Print Story Bonkers in the nut
Diary
By TheophileEscargot (Sat Nov 02, 2013 at 01:47:35 PM EST) Reading, Theatre, MLP (all tags)
Reading: "Broken Homes". Theatre: "Nut". Links.


What I'm Reading
Broken Homes. Fourth book in the "Rivers of London"/"Peter Grant" series about a magic-using London policeman. I got this on impulse after finding out the new Dresden Files was delayed till next year, but it's a good series in its own right.

A suspicious death takes Grant to a fictional tower block in Elephant and Castle. Seems loosely based on Erno Goldfinger and Trellick House, but the design is very different.

Good fun and entertaining. Not particularly outstanding though. Feels a bit like it's marking time until the big plot development at the end..

Theatre
Saw Nut at the National Theatre shed. Low-key play about a mental illness. Well written and well acted, with some plausible and sympathetic characters. It's a bit lacking in events though: it's an interesting slice of life, but doesn't have much resolution

Having watched "The Iron Lady" and a couple of other things lately, I found it pretty obvious from the start that some characters were figments of her imagination. The play might have more of a feeling of drama if that comes as a revelation.

Overall, well-crafted and compelling, but not essential watching.

Web
Socioeconomics. Against forecasts: "You don't expect a doctor to tell you what your next ailment will be". Supermarket wine price graphs. Fewer women in tech: female undergraduates fell from 37% in 1985 to 18% in 2010.

Video. 17th century London flythrough. Starships were meant to fly. Phone call through electromechanical exchanges. Tow truck fail. Wind sculptures.

Pics. Top of the Golden Gate Bridge. Dog catches frisbee in bullet time. Architects build yachts. Museum of Soviet arcade games.

Politics. Whatever happened to the Left's passion for full employment? (Not a great article, would be interesting to see exactly how unemployment went from moral failure to macroeconomic issue and back to moral failure again). British journalists lock each other up. Vigilantes murder man falsely accused of paedophilia. Universal credit: £120m could be written off to rescue welfare reform. Shoesmith didn’t get a fair hearing: FlipchartFT issues a well-deserved I Told You So.

Random. Tunnel under Mexican/US border. America needs fewer time zones.

Sci/Tech. Amazon supports Static JavaScript Apps.

Articles. Dr Who fan same age as series looks back. When and why did it become cool to use both straps of a backpack?

< Spy Scandal Part XXVII | Books I've Read This Year 2013 >
Bonkers in the nut | 20 comments (20 topical, 0 hidden) | Trackback
Broken Homes by Merekat (4.00 / 1) #1 Sat Nov 02, 2013 at 03:36:09 PM EST
In the main, I was disappointed. The architectural diversions were ok but mainly it existed to hang one plot item on. Not quite as irritating as George RR Martin as I didn't have to wade through as much shite to find this out.

Doctors do make forecasts by lm (4.00 / 2) #2 Sat Nov 02, 2013 at 04:31:06 PM EST
E.g. if you keep doing x, y is likely to happen or you should do a so that b is less likely to happen or you have a history of z in your family and c behaviors make z more likely so you should minimize c to make it less likely that you'll come down with z.

After all aren't vaccines recommended precisely on forecasting? I mean, when my daughter went to Nepal, the doctor recommended vaccines that don't usually get recommended stateside because she's more likely to encounter certain diseases in Nepal than in the US.


There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
As he says: by TheophileEscargot (2.00 / 0) #3 Sat Nov 02, 2013 at 05:21:43 PM EST
We must distinguish between forecasting and predicting. A prediction is a conditional statement, a hypothesis, whereas a forecast is an unconditional one. "If you cut prices, then ceteris paribus, demand will rise" is a prediction.
If the doctor said "you should get vaccinated because it reduces the probability of getting malaria", that's a prediction. If he said "You should get vaccinated because otherwise you'll catch malaria in Nepal", that's a forecast.
--
It is unlikely that the good of a snail should reside in its shell: so is it likely that the good of a man should?
[ Parent ]
I don't think he would word it that way by lm (4.00 / 1) #4 Sat Nov 02, 2013 at 05:55:43 PM EST
If you go back to the bright line that he draws, we can apply that to way that doctors talk about certain diseases. A prediction is a conditional statement, a hypothesis, whereas a forecast is an unconditional one. "If you cut prices, then ceteris paribus, demand will rise" is a prediction. "Demand will rise" is a forecast.

So: "If you take the vaccine, then you'll be less susceptible to malaria" is a prediction. And it is probably true whether you're in the US or Nepal.

"In Nepal, you'll be more likely to be exposed to malaria" is a forecast. Doctors make this sort of forecast all the time.


There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
[ Parent ]
Well by TheophileEscargot (2.00 / 0) #6 Sun Nov 03, 2013 at 02:00:41 AM EST
If we accept this distinction between a "forecast" as unconditional, and a "prediction" as conditional, then being "more exposed" would seem to fall into the conditional/prediction side of the distinction. Being "more exposed" is not really a state like "having malaria", it's an expression of the likelihood of entering the state of having malaria.
--
It is unlikely that the good of a snail should reside in its shell: so is it likely that the good of a man should?
[ Parent ]
Really by lm (4.00 / 1) #7 Sun Nov 03, 2013 at 06:46:57 AM EST
If that is the case, almost no one makes forecasts and almost everyone makes predictions.

For example, weather forecasters say tomorrow we have an 80% chance of rain. Stock analysts say that "conditions are ripe for this stock to go up".

But I think the "conditional" in the definition is supposed to mean mean something more like a material conditional in logic, i.e. an if/then statement.


There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
[ Parent ]
Exactly by TheophileEscargot (2.00 / 0) #8 Sun Nov 03, 2013 at 06:59:02 AM EST
That's the point he's making about some critiques of economics: "why didn't they forecast event X at time Y". Almost no other profession is expected to do that, barring astronomers and eclipses.
--
It is unlikely that the good of a snail should reside in its shell: so is it likely that the good of a man should?
[ Parent ]
Except, again by lm (4.00 / 1) #9 Sun Nov 03, 2013 at 07:30:03 AM EST
Doctors and other professionals do it all the time, e.g. "you have six months to live."

As an IT profressional, I've advised clients that sooner or later, probably sooner, someone is going to try to hack them.

Mechanics often say things like, "you're brakes are in really rough condition, they are going to fail"

All of these really should be couched in language of probability but even if they were, they'd still be forecasts rather than predictions in the scientific sense.


There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
[ Parent ]
Um, I'm confused by TheophileEscargot (2.00 / 0) #10 Sun Nov 03, 2013 at 09:30:50 AM EST
Earlier you seemed to be saying forecasts (as opposed to predictions) are almost never made. Now you seem to be saying they're made very often.
--
It is unlikely that the good of a snail should reside in its shell: so is it likely that the good of a man should?
[ Parent ]
I think you misread what I've written by lm (2.00 / 0) #12 Sun Nov 03, 2013 at 02:01:50 PM EST
It was prefaced with a conditional, "if that's what a forecast is ..."

Regardless, I've posted to Stumbling and Mumbling asking for clarification on the distinction that is being made. The original author probably has better insight into what was meant than either you or I.


There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
[ Parent ]
difference between forecast and prediction by bobdole (2.00 / 0) #13 Sun Nov 03, 2013 at 03:11:59 PM EST
differs between different research communities. The only real insight here lies with the original author.

Earthquakes is a great example, lots of forecasting - but no predictions (well accurate ones anyway). Similarly forecasting flu-season is common, but predicting when you will be infected by a particular strain is uncommon.

Weather on the other hand (even though it is often called forecasting) is a great example of predictions. When your weather reporter reports 40% chance of rain, it is because of inference based on data about time and space, not because it on average rains 40% of the time in city X. Though both might correct, but you would expect the former not the latter.
-- The revolution will not be televised.

[ Parent ]
I don't think weather forecasts are predictions by lm (2.00 / 0) #15 Sun Nov 03, 2013 at 04:34:09 PM EST
Or at least not according the stated definition. They aren't testable hypotheses. You can't run 8pm on Nov 3rd, 2013 by 100 times to see if it rains 40 of those times.

Rather, the meteorologist is offering a forecast by the definition put forth on stumbling and mumbling: tomorrow state x will obtain, where x is a 40% chance that rain will fall.


There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
[ Parent ]
but they are by bobdole (2.00 / 0) #18 Mon Nov 04, 2013 at 10:54:20 AM EST
most weather predictions aren't based of a single simulation, you do several simulations with slightly different input parameters and then some trained boffin will look at them and adapt them to the stuff "that only humans can do" (i.e. weed out the more unlikely scenarios).

Tomorrow comes around and you can test your hypothesis, 1 simulation predicted 1" of rain, 2 simulations predicted 4", and so on. The fact that they don't give a point estimate when presenting doesn't mean that there isn't one or several behind.

The fact that they give "chances of rain" is a way of conveying uncertainty more than "stumbling and mumbling" - the truth is that predicting the weather is hard and trying to make it sound like your prediction is spot on is misrepresenting your results (which does include uncertainty).

A forecast would be it is 30% chance of rain in London tomorrow, based on the fact that London sees on average 109 rainy days per year. In the long run, I would be correct - but in the short term I would be bested by most weather reporters (and meteorologists notice the difference).
-- The revolution will not be televised.

[ Parent ]
by that metric by lm (2.00 / 0) #19 Mon Nov 04, 2013 at 02:22:54 PM EST
Every forecast is a prediction. The example that was given, 'stocks will rise' (or whatever it was) is preeminantly testable in just that fashion. So if that's what was meant, the distinction refutes itself.

There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
[ Parent ]
well by bobdole (2.00 / 0) #20 Mon Nov 04, 2013 at 04:28:30 PM EST
if testable was the only condition - yes. However, I'm pretty certain nobody stated that was the only condition.
-- The revolution will not be televised.
[ Parent ]
meteorology by bobdole (2.00 / 0) #14 Sun Nov 03, 2013 at 03:15:29 PM EST
is a science mostly based on prediction... lots of the applied economics is prediction based as well, unless I've misunderstood investment banking it is usually based on predictions for particular objects over a particular time rather than a general forecast that the economy is growing/shrinking.
-- The revolution will not be televised.
[ Parent ]
San Diego by ucblockhead (4.00 / 1) #5 Sat Nov 02, 2013 at 10:13:21 PM EST
That Guardian map is funny, the way it labels just the city of San Diego.  The borders of the city are strange, and not very meaningful because there are a number of cities that used to be part of San Diego itself but self incorporated.

Those tunnels are discovered constantly, and part of the reason is that both Tijuana and the San Diego Metropolitan area run right to the border.

I don't see much point in changing time zones in the US.  The other coast would still be three hours different, and in any case, I deal with people in so many other zones that it seems it hardly makes a difference.

Unless you are talking a single time zone for the whole country like China.  That means going to work hours before dawn in the Winter and trying to sleep with the sun setting after ten in the summer.  Fuck that.

---
[ucblockhead is] useless and subhuman

Tow truck fail by Herring (4.00 / 1) #11 Sun Nov 03, 2013 at 01:46:59 PM EST
You'd be amazed at how often stuff like that actually happens.

We had a case of ... no, mustn't go there. But it took them an hour of searching in the dark with torches to find the car.

christ, we're all old now - StackyMcRacky

Post it in the Hole by wiredog (2.00 / 0) #16 Mon Nov 04, 2013 at 06:54:48 AM EST
For our (and GCHQ's) amusement.

Earth First!
(We can strip mine the rest later.)

[ Parent ]
I bet you say that to all the boys by Herring (4.00 / 3) #17 Mon Nov 04, 2013 at 06:56:52 AM EST
[n/t]

christ, we're all old now - StackyMcRacky
[ Parent ]
Bonkers in the nut | 20 comments (20 topical, 0 hidden) | Trackback