Birth Date Bias in Football - An AFS Special Survey












Football statisticians may have noticed an unusual distribution of birth dates amongst players and, in this AFS Special Survey, we examine this issueFootball statisticians may have noticed an unusual distribution of birth dates amongst players and, in this AFS Special Survey, we examine this issue.
Birthdates of Premiership players by month
163
122
131
92
102
87
76
121
262
230
224
169
We’ve used the AFS data set on 1779 English players attached to Premiership squads from the 1992-93 to the 2004-05 season. The dataset is available as in the AFS shop as AFS Data File No 4
The dates of birth have been banded by month giving the following results:
% of Births by Month - Comparison of English Premiership Players with the General Population
| Month of Birth | Number of English Premiership Players | % Players | No of Births in general Population (thousands)* | % General Population |
| Jan | 163 | 9.16% | 577.8 | 8.31% |
| Feb | 122 | 6.86% | 531.1 | 7.64% |
| Mar | 131 | 7.36% | 584.7 | 8.41% |
| Apr | 92 | 5.17% | 565.6 | 8.13% |
| May | 102 | 5.73% | 595 | 8.56% |
| Jun | 87 | 4.89% | 583.7 | 8.39% |
| Jul | 76 | 4.27% | 609 | 8.76% |
| Aug | 121 | 6.80% | 594.3 | 8.55% |
| Sep | 262 | 14.73% | 595.6 | 8.57% |
| Oct | 230 | 12.93% | 589.7 | 8.48% |
| Nov | 224 | 12.59% | 558.1 | 8.03% |
| Dec | 169 | 9.50% | 568.7 | 8.18% |
| Total | 1779 | 100.00% | 6953.3 | 100.00% |
*The number of births in the general population are total figures for 1993-2003.
Source: National Statistics website
Crown copyright material is reproduced with the permission of the Controller of HMSO
The distribution of birth dates of Premiership players shows clear birth-date bias with 14.73% of players born in September as against only 4.27% in July.
A similar pattern can be seen amongst the cohort of 196 current and ex-Premiership managers and caretaker managers:
| Month of Birth | No of Managers | % |
| Jan | 19 | 9.69% |
| Feb | 15 | 7.65% |
| Mar | 13 | 6.63% |
| Apr | 14 | 7.14% |
| May | 16 | 8.16% |
| Jun | 11 | 5.61% |
| Jul | 13 | 6.63% |
| Aug | 9 | 4.59% |
| Sep | 18 | 9.18% |
| Oct | 22 | 11.22% |
| Nov | 22 | 11.22% |
| Dec | 24 | 12.24% |
| Total | 196 | 100.00% |
Why is there such a bias towards players born in September, October and November? The main reason, is the start of the school year at the beginning of September. During their sporting careers the older individuals in each year-group are more successful at playing football due to having extra months of development which makes them mentally and physically stronger than their peers who may have been born up to 12 months later.
This bias creates a problem for recruitment of the best players to elite football as potentially great footballers born in later months of the school year are being lost to football at an early stage.
This problem has been long been recognised by the Football Association, following original research by Caroline Sharp of the National Foundation of Educational Research in the 1990s. (See, for example, Craig Simmons’ 2001 paper ‘Season-of birth bias in association football’ published in the Journal of Sports Science which found bias towards older players in selection to the Football Association National School.)
According to Craig Simmons of the Football Association, the FA now includes education on this topic in all courses relating to children at the front end of selection to the professional game. Raising awareness of this topic will be extended to the grassroots game, where research suggests the bias is less dramatic but involves a lot more children.
Other than researching and raising awareness of the topic, the FA has been examining other solutions to the problem including the possibility of basing training classes on a rolling birthday intake, so that each child will have experience as both the oldest and youngest in the class.
Splitting the season into six-month cohorts would be effective in significantly reducing the bias - with the best way of achieving this, according to Craig Simmons, being to keep existing season start date in September but to have two birth-registration dates.
The Future
It is expected that the new FA Grass Roots Four Year Plan, will raise awareness of this issue and changes implemented within football should help mitigate the effects.
Within education, it is worth schools considering adjusting their physical education programmes - in some cases by combining school sports programmes across a number of different schools in the same area or making use of local sports centres in order that six monthly sports cohorts are possible. Recent research quoted in The Mail on Sunday (22/05/2005) has shown that children born in September are around 20% more likely to achieve an A* or A grade than those born in August so the time may be ripe for political action to make six monthly cohorts the rule in education as a whole.
Planning Conception?
As things stand, if you wish your child to be a professional footballer you should consider plan the date of conception for late January or early February to maximise the chance of the child being born at the optimal time (!) With new measures this will hopefully not be necessary.
All in the Stars?
Given the monthly differences in numbers of premiership players astrological explanations may be tempting for some.
108
103
93
81
103
233
240
234
157
163
150
114
% Dates of Birth of Premiership Players by Birth Sign
| Birth Sign | No. Of Players | % |
| Libra | 240 | 13.49% |
| Scorpio | 234 | 13.15% |
| Virgo | 233 | 13.10% |
| Capricorn | 163 | 9.16% |
| Sagittarius | 157 | 8.83% |
| Aquarius | 150 | 8.43% |
| Pisces | 114 | 6.41% |
| Aries | 108 | 6.07% |
| Taurus | 103 | 5.79% |
| Leo | 103 | 5.79% |
| Gemini | 93 | 5.23% |
| Cancer | 81 | 4.55% |
| 1779 | 100.00% |
However, studies around the world have clearly shown where seasonal start dates are in January (or in April as in Japan) then the birth date pattern shifts so that it is always the older children who are advantaged.