*2 years ago*

*Recently Carl watched our PingSkills Show on the relative age effect and decided to investigate further. He used a spreadsheet of 17585 ITTF players. Here's an email he sent us with some interesting findings. Thanks Carl!*

Hi Alois,

I was able to get some results after quite some hours from analyzing the numbers. I decided to use the "ITTF Birthdays" spreadsheet, since I assumed that the 17585 players who's names are on that sheet are probably among the best in the world. This was a big array of data which was able to do the job. I will briefly explain to you what I chose to do to get to a valid conclusion.

The advantage for using that spreadsheet was that there is virtually no chance that the values we get from analyzing that data could be due to randomness, which allows to be much more certain of any potential values we get.

My purpose was to compare the average we get from the players' to the average that we would normally get in a normal population.

After some searching, I found that the average for the USA population, which, being big enough and well diverse, can be considered a good representation of the world's population. The average turned out to be 6.6127 Months.

I then proceeded to calculate the average for the 17585 players, and got an average of 6.10132.

At first, this might not seem a big difference, but in fact, after applying some formulas, it shows that 6.10132 is very different than 6.6127.

What helps us further appreciate the difference, is the ranking of the months according to births in each of them. With our reference population of the US, which is considered to be the "random" one, the order months sorted from most to least births is:

8 > 9 > 7 > 10 > 6 > 3 > 12 > 5 > 11 > 1 > 4 > 2

However, according to the birthdays on the ITTF list, we get:

1 > 8 > 7 > 3 > 5 > 2 > 9 > 10 > 4 > 6 > 11 > 12

So as you can see, the 1 moved all the way to first spot from 10th spot, which is huge!

Looking at the different proportions for each month also lets us notice that when comparing the "random" values to the ITTF ones, the first 4 months have seen a rise in their proportions, while the months 5 to 12 (except 7) have seen dips in their percentage. So for example, even though for the players, 9 still has more players than 4, the percentage for 9 has dipped, and that of 4 increased, but not enough to let 4 surpass 9.

So despite the fact that there are many more babies born in August than in January, the number of players born in January is still the biggest of them all!

This, of course, indicates that there is an outside factor that very much influences which players get into professional playing, which we can safely assume is the Relative Age Effect, since the early months, between 1 and 4, saw an important rise in their percentages compared to the actual population's birth percentage in that month.

When I first saw this result, I must say I was really surprised. I thought that the younger the player is, the quicker he can learn from those older players while having more time in the future to train, but it seems like it wasn't the case, and with a difference between 6.1 and 6.6, this shows that there is clearly an advantage for people simply born in earlier months.

What I plan on doing next is finding out how big or small that advantage actually is, and whether or not it could potentially be a factor to be taken into consideration when judging a player's performance.

Also, I am now very interested in doing the same analysis for multiple sports and even for school studies if I can find similar lists, and compare them with Table Tennis. Who knows, maybe this might lead to some interesting results about sports and performance overall.

Again, thank you so much for the video that you made, and for sparkling that small curiosity in me. Even if those results turn out to mean nothing, it's been an absolute pleasure being able to analyze them and share the results with you. I plan on continuing the analysis on Thursday after I get my exams out of the way, and will certainly keep you updated if you are interested!

With sincere appreciation,

-Carl H.

Become a free member to post a comment about this blog.

## Madhu Madhavan

Posted 1 year agoWhen I first stumbled on this article, I thought you were going to compare performance differences across age groups. Alas! that was not the case.