Harder, better, faster, stronger… forever?

I’m Julian, developer on Datawrapper’s app team. This week’s Weekly Chart shows what happens when a computer scientist takes a shot at interpreting athletic record data — garnished with some Daft Punk puns.

Around the world, people are obsessed with setting records. Measuring performance and trying to beat previous achievements seems to be a big motivation for humans, no matter what they compete in. But how long can humanity keep breaking records? There must be limits, right? Nobody will ever run beyond the speed of sound — after all, we’re still humans, not superheroes.

This chart shows how performance has improved over time several disciplines. On the x-axis is time — not every sport started measuring world records at the same time — and on the y-axis is the progress that's been made in closing the gap between the first world record and the one that stands today. For example, if the first long jump record was six meters, and today’s record is eight meters, then a seven meter jump would represent 50% progress towards closing the two meter gap. (Yes, that means we're comparing more kilograms lifted to more meters jumped to fewer (!) seconds run. Bear with me, it will get even worse soon.)

We can see several things in this chart already. New records are still being set, but in some disciplines, like the marathon, the margins of improvement seem to be getting slimmer and slimmer. Some events have stalled altogether: The current long jump world record has not been broken since 1991.

A note about the data

It's substantially harder than I thought to get accurate data on world records, especially since there can be disagreement on what score somebody achieved or whether certain conditions or equipment rule out an event from consideration. As a result, I had to make some subjective decisions in compiling and cleaning this dataset, so take it all with a grain of salt.

My world record data is built entirely from tables in Wikipedia articles as of this week. It’s comprised of world records for six disciplines, two for each category:

I only included male athletes in my analysis because there is significantly less historical data available on records from female athletes.

Can we beat the record one more time?

We can make the chart even more controversial by also messing with its time axis. In the version below, time on x-axis is also measured in relative terms: zero marks the spot in history where a discipline's first world record was recorded (so the year 1820 for the high jump but 1985 for the deadlift), and 100 is the present day, January 2024.

Before we analyze this chart, let’s take a moment to focus on what different shapes of lines in this chart really mean:

  • ➡️ A long, straight line from left to right indicates a record that hasn't been beaten for a long time.
  • ⬆️ A long, straight line upwards means somebody broke the previous record by a large margin.
  • ↗️ A consistent diagonal line means that there's a steady pace of frequent, small improvements in this discipline.

We could take this thought one step further and consider how each line compares to the perfect diagonal, y=x. If a discipline is consistently above that diagonal, it means that the pace of record-breaking has slowed over time; if it's below, that means the pace of record-breaking has increased. The closer it is to that diagonal, the more consistent the pace and margin of new records has been over time.

Cool, now that we've established ways to interpret this weird chart, let’s take a closer look.

The lines are somewhat distributed around the magical y=x diagonal. It doesn't look like we humans have stopped breaking records yet. In disciplines like deadlift and bench press, it even seems like we've been breaking them at accelerating pace. Going back to the previous chart reveals that those are younger disciplines, so they might very well stagnate in coming decades, as other have.

I find really fascinating that, at least once a discipline is established enough to have records measured on a global level, people find ways to break them over and over again. There seems to be some truth to "what gets measured gets managed" after all. Even more astounding to me is that these improvements happen at relatively constant pace and by somewhat constant margins. I was expecting log functions, not straight lines. So it really looks like we can still expect all of the records above to be overtaken at some point. The consequential question is:

How long can we continue to break records?

For some sports, newly developed techniques can lead to new records. This has been the case for the high jump, where the Fosbury Flop — a newly developed jumping technique — led to a series of new world records in after the 1960s, an effect we can observe in our chart. That happens for some sports... but haven't we probably figured out running by now?

Well, you can’t change running but you can change your shoes. Super Shoes lead to new records in marathons that were previously deemed impossible. Naturally, this spawned a debate. Isn’t that cheating? Sometimes, such situations can lead to a split where one sport becomes two, as is the case with raw and equipped powerlifting. So even if we might max out a certain discipline, we can slightly change the rules and push onward.

There are other ways to improve performance. The Tour de France, for example, is notoriously known for athletes usage of performance enhancing drugs. Circumventing doping tests has become a challenge on its own. We also known that genes play a role in athletic performance so in the future, biomedical advances might lead to records that look impossible today.

Still, for some sports we might not be able to further improve technique or come up with new drugs or gadgets to enhance performance. For them we have to pull our final, most powerful trick: measuring in more detail. We might have to aim for smaller and smaller improvements, but we will never truly reach any limit. Even if we have to switch our charts to a logarithmic time axis, competition can continue for a really long time.

I hope there were some interesting insights! Next Thursday, we'll be back with a Weekly Chart surprise. Until then, happy record breaking!