Every now and then you'll read or hear about some dramatic new (or, more likely, recycled old!) training method that gets huge gains in performances from some group of people.  This happens in most sporting fields.  There'll be some new product making big claims about 20% improvements in some measure of performance and so on .. Glossy adverts, flash websites making big claims, and they've got studies to back them up!

How does this happen and what does it mean, and how can you, as a skeptical cyclist who's seen it all before, make sense of it?

Leaving out sloppy science and badly designed studies paid for by companies trying to sell their product, let's see if we can make some sense out of training and performance.

Firstly, take a look at this graph :

 The X axis is time spent training, the Y axis is performance and hopefully you'll notice that there's a ceiling, that's your genetic potential.  Everyone has one of these, and ultimately that's what decides if you can be an elite athlete or not. The only way to go above your genetic potential is to get involved in doping.

The graph starts at "untrained novice".  In our context, this would be someone who's never ridden a bike in any sort of training sense but can ride without falling over.

As you can see from the graph, initial performance increases are rapid.  A novice, very quickly with good training, progresses and makes rapid gains.  This isn't unique to cycling, it's the same in weight training, running, rowing, you name it .. anything that has a significant physiological fitness component responds in this way to training.

What's interesting, and often misleading, is that even poorly designed training plans can lead to rapid initial advancement, it's call the 'Novice effect' which I've mentioned before.  It approaches the genetic limits slower than an optimally designed program would. 

An extreme example of this would be setting a couch potato up with a program of lap swimming as their cycling training - initially their cycling would improve but it would rapidly plateau, much more rapidly than a program of well designed cycling intervals would.  The novice effect is basically 'something is better than nothing' in this context.  This often leads to much confusion, as we, gullible humans, see what we did initially working (even if it's not optimal, how do you know?) and assume that it's the best way to train.  It might be, or it might not be, the novice effect can be deceiving to the unwary.  This includes a lot of exercise physiologists who use poorly designed studies and very unwise extrapolations to lead coaches and athletes down dead-end paths.  So-called 'evidence-based coaching' is fraught with peril, as so many studies are poorly designed and the subjects badly chosen.

What should, I hope, be obvious from this visual depiction of progress in the graph above, is that it's relatively easy to get significant improvements from untrained athletes, and also, that it's much harder to get improvements from athletes who are close to their genetic potential.  A 20% improvement in a novice is fast and easy, in an elite athlete close to their potential a 2% improvement may take months or years or never.

Marketing people love to use novice improvement to 'prove' that their product is better than everything else and thus, sell you something.  Sometimes, their product is very good, but the data is misleading if it's not viewed in the context of the above graph.