Graphic Methods for Presenting Facts (1914) | Hacker News Reader

Graphic Methods for Presenting Facts (1914)

archive.org

What an outstandingly beautiful book. The book is strangely familiar yet foreign at the same time - the graphics would work in any modern setting, yet the language is adorably dated. For some reason beautiful late 19th/ early 20th century books are astoundingly beautiful in their black and white simplicity, and I am always left yearning something of similar aesthetics in modern setting.

If you like this, you should probably get hold of a copy of 'The Visual Display of Quantitive Information' by Edward Tufte, which collects loads of similar examples from various points in history.

Yes, Tufte's work is very good. I also like his war against powerpoint. A futile war unfortunately. Sigh.

Speaking of beautiful books: I have a bunch of old radar books (I assume it was some engineer who passed away), from the 1940s to 60s, and although there are some bits that definitely haven't aged well they are mostly still useful.

The things that amazing though is just how nice to hold and read they are. They are astonishingly well typeset, legible graphs, silky paper etc.

I checked out some of the prices in the back and they are similar to what one would expect for a new hardback textbook today, sadly. So books have got worse apparently.

I especially love the drawings and the lettering. Modern graphics produced on computers are great, but as a culture we are losing the ability to produce graphics by hand.

Some time ago I think HN had links to New York Times graphics from the 1940's. Great stuff but I don't have the link.

The majority of plots are strictly black and white, and I'm interested in how legible they are even when they're showing lots of data.

I feel like there are some techniques that are lost when color is so easy to use.

Also there are some simple touches -- like gridlines that only go as far as they need to (e.g. page 23) -- that would be trivial for a hand-drawn plot but quite difficult in "modern" tooling.

Colour is easy if you have a colour printer. in 1914, b&w printing was significantly more common than colour for book production. The reason most old books have a "plates" section is that colour printing required higher quality paper and more expensive printing techniques.

You can roneostat b&w and you can hand-cut graphics onto a roneostat sheet.

If you've never worked with carbon paper, it was a messy business.

Hm, perhaps worded wrongly but I agree. I'm saying that the plots in this book were working under limitations (like black and white) and yet produced excellent results through good design and techniques that are less-used nowadays.

The author makes many mentions of color in the original charts he cites and how it was used to improve the designs which already work well in B&W.

Lots of professionals (e.g. Data Journalists) still do those finishing touches in tools like Adobe Illustrator.

Here's a NYT tool to convert d3.js graphics to SVG to allow manual editing: https://github.com/NYTimes/svg-crowbar

There's an old rule in graphic design that "it has to work in B&W first."

> Organization charts are not nearly so widely used as they should be. As organization charts are an excellent example of the division of a total into its components, a number of examples are given here in the hope that the presentation of organization charts in convenient form will lead to their more widespread use.

If Mr. Brinton were still alive today, he would be a very happy man...

Is there an equivalent source for presenting ideas? For my sins I prepare a lot of sales architecture slides and I often struggle to find diagrams beyond the smart art graphics in Office and Google Slides (as good as they are).

Some of the stuff in there is very good, and some is at the level of "chartjunk" as Edward Tufte calls it in his also very good series of books about graphical display of data. (his books are also worth a read)

(Chartjunk = plots or charts where flowery or gratuitous graphical embellishments, like the octopus, simply obscure the data being presented)

I am actually most surprised by the fact that a book was able to be produced with so many graphics at that time. I would have thought that such graphics were incredibly expensive or prohibitive to put in a book until later in the 1900s. I don't know why but I just had that impression.

The "chartjunk" here is generally presented as a counterexample. The octopus on p. 20, for instance, is explicitly called out in the caption as problematic because you can't accurately see the data.

For another example, you can take a look at page 39. You'd probably consider Fig. 40 to be "chartjunk," but it's there to contrast with Fig. 41, which shows the same information in a much better way.

I would say a careful read of the actual text suggests that most of the examples are intentional chartjunk, with great commentary on why they should be thought of that way.

In total, this book is very much like Tufte with as many bad examples as good -- on purpose.

octopus on page 20 knows how to party

Some of these diagrams are beautiful, I'd like to get a big print of routing diagram for materials and printed forms in a manufacturing plant.

Does anyone know of an OCR-ed version of the book?

https://archive.org/stream/graphicmethodsfo00brinrich/graphi... ?

Not pretty but the actual text seems fairly error-free.

Acrobat OCR'd it when I opened the PDF. I'm sure several websites could do the same.

Amazing. Cookbooks are still in fashion, too.

The fact is, those graphic methods can also be used to present lies.

people have been telling lies, damned lies, and statistics stories since the birth of statistics.

Upvoted for Twain.

Twain would have loved TV and the Internet. He'd be Colbert.

Reply on news.ycombinator.com