Posts from November 2023 (15)

November 28, 2023

United Rugby Championship Predictions for Week 7

Team Ratings for Week 7

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 12.01 10.73 1.30
Bulls 9.80 9.00 0.80
Munster 9.68 10.54 -0.90
Stormers 6.91 8.93 -2.00
Glasgow 6.79 4.19 2.60
Ulster 5.87 7.80 -1.90
Connacht 1.86 3.17 -1.30
Sharks 0.79 0.85 -0.10
Lions -0.48 -4.26 3.80
Edinburgh -0.81 -2.26 1.40
Benetton -2.10 -5.12 3.00
Cardiff Rugby -4.47 -5.16 0.70
Ospreys -6.44 -7.79 1.40
Scarlets -9.50 -2.23 -7.30
Dragons -12.16 -8.14 -4.00
Zebre -17.74 -20.25 2.50

 

Performance So Far

So far there have been 48 matches played, 31 of which were correctly predicted, a success rate of 64.6%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Cardiff Rugby vs. Stormers Nov 25 31 – 24 -9.90 FALSE
2 Edinburgh vs. Benetton Nov 25 22 – 24 6.70 FALSE
3 Ospreys vs. Scarlets Nov 25 31 – 9 2.10 TRUE
4 Lions vs. Zebre Nov 25 61 – 19 17.70 TRUE
5 Bulls vs. Connacht Nov 26 53 – 27 9.40 TRUE
6 Sharks vs. Dragons Nov 26 69 – 14 10.90 TRUE
7 Leinster vs. Munster Nov 26 21 – 16 4.20 TRUE
8 Glasgow vs. Ulster Nov 26 33 – 20 3.40 TRUE

 

Predictions for Week 7

Here are the predictions for Week 7. 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 Munster vs. Glasgow Dec 02 Munster 6.90
2 Bulls vs. Sharks Dec 02 Bulls 11.00
3 Cardiff Rugby vs. Scarlets Dec 03 Cardiff Rugby 7.00
4 Lions vs. Dragons Dec 03 Lions 15.70
5 Stormers vs. Zebre Dec 03 Stormers 28.70
6 Ulster vs. Edinburgh Dec 03 Ulster 10.70
7 Benetton vs. Ospreys Dec 03 Benetton 8.30
8 Connacht vs. Leinster Dec 03 Leinster -8.10

 

Top 14 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
Stade Toulousain 6.26 7.24 -1.00
Racing 92 5.44 3.42 2.00
Stade Rochelais 4.42 6.40 -2.00
Bordeaux Begles 4.24 4.07 0.20
Toulon 3.78 3.11 0.70
Stade Francais 2.72 2.62 0.10
Lyon 1.77 2.97 -1.20
Section Paloise 1.64 -0.44 2.10
Clermont 1.15 1.47 -0.30
Castres Olympique 0.99 -0.39 1.40
Montpellier -0.32 2.09 -2.40
Bayonne -0.99 -1.32 0.30
USA Perpignan -5.66 -3.64 -2.00
US Oyonnax -7.84 -10.00 2.20

 

Performance So Far

So far there have been 56 matches played, 42 of which were correctly predicted, a success rate of 75%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Bordeaux Begles vs. USA Perpignan Nov 26 46 – 22 15.60 TRUE
2 Lyon vs. Bayonne Nov 26 42 – 29 8.80 TRUE
3 Montpellier vs. US Oyonnax Nov 26 21 – 26 15.30 FALSE
4 Section Paloise vs. Stade Francais Nov 26 30 – 6 4.20 TRUE
5 Stade Toulousain vs. Clermont Nov 26 31 – 10 10.90 TRUE
6 Toulon vs. Castres Olympique Nov 26 41 – 19 8.40 TRUE
7 Racing 92 vs. Stade Rochelais Nov 27 32 – 10 6.50 TRUE

 

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 Bayonne vs. Montpellier Dec 03 Bayonne 5.80
2 Castres Olympique vs. Lyon Dec 03 Castres Olympique 5.70
3 Clermont vs. Racing 92 Dec 03 Clermont 2.20
4 Stade Rochelais vs. USA Perpignan Dec 03 Stade Rochelais 16.60
5 Toulon vs. Section Paloise Dec 03 Toulon 8.60
6 US Oyonnax vs. Bordeaux Begles Dec 03 Bordeaux Begles -5.60
7 Stade Francais vs. Stade Toulousain Dec 04 Stade Francais 3.00

 

Rugby Premiership Predictions for Round 8

Team Ratings for Round 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.

Current Rating Rating at Season Start Difference
Saracens 8.38 6.82 1.60
Leicester Tigers 4.39 6.13 -1.70
Sale Sharks 4.09 6.46 -2.40
Exeter Chiefs 3.58 -0.89 4.50
Harlequins 2.44 1.93 0.50
Northampton Saints 1.98 2.66 -0.70
Bristol 1.85 1.00 0.90
Bath 1.64 -1.26 2.90
Gloucester -5.53 -1.45 -4.10
Newcastle Falcons -12.07 -10.63 -1.40

 

Performance So Far

So far there have been 35 matches played, 24 of which were correctly predicted, a success rate of 68.6%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Northampton Saints vs. Harlequins Nov 25 36 – 33 6.20 TRUE
2 Sale Sharks vs. Bath Nov 25 11 – 9 9.90 TRUE
3 Saracens vs. Bristol Nov 26 39 – 31 13.50 TRUE
4 Gloucester vs. Leicester Tigers Nov 26 20 – 38 -1.00 TRUE
5 Newcastle Falcons vs. Exeter Chiefs Nov 27 14 – 20 -10.50 TRUE

 

Predictions for Round 8

Here are the predictions for Round 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.

Game Date Winner Prediction
1 Harlequins vs. Sale Sharks Dec 02 Harlequins 4.40
2 Bath vs. Exeter Chiefs Dec 03 Bath 4.10
3 Bristol vs. Gloucester Dec 03 Bristol 13.40
4 Saracens vs. Northampton Saints Dec 03 Saracens 12.40
5 Leicester Tigers vs. Newcastle Falcons Dec 04 Leicester Tigers 22.50

 

November 24, 2023

Detecting ChatGPT

Many news stories and some StatsChat posts have talked about detecting the output of Large Language Models. At the moment, tools to do this are very inaccurate.  Denouncing, for example, a student paper, based on these detectors wouldn’t be supportable. Even worse, the error rate is higher for people who aren’t native English speakers, a group who can already be accused unfairly.

We might hope for better detectors in the future.  If people using ChatGPT have access to the detector, though, there’s a pretty reliable way of getting around it. Take a ChatGPT-produced document, and make small changes to it until it doesn’t trigger the detector.  Here we’re assuming that you can make small changes and still get a good-quality document, but if that’s not true — if there’s only one good answer to the question — there’s no hope for a ChatGPT detector to work.  Additionally, we’re assuming that you can tell which random changes still produce a good answer.  If you can’t, then you might still be able to ask GPT whether the answer is good.

A related question is whether Large Language Model outputs can be ‘watermarked’ invisibly so as to be easier to detect. ChatGPT might encode a signature in the first letters of each sentence, or it might have subtle patterns in word frequencies or sentence lengths. Regrettably, any such watermark falls to the same attack: just make random changes until the detector doesn’t detect.

On the preprint server arXiv recently was a computer science article arguing that even non-public detectors can be attacked in a similar way. Simply take the Large Language Model output and try random changes to it, keeping the changes that don’t mess up the quality.  This produces a random sample from a cloud of similar answers. If there aren’t any similar answers accessible by small changes, it’s going to be hard for the AI to insert a watermark, so we can assume there will be.  ChatGPT didn’t actually produce these similar answers, so a reasonable fraction of them should not trigger the ChatGPT detector.  Skeptics might be reassured that the researchers tried this approach on some real watermarking schemes and it seems to work.

November 23, 2023

Whole lotta baseball

From Ars Technica (and while it’s not a story about baseball, it is is trying to use numbers to mean something)

It’s actually 162 regular season games a year for 30 teams which means, 2,430 games a year. That’s 32,805 hours of baseball based on the average length of a game lasting 162 minutes. The regular season is 185 days long, which equals 4,440 hours. So there’s more baseball than time.

These numbers struck me as wrong immediately.  If there are 32k hours of baseball in 4k hours of regular season, it means an average of eight baseball games being played at any hour of the day or night. Since there’s a maximum of 15 games being played simultaneously (because 30 teams), that would mean a full baseball schedule for an average of nearly 12 hours every day.  There is a lot of baseball, but not that much.  They don’t play at 3am, and they take occasional days off to travel.

So, let’s run the numbers:

  • 162 games by 30 teams is 162×15 games, or 2430 games.
  • Average game lasts 162 minutes. 162×2430 is 393660 minutes, or 393660/60=6561 hours.
  • 185 day season is 185×24=4440 hours

The total hours of baseball seems off. In fact, it’s off by exactly a factor of five, suggesting the story was working with 12-minute hours for some reason.  With 6561 hours of baseball in a 4440 hour season, we’re looking at about 1.5 baseball games simultaneously, averaged over the season, which is more plausible.

While we’re at it, we might want to check on the 162 minutes/game since it’s a bit suspicious for two unrelated numbers in the same calculation to both be 162.  It’s right, at least for 2023, though it’s down from over 3 hours the previous season.

November 21, 2023

United Rugby Championship Predictions for Week 6

Team Ratings for Week 6

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 11.92 10.73 1.20
Munster 9.76 10.54 -0.80
Bulls 8.55 9.00 -0.40
Stormers 8.18 8.93 -0.70
Ulster 6.64 7.80 -1.20
Glasgow 6.02 4.19 1.80
Connacht 3.11 3.17 -0.10
Edinburgh -0.11 -2.26 2.10
Sharks -2.23 0.85 -3.10
Lions -2.25 -4.26 2.00
Benetton -2.80 -5.12 2.30
Cardiff Rugby -5.75 -5.16 -0.60
Ospreys -7.91 -7.79 -0.10
Scarlets -8.03 -2.23 -5.80
Dragons -9.14 -8.14 -1.00
Zebre -15.98 -20.25 4.30

 

Performance So Far

So far there have been 40 matches played, 25 of which were correctly predicted, a success rate of 62.5%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Edinburgh vs. Bulls Nov 18 31 – 23 -6.90 FALSE
2 Ulster vs. Lions Nov 18 24 – 17 14.40 TRUE
3 Zebre vs. Cardiff Rugby Nov 18 22 – 22 -7.80 FALSE
4 Sharks vs. Connacht Nov 19 12 – 13 -1.40 TRUE
5 Dragons vs. Ospreys Nov 19 20 – 5 -1.80 FALSE
6 Munster vs. Stormers Nov 19 10 – 3 5.20 TRUE
7 Glasgow vs. Benetton Nov 19 26 – 12 12.50 TRUE
8 Leinster vs. Scarlets Nov 19 54 – 5 19.80 TRUE

 

Predictions for Week 6

Here are the predictions for Week 6. 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 Cardiff Rugby vs. Stormers Nov 25 Stormers -9.90
2 Edinburgh vs. Benetton Nov 25 Edinburgh 6.70
3 Ospreys vs. Scarlets Nov 25 Ospreys 2.10
4 Lions vs. Zebre Nov 25 Lions 17.70
5 Bulls vs. Connacht Nov 26 Bulls 9.40
6 Sharks vs. Dragons Nov 26 Sharks 10.90
7 Leinster vs. Munster Nov 26 Leinster 4.20
8 Glasgow vs. Ulster Nov 26 Glasgow 3.40

 

Top 14 Predictions for Round 8

Team Ratings for Round 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.

Current Rating Rating at Season Start Difference
Stade Toulousain 5.90 7.24 -1.30
Racing 92 4.93 3.42 1.50
Stade Rochelais 4.93 6.40 -1.50
Bordeaux Begles 3.82 4.07 -0.30
Stade Francais 3.36 2.62 0.70
Toulon 3.32 3.11 0.20
Lyon 1.56 2.97 -1.40
Clermont 1.51 1.47 0.00
Castres Olympique 1.45 -0.39 1.80
Section Paloise 1.01 -0.44 1.40
Montpellier 0.33 2.09 -1.80
Bayonne -0.78 -1.32 0.50
USA Perpignan -5.23 -3.64 -1.60
US Oyonnax -8.49 -10.00 1.50

 

Performance So Far

So far there have been 49 matches played, 36 of which were correctly predicted, a success rate of 73.5%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Bayonne vs. Section Paloise Nov 18 35 – 16 3.70 TRUE
2 Castres Olympique vs. Stade Toulousain Nov 18 31 – 23 1.40 TRUE
3 Clermont vs. Toulon Nov 18 27 – 30 5.50 FALSE
4 Stade Francais vs. Racing 92 Nov 18 9 – 13 5.90 FALSE
5 Stade Rochelais vs. Bordeaux Begles Nov 18 25 – 21 8.00 TRUE
6 US Oyonnax vs. Lyon Nov 18 38 – 20 -5.00 FALSE
7 USA Perpignan vs. Montpellier Nov 18 23 – 16 0.30 TRUE

 

Predictions for Round 8

Here are the predictions for Round 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.

Game Date Winner Prediction
1 Bordeaux Begles vs. USA Perpignan Nov 25 Bordeaux Begles 15.60
2 Lyon vs. Bayonne Nov 25 Lyon 8.80
3 Montpellier vs. US Oyonnax Nov 25 Montpellier 15.30
4 Racing 92 vs. Stade Rochelais Nov 25 Racing 92 6.50
5 Section Paloise vs. Stade Francais Nov 25 Section Paloise 4.20
6 Stade Toulousain vs. Clermont Nov 25 Stade Toulousain 10.90
7 Toulon vs. Castres Olympique Nov 25 Toulon 8.40

 

Rugby Premiership Predictions for Round 7

Team Ratings for Round 7

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
Saracens 8.88 6.82 2.10
Sale Sharks 4.80 6.46 -1.70
Exeter Chiefs 3.99 -0.89 4.90
Leicester Tigers 2.91 6.13 -3.20
Northampton Saints 2.30 2.66 -0.40
Harlequins 2.13 1.93 0.20
Bristol 1.34 1.00 0.30
Bath 0.93 -1.26 2.20
Gloucester -4.05 -1.45 -2.60
Newcastle Falcons -12.49 -10.63 -1.90

 

Performance So Far

So far there have been 30 matches played, 19 of which were correctly predicted, a success rate of 63.3%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Bath vs. Bristol Nov 18 20 – 19 6.60 TRUE
2 Sale Sharks vs. Newcastle Falcons Nov 18 40 – 22 24.50 TRUE
3 Leicester Tigers vs. Northampton Saints Nov 19 26 – 17 6.00 TRUE
4 Harlequins vs. Saracens Nov 19 10 – 38 4.70 FALSE
5 Exeter Chiefs vs. Gloucester Nov 20 25 – 24 16.80 TRUE

 

Predictions for Round 7

Here are the predictions for Round 7. 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 Northampton Saints vs. Harlequins Nov 25 Northampton Saints 6.20
2 Sale Sharks vs. Bath Nov 25 Sale Sharks 9.90
3 Saracens vs. Bristol Nov 26 Saracens 13.50
4 Gloucester vs. Leicester Tigers Nov 26 Leicester Tigers -1.00
5 Newcastle Falcons vs. Exeter Chiefs Nov 27 Exeter Chiefs -10.50

 

November 18, 2023

Bird of the Century

For years, Bird of the Year has been the only self-selected (‘bogus’, ‘straw-poll’, ‘unscientific’) survey endorsed by StatsChat.  The unique feature of Bird of the Year as a bogus poll is that no-one pretends it is anything else. The pūteketeke won Bird of the Century fair and square, due to John Oliver’s marketing efforts, and no-one seriously thinks this says anything real about the relative popularity of New Zealand birds.

The key takeaway from Bird of the Century is this is what bogus polls are like. All of them. When a bogus poll agrees with population opinion, it’s just an accident.  When someone claims to have information from a survey result, it’s always good to ask whether it’s the type of survey that’s more accurate than just pulling a number out of your arse, or not.

A couple of weeks ago, there was a widelyreported claim made by a spokesperson for AA Insurance that 53% of people in New Zealand wanted a ban on domestic use of fireworks. None of the media outlets asked (or reported asking) anything about how he got that number.  When a news report says something that’s attributed to an anonymous source like that, you want to know who is vouching for the credibility of the source.

I happened to see a post on social media by someone who had been in a survey that could have been the one quoted, which was run by my2cents.  I don’t know how good their surveys are, but they at least qualify as trying to get the right answer.  If that survey was actually the one reported by AA Insurance, it would be good to know.

In some contexts, such as election polling or policy decisions, you might want to know more about the methods used and the reputation of the pollsters.  Even in simple news reporting, though, it’s important to ask if this is the sort of survey that gives you information or the sort of survey that just gives you grebes.

November 14, 2023

United Rugby Championship Predictions for Week 5

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.

Current Rating Rating at Season Start Difference
Leinster 9.84 10.73 -0.90
Bulls 9.69 9.00 0.70
Munster 9.58 10.54 -1.00
Stormers 8.36 8.93 -0.60
Ulster 7.38 7.80 -0.40
Glasgow 5.87 4.19 1.70
Connacht 3.16 3.17 -0.00
Edinburgh -1.25 -2.26 1.00
Sharks -2.27 0.85 -3.10
Benetton -2.65 -5.12 2.50
Lions -2.98 -4.26 1.30
Cardiff Rugby -4.97 -5.16 0.20
Scarlets -5.94 -2.23 -3.70
Ospreys -6.65 -7.79 1.10
Dragons -10.40 -8.14 -2.30
Zebre -16.76 -20.25 3.50

 

Performance So Far

So far there have been 32 matches played, 20 of which were correctly predicted, a success rate of 62.5%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Zebre vs. Sharks Nov 11 12 – 10 -12.70 FALSE
2 Cardiff Rugby vs. Bulls Nov 11 12 – 18 -11.80 TRUE
3 Ulster vs. Munster Nov 11 21 – 14 -1.60 FALSE
4 Benetton vs. Stormers Nov 12 20 – 17 -8.90 FALSE
5 Ospreys vs. Glasgow Nov 12 23 – 31 -8.70 TRUE
6 Scarlets vs. Lions Nov 12 23 – 24 1.50 FALSE
7 Edinburgh vs. Connacht Nov 12 25 – 22 -1.30 FALSE
8 Dragons vs. Leinster Nov 12 10 – 33 -14.90 TRUE

 

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.

Game Date Winner Prediction
1 Edinburgh vs. Bulls Nov 18 Bulls -6.90
2 Ulster vs. Lions Nov 18 Ulster 14.40
3 Zebre vs. Cardiff Rugby Nov 18 Cardiff Rugby -7.80
4 Sharks vs. Connacht Nov 19 Connacht -1.40
5 Dragons vs. Ospreys Nov 19 Ospreys -1.80
6 Munster vs. Stormers Nov 19 Munster 5.20
7 Glasgow vs. Benetton Nov 19 Glasgow 12.50
8 Leinster vs. Scarlets Nov 19 Leinster 19.80