Predictions for the RWC Cup Semi-Finals
Team Ratings for the RWC Cup Semi-Finals
The basic method is described on my Department home page.
Here are the team ratings prior to this weekend’s games, along with the ratings at the start of the cup.
| Current Rating | Rating at Season Start | Difference | |
|---|---|---|---|
| New Zealand | 28.74 | 27.60 | 1.10 |
| England | 21.46 | 19.78 | 1.70 |
| South Africa | 20.59 | 18.51 | 2.10 |
| Ireland | 10.56 | 12.12 | -1.60 |
| Wales | 9.61 | 12.72 | -3.10 |
| Australia | 6.60 | 8.70 | -2.10 |
| Scotland | 6.14 | 5.44 | 0.70 |
| France | 5.26 | 7.11 | -1.80 |
| Argentina | 3.32 | 4.15 | -0.80 |
| Japan | -5.52 | -7.67 | 2.20 |
| Fiji | -5.54 | -5.27 | -0.30 |
| Italy | -9.53 | -8.57 | -1.00 |
| Samoa | -15.53 | -11.89 | -3.60 |
| Tonga | -15.83 | -20.46 | 4.60 |
| Georgia | -18.08 | -17.43 | -0.70 |
| USA | -18.64 | -16.38 | -2.30 |
| Uruguay | -30.89 | -35.67 | 4.80 |
| Canada | -34.63 | -32.45 | -2.20 |
| Russia | -36.32 | -36.78 | 0.50 |
| Namibia | -39.99 | -41.80 | 1.80 |
Performance So Far
So far there have been 41 matches played, 38 of which were correctly predicted, a success rate of 92.7%.
Here are the predictions for last weekend’s games.
| Game | Date | Score | Prediction | Correct | |
|---|---|---|---|---|---|
| 1 | England vs. Australia | Oct 19 | 40 – 16 | 12.90 | TRUE |
| 2 | New Zealand vs. Ireland | Oct 19 | 46 – 14 | 15.10 | TRUE |
| 3 | Wales vs. France | Oct 20 | 20 – 19 | 5.10 | TRUE |
| 4 | Japan vs. South Africa | Oct 20 | 3 – 26 | -18.90 | TRUE |
Predictions for the RWC Cup Semi-Finals
Here are the predictions for the RWC Cup Semi-Finals. 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 | England vs. New Zealand | Oct 26 | New Zealand | -7.30 |
| 2 | Wales vs. South Africa | Oct 27 | South Africa | -11.00 |
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 »
Why would South Africa beat Wales? In the last three matches Wales beat them.
5 years ago
Because their last game was nearly a year ago, when their ratings were very similar in my rating calculations (about 13). Since then Wales were about the same at the start of the RWC, but South Africa improved markedly (draw with All Blacks in NZ for one). Since the start of the RWC Wales rating has gone down, South Africa’s has gone up.
Some of the increased rating of South Africa has been due to big wins against lesser opposition. I do have a parameter to limit the effect of those wins on a team’s ratings however.
Many commentators have suggested South Africa have improved over the last year. Wales had some good wins but also some losses (2 to Ireland) and some close games against lesser opposition. I think they have been unimpressive in the RWC and their last try against France was clearly forward, even the action of the Wales player was ripping the ball in a forward direction.
My predictions are entirely statistical and objective but I do like to question their rationality because I could have made an error. In this case having looked at where the ratings came from, I don’t see a problem.
5 years ago
Thank you for the response! I was pretty much on the fence with this one. Although I don’t feel SAs performance has been that great either.
5 years ago
I see you have New Zealand beating England. For this match I was very unsure and I felt that England had improved considerably over recent years and also through this World Cup.
Also the last time they played New Zealand won by only one point!
5 years ago
As I have said before my predictions are based on a simple statistical technique, exponential smoothing, and as such are not subjective. Besides being a bit of enjoyment for me, they are message about the value of statistics in decision making.
One advantage of a statistical approach to forecasting rugby is that evidence is collected from all relevant games. Your feelings are unlikely to take into account all games, because it is simply hard to remember them all and then factor them in to your prediction.
Here are some points to consider. Firstly that win by the ABs was in England, and was nearly a year ago and my England rating changed by under 0.5. Both teams have changed since that game. England’s rating has improved from roughly 15 to 20 before the World Cup, and 1.7 since the RWC start. When examining how much England may have improved games I can point to are a draw with Scotland at home, and 2 losses to Wales since playing the ABs, but obviously some good performances.
As to the RWC itself, England beat Australia who never had the same 9 and 10 from one game to the next, with two intercept tries, one in junk time when the game was over. The ABs killed Ireland who had beaten them in two out of their last three matches. Beware of placing too much emphasis on the last available head to head results!
So, I will go with my predictions. That said they are only predictions and only reliable when considered over a large number of matches. Anything can happen in a particular match on the day (red card?).
As an ABs supporter I will be as nervous as anyone on the day.
5 years ago
I understand. Thank you for your response. I also hope New Zealand beat England.
Best wishes
Conor
5 years ago
Hi David, ave you any idea why there has been such a marked difference between correct predictions of the Mitre 10 Cup and the Rugby World Cup? Has this occurred because of the greater amount of games played resulting in considerably more factors at play for the Mitre 10 Cup which simply boils down to chance?
Regards Chris.
5 years ago
Interesting question. I think the Mitre 10 and RWC differences are explicable in part because many pool games are highly predictable and to get a different result is a major upset (Japan v Ireland, etc). Of the closer games, getting them right is a bit of luck. Note there were only three where my predicted margin was under 5. I haven’t looked but I suspect that the margins would be a lot smaller in the Mitre 10.
Over the years making these predictions, I expect in most rugby competitions to get a bit over 70% of results correct (winner only, not a given margin).
Curiously, the NRL is much harder. Except for this year, just over 60% would be typical. Why? Closer competition because of salary cap? Easier for one side to get a run on? My observation is that there a lot of unusual results, big turns around in performance. As a result I have always harboured suspicions that the NRL is not clean, and there are other reasons to suspect that.
5 years ago
Forgot to mention one difference between league and union. The set piece. In league it is a joke whereas in union it is important, and it would seem that ability at the set piece would be reasonably consistent over time, leading to more consistent performance.
5 years ago
Hi David,
I find your prediction very fascinating, can you please setup a tutorial, going through the process of predicting one game.That will be very helpful!
5 years ago
Hi David great predictions really intrested to see how the semi’s play out. Do these predictions take injuries or referees into consideration? For example SA have a terrible record under Jerome Garces but Wales have a few injury concerns.
5 years ago
They had the spider camera for that Welsh forward pass and from that angle it proved very inconclusive.
5 years ago
With one of the key players of South Africa, Cheslin Kolbe being unfit, how does it impact the prediction.
5 years ago
I will answer the last few comments here.
1. There is information about the method on my home page with the link on every post. I agree it is a bit brief and I think I will try and expand it. Only problem is the location of my home page has changed and I am not sure how to access it to edit it right now.
2. The only data relevant to my predictions are past results, namely the difference in scores of the two teams involved, and the home ground advantage. Home ground advantage was only relevant in this RWC in games where Japan was playing.
If you think the referee or the presence or absence of particular players is important, then you should adjust your personal view of the game margin as you see fit. I make no recommendations in that regard.
3. Wales final try was a clear forward pass. Check the starting position of the ball and its position when caught in relation to the five yard line. The camera angle shown at the time was poor and that TMO is as bad as Garces :-)
5 years ago