V/AFL CHARTS

This page contains charts with minimal commentary about miscellaneous topics that have grabbed my attention along the way.

All of the charts can be clicked on to access a larger version.

MARGIN PROFILES By TEAM By ERA

Data: All games from 2000 to R11 2016

Topic: Profile the game margins (ie own score less opponent score) for all games played by a team against all opponents during a particular era

Points of Interest

  • Certain teams have distinctly different-shaped profiles in some eras - for example, compare North Mebourne and the Western Bulldogs in the 2000 to 2016 era
  • One way of assessing the relative performance of any team in a particular era is to see how high its mean margin is compared to all other teams in the era. That mean is given by the notch inside the team's boxplot for that era.

PREDICTABILITY OF TOTAL SCORE BY ERA

Data: All games from 1897 to 2016 R8

Topic: Relationship between the actual Total Score and that predicted by MoSSBODS

Points of Interest

  • Less of the variance in Total Scores (9%) has been explained in the most-recent Era than in any before it
  • The mean absolute error (MAE) in this Era, however, has been lower than in either of the two preceding Eras
  • The spread of Expected Total Scores has vary quite markedly across Eras, narrowing especially in the most-recent Era

PERCENTAGE OF POINTS SCORED IN EACH QUARTER BY ERA

Data: All games from 1897 to 2015

Topic: Scoring in each quarter of a game as a proportion of the final score in that game

Points of Interest

  • The median for each quarter and for each era is around 25%, as we might expect
  • For the most-recent three eras, however, the tendency has been for there to be a slightly higher proportion of points registered in 3rd and 4th Quarters
  • The spread of percentages is quite large, with up to 60% and in some cases 0% of points in a single game being registered in one of the quarters

RELATIONSHIP BETWEEN CONVERSION RATE AND SCORING SHOTS

By Era

Data: All games from 1897 to R6 2016

Topic: Relationship between Scoring Shot production and Conversion Rate

Points of Interest

  • Mild evidence that teams generating more Scoring Shots tend to be more accurate
  • Small effect size
  • Not true in all eras
  • More prominent at higher Scoring Shot levels
  • Note that the spread of Conversion Rates reduces with the number of Scoring Shots (ie if we think about conversion as sampling from a Bernoulli distribution, we're seeing the effects of increasing the sample size, n).

By Team in the Modern Era

Data: All games from 2000 to R6 2016

Topic: Relationship between Scoring Shot production and Conversion Rate for each team

Points of Interest

  • Mild evidence that, for a number of teams in the modern era, their generating more Scoring Shots tends to be associated with their being more accurate
  • Small effect sizes
  • Not true for many teams
  • More prominent at higher Scoring Shot levels for those teams for which it is true
  • Again, note that the spread of Conversion Rates reduces with the number of Scoring Shots (ie if we think about conversion as sampling from a Bernoulli distribution, we're seeing the effects of increasing the sample size, n).

By Venue in the Modern Era

Data: All games from 2000 to R6 2016

Topic: Relationship between Scoring Shot production and Conversion Rate for each venue where at least 50 games have been played during the Era

Points of Interest

  • Mild evidence that, for a number of venues in the modern era, teams that generate more Scoring Shots tend to be more accurate
  • Small effect sizes
  • Not true for many venues
  • More prominent, if anything, at higher Scoring Shot levels for those venues where it is true
  • Again, note that the spread of Conversion Rates reduces with the number of Scoring Shots (ie if we think about conversion as sampling from a Bernoulli distribution, we're seeing the effects of increasing the sample size, n).