Questioning the Underdog Hypothesis #3 Ford & Williams

What is the Underdog Hypothesis and why question it?

Ford & Williams (2011)

A study of 205 award winning (e.g. MVP, Ballon D’Or) male athletes from football (FIFA & UEFA), ice hockey (NHL), baseball (MLB) and Aussie rules football (AFL), from between 1987 and 2007.

Crossed arms photo created by creativeart - www.freepik.com

Main findings:

1. When award winning athletes were compared to the source professional cohorts there were significant (p<0.5) differences between Birth Quartiles with Q3 & Q4 higher than expected and Q1 & Q2 lower than expected.
2. Actual %s were Q1 (21%), Q2 (24%), Q3 (29%), Q4 (25%), (H1 45%, H2 54%).

Let’s look further:

For me, two problems exist with this study in how the results are interpreted. Firstly, that the entire professional cohort is treated as a single entity (Q1 (31%), Q2 (27%), Q3 (23%), Q4 (19%)), both in terms of mixing the four sports but more importantly by age (an 18 yr old is the same as a 35 yr old). Secondly, the fact that RAEs can continue to decline during professional careers (which may not have been considered in 2011).

1. The average age at first award was 25.6 (SD 3.7). It is likely that the RAE profile of 22 to 29 yr olds (Prime?) differs from those in the 18-22 (Rookie?) and 30+ (Veteran?) age profiles. RAE can decline in the latter years of junior sport. It can continue this decline in the early years of senior sport. As no significant difference was found compared to the general population it is possible no significant difference would be found when comparing to a Prime population. Also multiple award winners would increase the average age profile.
2. We shouldn’t really be surprised when underrepresented quartiles provide more award winners as they are less likely to be diluted by poorer quality/potential players in overrepresented quartiles. 
3. Q3 was the highest quartile. Q3 and Q4 performed the same in comparison to the source cohort. Q2 performed slightly below expected and Q1 were far below expected results.

The authors suggest that younger players have developed more skill during their development due to the higher challenge of playing/training with relatively older players, namely the Underdog Hypothesis.

However Q4 don’t outperform Q3 and perhaps Q2,Q3 & Q4 are similar and Q1 make these Birth Quarters appear to outperform it by underperforming itself instead. Perhaps this is the Fall of the Top Dog (a continued RAE profile decline) rather than the Rise of the Underdog?

Notes:

No RAEs found when the four sports analysed separately.

Half of the participants were from baseball.

Relatively small sample size.

Paper

Series:

#1 Gibbs, Jarvis & Dufur (2012)

#2 Fumarco, Gibbs, Jarvis, & Rossi (2017)

#3 Ford & Williams (2011)

#4 McCarthy, Collins & Court (2016)