Only three teams in VFL/AFL history have trailed by more than three goals at Quarter Time in the Grand Final and gone on to win. The most recent was Sydney in 2012 who trailled the Hawks by 19 at the first break before rallying in the second term to kick 6.0 to 0.1, eventually going on to win by 10 points, and before that Essendon who in 1984 trailed the Hawks by 21 points at Quarter Time - and still trailed them by 23 points at Three Quarter Time - before recording a 24 point victory on the strength of a 9.6 to 2.1 points avalanche in the final term.
The other team was Carlton in the equally-famous 1970 Grand Final where they trailed the Pies by 29 points at Quarter Time, 44 points at Half Time, and 17 points at Three Quarter Time, before eventually prevailing by 10 points.
So, the first thing you want your team to do in a Grand Final if they're to go on and win it is to avoid trailing by a large amount at Quarter Time.
A small deficit is more manageable, though still not ideal, since winning Grand Finalists have been level or trailed by up to 18 points in only about 28% of all resolved Grand Finals, and in about the same proportion of Grand Finals from the modern era alone. In no era have many more than about one-third of winning Grand Finalists found themselves level or trailing by 3 goals or less at Quarter Time.
Slender leads at Quarter Time are fine - indeed, they're the norm, since more than half of all Flag winners have led by less than three goals at the first change and less than one in five had led by a larger margin. In the modern era fully 60% of winners have led by less than three goals at Quarter Time.
(The table below provides the overall and era-by-era Quarter Time data in the middle block of the five shown. The fourth and fifth blocks are for the winning team's position at Half Time and Three Quarter Time).
The picture at Half Time is a little different, though mostly in terms of the distribution of leads. It's still the case that only about one-quarter of eventual winners are level or trail at the main break (a little more than one-quarter in the modern era), but over one-third lead by more than three goals.
Three Quarter Time is not the time to be trailing. Only two teams have trailed by more than three goals at Three Quarter Time and gone on to win - Essendon, in 1984, as mentioned earlier, and the Hawthorn side of 1971, who trailed the Saints by 20 points at Three Quarter Time before eventually winning by 7 points. Just nine more winning teams in history have been level or trailed by any amount at Three Quarter Time, and only two other teams aside from Essendon since 1978: Carlton, who trailed Collingwood by 9 points in 1981, and Geelong, who trailed St Kilda by 7 points in 2009. Almost 50% of winning Grand Finalists have led by more than three goals at the final change, though this has been true of only one-third of the winners in the modern era.
In summary - and this data appears as the first three columns of the table above - about 70% of ultimate Flag-bearers have led at the end of the 1st Quarter, over three-quarters have led at the end of the 2nd Quarter, and 90% have led at the end of the 3rd Quarter.
Looking at the second block of data we see that, on average, winning Grand Finalists score more points than their opponents in any given Quarter about 70% of the time. Winners from the modern era, however, have outscored their opponents in 2nd and 3rd Quarters less often - only about 60% of the time - but have outscored them in 4th Quarters almost 90% of the time.
Finally, as another way of presenting the all-time data for winning teams' Quarter-end-by-Quarter-end leads, the table at right provides percentiles from the various distributions. So, for example, the first entry of -10 tells us that only 10% of winning teams have trailed by 10 points or more at Quarter Time, and the -7 in the neighbouring cell tells us that only 10% of winning teams have trailed by 7 points or more at Half Time.
As I wrote in the title for the corresponding blog post last season, to win a Grand Final you must first lead.