Speed is a key skill in today’s fast-moving and forever-changing world. However, most companies are not designed for speed – instead they are designed for efficiency as they typically need to cover a long distance. They end up wasting a lot of time in simply waiting for decisions, or some critical resources, or approvals from management, and so on. On the other hand, startups are designed for speed and don’t (need to) care (so) much for efficiency, because they must perform on a very short runway. They have a limited amount of time and money, and while they still have the funds to keep them going, they must make the best of it and keep experimenting till they discover repeatable, scaleable and sustainable way to make money. Speed is important to make the kill today, and if we survive to tell the tale, the efficiencies can always come tomorrow. Good or bad, that’s the way it works.
However, have you thought what exactly is the relationship between distance covered and the speed attained? Logically, it seems clear – we can always go fast over short distances, and must lose speed over longer distances. But is there any data to support this? And if yes, is that a straight line, a curve,…? Further, are there variations to it? Over the same distance, can we still go faster than what seems to be physical limits of a human being? For example, could working in teams make it faster?
I thought of examining the data.
The longer you go, the slower you get
Of course, you always knew it. But, do we have any data to support it? And, even if this were true anecdotally, is there any math behind the data?
I looked the men’s track and field world records (http://en.wikipedia.org/wiki/List_of_world_records_in_athletics) and plotted the speed vs distance for all distances from 100m to 100kms in running. The resulting curve looks like this:
It is interesting to note that the data is fairly consistent when it comes to human limits – there is an inverse relationship between the distance covered and the top speed achieved. In other words, the longer you go, the slower you get – even when you compare world records of different specialist runners. Of course, we might not be able to push it ad infinitum and achieve mach speeds or more by shortening distances to ridiculously low atomic distances, but then, who knows 🙂 – maybe there is some interesting problem waiting to be discovered when do do that!
So, what would be the most logical thing to do if you need to get there faster? No doubt – go short distances. What if the distance covered is more than what one person can sustainably complete – surely, we can’t just travel a fraction of it and stop? Let’s look at relay teams.
A relay team is faster than an individual
What happens when we take numbers from relay race and compare it with a single runner running (or swimming the same distance) alone? I took data from four men’s events and compared them side by side:
The data is limited only for short-distance relay races (and I was tempted to include data from ekidens and Swedish Relays, but maybe another day!). However, in each of the cases, the average speed went up (between 13% and 28%) when a relay team ran the same distance as an individual runner.
Again, is it logical to explain. One runner running the entire 400m will get tired, but each time a new runner comes and does next 100m of the relay race, he/she can put in fresh burst of energy and since they have to sustain themselves only for 1/4th of the total distance, they can be very fast – certainly faster than the individual runner who has already done first or second laps.
So, clearly, if we divide the tasks such that different team members can add value to it at different point in time, we can do the same work faster.
How about if we all did the same job together? For that, I had to leave athletics track and find something else where teams work together on the same task as the same time. Guess any sport that does it?
A simultaneous team gets faster when you add players
In rowing, there are multiple combinations that basically cover the same distance of 2km. Starting with a solo, it can go up to eight rowers, and the data gets interesting. Here is the data for sculls and coxless pairs:
So, when we add the number of people who are simultaneously working on the same problem, even when the problem gets bigger (weight of double scull is 2x of scull and that of quad is 4x, with the length going up by roughly 30% every time), we get faster! This seems like a great motivation for solving large problems as long as the logic of simultaneous players can be applied. Does this go ad infinitum? What about famous snake boat races that have upto 100-125 oarsmen in each boat (and 25 singers and four helmsmen too!). The top speed recorded is 20.8 km/hr, so clearly this speed starts to taper off at some point. But clearly, the speed does go up between 1 and 100+ oarsmen.
These were my data-backed insights:
- If you cover short distances, best is to go alone. You can go fastest.
- For short distances, when you want to go still faster than one individual, working in a relay format seems to be better than a single runner. This strategy might not work for very short distances, because the overheads in coordination and switching might offset any potential gains from relaying.
- If you cover long distances, best is to work in teams. You can sustain it farthest.
Guess what does it tell us about teams at workplace? Agile teams are small, typically not exceeding 7+/-2 members and they often work in pairs or might even swarm together to solve complex problems. And the data from sports seems to suggest that it work!
I’d love to know if you have data from your software teams – do they have data that supports – or contradicts – the data from sports? It might make for an interesting conversation…
(Originally published at https://www.linkedin.com/pulse/how-go-faster-than-you-can-tathagat-varma)