Posts filed under General (3156)

Team Ratings for Round 15

The basic method is described on my Department home page. I have made some changes to the methodology this year, including shrinking the ratings between seasons.

Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 104 matches played, 59 of which were correctly predicted, a success rate of 56.7%.

Here are the predictions for last week’s games.

Predictions for Round 15

Here are the predictions for Round 15. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Compare these two stories:

• NFL (US football) players are getting taller and heavier, with animated graph, by Noah Veltman
• Rugby players are getting bigger, by Peter Milburn

The second story actually gives more context and explanation, but the first one is (to me) more effective.  It also shows something surprising: the size distribution splits into separate modes in recent years, perhaps reflecting specialisation in playing positions.

The second story actually argues that there isn’t a similar divergence in builds of rugby players, so I went to look at the data (which involved scraping it off the NZ Rugby Museum website).  The pattern over time I get is (click to embiggen)

which suggests that rugby players aren’t just getting bigger, they are showing a little of the same separation into big and very big seen in the NFL players

The need for science communication:

 Stephen Curry, writing at The Guardian

Even so, I think we need to work on our relationship. Approval ratings may be high and over two-thirds of you may also be happy to leave it to the ‘experts’ to advise the government on science, but a similar proportion still believe that scientists don’t try hard enough to listen to what ordinary people think or to inform them about their work.

 

Robert Finn, writing at Scientific American

The journalist reached out to Dr. A and also to two other researchers (Drs. X and Y), who work in related fields, to get independent comment. Boy oh boy did Dr. X and Dr. Y comment, and those comments surely were independent, which is what any journalist wants. But in the same emails in which they eviscerated the study they also insisted that their comments remain off the record.

…

Because our sources said that their comments were off the record, we couldn’t use them in any way, and I can’t quote them here, not even anonymously. At this writing, the journalist has been unsuccessful in finding sources willing to offer on-the-record comments or criticisms of the study.

And, for some promising news, there is a new science column in the ChCh Press, that gives brief summaries of science stories over the week. It’s written by Sarah-Jane O’Connor, who is both a scientist with a PhD in Ecology and a journalist.

Keith Humphreys, writing in New York Magazine

A clever new study in the journal Addiction provides clues about who is worst at owning up to the full extent of their drinking.

The researchers surveyed over 40,000 people with standard alcohol survey questions about their quantity and frequency of alcohol consumption — “How many drinks have you had in the past month?” and so on. But in a smart twist, they then asked a more immediate question: “How many drinks did you have yesterday?”

I’ve written about this technique before; it can be very powerful, though it won’t help much if people are intentionally misleading you.

Thomas Lumley’s latest New Zealand Listener column points out that while people love the sound of public screening for disease,  it has a significant problem: Most people who are screened aren’t sick. And that’s when the spectre of false positives arises …

Read the column here: Failing the screen test

Team Ratings for Round 14

The basic method is described on my Department home page. I have made some changes to the methodology this year, including shrinking the ratings between seasons.

Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 99 matches played, 56 of which were correctly predicted, a success rate of 56.6%.

Here are the predictions for last week’s games.

Predictions for Round 14

Here are the predictions for Round 14. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

From Mark Hanna on Twitter, ACC expenditures on acupuncture (the sources are in the Twitter conversation)

The graph shows three things. Firstly, there was a review promised. Second, the expenditure on acupuncture is vastly different from the projections. Third, it’s getting to be a moderately large sum of money.

Comparing to the 2012 Pharmac report (PDF, p5), there were only seven prescription items where Pharmac spends more per year than the $17 million ACC spent on acupuncture that year (and one about the same). Things like diabetes test strips and the breast-cancer drug Herceptin.

I’m not as opposed as Mark to spending taxpayer money on the placebo effect, but at some point you have to wonder whether there might be more cost-effective ways to get it.

• A Lotto story in the Herald that isn’t untrue or importantly misleading. Yay!
• The Art of Risk: videos from Leeds University
• “Incidentalomas“: the problem with new medical screening technologies

Team Ratings for Round 13

The basic method is described on my Department home page. I have made some changes to the methodology this year, including shrinking the ratings between seasons.

Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 91 matches played, 51 of which were correctly predicted, a success rate of 56%.

Here are the predictions for last week’s games.

Predictions for Round 13

Here are the predictions for Round 13. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Even with large amounts of data, automated predictions must usually incorporate explicit or implicit prior understanding of the structure of the problem. “Look for anything” is not good enough: “anything” is too big.

Here, for your weekend light entertainment, are some examples where the prior structure was too strong or too weak:

The example that prompted this post, from the blog of Melville House Press, is about automated scanning of books to create digital editions

 in many old texts the scanner is reading the word ‘arms’ as ‘anus’ and replacing it as such in the digital edition. As you can imagine, you don’t want to be getting those two things mixed up.

A similar phenomenon was pointed out at Language Log a decade ago

Fear not your toes, though they are strong,
The conquest doth to you belong;

Daniel Dennett recounts two anecdotes of speech recognition, one human and one computer, which err in the opposite direction to the text recognition example. The computer one:

An AI speech-understanding system whose development was funded by DARPA (Defense Advanced Research Projects Agency), was being given its debut before the Pentagon brass at Carnegie Mellon University some years ago. To show off the capabilities of the system, it had been attached as the “front end” or “user interface” on a chess-playing program. The general was to play white, and it was explained to him that he should simply tell the computer what move he wanted to make. The general stepped up to the mike and cleared his throat–which the computer immediately interpreted as “Pawn to King-4.” 

And, the example that is frustratingly familiar to so many of us: mobile phone autocorrupt, which you can search for yourself.