February 23, 2021

Pro14 Predictions for Round 13

Team Ratings for Round 13

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.

Current Rating Rating at Season Start Difference
Leinster 17.60 16.52 1.10
Munster 10.72 9.90 0.80
Ulster 9.67 4.58 5.10
Edinburgh 3.52 5.49 -2.00
Glasgow Warriors 2.50 5.66 -3.20
Connacht 1.98 0.70 1.30
Scarlets 1.21 1.98 -0.80
Cheetahs -0.46 -0.46 0.00
Cardiff Blues -0.83 0.08 -0.90
Ospreys -1.93 -2.82 0.90
Treviso -5.84 -3.50 -2.30
Dragons -8.05 -7.85 -0.20
Southern Kings -14.92 -14.92 0.00
Zebre -15.17 -15.37 0.20

 

Performance So Far

So far there have been 71 matches played, 48 of which were correctly predicted, a success rate of 67.6%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Dragons vs. Leinster Feb 20 29 – 35 -20.30 TRUE
2 Glasgow Warriors vs. Ulster Feb 20 13 – 19 0.50 FALSE
3 Scarlets vs. Treviso Feb 21 41 – 17 12.60 TRUE
4 Ospreys vs. Zebre Feb 21 10 – 0 20.70 TRUE
5 Connacht vs. Cardiff Blues Feb 21 32 – 17 8.10 TRUE
6 Edinburgh vs. Munster Feb 21 10 – 22 0.30 FALSE

 

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.

Game Date Winner Prediction
1 Treviso vs. Connacht Feb 27 Connacht -1.30
2 Cardiff Blues vs. Munster Feb 27 Munster -5.00
3 Ulster vs. Ospreys Feb 27 Ulster 18.10
4 Edinburgh vs. Scarlets Feb 28 Edinburgh 8.80
5 Zebre vs. Dragons Feb 28 Dragons -0.60
6 Leinster vs. Glasgow Warriors Mar 01 Leinster 21.60

 

February 19, 2021

Bitcoin and EFTPOS

The combination of descriptions about NZ as a small, empty, remote island and news about Tesla buying lots of Bitcoin made me compare this graph of blockchain transactions per day (from Wikipedia, by user HocusPocus00)

with an infographic from paymentsnz for 2019.

There are about a third of a million Bitcoin transactions per day, and nearly 1.25 million Ethereum transactions per day.  How does that compare with the NZ economy?

We get through about a million direct electronic transfers per day  — when you send money from your bank account to someone else’s, like your plumber or a guy on TradeMe or splitting an accomodation bill with your mates.  There’s about 360,000 direct debit transfers — when Vodafone sucks your phone and broadband bill out of your bank account.  We have about 2.8 million EFTPOS debit transactions per day, and about 2.3 million credit card and PayWave transactions per day at point of sale (that was 2019; I’m guessing Paywave is up and EFTPOS is down in 2020).

Now, a single Bitcoin transaction can be more complicated than a single debit card purchase, and it might actually package a whole lot of transactions together (and, no, this isn’t quite like how a single supermarket debit card transaction might include toothpaste and cheese). It’s still a little surprising how few blockchain transactions there are.

February 16, 2021

Top 14 Predictions for Round 17

Team Ratings for Round 17

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.

Current Rating Rating at Season Start Difference
Stade Toulousain 8.62 4.80 3.80
La Rochelle 7.28 2.32 5.00
Racing-Metro 92 6.67 6.21 0.50
Bordeaux-Begles 4.90 2.83 2.10
Clermont Auvergne 4.85 3.22 1.60
RC Toulonnais 3.53 3.56 -0.00
Lyon Rugby 3.18 5.61 -2.40
Montpellier 0.55 2.30 -1.80
Stade Francais Paris -1.20 -3.22 2.00
Castres Olympique -2.04 -0.47 -1.60
Brive -3.08 -3.26 0.20
Section Paloise -4.00 -4.48 0.50
Aviron Bayonnais -6.03 -4.13 -1.90
SU Agen -12.65 -4.72 -7.90

 

Performance So Far

So far there have been 115 matches played, 75 of which were correctly predicted, a success rate of 65.2%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Aviron Bayonnais vs. Brive Feb 15 26 – 23 3.70 TRUE
2 Bordeaux-Begles vs. Stade Francais Paris Feb 14 44 – 6 9.00 TRUE
3 Castres Olympique vs. Montpellier Feb 14 48 – 17 1.70 TRUE
4 Racing-Metro 92 vs. Lyon Rugby Feb 14 34 – 26 10.00 TRUE
5 RC Toulonnais vs. La Rochelle Feb 15 11 – 29 3.30 FALSE
6 Stade Toulousain vs. Section Paloise Feb 13 31 – 9 15.60 TRUE

 

Predictions for Round 17

Here are the predictions for Round 17. 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.

Game Date Winner Prediction
1 Brive vs. Bordeaux-Begles Feb 22 Bordeaux-Begles -2.50
2 Clermont Auvergne vs. Aviron Bayonnais Feb 21 Clermont Auvergne 14.80
3 La Rochelle vs. Stade Francais Paris Feb 21 La Rochelle 12.90
4 Lyon Rugby vs. Stade Toulousain Feb 22 Lyon Rugby 0.90
5 Racing-Metro 92 vs. Castres Olympique Feb 20 Racing-Metro 92 13.70
6 Section Paloise vs. RC Toulonnais Feb 21 RC Toulonnais -1.20
7 SU Agen vs. Montpellier Feb 21 Montpellier -6.80

 

Rugby Premiership Predictions for Round 10

Team Ratings for Round 10

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.

Current Rating Rating at Season Start Difference
Exeter Chiefs 7.05 7.35 -0.30
Bristol 4.00 1.28 2.70
Sale Sharks 3.52 4.96 -1.40
Wasps 2.54 5.66 -3.10
Harlequins 1.35 -1.08 2.40
Northampton Saints -1.04 -2.48 1.40
Bath -2.01 2.14 -4.10
Gloucester -3.06 -1.02 -2.00
Leicester Tigers -5.61 -6.14 0.50
Newcastle Falcons -6.17 -10.00 3.80
London Irish -6.31 -8.05 1.70
Worcester Warriors -7.35 -5.71 -1.60

 

Performance So Far

So far there have been 48 matches played, 28 of which were correctly predicted, a success rate of 58.3%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Exeter Chiefs vs. London Irish Feb 14 26 – 3 17.10 TRUE
2 Gloucester vs. Bristol Feb 13 17 – 18 -2.80 TRUE
3 Harlequins vs. Leicester Tigers Feb 14 37 – 24 11.20 TRUE
4 Sale Sharks vs. Bath Feb 13 22 – 27 11.80 FALSE
5 Worcester Warriors vs. Wasps Feb 15 13 – 17 -5.70 TRUE

 

Predictions for Round 10

Here are the predictions for Round 10. 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.

Game Date Winner Prediction
1 Bath vs. Gloucester Feb 20 Bath 5.50
2 Exeter Chiefs vs. Northampton Saints Feb 21 Exeter Chiefs 12.60
3 Harlequins vs. Sale Sharks Feb 21 Harlequins 2.30
4 Leicester Tigers vs. Wasps Feb 21 Wasps -3.60
5 London Irish vs. Bristol Feb 22 Bristol -5.80
6 Worcester Warriors vs. Newcastle Falcons Feb 21 Worcester Warriors 3.30

 

Pro14 Predictions for Round 12

Team Ratings for Round 12

More weird stuff this week. A game that was previously listed as Round 12 is now listed as Round 14. Also because the game scheduled for last week was not played, my last week results refer to the previous week for which I have already published the result.

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.

Current Rating Rating at Season Start Difference
Leinster 18.18 16.52 1.70
Munster 10.20 9.90 0.30
Ulster 9.09 4.58 4.50
Edinburgh 4.04 5.49 -1.40
Glasgow Warriors 3.08 5.66 -2.60
Connacht 1.35 0.70 0.60
Scarlets 0.71 1.98 -1.30
Cardiff Blues -0.20 0.08 -0.30
Cheetahs -0.46 -0.46 0.00
Ospreys -1.46 -2.82 1.40
Treviso -5.35 -3.50 -1.80
Dragons -8.63 -7.85 -0.80
Southern Kings -14.92 -14.92 0.00
Zebre -15.64 -15.37 -0.30

 

 

Performance So Far

So far there have been 65 matches played, 44 of which were correctly predicted, a success rate of 67.7%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Dragons vs. Connacht Feb 06 20 – 30 -2.90 TRUE

 

Predictions for Round 12

Here are the predictions for Round 12. 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.

Game Date Winner Prediction
1 Dragons vs. Leinster Feb 20 Leinster -20.30
2 Glasgow Warriors vs. Ulster Feb 20 Glasgow Warriors 0.50
3 Scarlets vs. Treviso Feb 21 Scarlets 12.60
4 Ospreys vs. Zebre Feb 21 Ospreys 20.70
5 Connacht vs. Cardiff Blues Feb 21 Connacht 8.10
6 Edinburgh vs. Munster Feb 21 Edinburgh 0.30

 

February 15, 2021

Vaccines and testing

The current NZ Covid restrictions happened because someone at risk of infection in their job developed symptoms consistent with Covid and got tested.   With the first load of Pfizer vaccines arriving today, that raises a question about testing.

All the current vaccines seem to do best at preventing serious illness and death, and slightly less well at preventing mild illness. We don’t have really good data on preventing asymptomatic illness, but the information we have suggests that the trend continues and they’re somewhat less effective at preventing asymptomatic illness.  For most purposes that’s a great tradeoff — Covid is a global catastrophe only because people get sick and die, that’s how it’s different from the common colds caused by other coronaviruses.  For preventing border incursions things are more complicated.

When we vaccinate the first target group — people working at the border, the highest-risk healthcare workers, and their household contacts — we dramatically reduce the chance they will get sick or die from Covid, and we reduce, less dramatically, the chance they will get infected and pass the virus on.  We should expect fewer cases of the virus getting past the regular workplace testing — but when it does, it’s less likely to be stopped by someone stepping up and getting a test on their own.  Outbreaks are likely to be bigger by the time we see them, with a higher risk of needing lockdowns.

This doesn’t mean we shouldn’t vaccinate border workers. The whole point of the vaccine is to stop people getting sick and dying, and that’s how we should use it. The answer isn’t to use MIQ workers as coal-mine canaries. We do need to think about how to change testing to respond to the new circumstances.  Many experts have already been calling for more frequent testing of high-risk workers, ideally using new saliva-based tests to reduce the ouch factor.  The case for more-frequent tests will be much stronger as we progressively vaccinate the people at highest risk and become less likely to pick up outbreaks by waiting for people to get sick. 

February 9, 2021

Top 14 Predictions for Round 16

Team Ratings for Round 16

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.

Current Rating Rating at Season Start Difference
Stade Toulousain 8.18 4.80 3.40
Racing-Metro 92 6.78 6.21 0.60
La Rochelle 6.61 2.32 4.30
Clermont Auvergne 4.83 3.22 1.60
RC Toulonnais 4.23 3.56 0.70
Bordeaux-Begles 4.05 2.83 1.20
Lyon Rugby 3.07 5.61 -2.50
Montpellier 1.40 2.30 -0.90
Stade Francais Paris -0.33 -3.22 2.90
Castres Olympique -2.90 -0.47 -2.40
Brive -3.11 -3.26 0.20
Section Paloise -3.57 -4.48 0.90
Aviron Bayonnais -6.00 -4.13 -1.90
SU Agen -12.66 -4.72 -7.90

 

Performance So Far

So far there have been 109 matches played, 70 of which were correctly predicted, a success rate of 64.2%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Montpellier vs. Stade Francais Paris Feb 07 31 – 6 9.40 TRUE
2 Clermont Auvergne vs. Lyon Rugby Feb 06 26 – 18 4.50 TRUE
3 Castres Olympique vs. Aviron Bayonnais Feb 08 31 – 21 8.10 TRUE
4 Racing-Metro 92 vs. La Rochelle Feb 08 26 – 22 7.00 TRUE

 

Predictions for Round 16

Here are the predictions for Round 16. 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.

Game Date Winner Prediction
1 Aviron Bayonnais vs. Brive Feb 15 Aviron Bayonnais 3.70
2 Bordeaux-Begles vs. Stade Francais Paris Feb 14 Bordeaux-Begles 9.00
3 Castres Olympique vs. Montpellier Feb 14 Castres Olympique 1.70
4 Racing-Metro 92 vs. Lyon Rugby Feb 14 Racing-Metro 92 10.00
5 RC Toulonnais vs. La Rochelle Feb 15 RC Toulonnais 3.30
6 Stade Toulousain vs. Section Paloise Feb 13 Stade Toulousain 15.60
7 SU Agen vs. Clermont Auvergne Feb 14 Clermont Auvergne -9.90

 

Rugby Premiership Predictions for Round 9

Team Ratings for Round 9

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.

Current Rating Rating at Season Start Difference
Exeter Chiefs 6.69 7.35 -0.70
Sale Sharks 4.40 4.96 -0.60
Bristol 4.15 1.28 2.90
Wasps 2.67 5.66 -3.00
Harlequins 1.21 -1.08 2.30
Northampton Saints -1.04 -2.48 1.40
Bath -2.89 2.14 -5.00
Gloucester -3.20 -1.02 -2.20
Leicester Tigers -5.46 -6.14 0.70
London Irish -5.95 -8.05 2.10
Newcastle Falcons -6.17 -10.00 3.80
Worcester Warriors -7.49 -5.71 -1.80

 

Performance So Far

So far there have been 43 matches played, 24 of which were correctly predicted, a success rate of 55.8%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Bath vs. Harlequins Feb 07 15 – 28 2.00 FALSE
2 Bristol vs. Sale Sharks Feb 07 13 – 20 5.60 FALSE
3 Leicester Tigers vs. Worcester Warriors Feb 07 41 – 24 5.20 TRUE
4 London Irish vs. Gloucester Feb 07 32 – 26 1.10 TRUE
5 Newcastle Falcons vs. Exeter Chiefs Feb 07 9 – 15 -8.80 TRUE
6 Wasps vs. Northampton Saints Feb 07 17 – 22 9.80 FALSE

 

Predictions for Round 9

Here are the predictions for Round 9. 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.

Game Date Winner Prediction
1 Exeter Chiefs vs. London Irish Feb 14 Exeter Chiefs 17.10
2 Gloucester vs. Bristol Feb 13 Bristol -2.80
3 Harlequins vs. Leicester Tigers Feb 14 Harlequins 11.20
4 Northampton Saints vs. Newcastle Falcons Feb 14 Northampton Saints 9.60
5 Sale Sharks vs. Bath Feb 13 Sale Sharks 11.80
6 Worcester Warriors vs. Wasps Feb 15 Wasps -5.70

 

Pro14 Predictions for Postponed Matches

Team Ratings for Postponed Matches

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.

 

Current Rating Rating at Season Start Difference
Leinster 18.18 16.52 1.70
Munster 10.20 9.90 0.30
Ulster 9.09 4.58 4.50
Edinburgh 4.04 5.49 -1.40
Glasgow Warriors 3.08 5.66 -2.60
Connacht 1.35 0.70 0.60
Scarlets 1.16 1.98 -0.80
Cheetahs -0.46 -0.46 0.00
Cardiff Blues -0.65 0.08 -0.70
Ospreys -1.46 -2.82 1.40
Treviso -5.35 -3.50 -1.80
Dragons -8.63 -7.85 -0.80
Southern Kings -14.92 -14.92 0.00
Zebre -15.64 -15.37 -0.30

 

Performance So Far

So far there have been 64 matches played, 44 of which were correctly predicted, a success rate of 68.8%.
Here are the predictions for last week’s games.

 

Game Date Score Prediction Correct
1 Dragons vs. Connacht Feb 06 20 – 30 -2.90 TRUE

 

Predictions for Postponed Matches

Here are the predictions for Postponed Matches. 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.

 

Game Date Winner Prediction
1 Dragons vs. Edinburgh Feb 13 Edinburgh -6.90
2 Glasgow Warriors vs. Treviso Feb 14 Glasgow Warriors 14.90

 

February 2, 2021

Vaccines new and very new

First, two new vaccines.  Johnson & Johnson and Novavax put out press releases about their vaccines last week. These are two vaccines that NZ has agreements to buy.  The Novavax vaccine is (a slightly modified version of) the viral spike protein. The mRNA vaccines (Moderna, Pfizer) get your body to make this protein, and the DNA vaccines (AstraZeneca, J&J) get your body to make mRNA to make the protein; a protein vaccine cuts out the middleman.   Novavax are reporting good results, estimating 95% efficacy against the original variants of SARS-2-CoV [?Covid Classic], about 85% against the B 1.1.7 (“UK” variant) and about 60% against the B 1.351 (“South Africa” variant). That’s as good as the mRNA vaccines against the original strains. It might well be as good against the new variants — we don’t have direct randomised trial estimates of effectiveness for the new variants and the mRNA vaccines, but we have antibody tests that suggest substantial but lower protection. Johnson & Johnson didn’t provide much data, and they aren’t claiming as high effectiveness, but they are claiming to prevent serious disease and death, and it’s a single-shot vaccine. We will presumably see more information from both companies when they are evaluated by the FDA and its European counterpart, as we did for earlier vaccines.

Next, the AstraZeneca vaccine in people over 65.  The situation has clarified a bit, though it’s still not clear what number Handelsblatt were quoting.  There is very little Covid data (so far) on people over 65, because there were very few in the trial.  This will improve somewhat when data are released from trials in India and the US.  There is good reason to expect that the vaccine would work in people over 65, but very little direct experience.  In a usual drug-approval situation there is a very strong status-quo bias: the status quo is often not that terrible; the drug doesn’t realistically promise a big improvement; you don’t want to give drug companies incentives to cut corners in trials. Usually it’s better to postpone a decision for a year or so to get better information. With Covid, the status quo is a massive human and economic disaster, and the balance of risks is very different. Leaving people unvaccinated is dangerous, epidemiologically and to the global economy as well as to them personally, so it’s not clear whether the ‘safe’ approach for regulators is to approve the vaccine for all ages or carve out an over-65 exception. It probably depends on the country.

Finally, arithmetic. Israel has been doing well at delivering vaccines, and we’re starting to see rates of infection after vaccination. These aren’t straightforwardly comparable with vaccine efficacy numbers.  A story in the Jerusalem Post has the headline Just 0.04% of Israelis caught COVID-19 after two shots of Pfizer vaccine  and goes on to say (emphasis added)

According to the studies conducted by Pfizer, the vaccine had an efficacy of about 95%, which is considered very high. The Israeli data appear to confirm the inoculation’s effectiveness, showing an even more promising result.Later in the day, Maccabi Healthcare Services – one of the country’s four health maintenance organizations – released the first results of the vaccination campaign of its members, with the organization also comparing the data to a control group that did not get inoculated.
Some 248,000 Maccabi members were already a week after the second shot as of Thursday. Of those, just 66 got infected with the virus, the majority of them over the age of 55 and about half of them with preexisting conditions. All those infected experienced only a mild form of the disease, and none were hospitalized.Over the same period of time, some 8,250 new cases of COVID-19 emerged in the control group of some 900,000 people having a diverse health profile. Those who were not inoculated were therefore 11 times more likely to get the disease than those who were immunized, showing 92% effectiveness.

95% effectiveness means you’d expect 5% of vaccinated people to test positive as infected if 100% of unvaccinated people did. Or 0.05% test positive as infected if 1% of unvaccinated people did. But 1% is a very high rate — the US, at its very worst, wasn’t getting close to 1% of the population as new cases in a week.