I think one of the problems with science journalism is that before a science story can be reported in the news media, someone has to convert everything from metric to journalist units. But some recent work may allow us to do science directly in journalist units, thereby making scientific papers immediately understandable to laypeople. According to a throwaway letter in the Guardian

1 billion Hiroshimas = 1 Isle of Wight × 20 speeding bullets

This is based on a G2 article about the asteroid that wiped out the dinosaurs that described the asteroid’s mass, energy and speed in those terms.

Unfortunately, the equation is wrong, as you can’t multiply speed by islands to get explosions. But it’s not far off. In fact, kinetic energy K = ½mv², where m is mass and v is speed — so actually

1 Chicxulub asteroid strike = 10⁹ Hiroshimas = ½ Isle of Wight × (20 × speeding bullet)²

I’ll forgive the correspondent the factor of 2, but he should have known that the speed needed squaring given that the only other sentence in his letter was “Who needs E=mc²?” Honestly, it’s as if some people have no grasp of dimensional analysis at all.

In fact, this is also wrong, because the Isle of Wight is more correctly a unit of area, not mass, so to use standard journalism units, we should really write

1 Chicxulub = 10⁹ Hiroshimas = ½ (Isle of Wight^³⁄₂ ρ_rock) × (20 × speeding bullet)²

Better still, that should be

1 Chicxulub = 10⁹ Hiroshimas = ½ (Isle of Wight^³⁄₂ ρ_rock) × (21 × speeding bullet)²

as the rock hit 20 times faster than a bullet, not 20 times as fast as one.

And we can test this hypothesis simply by typing “(1/2) * (density of rock * (isle of wight area)^(3/2)) * (21 * speed of bullet)^2″ into Wolfram|Alpha. It returns the figure 5.023×10²³J, and if you click on that figure, it rephrases it to “≈ 1.005 × estimated energy released by the Chicxulub meteor impact”.

Let’s just bask in the impressiveness of that for a moment.

Done basking? Then it’s time to admit there are a few problems with this. Alpha cites this as 8 billion Hiroshimas, not one billion. Alpha also takes ‘a bullet’ to be a rimfire .22LR usually deployed against small pests and tin cans, whereas Dr Collins appears to favour the somewhat meatier M16 assault rifle. Maybe that’s standard for a speeding bullet. Also I assumed the asteroid was a sphere that would cover an area of land equal to one Isle of Wight. In fact the Isle of Wight is long and thin so if we spun it around its major axis it would be a bit lighter than this; equally we could attempt to estimate the mass of the Isle of Wight and that could go either way.

The point is that you absolutely can do science in these units. They totally work. We use metric instead only because the numbers are easier — 1 Joule is 1 kilogram metre per second squared, avoiding having the annoying factor of 21 kicking around that the journalism units version above does. (I’m not going to quibble about the billions, though, as you only need to define the ‘gigashima’ to make that go away.)

To make life easier for anyone choosing to do science in journalism units, I have identified some relationships that may prove useful:

• 1 coal-fired power station ≈ 1 Hiroshima per day
• 1 thickness of human hair ≈ 1000 Olympic swimming pools per area the size of Wales
• 1 weight of a double-decker bus ≈ 1 Hiroshima per distance to the moon and back

All of these are approximate, but they’re all exactly true for at least one combination of reasonable guesses, so all we have to do is identify a mutually-convenient set of plausible values, then agree to use it forever. We can’t really fiddle with

• 1 Isle of Wight = 381km²
• 1 distance to the moon and back = 3.85×10⁸m
• 1 Wales = 20,779km²

and we know that defining

• 1 speeding bullet = 340ms⁻¹
• density of rock = 2.65g/cm³
• 1 Chicxulub asteroid strike = 5.023×10²³J

gives us one neat relationship. Let’s add to that

• 1 Hiroshima = 1 Chicxulub ÷ 8 billion = 6.27875×10¹³J
• 1 coal-fired power station = 1 Hiroshima ÷ 24h = 726.7MW
• 1 double-decker bus = 1 Hiroshima ÷ 1 moon and back = 8.416 tons
• 1 Olympic-size swimming pool = 2,500m³
• 1 thickness of a human hair = 1000 Olympic pool ÷ 1 Wales = 120.3μm

Now all our relationships are spot on, and we can hopefully get on with doing some science with journalism units. At least, science that involves using coal to power buses to the moon.

Ie, the best science.

It’s a common misconception that the Qwerty keyboard is designed to slow users down to prevent typewriters jamming. It fact, it’s designed to keep commonly consecutive letter pairs apart, so that two adjacent levers won’t collide.

(A more fun, but irrelevant, Qwerty story is that it is also designed such that the word ‘typewriter’ is all on the top row, to make demonstrating it easy. This story, if true, is itself fun but sucks all the fun out of the fact that the longest word that can be typed on the top row of a typewriter is ‘typewriter’. One of these is a fun fact, but I’ve no idea which.)

Nowadays, obviously, there are no swinging arms to collide, so we want the commonly-used keys to be reachable, and if possible to alternate hands as much as possible. Dvorak and Coleman have each had a stab at designing a better layout, but both aimed at the computer keyboard.

But increasingly, I type on my phone, using one very mobile thumb. I can get to any point on the screen, more-or-less right away — but sometimes I miss, and usually the phone figures out what I meant and autocorrects it. So maybe the most important thing about any given keyboard layout is how likely it is that a typo will result in a real word that the phone isn’t to know isn’t what I meant.

I wondered if suddenly Qwerty might be optimal again — separating pairs of letters that can be swapped to make another real word and that appear next to each other in English words aren’t totally different goals. So I thought I’d investigate.

So first I loaded the CSW12 Scrabble word list, and worked out a big table of how many places in the list you can replace each letter with each other letter to create a new word.

As you can see, the letters involved in typos that are genuine words are also the most common letters — except C and P. (The frequency values are on an arbitrary scale to match the typo figures.)

Then I wrote a Python routine to generate a ‘badness’ score for each layout, which is the total number of words you can make by replacing a letter of another word with one of the six keys adjacent to it. Running it on 10,000 random layouts, the average badness is around 83,603, with a standard deviation of 14,024.

Here are some other layouts I tried:

(Predictable answer to question in title: “haha, no”.) Alphabetic uses the same key layout as Qwerty: 10 on the top row, 9 on the second and 7 on the bottom. The ‘best I found’ layout was derived from a random board on that Qwerty grid (since actually Dvorak and Coleman don’t really fit on a phone), by swapping letter pairs at random and keeping the change if it seemed to work. (This is called a ‘genetic algorithm’, albeit a crude one.) I think I did 5,000 steps, five or six times. Here’s the layout it found:

The most obvious thing it’s done is put S (the most typoable letter) in a corner and shoved Q up against it. Another potential improvement to the model is to account for second-nearest neighbours — since flagging an error but correcting it to the wrong thing isn’t much better than missing it.

Another thing it’s done is put all the rarest letters in the middle where they have lots of neighbours — almost precisely the opposite of what Dvorak and Coleman did. Which makes sense, both intuitively and because all the standard layouts are in the worst 5% of all layouts (assuming normal distribution).

Anyway, I think we can all agree this is plainly the best possible keyboard layout for smartphones, and we should name it Taylak and petition Apple and Google to include it as the default for everything ever. I certainly can’t imagine how using the same layout on phones and computers could possibly be more desirable than this.

Here, to end on, is the worst layout I could find, with 204,290 = μ + 8.61σ possible real-world typos:

Nobody use that layout.

On July 5 2010, Total Film fooled a lot of people into believing that that was the day Marty McFly and Doc Brown visited in Back To The Future Part II. In fact it was October 21 2015.

Yesterday, on June 27 2012, Simply Tap fooled a lot of people into believing that that was the day Marty McFly and Doc Brown visited in Back To The Future Part II. In fact it was still October 21 2015.

To punish people for falling for such nonsense, I propose more of these, increasing in frequency as we approach the actual date, so that when it actually happens, nobody believes it. But when?

The first of these errors was 1934 days premature. The second was 1211 days premature. (I’m not counting the copycat hoax on July 6 2010.) I think the reason these hoaxes are so seductive are that $\frac{1934}{1211}=1.60$, and that’s very close to the golden ratio, $\phi$. $\phi\approx1.62$, and has the lovely property that $\frac{1}{\phi}=\phi-1$ (or, $\phi^2=\phi+1$). It’s the only number of which that’s true, and it often appears in nature, art and architecture.

(Or, if you prefer, reports of $\phi$ appearing by accident are mostly coincidence and optimistic rounding, and so is this. I’ll leave that decision to you.)

Continuing the Golden Cascade of Back To The Future Hoaxes, we should have the next on September 23 2013. There will be two in 2014: on July 4 (when the alien mothercraft destroys the Hill Valley town hall) and December 28. The hoaxes will have to come thick and fast in 2015: April 18, June 27 (like this year), August 10, September 6 and 23, and October 4, 10, 14, 16, 18, 19 (twice), and then 25 separate hoaxes on the day before Future Day.

To be honest I suggest we use areweinthefutureyet.com for those ones.

I spend a lot of time getting cross at crappy interfaces on software, but the fact is that real life objects are just as bad. I’m typing this at a laptop, for example, with a trackpad. And while Apple have multitouch, click-sensitive trackpads that make sense, this one scrolls using the right hand half a centimetre of pad, which is visually and tactilely indistinguishable from the rest of it. It’s the little things, like Jeff Atwood’s cat feeders, but it’s also the really big things. Our old DVD recorder which asked, when you put a DVD in, if you would like to “access the disc contents” — and to get to the DVD menu, you had to say ‘no’. Toasters which measure time in completely arbitrary units, at least as far as I’m aware. Washing machines that have a key to their own interface on the front. TVs that have a hundred buttons to do basically one thing and you still have to tap ‘circle with an arrow’ four times to watch a DVD.

It’s like people don’t learn. I remember when I could operate a microwave oven by typing in a time and pressing ‘cook’. Ryan North has one where you can just bark numbers at it. And yet this is the thing we have at work:

Obviously that’s absurdly overcomplicated for a device which, much as manufacturers kid themselves otherwise, has only two modes: on and off. (Really, who wants to ‘slow defrost’ anything?) But my main objection is that I have had a bagel spinning around in there for ages, with the microwave all humming to itself and lit up, and without the bagel getting so much as lukewarm. Apparently it only heats food if you set the timer. Where exactly the clue to this is supposed to be is not clear. I assumed when a microwave lit up and span your food round it was cooking. Apparently this one also has a ‘shop demonstration’ mode.

Because the fact is that outward appearances matter. I forgive the absurd obsession of washing machine manufactures with putting any and all options on dials because that at least makes their products obviously washing machines. Humans in Western society have loads of cues and associations built in — say, we pull doors with handles — and it’s daft not to take advantage of them and insane to actively work against them. Putting a handle on a push-only door will confuse and annoy, and this… This is the sink in a bar in Manchester:

Perhaps they considered it designery and artsy, but in fact what they have done is add taps to a urinal and call it a sink. The thing is that I took it for a urinal when I first saw it, and while I noticed the taps fairly quickly, I don’t trust the entire inebriated male population of Manchester to all have done so on any given night. That is a sink which I confidently predict has been pissed in and that is enough to put me off washing in it.

What you need, when cleaning yourself, is a clear interface. So don’t get this brand of shower:

I put it to you that it’s less than totally clear which way is ‘hot’ on this dial. The arrow points right, but it’s on the left. If you’re standing in a room full of steam and you wear glasses, that’s fairly dangerous. And it’s obvious what the dial does, so all they had to do was convey which way was which. That is literally the smallest amount of information it is possible to encode. And still it’s ambiguous.

Even stranger is the dimmer switch in our new conference room.

This has two buttons, one at the top of the switch, and one at the bottom. Here is how I assumed it would work:

• Pressing the top switch would make the lights brighter.
• Pressing the bottom switch would make the lights dimmer.

Here is how it works:

• A short tap of the bottom switch turns the lights on or off.
• A long hold of the bottom switch makes the lights brighter.
• A second long hold of the bottom switch makes the lights dimmer. This alternates between brighter and dimmer.
• The top switch exists only to make the object resemble a lightswitch.

The upshot of this is that we all look stupid when presenting the work of the Unit to outsiders, because we are supposed to be carefully controlling illumination in clinical trials and can’t operate our own office lights.

And I suppose I just thought that stuff like toasters and lights would have been around for long enough by now for us to have basically mastered the art of making them usable. We don’t have Steve Jobs around any more to fix this stuff up for us, so perhaps it would be a fitting tribute to him if we all stopped making watches that require one long and fourteen short button presses, each with its own high-pitched beep, just to correct for daylight savings, hmm?

Yesterday, I saw Betty Boop and the Little King. It was made in black and white in 1939, and I mention it because there’s a rather impressive 3D effect at around 1:50 (then again later).

Here’s a link from Archive.org if you prefer, but I had less luck with their embedder.

What I like about it is how it was achieved. It was done the same way all visual effects in the 30s were done: they built the thing and pointed a camera at it. That approach wouldn’t even occur to us now.

They had an entire rotating set built to create 30 seconds or so of footage.

I really like a lot of the CGI stuff we have now, and when it’s used well it’s just flawless, but it’s almost impossible to feel a sense of wonder at a visual effect in a modern film because it’s sort of clear how all of them were done. And there’s an elegance to the 1930s solution that the modern one lacks.

Apparently, there are plans to make a TV improv comedy show where viewers can tweet ideas.

This sort of thing annoys me. Partly it’s because the audience at a studio recording of @cuff (formerly “Fast and Loose”, formerly “Mock the Week”, formerly “Whose Line Is It Anyway”) is already sufficient to generate ideas, and I suspect that the hassle-factor involved in getting tickets and physically to a recording will increase the average quality of suggestions, too. If you have a room full of self-identifying improvised comedy fans, who have to shout ideas in person in front of seven professional comedians, then they’re surely bound to give you funnier ideas than Jeff and Tina McRandom sitting at home tapping out “now do it like a horror movie” and hitting “send” — not least because Jeff and Tina will post their first thought without knowing that (or wondering if) it’s horrendously clichéd.

Also, once you get a million people tweeting ideas, you can pretty much do whatever you were going to do anyway and rationalise it post-hoc by searching the Twitter stream for the ten or twenty people who will inevitably have suggested it. It would be a lot of fun to pre-record an hour’s scripted and predictable comedy, then show it as a “Twitter experiment”, with a live ticker of tweets from real users whose uninspired suggestions best matched the next bit of the script.

But mostly, they annoy me because the use of Twitter is always, always superfluous. The idea “improv show where the audience can send in ideas” is not something that became possible when Twitter launched. It was possible the moment SMS messages became commonplace. It was possible the moment webforms became a thing. Pretty sure you could do it with telephones, or fax, and other technologies that predate television. It’s an idea that has been easily doable for something like fifteen years, and yet somehow producers can’t get excited about it unless it involves Twitter or Facebook — and preferably the former because that has more jargon. And they always talk about “technology” as if sending 140-character messages to a TV studio requires some kind of fucking flux capacitor that only the mad scientists at Twitter have access to. It doesn’t. It requires a webform. It’s obvious that the creative process is less “think of a great concept, then figure out which technology could make it happen” than “think of something popular, then flick through the Radio Times until you find something to bolt it onto”.

Ultimately, though, to the average viewer, @cuff will be indistinguishable from Fast and Loose: many will be using iPlayer or watching repeats, so won’t be able to post ideas; the majority of those who can probably won’t bother; the majority of ideas that are posted won’t be read because there will just be too many to track; and the majority of ideas that are read won’t be used. So what’s the point?