistics as follows:

Green – Premier (1st)
Blue – Runner up (2nd)
Purple – Eliminated in the Preliminary Final (3rd)
Orange – Eliminated in the Semi Finals (5th)
Red – Eliminated in the Elimination Finals (7th)
Plain text – outside the finals, but not in the bottom four
Black bold – Bottom four

Table 1: First goalscoring results for each season

Overall, in the 1,373 AFL matches played from 2008 to 2014, the team that scored the first goal in a match won 812 matches. This is a success rate of 59.14 %[2]. A long way from the first goal being the ‘Kiss of Death’!

The winning percentage of opening goalscoring teams fluctuated between approximately 56 % and 64 % over the seven season period. The winning percentage peaked at over 64 % for the 2014 season. Table 2 sets out the winning percentages for each season, including a moving three season average winning percentage.

Table 2: Winning percentage of first goalscoring teams – All games

The moving average of the winning percentage for first goalscoring teams demonstrates that there has been a gradual increase in this statistic in recent seasons. This could be attributed to some degree to the impact of the new clubs, Gold Coast and Greater Western Sydney, who battled to have an impact in their first two or three seasons in the competition. It is arguable that the gulf between the stronger teams and the lesser teams has widened in recent seasons. If this is the case, then it is a reasonable proposition that the stronger teams are more likely to score the first goal and to go on to win a game than would be the case in a season that is more even, thus bolstering the winning percentage of first goalscoring teams. Table 3 explores some measures of the evenness in the competition in the 2008-2014 period.

Table 3: Measures of difference between stronger and lesser teams.

This table demonstrates a shift towards “dominant teams” in a season, that is, teams that have won at least 16 games in a home and away season. The number of “cellar dwellers” – teams that won no more than six games in a home and away season – usually hovered around three, apart from a spike of five teams in 2012, which was GWS’s first season in the competition and Gold Coast’s second season. The data regarding the win-loss record needed to qualify for the finals is inconclusive. A record of 12-10 has usually been enough and the increase in the dominant teams has not resulted in a sustained uplift in the win-loss record needed to make the eight. A notable exception was GWS’s debut season of 2012, where North Melbourne’s record of 14-8 only netted eighth position after the home and away rounds. In that year, St Kilda’s record of 12-10 and a percentage of over 123 % landed them in ninth place. In most seasons, such a performance would be enough to make the finals or perhaps snag seventh position.

The most decisive evidence of the widening gulf between the stronger teams and the lesser teams is the large increase in the number of 100 point blow-outs. There were only 10 of these results in the 528 home and away games played in seasons 2008-2010, a proportion of less than 2 % (1.89 %). In contrast, following the introduction of Gold Coast in 2011, seasons 2011-2014 produced 34 home and away games with a margin of at least 100 points from the 781 home and away games played. The proportion of these blow-outs is over 4 % (4.35 %), more than double the proportion for 2008-2010.

This data provides cogent evidence of a greater disparity in recent seasons between the stronger and weaker teams. A reasonable inference may be drawn that this trend is a major factor in the overall increase since 2011, especially from 2012 onwards, in the winning percentage of teams that opened the goalscoring in a match.

An interesting aspect of this analysis is first goalscoring results in finals. The winning percentage of first goalscoring teams is examined in Table 4.

Table 4: Winning percentage of first goalscoring teams – Finals

For finals, the winning percentage is 53.13 % over the period. This is more than 6.0 % lower than the winning percentage for all games. It is understandable that the extent of the first goal advantage is less for finals because of the higher overall standard of the competing teams and a smaller difference in class between the best team and the worst team. An interesting facet is the wide range of results, with the winning percentage ranging from approximately 33 % to 80 %. Again, this can be attributed to a smaller difference in the quality of the competing teams. It is a reasonable argument that, where two teams are evenly matched, there will be a greater degree of uncertainty of the first goalscoring team going on to win the match than where there is a vast difference in the quality of the combatants.

In the eight Grand Finals played in the period, the first goalscoring team has won four times[3], lost three[4] and drawn one[5]. The closeness of the win-loss ledger of 4-3 in this sample of Grand Finals is consistent with the theory that the winning percentage of first goalscoring teams should be lower, or closer to 50 %, when the two competing teams are closer in overall capability. As the events of VFL/AFL Grand Finals over the years are reasonably widely documented, it would be an interesting analysis to examine a large data set of Grand Finals and to consider the win-loss record of first goalscoring teams. I will leave this analysis for another day.

2. RELATIONSHIP BETWEEN FIRST GOALSCORING AND OVERALL PERFORMANCE

The starting point of this section is, for each season, the clubs that were ranked in the top eight for the number of games in which they scored the first goal. This analysis is set out in Table 5, which applies data from Table 1. Table 5 examines how many of these top eight ranked teams made the finals. In terms of ranking clubs in the first goalscoring tables, if two or more clubs scored the first goal in the same number of matches, then the teams were first ranked by the number of games won where the first goal was scored. If two or more clubs were also equal on this criterion, each team’s ladder position after the finals was used to determine their ranking for this purpose.

Table 5: First goalscoring teams – Top eight rankings in each season (all matches)

This table shows that success in first goalscoring has a strong correlation with overall team success: 44 out of the 56 teams (78.57 %) that ranked in the top eight first goalscorers in 2008-2014 also made the finals. Another illustration of this trend is that there were only two teams that finished in the top four (after the finals) – Western Bulldogs (4th in 2009) and Hawthorn (1st in 2014) – that did not make the top eight rankings for first goalscoring. In other words, 26 out of 28 top four finishers in the period finished in the top eight first goalscorers. In relation to premiership sides, the first goalscoring rankings were: 2008 – Hawthorn – 8th; 2009 – Geelong – 1st; 2010 – Collingwood – 1st; 2011 – Geelong – 2nd; 2012 – Sydney Swans – 5th; 2013 – Hawthorn – 1st; 2014 – Hawthorn – 9th.

A flaw in this analysis is that it favours teams who have played in the finals, as such clubs have more games from which to score the first goal. If the data is limited to home and away games, then the revised results are as set out in Table 6.

Table 6: First goalscoring teams – Top eight rankings in each season for home and away games.

When the data is limited to home and away games and a level playing field is achieved, the correlation is less: 38 out of 56 finals places, or 67.86 %, compared to 44 out of 56 finals places, or 78.57 %, when all premiership matches are considered. The percentage has fallen by nearly 10 percentage points, although the percentage of top eight first goalscoring teams that have made the finals of 67.86 % is still reasonably strong. For home and away matches only, the link between top eight first goalscoring finishes and top four placings after the finals is significantly weaker. There were seven out of 28 (25 %) top four finishers that did not also make the top eight rankings in first goalscoring in the relevant seasons: Hawthorn (1st in 2008), St Kilda (4th in 2008), Western Bulldogs (3rd in 2009), Hawthorn (3rd in 2011), West Coast (4th in 2011), Geelong (3rd in 2013) and Port Adelaide (3rd in 2014). Premiership teams achieved a good level of success in first goalscoring. The results for premiership teams were: 2008 – Hawthorn – 10th; 2009 – Geelong – 1st; 2010 – Collingwood – 2nd; 2011 – Geelong – 3rd; 2012 – Sydney Swans – 5th; 2013 – Hawthorn – 1st; 2014 – Hawthorn – 7th. Overall, these results are approximately the same as the results for premiership team rankings for all matches.

Another measure of the relationship between first goalscoring data and overall performance is to examine the variations for each season in each club’s first goalscoring ranking and their final ladder position. For each season, the total amount of variation, or the variance, will be calculated and will be divided by the number of clubs to determine the average variance, or the standard deviation. Applying statistical principles, the amount of the variation for each item, rather than whether the variation is positive or negative, will be examined. The standard deviation will be a measure of the connection between first goalscoring performance and overall performance. To achieve a level playing field with this analysis, only home and away games have been counted. Of course, the lower the standard deviation for each season and for the 2008-2014 period as a whole, the stronger the connection between first goalscoring data and overall performance.

Table 7 sets out the data. The standard deviation is also worked out for each club to get a picture of how the club’s first goalscoring performance tracks with its overall performance.

Table 7: Analysis of the correlation between first goalscoring rankings (home and away only) and overall performance.

Legend:

Club name abbreviations are self-explanatory

20XX FGR = First Goalscoring Ranking (FGR) for season 20XX

20XX VAR = Variance between FGR and final ladder position for season 20XX, equals FGR minus final ladder position

SD = Standard deviation

This table shows that there is only a mild connection between a club’s first goalscoring ranking for home and away matches and their final ladder position. The average variance was 3.42 ladder positions, which provides some predictive value but is not overly precise. For the statistically minded, looking at all variances as positive values, the median variance is 3. This indicates that the high variances have only had a small effect in skewing the average figure of 3.42.

There are 119 data items of variance, which can be distilled as follows to get a better sense of the relationship between first goalscoring ranking at the end of the home and away season and final ladder position:

FGR and final ladder position the same = 19 (15.97 %)
FGR and final ladder position differ by 1 to 5 places = 73 (61.34 %)
FGR and final ladder position differ by 6 to 10 places = 23 (19.32 %)
FGR and final ladder position differ by 11 or more places = 4 (3.36 %)

Applying these results, a team’s first goalscoring ranking at the end of the home and away season has been a correct predictor of that team’s final ladder position about 16 % of the time, or about once from six cases. If we examine the data in the 1 to 5 places further, we find that a team’s first goalscoring ranking will be the same as, or one, two or three places different from, the team’s final ladder position in 69 out of the 119 instances. In other words, based on the 2008-2014 data, if you took a club’s first goalscoring ranking at the end of the home and away season, then the club will finish within three ladder positions of that ranking 57.98 % of the time. This is slightly more than four times out of seven, better than even money odds.

The performances of one traditional club – Melbourne – in first goalscoring rankings have very closely tracked the final ladder positions. The first goalscoring rankings of new teams Gold Coast and Greater Western Sydney have also been close to the final ladder positions. There was also a reasonably correlation of results, between a standard deviation of 2.00 and 2.50, for Adelaide, Collingwood, Geelong and the Brisbane Lions. Richmond is the club with the greatest disparity of results. Surprisingly, Hawthorn’s first goalscoring rankings were markedly different from their final ladder positions, producing a standard deviation score of 4.57.

4. Club data – Comparison between first goalscoring performance and overall performance

This part of the paper carries out a more detailed analysis of each club’s first goalscoring results to their overall performance.

Table 8 sets out the following items of information:

A) The number of games in which the club scored the first goal. This information is lifted from the column on the extreme right in Table 2. To improve comparability, the totals for Gold Coast and Greater Western Sydney will be adjusted[6] so that their results are “extrapolated” over a period of seven seasons (2008-2014). This also applies for item (c), below;

B) Based on the data in (a), the overall club’s ranking in first goalscoring for the period. In the event of two or more clubs scoring the first goal in the same number of games, the number of games won will be the tiebreaker;

C) Home and away games only – the number of games in which the club scored the first goal. The total number of home and away games played in the period is 88 for Gold Coast, 66 for Greater Western Sydney and 154 for all other clubs;

D) Based on the data in (c), the club’s overall ranking in first goalscoring for the period. In the event of two or more clubs scoring the first goal in the same number of games, the number of games won will be the tiebreaker;

E) The club’s overall performance tally. This is simply the total of all ladder positions in the period. Naturally, the lower the total, the better. To improve comparability, the totals for Gold Coast and Greater Western Sydney will be adjusted[7] so that their results are “extrapolated” over a period of seven seasons (2008-2014). The best possible overall performance tally is seven and the worst possible overall performance tally is 119 (16th in 2008-2010, 17th in 2011 and 18th in 2012-2014);

F) The club’s ranking in the overall performance tally. In the event of a tie, the number of matches won will be used to separate the teams; and

G) The variance, or difference, between the club’s first goalscoring ranking for home and away games and the club’s overall performance ranking. This will be used as a point of comparison later in relation to the link between first goalscoring in finals matches and overall success.

Table 8: Club results: Comparison between first goalscoring performance and overall performance.

As a starting point in looking at Table 8, the first goalscoring rankings for all premiership matches are quite similar to the first goalscoring rankings for home and away games. The key takeaway from the table is that there is a strong level of similarity between the first goalscoring rankings for home and away games and the overall performance rankings. Indeed, this is a major conclusion in this analysis as a whole.

We will now take a closer look at the data in the table and draw out more important observations.

Firstly, for 14 out of the 18 clubs, the first goalscoring ranking for home and away games was within three placings of the club’s ranking in the overall performance tally. If we expanded the analysis to first goalscoring for all premiership season matches, 17 out of the 18 clubs had first goalscoring rankings that were within three placings of the club’s ranking in overall performance. Indeed, for all premiership season matches, four clubs had the same overall first goalscoring ranking and ranking in overall performance.

Secondly, 11 out of the 18 clubs had an overall first goalscoring ranking in home and away matches that was within two placings of the club’s ranking in overall performance. This figure improved to 12 clubs when the data for all premiership season matches was analysed.

Next, the club with the biggest discrepancy in rankings was clearly Richmond. Their overall first goalscoring ranking was sixth, both in all matches and home and away matches only, in sharp contrast to their overall performance ranking of 14th. North Melbourne also had a significant variance: an overall first goalscoring ranking of fifth, improving to third in home and away games, and an overall performance ranking of eighth. Western Bulldogs and Carlton had mild differences in their rankings. The Bulldogs ranked 16th in overall first goalscoring for home and away games (13th in all matches), but their overall performance ranking was six placings higher at 10th. Carlton were placed 11th in overall first goalscoring rankings for home and away matches (10th for all matches), but their overall performance ranking for the period was seventh.

An important feature of the club results table is that the five most successful clubs in 2008-2014 filled five out of the top six club rankings in first goalscoring (using home and away games only or all premiership matches). The five most successful clubs of the period are the premiership teams of Hawthorn (2008, 2013 and 2014), Geelong (2009 and 2011), Collingwood (2010) and the Sydney Swans (2012), plus St Kilda, which finished runner-up in 2009 and 2010. North Melbourne were the exception, finishing third in the home and away first goalscoring rankings, fifth in first goalscoring for all games, but only eighth in their overall performance ranking.

Applying the home and away basis for overall first goalscoring rankings, the positions for the top six first goalscoring clubs were: Collingwood (1st, 3rd in overall performance), Geelong (2nd, 1st in overall performance), North Melbourne (3rd, 8th in overall performance), Hawthorn (4th, 2nd in overall performance), St Kilda (5th, 5th in overall performance) and Sydney Swans (6th, 3rd in overall performance).

The key takeaway is that, in the long-term, clubs that frequently score the first goal of a match tend to be the most successful clubs.

For clubs in positions seven to 13 in the home and away first goalscoring rankings, there were moderate variations between the first goalscoring ranking and the overall performance ranking. Three out of these seven teams had ranking variances of two or less. At the bottom end of the spectrum, three out of the five clubs ranked 14th to 18th in home and away first goalscoring rankings were also ranked in the bottom five in overall performance , with Port Adelaide and the Western Bulldogs being the exceptions. The bottom two clubs in home and away first goalscoring, Gold Coast and Melbourne, ranked 16th and 17th, respectively, in overall performance. Consequently, we may conclude that the clubs that are the least frequent first goalscorers tend to be worst performed clubs overall.

5. Club data – First goal in finals and correlation to overall performance.

An interesting dimension of the first goal data is whether first goalscoring in finals matches correlates with overall team success and the strength of this correlation compared to the corresponding correlation for home and away games.

Table 9 sets out, for the clubs that qualified for at least one finals series in the period, their first goalscoring data in finals games and their overall performance. Gold Coast, Greater Western Sydney and Melbourne are not included in this table as they did not play in a finals match in 2008-2014. In determining the first goalscoring ranking for finals where teams have scored the first goal in the same number of finals, the number of finals won on these occasions will be the first tiebreaker. If two or more clubs have won the same number of finals where they also scored the first goal, then the teams with the least number of finals played will rank higher. If there is still a tie, then the number of matches won overall in the period will be used to separate the teams.

The finals ranking is based on the number of finals won and, in the event of a tie, the winning percentage. If teams have the same winning percentage in finals, then they will be separated by the number of finals played, with the higher number of finals played, the better. If teams are still level, then the team’s best result in finals will be used as the determinant of finals ranking.

Table 9: First goalscoring in finals.

The finals rankings for teams in first to sixth positions are very similar to the overall performance rankings for these six clubs. This is a predictable outcome, as winning a high number of finals games is conducive to winning a premiership. Unlike the old Final Four and Final Five systems, clubs in recent times have to win three finals if they finished in the top four to claim a premiership. Four finals wins on the trot is necessary if a team finishes between fifth and eighth. This has yet to be achieved since the current finals system was implemented for the 2000 season. Adelaide won four consecutive finals in their premiership year of 1997 under a different version of the final eight system.

An interesting aspect is the strong correlation between first goalscoring ranking in finals and finals ranking. For four out of the 15 clubs that saw finals action in 2008-2014, the respective rankings were the same. The maximum variance in rankings was three (for three clubs), the total variance was 20, producing a standard deviation of 1.33. This is significantly lower than the standard deviation of 2.56 positions in relation to the correlation between first goalscoring rankings for home and away games and overall performance ranking: please refer to the extreme right column in Table 8.

Next, we will examine the nexus between first goalscoring rankings in finals and overall performance. This analysis does not include the bottom three clubs in overall performance ranking – Gold Coast (16th), Melbourne (17th) and Greater Western Sydney (18th) – as these clubs did not play in a finals match in the period. In any case, their exclusion from this part of the exercise is neutral because the three clubs occupy the bottom three positions in overall performance. The variance column in the extreme right hand column of Table 9 shows how the first goalscoring rankings for the 15 clubs in final compare to their overall performance ranking. The differences in the respective rankings were mild. Only three clubs – Adelaide, Port Adelaide and West Coast had major ranking disparities. The standard deviation result of 2.20 means that there is a reasonably close connection between a club’s first goalscoring ranking in finals matches and the club’s overall performance ranking. Rounding numbers, this means that, on average, a club’s first goalscoring ranking in finals will be within two places of its overall performance ranking. The standard deviation of 2.20 for finals games produced a moderately lower difference of rankings, or a stronger correlation, than the relationship between first goalscoring ranking in home and away games and overall performance ranking, where the standard deviation was 2.56.

An interesting dimension of this analysis is the performance of the top six clubs in the period: Geelong, Hawthorn, Sydney Swans, Collingwood, St Kilda and Fremantle. This set of six clubs are the top six clubs in the overall performance rankings. The set covers all seven premierships won in the period and, indeed, all runners-up in the period have come from this group. For this group of six clubs, the total variance in first goalscoring rankings for home and away games compared to overall performance rankings (see Table 8) was only 12, producing a standard deviation of 2.00. For finals matches only (refer to Table 9), the total variance for the six clubs was a mere eight, resulting in a standard deviation of 1.33.

The key takeaway from this result is that the clubs that are consistently the best at scoring the first goal have usually the best performed clubs in the period (with the exception of North Melbourne). This tendency is stronger in finals matches. The results for clubs out of the top six have been a mixed bag. However, the teams that score the first goal in a match less frequently tend to be worst performed teams overall.

6. Inferences from the data

The first goalscoring data for seasons 2008 to 2014 supports the findings set out below.

A) Teams that score the first goal have a sound success rate overall, approximately 59.14 % over the period. In the last three seasons, the winning percentage has been 61.84 %. For a betting person, the longer range percentage of 59.14 % could be applied and translated to a price of \$1.69. This could be used as a benchmark. After the first goal of a match is scored by a team, a punter could finish in front in the long-term by backing that club if their price is higher than \$1.69. Conversely, the punter could lay off the club at a price of less than \$1.69. A risk inherent in framing markets of this kind is that future results may differ significantly from first goalscoring trends experienced in the 2008-2014 period.

B) First goalscoring clubs in finals matches have been successful more often than not, but to a lesser degree than clubs in home and away matches. The winning percentages in the period are 53.13 % for finals (which equates to a betting price of \$1.88) and 59.43 % for home and away matches (which equates to a price of \$1.68)  [Not that anyone’s encouraging wagering, of course. – Ed].

C) Clubs that have scored the first goal most frequently in a season have had a reasonable strike rate in making the finals. On approximately 67.86 % of cases in the period, a club that ranked in the top eight for first goalscoring rankings for a season made the finals.

D) Overall, there is a mild connection between a club’s first goalscoring ranking in the home and away rounds and the club’s final finishing position. The average variation in rankings, or the standard deviation of the rankings difference, is approximately three places. If you took a club’s first goalscoring ranking at the end of the home and away season, then the club will finish within three ladder positions of that ranking approximately 57.98 % of the time.

E) For the period of 2008-2014 as a whole, first goalscoring rankings for home and away games were reasonably similar to clubs’ overall performance rankings. For 14 out of the 18 clubs, the first goalscoring ranking for home and away games was within three placings of the club’s ranking in the overall performance tally. If we expanded the analysis to first goalscoring for all premiership season matches, 17 out of the 18 clubs had first goalscoring rankings that were within three placings of the club’s ranking in overall performance. The standard deviation of the differences in rankings, i.e. the difference for a club between the first goalscoring ranking in home and away games and the overall performance ranking, was 2.56 positions.

F) The clubs with the best overall record in first goalscoring have been the most successful overall, with the exception of North Melbourne (eighth most successful in the period). The five most successful clubs in 2008-2014 were five out of the six highest ranked clubs in first goalscoring (using home and away games only or all premiership matches). If this analysis is applied to the top six performed clubs in the competition in the period – Geelong, Hawthorn, Sydney Swans, Collingwood, St Kilda and Fremantle, the standard deviation of the difference in rankings between first goalscoring in finals and overall performances is only 1.33, which is approximately one ranking position only.

G) The clubs with the worst overall record in first goalscoring in home and away games tended to be the least successful overall.

H) The clubs “in-between”, i.e. ranked between 7th and 13th in first goalscoring in home and away matches, had mixed results as there were some moderate variations between the first goalscoring rankings and overall performances rankings.

I) There is a strong correlation between first goalscoring in finals matches and winning finals matches, or the ‘finals ranking’. The standard deviation of the difference in rankings for these two measures was only 1.33 positions.

J) There is a moderately strong connection between a club’s first goalscoring ranking in finals matches and the club’s overall performance ranking. The standard deviation of the difference in rankings for these two measures was 2.20. Rounding numbers, this means that, on average, a club’s first goalscoring ranking in finals will be within two places of its overall performance ranking.

[1] Thank you to Paul Jeffs for his permission in using data from afltables.com for the purposes of this article. This website sets out in chronological order each score in every AFL game from 2008 to 2014. After reviewing the Scoring Progression table for each and every match, I collected first goalscoring data and collated it for the purposes of this article.

[2] All percentages in this article are rounded to two decimal places, unless otherwise stated.

[3] Max Rooke for Geelong in 2009, Tyson Goldsack for Collingwood in the 2010 Replay, Travis Varcoe for Geelong in 2011 and Jack Gunston for Hawthorn in 2013.

[4] Tom Lonergan for Geelong in 2008, Xavier Ellis for Hawthorn in 2012 and Josh Kennedy for Sydney Swans in 2014.

[5] Darren Jolly for Collingwood in the 2010 draw.

[6] For Gold Coast, the result is multiplied by 7/4. For Greater Western Sydney, the result is multiplied by 7/3. This to extrapolate the data to seven seasons from four seasons (for Gold Coast) and from three seasons (for Greater Western Sydney). The resultant total is rounded.

[7] The extrapolation method is complicated because of the transition in the number of clubs from 16 to 18 in seasons 2010 to 2012. For each club, we express the finishing position for each season as a proportion of the number of teams in the competition. The results are averaged. This average proportion is multiplied by 16 for both clubs for 2008 to 2010 and, for Greater Western Sydney, by 17 in 2011. The extrapolated finishing positions for each club are then added to their actual finishing positions to reach the adjusted total for the period.

[8] Won a total of 83 games in the period.

[9] Won 69 games in the period when they scored the first goal.

[10] 43 of these home and away games were won.

[11] 48 of these home and away games were won.

[12] 35 for four seasons, multiplied by 7/4, equals 61.25, which was rounded to 61.

[13] No finals appearances. Data the same as for all matches.

[14] 27 for three seasons, multiplied by 7/3, equals 63.

[15] No finals appearances. Data the same as for all matches.

[16] Won 76 games in the period when they scored the first goal.

[17] Won 52 games in the period when they scored the first goal.

[18] Won a total of 80 games in the period.

[19] Won 34 games in the period when they scored the first goal.

[20] Won 55 games in the period when they scored the first goal.

[21] Won 39 games in the period when they scored the first goal.

[22] Reached the Preliminary Final in 2012.

[23] Won 66 matches in the period.

[24] Reached the Semi-Final in 2011 and 2013.

[25] Won 83 matches in the period.

[26] One Preliminary Final appearance (2014), lost by 71 points.

[27] Won 84 matches in the period.

[28] Won 79 matches in the period.

[29] One Preliminary Final appearance (2011), lost by 48 points.

1. Hi James, I went back and had a look at this on the back of the Round 1 results.

3 matches out of eight won by the team who scored the first goal, but Essendon v Sydney and Adelaide vs North the only games where a 2014 finalist vs a non-finalist didn’t score first and win.

I love this!

2. Vee says

Love this article. It’s a great, in depth look at an interested theory. Well done James.

3. James Grapsas says

Hi everyone

After 7 rounds of the 2015 season, the winning percentage of first goalscoring teams is approxmately 56 % (35W, 28L). We are nearly one-third of the way through the season, with 63 out of the 207 games played (assuming no Grand Final replays!).

After round 6, the winning percentage was tracking at the 2008-2014 average of 59 % (32W, 22L). The round 7 results, in which only three out of the nine first goalscoring teams won, reduced the winning percentage a fraction…

4. James Grapsas says

Hi all

We are about halfway through the season now: 102 games have been played out of the total of 207 for the year (assuming no GF Replay is needed).

The winning percentage of first goalscoring teams is approximately 58 % (59 wins, 43 losses). This is only 1 % less than the average of approximately 59 % for the 2008-2014 period studied.

5. James Grapsas says

Good afternoon all

We are almost at three-quarter time of the season: 152 games played out of 206 (taking into account the cancelled Adelaide vs Geelong match from round 14 and assuming no GF Replay will be required).

The winning percentage of first goalscoring teams has climbed to 62.50 % (95 wins, 1 draw, 56 losses). This is about 4.5 % higher than the average for the period that was studied for the purposes of the article.

Since the last check of the stat after round 12, the teams that have scored the first goal have won 36 out of the last 50 games played (72 % winning record in that time).

6. James Grapsas says

Sorry – correction to my last post. The percentage of 62.50 % for the 2015 season to date is about 3.5 % higher than the average (about 59 %) for the 2008-2014 period that was examined in the article.

7. James Grapsas says

I have collated the final set of stats for the 2015 season. Round 19 to the Grand Final consisted of 54 games, in which the first goalscoring team won 32 times (with one draw). In the finals, the first goalscoring team won three games and lost six. In the Grand Final, West Coast’s Luke Shuey kicked the first goal in his team’s eventual 46 point loss to Hawthorn.

The 2015 tally for first goalscoring teams was: 206 games played, won 127 (61.65 %), lost 77, drew 2.

This brings the 3 year moving average winning percentage (2013-2015 seasons) to 61.77 % (620 games played, won 383, lost 233, drew 4).

For the 8 season data set of 2008-2015, the grand totals are:

* Overall – 1,579 games played, won 939, lost 625, drew 15. Winning percentage is 59.47 %.

* Home and away games – 1,506 games played, won 902, lost 590, drew 14. Winning percentage is 59.89 %.

* Finals – 73 games played, won 37, lost 35, drew 1. Winning percentage is 50.68 %.

* Grand Finals – 9 games played, won 4, lost 4, drew 1. Winning percentage is 44.44 %.