Surrendered Leads
/Last weekend, thinking back over the season so far, I couldn't recall a game where a team had come from a long way behind in the final term. In fact, analysis revealed, no team that's trailed by more than 15 points at Three-Quarter Time this year has gone on to win. So my memory's still okay then.
That maximal 15-point deficit seemed low to me and got me to wondering how it compared historically.
The chart above provides, by season (and by era, as defined in this blog and denoted by colour), box plots of the winning teams' leads at Three-Quarter Time and shows that the median lead - signified by the notch in each box - has been low this year relative to recent seasons. It also shows that the range of leads has been very small in historical terms. More particularly, the smallest lead - the end of the lower "whisker" - has been higher in 2014 than it's been in every season since 1963. (Note that I've excluded all drawn games from the analyses throughout this blog.)
We can also say that teams that have led at Three-Quarter Time have only rarely been run down this season. The proportion of teams that have suffered that fate is given by the fraction of the distribution of winning team leads that are negative. We can guess at that proportion from the chart above by noting that, since the lower edge of the box relating to 2014 sits above the 0 point on the chart, it must be less than 25%, but to get the actual percentage you'll need the chart below.
The actual figure for 2014 is about 13%, which is a couple of percentage points lower than the figure for 2013, but not all that low in the general context of the modern era. In 1994, for example, the low point for the era, only about 8% of teams that led at Three-Quarter Time were subsequently beaten.
If we look at the same data for winning team leads at Half-Time we find that, by historical standards, Half-Time leads have been close to average so far this season, though the range of typical leads has been quite narrow - about 50% of leads have been in about a 20 point margin.
Looking next at the proportion of teams run down after leading at Half-Time we find that the figure for this season is very low in historical terms at about 16% - the second-lowest figure recorded in the current era. In other words, only about one team in six that's led at Half-Time has gone on to lose this year. You need to go back to the very earliest VFL history to find percentages consistently as low as that for entire seasons.
Finally, reviewing the Quarter Time leads of winning teams in 2014 we find that the median of those leads is in line with history, especially that of the modern era, and the range of those leads, though at the smaller end, is also consistent with historical precedent.
Winning teams have tended, throughout VFL/AFL history, to assert their superiority early on in contests, so much so that those teams that have led at Quarter-Time Time have been defeated less than one-third of the time - less than one-quarter of the time in 2014. Here too this proportion is quite low by historical standards.
In summary, median leads of winning teams at the ends of quarters have been at about historical levels this year - perhaps a bit below average if we look only at leads at Three-Quarter Time. Further, the proportion of teams coming from behind to claim victory has been relatively low this year in historical terms, especially if we narrow the discussion and talk only about Quarter-Time and Half-Time leads.
Just about any lead has been a good lead in 2014.
