Super 15 Predictions, Round 6
Team Ratings for Round 6
This year the predictions have been slightly changed with the help of a student, Joshua Dale. The home ground advantage now is different when both teams are from the same country to when the teams are from different countries. The basic method is described on my Department home page.
Here are the team ratings prior to Round 6, along with the ratings at the start of the season.
| Current Rating | Rating at Season Start | Difference | |
|---|---|---|---|
| Chiefs | 8.92 | 6.98 | 1.90 |
| Crusaders | 7.80 | 9.03 | -1.20 |
| Brumbies | 4.68 | -1.06 | 5.70 |
| Stormers | 3.09 | 3.34 | -0.20 |
| Sharks | 2.56 | 4.57 | -2.00 |
| Hurricanes | 2.21 | 4.40 | -2.20 |
| Bulls | 2.03 | 2.55 | -0.50 |
| Blues | 0.26 | -3.02 | 3.30 |
| Reds | -1.69 | 0.46 | -2.20 |
| Cheetahs | -4.20 | -4.16 | -0.00 |
| Highlanders | -5.64 | -3.41 | -2.20 |
| Waratahs | -6.54 | -4.10 | -2.40 |
| Force | -8.60 | -9.73 | 1.10 |
| Kings | -8.89 | -10.00 | 1.10 |
| Rebels | -10.77 | -10.64 | -0.10 |
Performance So Far
So far there have been 28 matches played, 19 of which were correctly predicted, a success rate of 67.9%.
Here are the predictions for last week’s games.
| Game | Date | Score | Prediction | Correct | |
|---|---|---|---|---|---|
| 1 | Highlanders vs. Hurricanes | Mar 15 | 19 – 23 | -5.60 | TRUE |
| 2 | Waratahs vs. Cheetahs | Mar 15 | 26 – 27 | 2.20 | FALSE |
| 3 | Kings vs. Chiefs | Mar 15 | 24 – 35 | -14.30 | TRUE |
| 4 | Crusaders vs. Bulls | Mar 16 | 41 – 19 | 7.40 | TRUE |
| 5 | Reds vs. Force | Mar 16 | 12 – 19 | 12.50 | FALSE |
| 6 | Sharks vs. Brumbies | Mar 16 | 10 – 29 | 5.90 | FALSE |
Predictions for Round 6
Here are the predictions for Round 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 | Chiefs vs. Highlanders | Mar 22 | Chiefs | 17.10 |
| 2 | Crusaders vs. Kings | Mar 23 | Crusaders | 20.70 |
| 3 | Reds vs. Bulls | Mar 23 | Reds | 0.30 |
| 4 | Force vs. Cheetahs | Mar 23 | Cheetahs | -0.40 |
| 5 | Sharks vs. Rebels | Mar 23 | Sharks | 17.30 |
| 6 | Stormers vs. Brumbies | Mar 23 | Stormers | 2.40 |
| 7 | Waratahs vs. Blues | Mar 24 | Blues | -2.80 |
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 »
Measuring “Performance so Far” by looking at predicting the correct winner is not really a useful measure. Some games are too close to call, while others have clear winners. Even a novice could get decent projections. I would rather suggest that you add 2 additional measures of “Performance”:
1. The mean (average) absolute error
2. The median absolute error per round
In my opinion the median would give the best measure of performance for the following reason. In some games one team dominates to such extent that you get run-away scores. These should be regarded as statistical outliers.
12 years ago
Louis,
I agree in principle but not in detail.
Firstly, the whole point is to predict wins, so you’d want a metric that focuses on small margins. That is, it’s much worse to get a -1 margin as +1 than to get a -5 margin as -7. The mean is too sensitive to extreme results, and the median is insufficiently sensitive to non-extreme results. [updated to add: I’m arguing for some sort of trimmed mean here]
Secondly, the model is already optimising based on squared error, and so based on the mean, so it should look unfairly good there.
And finally: if you want the mean or median prediction, no-one’s stopping you :). You can even post them here in the comments for everyone to read and compare.
12 years ago