Metagaming

I'm guessing that most people won't have a clue what I mean.  Not just in general but specifically to the title of this blog.  My initial thoughts before writing this blog are to meander through the power of the mind, the placebo effect and then neatly on to having a pop and my board gaming buddies. At least one of my buddies is one of the few people likely to read my blog.  Funny, this statement put me in mind of Bilbo Baginses speech at his n'hundred'n'nt'elevntieth birthday:

"I don't know half of you half as well as I should like, and I like less than half of you half as well as you deserve."

The picture is of tangential relevance only to the blog but over on my gaming blog, in order to find an illustrative picture to add "flavour" to the text, I tend to search google with the theme of the blog and use the preface "fine art".  So this is an image of "fine art statistics".  Nice.

Meta- can be used as a prefix to mean "beyond" or in some sense relating to the subject it prefixes.  Metadata is an integral part of my working life and it appears metagaming is equally entwined within my social life and this got me thinking about how displaying results in an intuitive visual way, allowing us to truly see the trends, actually encourages us to want to influence them.  For example, I currently weigh just over 13st (I use *st* as any US readers will need to get their calculators out and then remember whether its 16 or 14 for the multiplier and will probably give up - as I do when someone says 180lbs - doh) the mere fact of recording my weight makes me want to change it, thankfully I tend to forget and then about once per week I have a disappointing few hours whilst my brain endeavours to erase what I just saw when I stood on the scales. 

[note to self, its long enough after the theft of your beloved Commencal Meta 5.5.3 to buy a new bike - quick before you burst].  

I know that if I put up a chart and routinely measured my weight I would be motivated to make the trend move in the right direction.  

Back to games.  

A few weeks ago I programmed a spreadsheet to use an adapted chess ranking formula to rank the performance of my gaming buddies and me - more detail here - the existing method was not really capturing the performance overall.  The method allowed me to enter the results for the previous few months and to display a performance ranking graph.  What is interesting is that this does bring out the metagamer.  One player has consistently won, or more accurately, consistently won when it mattered, and maintained his #1 Seed ranking since the very beginning.  Now I must add that there are many inadequacies with the scoring system, for example if we play only games that one person does well at then they will rise to the top, this wouldn't happen in chess as it's the same game for each calculation of rank.  It is like saying the top seeded tennis player will be determined only on the speed of his first serve - this would favour fast first servers.  In the long run, as long as the games we choose aren't selected any differently due to the result, this should average out (regress to the mean) and there will not be a systematic bias towards a particular game type.  But here in lies the problem, we are only human, one guy is winning so we should all try to beat him and choose games he is no good at!  I'm not advocating this stance or accusing anyone of saying it,.... er .... ok I guess I am accusing anonymous of saying it but the point stands.  We are only human, we see rankings and we start to want to improve them.  We automatically want to play the metagame.  

The placebo effect is real.  Whilst it may not be a real treatment or any treatment at all the effect does exist.  In many cases it can be put down to the added attention of the staff running a clinical trial but not always.  It can be confounding factor in so called complementary medicine reported success.  However, the fact remains that we have the power within our minds to affect our health.  As a scientist, in a medical related discipline, I believe it is important to qualify that I am not advocating complementary medicine, just noting that we can appear to influence our own health/bodies.  I found it tragic just recently to read that the fabulously intelligent and inspirationally creative Steve Jobs probably accelerated his early death by trusting in "magic" medicine.  

We should all start collecting statistics about our health and recording them on a personal dash board.  Then we wouldn't be able to help ourselves, we'd have to win - or go to the health professionals when we were clearly loosing.

Lies, Damn Lies & Statistics

This much quoted phrase, with an inherent progression starting at a low point, developing to an even lower point and then ending with statistics, is (according to wikipedia) potentially un-attributed – or at the very least – avoided attribution.

On the today programme on radio 4 this morning, Francis Maude the Minister for the Cabinet Office and Paymaster General, was mauled by Evan Davis.  The opponent in the debate was Mark Serwotka, general secretary of the Public and Commercial Services Union, who was on the show defending the apparently indefensible strike – he sat back and let Evan do his work.   The weapon of choice for Evan you ask? Statistics.  Or more correctly reported facts.  Maude was quoting the Hutton Report, saying that pensions have been shown to eventually become unaffordable and that if we don’t act now with public sector pensions, future generations will have an unmanageable tax burden.  Evan, having asked him if he had read the report, asked a direct question about an apparent misquote, saying he had used <Ctrl>+F to search for the word!  He then went on to quote the actual numbers from the report that seemed to contradict the attributed “direct” quote.  A bloodbath of squirming and smirking followed, where I was left feeling let down by the government minister but thoroughly impressed with the quality of journalism on Radio 4 – top job guys.

The Devil may care

I like the new(ish) stats page on blogger. If I'm honest I like stats. As mentioned before (http://jims-soliloquy.blogspot.com/2010/04/big-brother.html), periodically I have a nosy at who, if any, are visiting my musings. I paticularly like the search terms that people have entered immediately prior to selecting my blog from the list presented by their search engine (not shown here). Anyhow, today I noticed the portentous count that had been reached for my UK total.

Janus Tendencies

So today I had a quick tidy up of the blog setup and I remembered that I had stopped recording, on board game geek, the games that I have been playing this year.  Shame really, I think there is a little Janus in all of us at this gateway to the new year and it would have been nice to examine the statistics like some latter-day sooth sayer and predict what it means about the changing nature of me and what may happen in 2011.

Note-to-self, must play (and record) more games.

No Crib for a Bed

Last night we were lucky enough to be invited to stay for supper with some great friends following the annual Oakley Court carol service.  After eating we were casually invited to play some cards.  Fabulous.  The crib board was produced and the games began.  I took an absolute drubbing - not helped by playing against a Jedi Crib Master - "these are not the cards you're looking for".

Humiliated and dejected I decided to take on my iphone to claw back some self respect.  During this first game, I was taking the lead and feeling very confident and then I was presented with this hand.

So obviously I could sit and work out the mathematical possibilities and the appropriate probability distribution but I was taking a, ahem, 10 minute break and so I made a snap decision.  I would discard the 4 and the 3 since the resulting hand was worth 14 in its own right.

Blast, who'd've thought it, I should have kept the 4 and the 3.  This meant that my hand is now still only worth 14 but if I'd thrown the 10 and a 5 into my crib I'd have had a hand worth 18 and a stronger seeded crib.

Adding insult to injury my crib was only worth 4 points but had I have thrown in the 5 and 10, it would have been worth 8 making my total score 26, a full 10 points more than my actual score of 16.  I guess an implicit, Jedi like understanding of probability is what one must develop to make the more likely (to succeed) selection, repeatedly.  Still I beat the iPhone - scant satisfaction though, as I know deep down that Master Sandy would have punished the choice mercilessly.  I'll lay down my sweet head......

Beware what you wish for...

Taking the BBC results table, which still has one constituency to declare, and using some Exel™ jiggery-pokery you get this table.  That is, taking the popular vote and allocating the proportion of seats dependent on the proportion of the electorate that voted for that party makes interesting reading (don’t forget the rounding!).  Bare in mind that I have not read any papers on how proportional representation actually works, or indeed if people knew that we didn’t have a first past the post mechanism would they vote differently?  I’m sure they would.  So this table is not necessarily representative of a PR Election but it does make you think about the level of negotiation needed to reach agreement.

Swing-o-meter

http://www.telegraph.co.uk/news/newstopics/politics/2432632/UK-General-Election-2010-political-map.html

Given that we need a 19.5% swing to liberals to get a hung parliament, is there really any chance of electoral reform?

I tried to post this yesterday but forgot to press the go-live button on the blogger dashboard. Less than prophetic in the most depressing way. Don’t try pressing the reset map button, it doesn’t work. Unfortunately we have all voted and the count is coming in. After all the haggling and horse trading is complete the botched government that is left will almost certainly be a travesty to the Popular Vote. Liberal Democrats have 8% of the seats from 23% of the electorate voting for them.

  http://news.bbc.co.uk/1/shared/election2010/results/

Jimmymandering

Not sure if any one else has seen it or not, but there is a general election in the UK on 6^th May 2010.  It’s clear that voter apathy has lead to poor turnout at general elections in the UK previously.  This graffiti-pedia page has the UK ranked joint 25^th.  My personal voter turnout has been just as poor, as until recently I lived in the constituency of Henley with such Conservative luminaries as Michael Heseltine and Boris Johnson – so I felt my vote would not make any difference either way against the swathes of Tory Heartlanders.  Currently the MP is John Howell, no I’d not heard of him either.

 

Last year we moved home only 4.3 miles away from where we used to live.  It struck me that we may just have moved far enough to be out of the Henley constituency and be in a position where at last my vote would actually count! So after a little bit of research I discovered I have indeed shifted from the Henley constituency to the Wantage constituency.  With a renewed vigour I started to brush up on the political agenda’s of each of the main parties and then looked for the current incumbent.  It turns out to be Ed Vaizey  a conservative, onetime close pal of the neighbouring Cameron family until a recent faux pas.  More conservative than Margret Thatcher, so again I think my vote has become redundant. 

 According to the Telegraph it would take a swing of 19.5% to switch from Conservative to Labour in Wantage – it would have switched to Liberals first clearly but still a similar swing is required.

This is the spread of votes at the last election in Wantage:

During my recent fervour of interest in government and politics I downloaded the 75Mb conservative manifesto, so perhaps I’ll précis it for you in the next blog.

Big Brother is watching you

The ubiquitous reference to an authoritarian regime is perhaps overkill for this blog entry but I do feel kind of guilty.  It is apparent, from most people that I talk to, that in general users of the internet are not aware of the level of information that is routinely collected on them and stored for posterity.  Well, maybe we're all aware we just, in that typically human way, manage to ignore it somehow.  We can switch off that conscious part of the brain and hide from big topics like death, homework, tax returns and in my case the dishwasher.  My original intention for this blog was not an attempt to generate followers or readers but it does tickle me to think that others read it and I often look to see who has.  Recently I had my first Hawaiian IP address.  You can see on the map above there is a dot.  I can zoom into this and show the detail, this map shows the last 100 unique IP addresses that viewed my site.  If I paid money, there would be more detailed information available, so advertisers and others may well be getting significantly more info from each of us than this.  I have had a different - and less user friendly - counter running since the launch of my site.  Thanks to a tip from Simon. Currently, there have been 1066 unique hits and 1063 visitors to my site - excluding the IP addresses that I usually connect from.

The simple screen dump of info from the Sitemeter site looks like this:

This hawaiian visitor was on a peak day..... (UTC -11hrs = 08:08:22 am Mar 31st)

Lest you misunderstand my motives here, I should probably add that what I am attempting to do in "broadcast mode" is what I have been doing to certain individuals on a 1:1 basis.  Namely, inform them about the quantity, nature and automation of the data that is available on all of us.  So there you go.

You have a little peek at me and I'm peeking back through the curtains ;-)

Don't do the crime

..... if you can't do the time.

So, I guess it's in some way inevitable with modern cars, no training for 20-25 years and a certain complacency that we all seem to acquire with driving.  I've been caught speeding.  The social acceptability of this crime is clearly changing.  Depending on context and era, certain crimes are deemed acceptable one moment and then completely unacceptable the next.  At some point in this nation's history, killing a Scotsman would have won me many accolades right now - unless he was in downing street, I'm guessing there would be a public outcry.  During my lifetime, I remember chuckling with Pop (my grandfather) about how he seemed to drive better when he'd had a couple of beers!  This would be unconscionable in today's society only 25 years later.  Thus I hang my head in shame, I'm not proud, yes I was clocked speeding.

Before I go on, there is a brief anecdote about the actual incident that, should your opinion of me not be on the floor already, will make it fall further.  For fear of turning into Ronnie Corbett I'll be brief.

I was heading to Basingstoke one morning, not far from my home, entering the village of Moulsford.  This village has two schools directly on the main road, one of which Will went to for many years.  Quite rightly the speed is 30mph, as I approached the village, as ever, I slowed to 30mph.  When I rounded the slight bend I saw a mobile speed van parked in the lay-by.  I checked my speedo, c. 30mph good, so I piously drove by with a self-satisfied grin on my face.  I then proceeded to text as many contacts in my mobile phone as I could find who I knew would be driving through Moulsford.  I have left this purposely ambiguous as to whether I was driving or not as the law on mobile phone usage has changed recently.  Plausible deniability and self incrimination are not so comfortable bed fellows.

So it was with much swearing and protestation as to my innocence that I read the notice that I had been clocked doing 34mph on that very day!!  After my ranting affirmation that I would take this to the European Courts for a ruling had subsided, eventually I was calm enough to discuss the document with Anne.  There was an opportunity to opt for paying £60 to go on a "Speed Awareness Course".  No need to tell the insurance company, no need to get 3 [more] points on my licence.  It sounded too good to be true, just the catch of course - it was a 4 hour session.  Still it had to be better than the points, right?

Actually, it was much better.  Clearly, there have been many punters through the doors of this "voluntary" course.  The course presenter was very good, he had heard it all, he knew what people were going to say before they said it, and he had all the answers.   Interestingly I can honestly say the 4 hours flew past and I feel much better informed about all the nonsense that we carry around with us about speeding, speed cameras and road signage.  I only hope that the shift in attitude that I have felt myself over the last few years has just taken a jump to a level that will help decrease road deaths.  What was quite worrying was that many of the attendees when asked to select answers from a multiple choice quiz about the highway code, failed.  If they were to take their driving test today, they would not pass.

Some thoughts:
In 2007 There were 247,780 road casualties in the UK.  There were 27,774 serious (life changing) injuries and critically there were 2,946 fatalities.  That is around 9 people every day die on our roads.
OK that's still better than the 70's proportionately.  Today 3k deaths out of 34M drivers and in the 70's it was 9k deaths out of 12M drivers.  We have improved from 9 deaths per 10,000 drivers to 1 death per 10,000 drivers but I don't think we should get complacent, that is still 9 people dying per day with the highest proportion of child deaths in Europe.

When the police attend an accident they have to choose from 54 categories as to what they think the cause of the accident is.  25% have excessive speed as the primary choice.  Over 75% have it as one of the top 3 reasons.  Speed is a major cause of accident and therefore death on the roads.

Roads are classified as Urban, Rural or Motorway there are some counterintuitive accident and death rates associated with this.  Where do accidents occur? Motorway 4%, Rural 25% and Urban 71%. Where do people die? Motorway 6%, Rural 54% and Urban 40%  The high rates of accident are simply related to the times that you are required to make a decision, on a motorway its rare in the town its common.  When we make decisions we are prone to errors.  The high rates of death in the town are the pedestrians, children and cyclist we kill.  On the rural roads it's just the myriad inanimate and animate objects that we are able to hit at high speed.

Conclusions:
Attitudes to speeding are changing - mine most definitely has, your tax payers contribution to my course has paid off.  Education and training surrounding driving is woefully inadequate - we should all be electing to do this type of course.  The driving test and maintaining your knowledge of the highway code should be tested much more stringently and regularly.

Information Error

There is a distinction between data and information.  Similarly there is a distintion between information and considered judgement.

With data such as: total cost of the prison service and the number of prisoners currently incarcerated:-

We get information like: It costs £40,000 per prisoner per year to keep them in prison.

We then get opinions like: The death sentence saves money and Newspapers call for the the return of hanging.

Previously:

When someone has collected the majority of the pertinent data, if not all of the relevant data; they have agregated it and sumarised a balanced view of the available information; and then they have made evidence based judgements and drawn challengeble conclusions, then they can call themselves a journalist and they are worthy of having their report or article published.

You may not have liked someone's conclusion after they had given a subject this 'considered judgement' but at the very least you could evaluate their reasoning and formulate a plausible argument.

Recently:

The development of the pace and currency of 24hr "news" channels initially brought a concern about the emergence of erroneous stories.  For me this has diluted a little now and my frustration is more centred around making "news" stories to fill the airtime - BREAKING NEWS: it's snowing in London

- rather than making news up.

In general, these stations have a prodution team and an editorial team ensuring there is a considered approach. They don't get this right all the time and I do wonder at the value of all the pseudo-news pumping out whilst they await a real news event. However this concern is nothing compared to the very recent phenomenen of rapid transmission of data, information and ill considered opinion on a worldwide platform via social networking sites like Twitter and Facebook.

I note here that I am fully aware that I am an accomplice in this global pumping out of "news" but hopefully you can see I am considering my role in this phenomenon.  I became aware of the potential for poor judgement on misinformation by us (the general public) when I recieved a very well meaning email request to forward a photo of a kidnapped boy. It turns out, with a little research, that it was an urban myth and actually a spam email chain.  Email though isn't as immediate or interconnected as social media sites.

Today:

Recently (as you know if you follow me on Facebook, Twitter or Linkedin) I've just started reading the Gruaniad on an app on my iPhone and there was an article about a Guatemalan lawyer having a post death video uploaded on to YouTube that surprisingly predicted his death.  This article was in the technology section during the week. Yes, tech is one of my favourites, as is politics.

Apparently, he was murdered and at his funeral a video was played that had him predicting his death at the hands of a corrupt regime. Worldwide public opinion swung in to action and the Guatemalan President was nearly twittered off his perch and YouTube'd into iNfamy.

The dust settles, time passes, journalism investigates (and let's face it CIA covers-up and Politico-doctors spin) and we start to see a considered opinion that is opposite to our twit-gut response.

I don't wish to comment on the Guatemalan lawyer specifically as I have only read the linked article. I am concerned though that it is indicative of a trend in global media, of humanity's desire to achieve prescience.  To be the first to know, to tell everyone that they new before, despite the need to consider facts and context prior to forming a judgement.  I'm not sure this trend can or should be halted, I just hope as a society we are able to handle this new toy.

I would liken it to giving someone a box of matches. With the right guidance, it could be a force for good. Without advice, socially accepted norms and a desire to learn, and even without any nefarious motive- it could cause a destructive inferno.

Publication Bias

Are journalists aware that their profession's propensity for choosing a fact to support their story is part of the theory of statistics? To be fair, it's a general foible of mankind, it's just that journo's do it with such volume. So clear is it that these "so called" peer reviewed quality journals are selecting positve results to publish, you have to adjust for it in any analysis of published data.

Outside of the scientific press this regurgitating of apparent facts to support whatever axe is currently being ground is sufficiently prevalent now that it is commonly considered as absurd. "Boffins say xxxx causes cancer shocker!" or "innoculation with xxx gives your kid nasty disease yyy instead".

It's no surprise then that we now hear that "if only the government had listened to the experts before Christmas and bought more Grit, we wouldn't be in this mess".

From the Daily Mail

Edmund King, president of the AA, said he had written to ministers on November 16 to warn about inadequate salt supplies, only to be accused of ' scaremongering'. 


'I'm afraid what we predicted has come true and it is pretty scary on the roads out there and on the 
pavements where people are getting injured.


Hey, I know, why didn't we create a  European grit mountain in Surrey, then we'd have enough to make sure all the whinging daily mail readers were able to go to their Tuesday night whist drive. Oh, yeah that's it, we've got a European debt mountain to dwarf Everest!

What about all the other things people were saying before Christmas? If it hadn't have snowed and we'd spent £15 g'dzillion on grit mountains we'd have stories like "£15 g'dzillion wasted on grit, and pensioner starves" or "£15 g'dzillion wasted on grit, and immigrants still pouring in from Eastern Europe" or "£15 g'dzillion wasted on grit, and hard working tax payers don't get life-saving expensive drug".

I suggest that when it snows journalist should be forced to stop work and go and play in the snow like normal people - also it should be allowed to throw special snow balls with stones in the middle if the journalist is identifiably from the Daily Mail.

 Rant over.