Nothin' matters and what if it did?

10:03 am

My opinion of Ana Marie Cox continues to deteriorate (see this earlier post). In this video, I'll save for another day my rant about the need to watch videos to keep up with the exchange of ideas, she argues that people do journalism because they love it, so we needn't worry too much that newspapers are dying. Isn't that like arguing that we should just shrug if all the universities were being closed because professors and scientists do research just because they love research, not because they're getting paid?

07:29 am

Lately when I find interesting articles or have notable observations I find myself tweeting or posting the link, maybe with a short comment, on Google Reader (or on Facebook if I think a lot of people will be interested). I find that somewhat regrettable. When I post(ed) them here I usually try/tried to find a bit more to say in terms of commentary. As the internet gets better and better, we get dumber and dumber.

(update: edited 'someone' --> 'somewhat')

10:44 am

In other news, this "climategate" thing has been pissing me off. Partly because the scientists from whom the emails were stolen were being political, stupid and imprudent, but mostly because it's being used, entirely predictably, to quickly jump to the wrong conclusions. I'm trying, sometimes unsuccessfully, to avoid getting into discussions about it. People who care about the truth and the science and hadn't made up their minds beforehand are able to figure out that this is not nearly as nefarious as skeptics are attempting to spin it. My impression is that those claiming that these emails are a disproof of AGW are not those inclined to take the science seriously or those who already made up their minds long ago. It's hard to find any evidence against AGW in this controversy. So I'm trying to avoid the debate in much the same way that I avoid the "birther" debate or the evolution debate. But in the meantime, the Northwest Passage is now passable, we're rapidly losing sea ice and temperatures increase and we keep twiddling our thumbs and feeling reassured because Matt Drudge laughs when global warming press conferences occur during a snow storm. I can't disprove determined scepticism so for my own piece of mind I'm going to try to leave it alone. But my favorite objection has been the one about the "politicization of science". That climate change sceptics can say this with a straight face is remarkable.

10:34 am - Weekend report

Were I to list some of the things I love about America, the way that they celebrate Thanksgiving would definitely be at the top of the list. I really love the four day break, much more of a break than a simple long weekend.

( Read more... )

02:15 pm

I cooked a turkey yesterday, the full deal, from making stuffing to trussing it up, basting every 20 minutes, and even making gravy. It all seemed to come together successfully and nobody died of salmonella poisoning, but I wonder if a vegetarian preparing a turkey is a bit like a pro-lifer giving a friend a ride to an abortion clinic.

01:16 pm

Yahoo claims that Real Salt Lake is the first professional North American sports team to win a title despite having a losing record in the regular season. (http://sports.yahoo.com/video/player/news/Sports_Minute/16744180). I know that that's false because the BC Lions won the Grey Cup in 2000 with a losing record in the regular season.

11:39 pm - Customer Service

Today we tried to go to the movies, twice. We drove to the theater for a 1 pm showing of "The Blind Side" but after we waited twenty minutes (without even a preview) they canceled the showing due to "technical difficulties". We were given coupons to come back for another show, somehow got 5 rather than 4. There was another showing at 2:20, so we exchanged our coupons for tickets to that and hung out at a book store until then. Unfortunately, a number of the other people must have had the same idea, because the theater was packed. We couldn't even find two seats together, let alone four. So we went to get yet another refund. The guest services person felt bad and gave us extra reimbursement tickets, but sort of a frustrating afternoon of attempted movie watching.

Then at 5 I took Jordan to his jujitsu class. Grace called in a Chinese food order and I was to pick it up on the way home. Well, Jordan and I walked into the Chinese food place and the guy behind the counter said, "ah, you're here, your order is ready" without my giving a name or anything. I was a bit surprised that he knew me, since I don't buy there very often and haven't been recently but when I commented saying "wow, you remember me?" He went on and on about how long he's been in the business and how good he is at remembering customers, etc. Okay, I guess, but I should have been more sceptical when the price of the order was only $18. That's a bit low for food for four, but I didn't think too much of it. Perhaps G, was planning to supplement with some other food. In any event, easy to guess what happened, right? I got home and the order was wrong. I hardly had the heart to call the restaurant and burst the guy's bubble about his excellent memory, etc., but G. had no such qualms. In any event, it seems they'd already discovered their mistake. Impressively, they sent over a driver with our food, at no extra cost, and left us with the order we'd picked up (what would they do with it, right?) So, I guess we have some chinese food leftovers to work through.

11:11 pm

One has only to follow the controversy that has exploded over the recent breast cancer detection recommendations to understand why health care is so expensive in America and why we should worry about the possibility of fixing it. The recommendations, that we stop, among other things, doing annual routine mammograms of women in their 40s, make, to my mind, good sense. Current practices in women under 50 are minimally effective, they cause relatively large numbers of false positive, and unnecessary pain and suffering. This article from the NYT, (thanks for the pointer, _rck_ ) which also argues that part of the issue may lie with the problematic way in which we conceptualize cancer, summarizes nicely:

And note that $5 billion is spent every year in the US on mammograms. And yet, if I'm to understand some people, these recommendations are a government plot or insurance company collusion or something. But imagine some new screening test that came out that had that kind of success rate. Would we really argue that it would be a wise thing to implement as part of one's health regimen, an unreliable test with a miniscule chance of detecting something serious? Wouldn't we rather save the test for specific circumstances, patients in high risk categories, for example?

So in an insurance system, in which costs are shared, it's not really fair or prudent to spend large amounts of money on something that is so ineffectual, particularly not when costs are already overwhelming and make insurance unaffordable. And yet, people are angrily insisting that their health insurance plans continue to cover this procedure as if the insurance company is obligated to fund every procedure regardless of how expensive or ineffective it is. We seem to be at something of an impasse. We, i.e., the pool of insured people, can't continue to fund every procedure regardless of how expensive it is, in terms of money or misdiagnosis, on the premise that it might save "even one life". We're at something of a crossroads, I think. Either people are going to have to change their mindset of thinking that they can/should/must get every single diagnostic and/or preventive measure accessible to them in the interests of keeping a shared resource affordable, or we're going to have to find ways to make the cost/benefit realities more stark for people, e.g., market based solutions as discussed here. You really want an annual mammogram despite good general health and no family history? Sure, but you have to pay for it, and the costs of ruling it out should a false positive come up. Not to mention radiated related health risks.

07:56 am - The Stupak Amendment

I haven't been following the discussion on the Stupak amendment very carefully, so I may have missed some subtleties, but I'm a bit confused by claims I've encountered that it somehow violates a woman's right to choose. Surely acknowledging or establishing the right to do X isn't accompanied by an obligation to have access to X paid for by the government, i.e., by taxpayers for whom X violates their ethical principles. People opposed to drug laws or alcohol prohibition aren't arguing that the government has an obligation to provide drugs or alcohol for its citizens, are they? If I don't believe the government should forbid the eating of meat, does it follow somehow that I should also believe my tax dollars should go towards subsidizing cattle farms?

x-posted to blogspot.

01:35 am

I like Ana Marie Cox, but I'm fairly certain the review she wrote of Palin's book in the Washington Post yesterday is the worst book review I've ever read in any sort of professional publication. Roughly, it says: "It was pretty long, I didn't finish it, but she did have some stuff to say about smoking that I didn't agree with." It may be deadline, but if you have nothing written, you have nothing written, don't just hand in your notes.

10:25 am - Making web pages readable.

I installed this tool recently and it has been working very well for me:

http://lab.arc90.com/experiments/readability/

It's some javascript that essentially reformats your page without all the jazzy boilerplate, going well beyond just stripping away ads, and regenerates just the content of the page according to font and margin sizes that you've preset. It's even better, usually, than getting the printable format for the page. (You do have to click it each time you open a new page, but I don't mind that, makes it easier to tell if I'm losing content.) It works well for most pages, although, perhaps tellingly, it reduces one's Facebook news feed to nothing.

07:50 am

There appears to be fairly broad consensus that Bill Belichik made a terrible call Sunday night by having his team try, but failing, to convert on fourth down, and thereby retain possession, rather than punt the ball down the field and let the defense take over. Alas, nobody bothers to present any evidence as to why precisely, except to assert that it was crazy/risky/stupid, whatever. (read the excoriations from Michael Wilbon and Michael Silver. Insults without evidence appear to be turning into sort of a MO for Wilbon, see my comments on his article about Rush Limbaugh) In any event, Tony Dungy said, "you have to go with the percentages". What do the percentages say?

Unfortunately, it's tough to get numbers. (In my garage I'm going to build an ontologically based data drilldown tool to sell to the NFL , NBA, MLB and maybe the NHL, It will allow coaches to quickly query on, say, [% of drives carried out by [Conference member/team/QB] from within the ?X yard line that result in [points --> TD / FGs] ] complete with confidence intervals, etc. ) But let's approximate. I think about 50% of 4th and close attempts make it. If Brady is the QB, it's probably more like 60%. And about 50% of drives starting in the red zone (20 yards from the end zone) result in TDs, let's say that drops to 40% if you're starting from the 28. So, the probability that going for a 4th down conversion would result in a Colt TD? I'd say it's .40 (that they'd not successfully convert) * .40 (probably of Colts scoring a TD when at the Patriot 28), or about a 16% chance.

That's actually not great chance but I'll bet it's higher than the probability of Manning getting a TD starting at his own 20, about where he'd be if they punted, within two minutes. I'm not sure what the chances are there, let's say they're 5-10%. Okay, so it looks like a bad call, but not a crazy bad call. Let's also factor in other considerations. Does Manning do particularly well at the end of a game? Unlike other sports, there does actually seem to be a phenomenon of some QBs being particularly effective under pressure. Another consideration? I read somewhere that the Patriots had a plan in mind to draw the Colts into taking a penalty, a plan that failed but presumably had some non-zero chance of success. These considerations push the percentages of either choice closer together. Already the difference between 5-10% and 16% seems to be well outside the "he clearly made a bad call" range. I'm sure he had some additional information or reason to believe that the latter number was actually smaller. So, it's not clear to me that Bellichik's call was particularly bad, I think the evidence argues against it, but only a small amount of evidence against the decision, could have sufficed to make it a reasonable decision.

Update: Ah, others who think Belichik wasn't so crazy: link to NFL Advanced Stats blog. It says the Colts would have likely got the ball at their 34, not their 20, as I assumed, and that the probability of a TD from there is 30%, considerably more than I guessed/assumed from the Colts 20.

04:05 pm

I've been reading a few discussions as to whether or not the Ft. Hood killer's actions constituted an act of terrorism. What I haven't seen is a clear explanation of why it matters. I suppose some of the emotion arises because the last administration pointed to terrorism in justifying a lot of the nasty things it did. But I really don't much see the point of trying to determine whether or not a notoriously underspecified concept, C, applies to a particular situation, S, unless the fact that S is an instantiation of C can/will or might be used to justify some extraordinary action. Are we afraid and or hoping that that might happen here? If so, what exactly might that be?

08:41 pm - Google Suggested Searches

A couple of Slate articles (link 1 and link 2) about Google's search suggestion have been getting a lot of amused attention. (I'm not sure why all the sudden attention, it's been around for quite a while.) But I'm sceptical about what the initial article claims about how the search suggestion algorithm works. It claims that the Google engineers have confirmed that "what we're looking at in the toolbar is, essentially, a list of the 10 most popular queries that start with a given prefix"

Try some examples. The fourth suggestion when I type 'which' is 'which of the famous corey boys starred in more than one vampire movie?' There must be many more popular searches starting with 'which', e.g., 'which [some general consumer product] is the best' or the like. When I try 'is' i get results that include 'is wendy williams a man?" Or try 'Canadian', the top suggestion I get is 'canadian conservative commentator rick barbara'. It seems unlikely to me that that's one of the 10 most frequent searches starting with 'Canadian' in the last n hours. That said, sometimes digging further into some of these examples that look unlikely as popular searches turn out to be notable phrases on the web. For example, the corey boys question has been asked in exactly that form on Wiki Answers, so I suppose it's possible that it's pure popularity.

Anyway, I'd guess that they're doing some term frequency/inverse document frequency analysis to determine how interesting the searches are as well and probably using that plus popularity to generate the suggestions. If so, that would, of course, affect the kinds of conclusions we can draw from this information. I'm not sure we can claim that we're getting a "sliver of the collective mind", they may very well be slivers of the mind of very few individual users.

Of further potential interest to sociologists or anthropologists or whatever, the suggested search gives us a peek into what kind of searches Google perceives as being potentially inappropriate or offensive. It won't give suggestions for some obvious terms, e.g., 'fuck, 'cock', 'shit'. It's also unwilling to give suggestions for more clinical terms like 'penis', 'vagina', 'masturbation' and 'anal'. That surprises me a bit, you'd think they'd be capable of doing something less heavy-handed and leave in the obviously inoffensive suggestions, e.g., 'vagina monologues' or 'anal cancer'. Yahoo search suggest does a much better job of being able to offer inoffensive suggestions in these circumstances. 'testicle' and 'sex' both seem to get further suggestions and, curiously, so does 'fart' and 'necrophilia'. Most non-sexual, non-curse terms get suggestions, e.g., 'dead baby', 'slitting', 'gory', 'slasher', 'stabbing pic', 'festering'. An interesting exception is 'pus-infe', but it works if you leave out the hyphen so it may be just a syntactic quirk.

UPDATE: Pogue shares my scepticism about this being just popular searches: http://pogue.blogs.nytimes.com/2009/11/12/fun-with-googles-auto-suggestion-feature/

04:14 pm - this 'n that

Just two quick unrelated things:

(i) Let me start with a sweeping generalization.  It seems to me that Canadians make a much bigger deal of remembering their war dead, and do so more solemnly,  than do Americans.  Veterans Day and Memorial Day strike me as a bit of a joke around here and poorly observed.  But I have recollections of attending some fairly moving and dignified Remembrance Day ceremonies, either at local cenotaphs or at my school and more recently at the Canadian embassy.    And I also remember most people wearing poppies during the days preceding Nov. 11, I have yet to a single one here this year, (although Grace and I still try to remember to wear one).  If I'm right, and my evidence is admittedly largely anecdotal, isn't this a bit counterintuitive given all the stereotypes of both countries and the fact that the US has participated in far more military action since WW2? 

(ii) Why on earth doesn't ESPN broadcast the final table of the World Series of Poker live?  I'd happily watch many hours of that.  But now that I know the outcome and because heavily edited poker broadcasts tend to focus only on exciting hands and communicate very little about general strategy, I'm unlikely to watch much of the broadcast that starts tonight.

09:58 pm - How to prevent mass murders

Whenever these Columbine/Austin clock tower/Ft. Hood sorts of events occur, the media orchestrates displays of handwringing and bewilderment about what went wrong and consternation about how we might prevent such things from happening in the future. But isn't the solution, although not easily implemented, fairly obvious? I believe that most of these nutbars who shoot up restaurants or army bases or schools or office buildings before taking their own lives, directly or by "death by cop", wouldn't bother to do so if they knew that despite the horrific actions they'd die in relative obscurity.

If I'm right, this imposes an obligation on the news media and news consumers to stop providing that which motivates the actions of these killers, i.e., fame. News media should just stop reporting the details about these kinds of killers. They could report the crimes but leave out the killer's name and details about the killer's personal life, focusing instead on the nature of the crime and the victims. Other than local media that may have family members of the killer in its audience, how is it in the public interest to learn the details of the private lives of these killers? In discussing these people ad nauseam is the media doing anything other than suggesting to those who are leading failed insignificant lives, that this route at least offers them an opportunity to matter and be noticed?

The obligation also falls on the news "consumers". We should stop seeking out and paying attention to such details and perhaps also join together to boycott news organizations that publish those names and details and the companies that sponsor them.

I'm not suggesting a legal ban, but a voluntary ethical code, based on the same kinds of principles that prevent news media from explaining how to build bombs or leaving out certain details of crimes or failing to publicize the names of the victims of some crimes. This information wouldn't have to be top secret, it should remain available to people. Psychology researchers and criminologists, for example, should continue to access it. But if people actually had to go to police information sources and the media failed to broadcast it, I suspect that the main motivation for these crimes would be eliminated.
x-posted to blogspot.

10:26 am - Off Year Elections

It's hard to know what last night's election results portend for the Democratic or Republican party. On the one hand, we see Virginia going, very heavily, Republican, on the other, the Democrats won a seat in NY that had been Republican since the Bronze Age.  I won't try to interpret these  results in terms of what voters think of the Obama administration, possible to spin it either way,  I suppose, but I think it is notable that both losers, in VA and in NY-23,  seemed to have made a point of distancing themselves from their party.  Deeds is a very right wing Democrat and purposefully distanced himself from Obama, and Scozzafava is fairly left wing, relative to Republicans, and ultimately endorsed the Democratic candidate. 

So, what are we to make of these things?  I suppose the simple lesson is that one ignores one's base at one's peril.  Triangulation only goes so far; cynical attempts to grab the swing voters can backfire.  Truth be told, I was, in an odd way,  pleased to see the third party candidate come as close as he did in NY-23.  Not, of course, because he was such a right winger, but because it showed that politics haven't become a matter of simply "supporting one's team", that principles and ideas still matter to some voters.

Tangentially, I think it is notable that the WaPo endorsed Creigh Deeds shortly before the election. At that time polls showed Deeds behind but not 18 points behind.  It makes one wonder what a newspaper endorsement is worth these days.  This certainly didn't give much evidence that it helps.

x-posted to blogspot.

11:23 am

The semantic web has been in development for what, 10+ years?  After all that time, here's the ontology constraining one of the main implementations to date: http://www4.wiwiss.fu-berlin.de/dbpedia/dev/ontology.htm.    (Thanks for those n-ary quantifiers, W3C.)

10:50 pm - Spinning the Semantic Web

Last week I attended the International Semantic Web (SW) Conference, ISWC 2009.  The semantic web project is one that has interested me for a long time because it would be a large scale knowledge representation implementation and because it involves standardizing languages and approaches for doing so.  But the semantic web has been taking its sweet time in catching on.  Interestingly, I attended a few sessions of the ISWC in 2002 or 2003 and there Tim Berners Lee claimed that we were just on the cusp of having it catch on and that it was picking up speed just the way the web originally did.  I think that now, in 2009, some momentum is finally beginning to gather. dbpedia is a SW version of, essentially, Wikipedia and there is a way to query it using the SW query language, SPARQL, and the NY Times is "semantic webifying" itself, but the SW has not caught on at nearly the same speed as the web did and I think it is useful to ask why.  Some of my thoughts:
( Read more... )

Anyway, I think the SW will catch on and is catching on, but I think it could have been happening much more quickly if people had mainly concerned themselves with making it useful and workable and less with exploiting it as a funding tool for interesting but ancillary AI problems.

(UPDATE: Some of paragraphs 1 and 2 revised for readability.  (yeah? well you should have seen them before the edits.) )

11:02 pm

I saw an ad today on CL  for two tickets to tonight's hockey game for $15, that was for both tickets.  I quickly responded and got 'em!  The seats turned out to be at the very top of the building but not too far from center ice, so not as bad as it might sound and at less than the price of a movie, well worth it.  Game was okay, although the Caps lost to the lowly Islanders in OT.