Stats Chat

Team Ratings for Week 15

The basic method is described on my Department home page.
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 112 matches played, 81 of which were correctly predicted, a success rate of 72.3%.
Here are the predictions for last week’s games.

Predictions for Week 15

Here are the predictions for Week 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.

Team Ratings for Round 21

The basic method is described on my Department home page.
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 140 matches played, 105 of which were correctly predicted, a success rate of 75%.
Here are the predictions for last week’s games.

Predictions for Round 21

Here are the predictions for Round 21. 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.

Team Ratings for Week 8

The basic method is described on my Department home page.
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 34 matches played, 24 of which were correctly predicted, a success rate of 70.6%.
Here are the predictions for last week’s games.

Predictions for Week 8

Here are the predictions for Week 8. 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.

Team Ratings for Round 14

The basic method is described on my Department home page.
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 65 matches played, 43 of which were correctly predicted, a success rate of 66.2%.
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.

Team Ratings for Round 5

The basic method is described on my Department home page.
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 32 matches played, 18 of which were correctly predicted, a success rate of 56.2%.
Here are the predictions for last week’s games.

Predictions for Round 5

Here are the predictions for Round 5. 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.

Team Ratings for Week 5

The basic method is described on my Department home page.
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 27 matches played, 17 of which were correctly predicted, a success rate of 63%.
Here are the predictions for last week’s games.

Predictions for Week 5

Here are the predictions for Week 5. 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.

Nature magazine (yes, that one) has a headline 75% of US scientists who answered Nature poll consider leaving.

This is a more honest phrasing that one usually gets in bogus poll headlines, but it’s still bogus.  The ‘poll’ was

Responses were solicited earlier this month on the journal’s website, on social media and in the Nature Briefing e-mail newsletter. Roughly 1,650 people completed the survey.

The other problem with the poll is that ’emigrational ideation’, as one might call it, is a very low bar. Lots of people consider doing things. Many fewer actually do them. People having been claiming they are considering leaving the US for a long time, and we don’t know from the ‘poll’ how the current numbers compare to the usual state of affairs. Remember this cartoon, from 2006:

In fact, I think it’s quite likely that there are more scientists than before actively looking into leaving the US.  Anecdotally, we are getting more enquiries here and more US applicants to advertised positions.  You might think this sort of anecdata is a pretty weak basis for conclusions . You’d be right, but it’s still better than a bogus poll.

Q: Did you see ebikes are good for your brain? If you’re old.

A: How old?

Q: Over 50

A: That’s not… Ahem. That’s a very inclusive definition of  ‘old’, isn’t it?

Q: Does it work?

A: Who knows? It might.

Q: Does this research show it works?

A: What would we want to check?

Q: Mice or people?

A: Good. People, not mice.

Q: Random allocation of people or just correlations?

A: Yes, that’s right.  This one is … somewhere in between.

Q: What’s in between randomly allocating people and just seeing what they choose? You flip a coin, but with your eyes closed?

A:  Well…

Participants were pseudo-randomly assigned to one of three groups: pedal cycling, e-bike or non-cycling control groups. Priority was given to filling the cycling spots then controls were recruited to match sample characteristics. The control group were recruited after the experimental groups had started to be run so that we could match age and gender in the control group with those participants in the experimental groups. The control group were aware that they would not be cycling during the trial and those in the experimental group were all re-engaging with cycling.

Q: So, comparing people who signed up to do cycling with people who signed up to avoid cycling and other exercise.

A: Yes, but in the context of cycling research, not just random people asked to not do any extra exercise.

Q: And did the cycling group have better brains?

A: They measured a whole lot of things. Eight different multi-question assessments. Some of those improved and some didn’t

Q: Is that what they expected?

A: Not really, no. They expected improvements in a lot of the things that didn’t go up

In line with our predictions, we found trends for improvement in executive function in the Stroop task and letter updating task in both cycling groups compared to baseline and the non-cycling controls. We also found improvement in speed of processing for go trials in the Stop-It signal task only for e-bike participants during the intervention. Measures of memory and spatial functioning did not show an effect of cycling. Furthermore, we found increases in self-reported mental health on the SF-36 health survey for only the e-bike cycling group. Despite strong evidence from previous studies for an increase in well-being after exercise and the impacts of the outdoor environment on this aspect of mental health, we did not find increases on the PWB, SL or PANAS questionnaires.

Q: Could it just be the study was too small to give reliable conclusions?

A: Well, if that’s the explanation, then the study is too small to support press releases and media attention.

Q: So e-bikes are bad?

A: E-bikes are a good form of low-carbon transport, especially in hilly cities with nice climates, and let you get some, but not too much, exercise.   I don’t think they need special effects on the brain to be popular.

Via the US website Electrek, which does news about electric vehicles and adjacent subjects, a story about bogus polls.

A bogus online clicky poll in Germany got 100,000 clicks and found that 94% of the time that a button was clicked, it corresponded to “absolutely not willing to buy a Tesla”.  Electrek calls this “100,000 Germans”, based on very little evidence.

The poll kept running.  By the time it had 470,000 clicks, only 29% of the clicks corresponded to “absolutely not willing to buy a Tesla”.  Quite a lot of these clicks — 253,000 — came from just two IP addresses in the US. A bit of maths shows that the clicks from the two US addresses made up about three-quarters of the pro-Tesla clicks.

The magazine, T-Online, was forced to the shocking conclusion that its meaningless customer-engagement exercise had been manipulated by someone else’s meaningless customer-engagement exercise.

Team Ratings for Week 4

The basic method is described on my Department home page.
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 19 matches played, 11 of which were correctly predicted, a success rate of 57.9%.
Here are the predictions for last week’s games.

Predictions for Week 4

Here are the predictions for Week 4. 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.