On Learning and Feedback Loops

Those who are the best are usually those who learn the best and quickest

Source

A little while ago I was in a conversation with an exceptionally smart coach and we discussed the relative glacial pace of evolution of methodologies (and research) in the world of marathon training. This was in contrast to the world of 100m running. Whilst we almost certainly overlooked much of the nuance, the factor we highlighted as the most impactful in this was feedback.

Unfortunately, the nature of marathon running means that feedback is both worse and less frequent which, as covered here, means learning is harder. Feedback is worse in that there is so much noise in the feedback - so many factors contribute to a successful marathon that the feedback can be muddied by these (see also these dichotomies that marathon times are prone to). Additionally, the physical toll a marathon takes on the body means that running them is infrequent. Historically this has been done twice yearly, more recently perhaps a little more frequently with the advent of ‘supershoes’.

Aren’t Training Metrics Helpful in this Feedback?

They certainly can be. But it’s important to remember that the goal is to finish the race fastest, not train the best. Similarly, as mentioned previously, the body is a complex system and ultimately we are mostly interested in the inputs and outputs, not necessarily surrogate markers or intermediaries.

It is possible, using an analysis of weak points, to address limiting factors and test this in an intermediary fashion. In the 100m this may be starts (looking at 10m and 30m times) for example. In the marathon this is often broken into increasing speed in shorter distances vs extending speed into longer distances. In the former racing those distances makes sense, in the latter people need to be a little more creative in terms of testing and feedback outside of D-Day given the potential fatigue from those efforts.

As a jumping off point and an example, the above has probably done all it will do for those not interested in the events specifically. But to further reinforce the importance of timely feedback and tightening the feedback loop continue reading.

Feedback in the Wild

Ever notice how Apple releases a product and swiftly releases an update? Yeah, that’s the feedback everyone sent it about bugs etc. They not only embrace this, they plan for it - probably releasing products a little earlier than they would otherwise.
Whilst not technically the same, this is related to some concepts that are very common in the technology world (where I have recently spent quite some time).

Minimum Viable Product

This is one of those ‘does what it says on the packet’ type of things. In essence, when it comes to features, products etc in the technology world, companies will often look for feedback based on the Minimum Viable Product (MVP). The MVP is about not sinking resources into something that may not be worthwhile in the medium to long term. Generally it involves putting in the least amount of work to get something into the public for feedback before iterating on it.
As mentioned, this may be a feature, a product (or even a whole company). Really this is about validating the market - “is this something that will work”. At the same time, there is feedback on the feature/product etc given it is still very early on.

The key here is the mentality more than the theory; let’s get something out soon to get feedback. Realising that feedback is the most potent learning tool we have and the key to unlocking performance (of any kind).

A/B testing

This concept is almost certainly more familiar to the majority of readers and involves producing two version of the same thing to be used on half the market/audience to see what performs better. Rather than running them in series we run them in parallel for faster, more potent feedback. Let’s not agonise on which is better, or get opinions from people in the wild, let’s stress test them both in the real world, learn quickly and iterate.

The Masters of Feedback

For those unfamiliar with “Mr Beast” he’s worth looking into. In short he’s the world’s largest and most successful YouTuber. What has this got to do with feedback? He’s absolutely maniacal about iterating on what is working, A/B testing etc his way to a great deal of wealth all via YouTube. He famously obsesses over his thumbnails for videos, that’s not the title, or the video or anything yet, JUST the thumbnail. This is a great example of yearning for feedback and leveraging this into learning to ultimately learn better than others and thus ‘win’.

Someone probably more familiar to readers is Tim Ferriss, who amongst other things may be one of the world’s best when it comes to eliciting feedback. In a real meta phenomenon, he has iterated on this process of asking for feedback over years and honed the craft. Some of my personal favourites amongst his feedback requests are are asking for feedback on a 1-10 scale, but not being allowed to say 7 (thus eliminating people’s ability to fence sit), or when looking at his book manuscripts asking for which 10% he should eliminate if needing to remove 10% and which 10% he should most certainly keep (enforcing people to pick and favourite and least favourite portion for him to examine). These two approaches are both variations of removing people’s inherent will to not offend by reframing the constraints. They force people to give feedback in a helpful way to the receiver whilst simultaneously removing some of the social niceties that can cloud good feedback.

Charles Darwin’s theory of evolution is increasingly more misquoted, bastardised etc. But one the big take away from it for me is that the most adaptable organisms thrive, and adaptation is about learning what is needed and evolving towards that (or at least it is in a modern world over a single lifetime). In essence, at least in my mind, this then boils down to: those who learn quickest are the best able to thrive. This is especially true in a time where Moore’s Law is showing itself ever more and technology is accelerating almost exponentially.

So whilst feedback can be uncomfortable, it is the only way we will navigate the dynamic nature of the modern world. The future will be built by those who learn quickest and iterate best.