Posts by Chris Miller

There is: Muphry’s Law

(also known as Hartman's Law of Prescriptivist Retaliation): "any article or statement about correct grammar, punctuation, or spelling is bound to contain at least one eror".

@hrrrrrrrrrmmmmmmm

Pity, you were doing quite well until your last paragraph

If there were anything I could do to help (rather than just get in the way), I'd be prepared to assist at my local nuke plant - knowing that I would be monitored for radiation exposure and taken out once I'd exceeded the safe limits (250mSv*, only permitted in situations where I would be contributing to saving a life). Meanwhile, due to the "slightly" (slightly!!) exaggerated statements from the press, people in desperate need because of a once in a millennium tectonic event are being deprived of supplies because drivers won't deliver to areas where they fear 'hazardous' levels of radiation.

Very few things in the real world are "absolutely safe" (as that BBC savant, Jonathan Dimbleby, kept asking on 'Any Questions' yesterday) - the sensible question is whether nuclear is safer (and/or less damaging to the environment) than other forms of energy production - to which the answer (on most bases) is 'yes' - 200 times safer (based on deaths per kWh) than wind power, for instance.

Xkcd (natch) has an excellent guide to radiation, here:

http://xkcd.com/radiation/

* this level of exposure probably (best estimates suggests) raises your lifetime cancer risk by 1 or 2% - but that risk is something like 20-25% to begin with.

and another thing

One significant difference between being an employee and being self-employed (whether via Schedule D or setting up a limited company) is that you can offset expenses against earnings, so items such as travel to the office are paid out of gross income. Now, in most European countries (France and Germany, not sure about others) ordinary employees can also offset travel costs. So your season ticket in France (which will cost << half what it does in the UK, to begin with) is paid for out of gross income. If employees want to complain about the unfairness of the tax system, I reckon this would be a good place to start (and it would encourage the use of public transport, and so tick the 'green' box as well).

The dollar costs are per TPM

So the SPARC system costs ~$20 million, while the Proliant is 'only' ~$2 million. Whether the Proliant could be scaled up 9x without substantially increasing the cost per TPM is an open question - my guess is it would still cost <$1.

"work as hard as people in communist Russia"

"We pretend to work, and they pretend to pay us."

Geordie/Scots

If you want to compare these two, then you'll need to specify which type of Scots. Official Scots as recognised on the Scottish Parliament website* is really Lallands - the Ayrshire speech in which Rabbie Burns wrote his "Poems in the Scottish _Dialect_" (my emphasis). But why insist on this version of 'Scots' as opposed to say Aberdonian or Midlothian (both closer to Geordie)?

* http://www.scottish.parliament.uk/vli/language/scots/index.htm

"hence I lost all my data"

No, you lost all your data because you failed to make a complete backup before attempting to install a new OS.

@Doc

I assume your doctorate isn't in statistics or a related discipline.

You're right that applying broad statistics to individuals doesn't work, and if you operated an insurance business on only 2 individuals you'd risk severe difficulties. But insurance companies insure large numbers of individuals and so the statistics are (almost) certain to work for them.

Just as when I toss a fair coin I know there's a 50% chance of heads - but on any individual toss I can't predict the outcome. If, however, I toss the coin a thousand times I know that it's (almost) certain that I will see between 400 and 600 heads. If I insure a thousand each (comparable) male and female car drivers, I will almost certainly pay out more in claims on the male drivers.

Still not understanding

Rob, you appear to be really interested in this topic - if so I would advise you to attend a statistics course and learn about Bayesian inference (easy to learn, much harder to understand). I repeat, there is no doubt that males experience more motor claims than equivalent females. The null hypothesis (that males and females experience the same rate of claims) is easily demonstrated to have much less than one in a billion chance of being correct. Your alternative suggestion of dividing the population by name into A-M and N-Z would just as easily be proven false.

Now, it's always possible to say "maybe there's a hidden variable - perhaps shoe size or length of hair is the real determinant" and (like any counterfactual) it's impossible to prove with 100% certainty that this isn't the case. But you can be sure that a whole lot of actuaries, who have to know a great deal about multi-variate analysis, have looked into such possibilities. After all, if you could find one that worked, you could make a lot of money by providing special rates for people with small shoe sizes or long hair.

But don't simply take my word for it. David Spiegelhalter* has written far more eloquently than I about the lunacy of this decision. Unfortunately, it's behind The Times's paywall:

http://www.thetimes.co.uk/tto/opinion/columnists/article2925741.ece

* Winton Professor of the Public Understanding of Risk at Cambridge

http://www.statslab.cam.ac.uk/Dept/People/Spiegelhalter/davids.html

Another vote for unpasteurised

When I were a lad, we had milk delivered from the local farm - 'green top', which meant unpasteurised from TT (tuberculin tested) cows. Unpasteurised milk is to pasteurised as pasteurised milk is to UHT.

Just for clarity (I agree with the principle)

It's not true that you can turn right at reds *throughout* the US - it varies from state to state. Try it in the wrong spot and you're very likely to get a ticket.

@Vladimir

Your presumption is wrong, I have made no claim about harmful or beneficial effects of pornography. What I have said, although many seem incapable of reading or comprehension, is that I am not surprised that someone who makes a living from prostitution and pornographic writing would claim that pornography is not harmful. No doubt Richard Desmond could be called upon to second the motion.

@Sorry ...

No, I haven't particularly noticed an increase in people carrying DSLRs - certainly for every one I see, I see 10 compacts. Maybe I don't get to the right 'events'. And if you're thinking of spending a grand on a body + a similar amount on lenses (hardly my idea of a mid-level system), I suggest that ElReg probably isn't the place to look for camera reviews - $deity knows there's no shortage of specialised camera sites.

I used to be a big fan of 35mm SLRs, with multiple bodies and lenses. Now I just take my compact superzoom. Maybe there's a few percent loss of quality, but I can have it with me all day without my back complaining about lugging a 6kg bag around with me.

Read up about the Medieval Warm Period

When there was no (or very little) polar ice about for a couple of centuries. Did Polar Bears become extinct as a result? No they retreated to the land, where there was still plenty of ice. This was 'only' 1,000 years ago.

A climate model that predicts a global meltdown based on the loss of Arctic ice cover is demonstrably incorrect, since it contradicts the historical record.