Super 15 Predictions for Round 12
Team Ratings for Round 12
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 | |
|---|---|---|---|
| Hurricanes | 15.41 | 16.18 | -0.80 |
| Crusaders | 14.50 | 15.23 | -0.70 |
| Highlanders | 11.01 | 10.29 | 0.70 |
| Chiefs | 9.89 | 9.29 | 0.60 |
| Lions | 8.93 | 13.81 | -4.90 |
| Bulls | -0.05 | -4.79 | 4.70 |
| Stormers | -0.31 | 1.48 | -1.80 |
| Sharks | -0.59 | 1.02 | -1.60 |
| Blues | -1.96 | -0.24 | -1.70 |
| Waratahs | -2.18 | -3.92 | 1.70 |
| Brumbies | -2.53 | 1.75 | -4.30 |
| Jaguares | -3.12 | -4.64 | 1.50 |
| Reds | -8.71 | -9.47 | 0.80 |
| Rebels | -9.87 | -14.96 | 5.10 |
| Sunwolves | -17.80 | -18.42 | 0.60 |
Performance So Far
So far there have been 67 matches played, 48 of which were correctly predicted, a success rate of 71.6%.
Here are the predictions for last week’s games.
| Game | Date | Score | Prediction | Correct | |
|---|---|---|---|---|---|
| 1 | Hurricanes vs. Sunwolves | Apr 27 | 43 – 15 | 38.50 | TRUE |
| 2 | Stormers vs. Rebels | Apr 27 | 34 – 18 | 13.20 | TRUE |
| 3 | Reds vs. Lions | Apr 28 | 27 – 22 | -16.20 | FALSE |
| 4 | Blues vs. Jaguares | Apr 28 | 13 – 20 | 6.80 | FALSE |
| 5 | Brumbies vs. Crusaders | Apr 28 | 8 – 21 | -13.00 | TRUE |
| 6 | Bulls vs. Highlanders | Apr 28 | 28 – 29 | -7.90 | 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 | Chiefs vs. Jaguares | May 04 | Chiefs | 17.00 |
| 2 | Rebels vs. Crusaders | May 04 | Crusaders | -20.40 |
| 3 | Waratahs vs. Blues | May 05 | Waratahs | 3.80 |
| 4 | Hurricanes vs. Lions | May 05 | Hurricanes | 10.50 |
| 5 | Stormers vs. Bulls | May 05 | Stormers | 3.20 |
| 6 | Sharks vs. Highlanders | May 05 | Highlanders | -7.60 |
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 »
I am a new subscriber and I am interested in how the ‘current ratings’ are arrived at – for example they obviously take into account the teams win loss records and margins, but what else? Is there anywhere on the site that tells us this – I am interested in the factors being considered – for example are home advantage, previous week as a bye or the strength of this weeks team selection taken into account or not?
The other question I have is that current ratings are listed but they do not seem to solely determine the weeks predictions. For example this week the Blues are ahead of the Warratahs in terms of ratings but the Warratahs are predicted to win. How does this work?
Thanks for the service provided,
Kevin
7 years ago
Only two things go into the predictions, the ratings and home ground advantage. There are two different values for home ground advantage, one for teams from the same country, one for when the teams are from different countries. Ratings are based on score differences in previous games and are updated using exponential smoothing after each round.
Waratahs are at home, Blues are from NZ playing in Australia so home ground advantage is higher than if they were from Australia. Home ground advantage makes the difference here. Waratahs and Blues ratings are pretty similar.
7 years ago
One thing I find most interesting is that one might consider this methodology common sense – a team’s performance being based upon their historical performance against teams vs the caliber of those teams. This common sense is backed up by online pundit picks, i.e. from website superbru which publishes the pick distribution. More often than not in Superbru’s case, the team picked by your algorithms are also the favourite picked under popular vote (50%+ of 190k players picked that team).
However, I noted that earlier in the season your algorithm picked the Jaguares to win a game that some pundits gave no chance – only 3% of players picked the Jaguares to win – and they did indeed win.
It would be interesting to see historically how the algorithm compares to popular vote, especially for close calls. For example, it almost doesn’t require statistics to claim that the Crusaders are going to beat the Sunwolves. However, it would be interesting to analyse cases where the popular vote vastly differs from your algorithm. In the case of the Rebels at the start of the season it makes sense – the Western Force dissolved and the best players mostly went to the Rebels, allowing them to propel up the table. But the case for the Jaguares is much less clear cut.
7 years ago