h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

I really think putting formulas on the picture itself will make a lot of people turn off in shock - but maybe I have an underrated opinion of humanity...

Barton, I think you're on to something. (or is that "onto something"? nah.)

It seems that it boils down to pedagogy vs informativeness, as always. Let's get ready to rumble...

Generalised formulae: Easier to read, easier to understand, easier to remember.

Specific formulae: More informative (although that's probably subject to discussion), more balanced. However, readers might "turn off in shock".

*ding ding*

It's a really hard choice. Although it might look as if pedagogy is winning, I think readers who totally despise mathematics would probably just look at the pretty picture and not care at all about the formulae.

"Looks like the tables are turned, Mike."
"Sure does, John. Informativeness is making a great comeback, folks, and Pedagogy is really taking a beating!"

On the other boxing glove ("Ow! That has ~got~ to hurt!"), readers with some knowledge of maths -- ie the average reader -- will probably look at those formulae and wonder what the heck they are. ("Sure did, John. Pedagogy is back in the game.") I think generalised formulae would help them understand what's going on.

Unless anyone else has some good arguments for specific formulae, it looks as if Pedagogy will win this one. Probably a TKO. (ahh... good ol' 8-bit NES "Punch Out"... )

It should probably be noted that I do not condone boxing, fighting, violence of any kind, or illegal distribution of Nintendo Entertainment System ROMs. "Punch Out", "NES" and "Nintendo Entertainment System" is a registered trademark of Nintendo. I am not in any way affiliated with Nintendo, and the opinions I express do not reflect those of Nintendo, h2g2 or the BBC.

Would it help at all to use y=tanh(t) for the sigmoid, possibly with appropriate translations (like y=1+tanh(t-1) or so)? It would make the equation simpler looking, but a lot of people won't know a hyperbolic tangent from Adam. I guess those are the people who aren't going to look at the formulas anyway... It's just the simplest equation I know for a sigmoid, so I thought I'd mention it.

GTB

For what it's worth, *I* think y=1+tanh(t-1) is simple looking enough and even if they don't know when to use a hyperbolic tangent (and I am one of those 'they'), they can still plug it into their spreadsheets and play with it which is at least part of the point. (Please note I have NOT performed that exercise yet and I have no idea how it scales with the rest -- hey, it's not MY article. I'm just a critic in the bushes.)

Oh! I'm glad you didn't say that I was on something, too.

Barton

The tanh() function would look like this in order to fit in:

3.8 * tanh(1.2 * x) + 4.1

Might be an improvement, but I'd have to do the curve pics all over again, since the tanh levels out at y = 8, and my current sigmoidal only goes up to about 4. This also means that the hyperbolic curve will look even weirder and the exponential curve will look lamer.

Frankly, I don't think it's worth it. I'd rather opt for simply skipping the formulae and just show what the curves look like.

Try 2.5 + 2*tanhx

On my graphing utility, it looks pretty good next to the other two, and it levels off at 4.5

It doesn't exactly go through the point (-1,1), but it goes through (-1, .977), which is really close.

I'd say include the formulae in a footnote. Those who hate them will have no trouble ignoring them, and those with a mathematical bent will be pleased to find them at all. It would be irritating if they're just not included.

GTB

BTW, nice graphs

Is that a PI? superscripted??? How do you make my browser do that when we aren't supposed to have GuideML or HTML in this section?!?
Razterfrazl Nurdle Draster Frotzer . . .

Both of those formulas would probably be better in footnotes, I'm afraid. But, how about pretty little duckies on the border?

Barton

2.5 + 2 * tanh(x) doesn't look very good to me. However, I have to thank you for the tip, because I get a rather good fit with 2.9 + 2.5 * tanh(x), which is nearly as simple. It goes up to y = 5.4 or something, so maybe I could just accidently lose the "scientific-looking" graph and pretend that the sigmoid was 2.9 + 2.5 * tanh(x) all along.

Actually, I'm not that lazy. I've made a new graph using the tanh(x) curve, and I must say that I rather like how the sigmoid turned out. Now you can really see how it levels off.

Thanks for the footnote tip, that might work.

Garge? Duckies?! Barton, you're scaring me.

Gargle? Duckies?! Barton, you're scaring me.

Hm. I blame an anomaly in the time-space continuum.

Yes, I like the new graph. It's easy to talk about how the two graphs can be taken for each other in early stages but can be seen to diverge.

Sorry about the there was a kink in my oxygen line.

Also sorry about the superscripted pi business too. It turns out that a splat(*) looks a lot like a superscripted pi in this font and late at night with my contacts fogged I thought . . . well -- never mind.

And, what's wrong with duckies?

Barton

Not withstanding how singularly thick it made them feel, the powers that be pretty big but not as big as the powers that be the directors of the BBC like this entry, and are happy for it to enter the guide.

Please be patient while it goes through the editorial process - this'll take time, particularly if you get the picture you're after - I'll be crossing my fingers for you.

On a side note, since Sam did mention that it made him feel a little thick, I think that's a killer blow struck for pedagogy!

Hooray!

Barton

Woah. This is my first edited entry! I can't even begin to describe how cool this feels. :D I wish there were , and smileys...

YAY!!

Actually, it ~will~ be my first edited enty, unless the Earth is being demolished sometime in the near future. But that's not the point of this posting.

I'm just wondering whether I should remove all but the best graph, ie the tanh(x) graph. I wouldn't want to confuse the artists...

Also, if they do make a graph or just stick mine in there, I need a way to insert a few comments about the sigmoidal curve and a footnote with the formulae.

How easy/hard will it be to discuss these things with the sub-ed?

depends on the sub-ed. Some are nice and check out the forums of entries they sub-ed. Some are nasty and just grunt a lot. I'm afraid it's rather luck-of-the-draw - they *ought* to check this thread and others on the entry, but sometimes they get too busy...

regardless of which, if you post to the attached forum when it becomes "pending", then the inhouse team will see what you write and fix it appropriately, even if the sub-ed doesn't.

There's more about the editorial process here: http://www.bbc.co.uk/h2g2/guide/SubEditors-Process

Thanks for the info.

Editorial Note: This thread has been moved out of the Peer Review forum because this entry has now been recommended for the Edited Guide.

If they have not been along already, the Scout who recommended your entry will post here soon, to let you know what happens next. Meanwhile you can find out what will happen to your entry here: http://www.bbc.co.uk/h2g2/guide/SubEditors-Process

Congratulations!