RWC 2023 Predictions for Week 2
Team Ratings for Week 2
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 | |
|---|---|---|---|
| Ireland | 17.04 | 15.46 | 1.60 |
| South Africa | 16.79 | 16.35 | 0.40 |
| New Zealand | 15.78 | 16.32 | -0.50 |
| France | 13.02 | 12.48 | 0.50 |
| Scotland | 6.37 | 6.81 | -0.40 |
| England | 4.88 | 4.22 | 0.70 |
| Australia | 3.32 | 2.63 | 0.70 |
| Wales | 1.22 | 0.75 | 0.50 |
| Argentina | 0.57 | 1.23 | -0.70 |
| Fiji | 0.11 | 0.58 | -0.50 |
| Samoa | -0.50 | -0.50 | -0.00 |
| Georgia | -3.25 | -2.56 | -0.70 |
| Japan | -5.19 | -6.10 | 0.90 |
| Italy | -7.32 | -8.84 | 1.50 |
| Tonga | -11.38 | -11.38 | -0.00 |
| Portugal | -13.61 | -13.61 | 0.00 |
| Namibia | -14.70 | -13.18 | -1.50 |
| Uruguay | -15.47 | -15.47 | -0.00 |
| Chile | -16.05 | -15.13 | -0.90 |
| Romania | -18.60 | -17.02 | -1.60 |
Performance So Far
So far there have been 8 matches played, 8 of which were correctly predicted, a success rate of 100%.
Here are the predictions for last week’s games.
| Game | Date | Score | Prediction | Correct | |
|---|---|---|---|---|---|
| 1 | France vs. New Zealand | Sep 08 | 27 – 13 | 3.20 | TRUE |
| 2 | Italy vs. Namibia | Sep 09 | 52 – 8 | 4.30 | TRUE |
| 3 | Ireland vs. Romania | Sep 09 | 82 – 8 | 32.50 | TRUE |
| 4 | Australia vs. Georgia | Sep 09 | 35 – 15 | 5.20 | TRUE |
| 5 | England vs. Argentina | Sep 09 | 27 – 10 | 3.00 | TRUE |
| 6 | Japan vs. Chile | Sep 10 | 42 – 12 | 9.00 | TRUE |
| 7 | South Africa vs. Scotland | Sep 10 | 18 – 3 | 9.50 | TRUE |
| 8 | Wales vs. Fiji | Sep 10 | 32 – 26 | 0.20 | TRUE |
Predictions for Week 2
Here are the predictions for Week 2. 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 | France vs. Uruguay | Sep 14 | France | 35.50 |
| 2 | New Zealand vs. Namibia | Sep 15 | New Zealand | 30.50 |
| 3 | Samoa vs. Chile | Sep 16 | Samoa | 15.50 |
| 4 | Wales vs. Portugal | Sep 16 | Wales | 14.80 |
| 5 | Ireland vs. Tonga | Sep 16 | Ireland | 28.40 |
| 6 | South Africa vs. Romania | Sep 17 | South Africa | 35.40 |
| 7 | Australia vs. Fiji | Sep 17 | Australia | 3.20 |
| 8 | England vs. Japan | Sep 17 | England | 10.10 |
David Scott obtained a BA and PhD from the Australian National University and then commenced his university teaching career at La Trobe University in 1972. He has taught at La Trobe University, the University of Sheffield, Bond University and Colorado State University, joining the University of Auckland, based at Tamaki Campus, in mid-1995. He has been Head of Department at La Trobe University, Acting Dean and Associate Dean (Academic) at Bond University, and Associate Director of the Centre for Quality Management and Data Analysis at Bond University with responsibility for Short Courses. He was Head of the Department of Statistics in 2000, and is a past President of the New Zealand Statistical Assocation. See all posts by David Scott »
Hey David, 100% success on week 1. Well done. I will be following your blog for the duration of the tournament. Do you have any thoughts on the points difference and how you might improve that side of your predictions?
8 months ago
Unfortunately there is not much to be done. Note that the points differences are the mean expected differences in the scores. They depend on the ratings which as I have indicated in my first post are quite tricky to evaluate, much more so than for a regular competition. They can be expected to improve over the course of the tournament but are unlikely to change by a large amount. The big blowout differences in scores could potentially change the ratings of the team involved except for the fact that there is a limit on how much an individual game can reasonably change ratings. This is because the model I use is not a straightforward exponential smooth: it has to account for the fact that the distribution of the score differences is heavy-tailed, that is, extreme differences are more common than if the distribution was a normal distribution. The heavy tailed property of the score differences also means that the variability in scores is great so even if the ratings are accurate so that the prediction of the mean difference is good, the difference between the mean and the actual difference can be very large. In this competition where there are teams which are badly mismatched (Ireland v Romania), it is likely that there will be a blowout. Note also that there can also be some upsets. I expect that there most likely be at least one in pool play.
When it comes to the finals, the likelihood is that scores will be closer. Not sure about the final though because of the very uneven draw: the finalist on one side of the draw is likely to be far stronger than the one from the other side of the draw.
8 months ago
Thanks David. Some big mismatches in this first weekend with Ire v Rom & Ita v Nam. I am really excited to see if the points accuracy improves as we progress through the tournament. Thank you for your insight.
8 months ago
Do you think based on your predictions for round 2 we might have any upsets in the liking of Fiji and tonga matches.
I would love to hear your thoughts.
8 months ago
Fiji would seem the most likely this week, but Japan is a possibility too. Fiji could well have knocked over Wales in their game. Longer term, Fiji and Samoa in the weaker pools would seem to be the most likely to cause an upset.
8 months ago