Hacker News Reader

Comment by pithon | Hacker News Reader

Why zoom out and lose the fidelity? So people who don't actually pay any attention to the axes/labels/etc. can walk away with a slightly better understand of the graph with their casual glance than they otherwise would have?

View on HN

They won't walk away with better understanding, they'll walk away with worse. It's better to not understand a graph at all than to be mislead by it.

That's a stupid highbrow attitude. I suppose the same concept applies to any form of information? If we don't understand the whole story, it's not worth knowing any part of it?

It's disinformation. It's like when the meteorologist shows a map of temperatures across the region, not mentioning it is the 'feels like' temperature, an arguably useless and misleading number made up of the actual temperature and the wind speed. Just tell me the temperature and wind speed - let me be the judge of what it 'feels like'. Maybe I wear a windbreaker.

For sub-freezing temperatures with strong wind there is most definitely a point in mentioning wind chill factor. It keeps inexperienced people (i.e. tourists) from flooding the ER with frostbites. Otherwise I agree "the temperature it feels like" is silly.

It looks like you have an incomplete understanding of the meaning of 'mislead'.

I agree, if the interesting range of truck orders was 40000-50000 and the fluctuations between those values is the interesting information. However, here the article just mentioned the nose-dive to nearly zero at 2008, and then shows the graph where the number again dives to the x-axis. Then you take a second look and note it's a minor dip compared to 2008. The "interesting" range of truck sales had already been established in the article as 0-50000.

Arguing that axes should never be cut is silly indeed, but axes should also show the relevant ranges for the context (here the 2008 values for both x and y would make today's fluctuations look minor).

Reply on news.ycombinator.com