Hacker News Reader

Comment by tacitusarc | Hacker News Reader

I disagree with the "steep learning curve" point. The X axis is acquired knowledge, and the Y axis is the effort required. I don't know why people assume the X axis is time. Not all graphs are temporal.

View on HN

It's standard practice to plot independent variables on the x-axis and dependent on the y. To my knowledge, this is how learning curves are typically plotted as well, and the use of "steep" is usually a misnomer.

Is there a reason you believe people have this unintuitive plot in mind, rather than simply conflating with the difficulty of physically scaling a steep slope?

The independent variable is how good you're trying to get, and the dependent variable is how much effort it costs.

> a reason you believe people have this [...] plot in mind

Sample size N += 1, HTH, HAND.

I disagreed at first, but realized what they were trying to convey. Specifically "learning curve" is a defined term in learning theory that represents the accumulation of knowledge. Like the the "forgetting curve" [1], the use of "curve" represents at each learning opportunity (or forgetting). "Learning curve" between intervals would be more akin to "learning gains".

So a "learning curve" that is "steep" implies that learning occurs rapidly; a "forgetting curve" that is "steep" would be "in one ear and out the other".

Super pedantic and I don't plan to change my use of steep learning curve, but I get where the intention is coming from.

[1] https://www.psychestudy.com/cognitive/memory/ebbinghaus-forg...

Does the steepness of the curve happen earlier or later in the timeline?

I imagine that one could spend a long time learning the fundamentals before getting to a stage where the proficiency escalates quickly i.e. where the steepness occurs. But the long time getting there is what the term "steep learning curve" is about. Like the opposite of taking a long run at jumping off a cliff. I'm sure that this is easier to show with a graph.

Reply on news.ycombinator.com