V/AFL Player Names 1897-2018

/

(This post is an updated version of one that first appeared in February of 2018, and now includes data for the 2017 and 2018 seasons.)

Today we'll use player data from the fitzRoy package in R to investigate the first names of players in the 15,402 games completed to the end of the first week of the 2018 Finals.

For this analysis I'll treat common variants of names as if they are the same so that, for example, Jarrad, Jarred, Jared, Jarrod, Jarryd, and Jaryd - all of which have appeared in at least one game - are treated as being the same name. Some variant sets, like this one, are fairly obvious, but others are more subjective.

Possibly the most contentious groups I've created are {John, Johnny and Jon}, the latter of which you could argue is actually a variant of Jonathan, and {Jack, Jackie, Jacky, Jock and Jockie}, which I've treated as its own group despite the common practice of some Johns being known as Jack.

Let's start by looking at the most common first names across history, where commonness is measured by the number of games played by anybody with that name. So, for example, if there were three Johns in a single game, they'd add three to the tally of games for that name. 

We see that the Jack and variants grouping (which I'll henceforth refer to as {Jack} to signify that it refers to a set of name variants, a practice I'll also use for other groups) has the most appearances with 21,670, or about 1.5 per game. Next is {Bill} with 17,044 appearances, or about 1.1 per game. The only other group with more than 15,000 appearances is {John}, which has 16,362 appearances, or also about 1.1 per game. 

Only nine name groups have an appearance count above 10,000, {Matthew} just attaining that mark in the current season.

The popularity of particular names has ebbed and flowed quite markedly across the years - as you can see, for example, for {Jack} above. As a result, different names appear among the set of most popular names when we look at different eras.

(Click on an image to view a larger version of it.)

One striking feature of these charts is the fact that there is one standout name group in every era: {Bill} in the 1897 to 1929 era, {Jack} in 1930 to 1959, {John} in 1960 to 1989, and {Matthew} in 1990 to 2018.

We can see how particular names tend to dominate within eras for contiguous periods of time by analysing the name with the most appearances in each season individually. 

{Bill} dominated for the first 19 seasons from 1897 to 1915, before yielding to {Charlie} and then {Harry} over the next two seasons. 1918 and 1919 saw a {Bill} resurgence before {Jack} took top spot in 1920 as something of a practice run for later. {Harry} and {John} next had a couple of year stints each prior to {Jack} returning to the top in 1925 and staying there until 1953 - a remarkable 29-year epoch, still the longest stretch of domination by any name.

{John} then built his own dynasty, leading for 20 seasons from 1954 to 1973 (which included a record 596 appearances in 1962), after which Peter and {Mike} dominated until 1989. After that, {Stephen} (1990 to 1992), Paul (1993) and {Anthony} (1994) all had short stints. {Matt} then defined a fourth dynasty, leading for 18 years from 1995 to 2012, before {Tom}, then {Matt} then {Jack}, then {Tom}, and then two more {Jack}s rounded out the sequence. It's hard to know if we're headed for another dynasty, or whether the single year reigns marking 2013, 2014, 2015 and 2016 will be the new norm.

In total, across the 122 seasons, only 13 different name groups have dominated for one or more seasons, and seven of those groups have led for fewer than five seasons:

  • {Jack} 33
  • {Bill} 21
  • {John} 20
  • {Matt} 19
  • Peter 9
  • {Mike} 7
  • {Harry} 3
  • {Stephen} 3
  • {Tom} 2
  • {Jim} 2
  • {Charlie} 1
  • Paul 1
  • {Anthony} 1

{Bill} reached its zenith in a 1909 clash between Essendon and Richmond in which 12 Bills took part, six on either team. That is the most for any name group in any single game.

Since only 18 players per game were named for teams in that era, that means exactly one-third of the players who took part in that match were named Bill or Billy.

I presume either that nicknames were heavily used on that day or that there were a lot of kicks and handpasses going to the wrong target.

 

 

Since 1980 the most namefellows in a single game is nine, which is the {Jack} count in a 2018 clash at the MCG between Richmond (three Jacks) and St Kilda (six Jacks). On that day St Kilda also had a Jake who's not included in the tally.  

With 22 players named in each squad, that represents only a slightly less impressive 20.5% Jack content on that day.

There were two other games in 2018 where the {Jack} count was eight, and there was a game in 1990 where the {Stephen} count was eight.

Across the entirety of V/AFL history, there have been 21 games where there were 10 or more players with the same name variant, including {Jack} counts of 10 in two different games, one at Princes Park, and the other at Punt Road, on the 27th of July, 1940.

We can also investigate the popularity of individual name groups within different seasons by calculating the proportion of games in a season that included at least one representative from a given group. In the table below we see the results for those name groups that appeared in every game in a particular season, or in every game but one.

We see from the first entry that {Bill} was the first group to have a representative in every game of a season, achieving this feat in 1901. Further down we find the record for the {John} group, which includes the extraordinary strings of appearances from 1960 to 1965. Every VFL game during that period included at least one representative from the {John} group.

The latest entry in the table is for the {Michael} group, which had at least one representative playing in all but one of the games in the 1989 season, missing only the 30 July game between Footscray and West Coast at the Western Oval. In that game, Footscray was missing Michael McLean from the previous week's drawn game, and West Coast was missing Michael Brennan from the previous week's win over Sydney.

As one final way of assessing the popularity of different first names across all of V/AFL history to 2018, let's calculate the proportion of games in which at least one of the participants had a particular first name or variant.

The winner on this metric is the {James} group, which has had a representative in almost 60% of games across history. Next is a name group we've not previously discussed, {Bob}, which has had a representative in 57% of all games. Third is {Mike} at just over 53%, one of only three more groups that have appeared in more than 50% of games.

The {Jack} group, despite its dominance in terms of cumulative appearances, is only 6th on this metric.

For some name groups, it's been a while since we saw a representative take the field and in the chart below we show all the groups that have been gameless since at least the end of the 2013 season and that had, previously, had at least five different players sporting the particular first name or variant.

Edgar and Ned have been absent longest with neither name running around a field with a guernsey on since 1928. {Ron} is perhaps the heaviest-hitter to have gone silent in recent years, it having not been sighted since 2004 before which 147 different players had been part of the group.

There are 266 first names and variants that have belonged to a single player, but only 19 that have belonged to a single player who played just a single game (so far).

My favourite amongst them is Hurtle - a fine name for an athlete from any area, surely - but the modern-day marketing possibilities for someone named Urban would be irresistible.

I would wager significant money on his nickname being either of "Keith" or "Legend".

By far the majority of these names come from the first 50 or so years of history, though we do have three from the last decade.

Let's finish by analysing the diversity of first names in individual games and seasons, which we'll do by first defining a players per name metric. It's calculated by dividing the total number of players in a game by the number of unique names. 

So, for example, if there were 30 distinct name groups across two teams of 22 players, the players per name metric for the game would be 44/30 = 1.46. This metric cannot fall below 1, which it does when every first name in a contest is unique. Its theoretical maximum value is the number of players in the game but, practically, it's rare to see values above 2.

We can see from the chart that the average value of this metric was high in the early years of football, signifying that a lot of players came from a relatively small pool of name groups at that time, and that it has been falling from about the early 1980s to its modern-day, historically low values. A value of 1.3 is equivalent to about 34 unique first names across the teams' 44 players. On that basis it's fair to claim that we're seeing a greater diversity of first names in the AFL these days.

The best exemplar of this diversity is the 2012 Round 6 game at Docklands between Carlton and GWS in which there were 43 different first names across the 44 players - the only match a Tom with a Tomas.

(One other interesting aspect of this chart is that it shows the change in the number of named players per game across history, with each era appearing as its own strata).