NRL 2016 Season analysis

9885ec3187ff9cd0ee98ae23cd040c35The rugby league NRL season 2016 starts tonight with the Eels taking on the Broncos, and once again Sportpunter will be delivering predictions for the upcoming season.
We are excited again with our model, as it has shown over the past two years to be very profitable since we introduced our player rating system. In fact the last 2 years we have made 5.4% ROI from 291 bets, and we hope that that number will continue into 2016.

Subscriptions are available with prices as shown here, and the betting histories are shown here:

But first, let’s do some analysis on how the model has gone, specifically looking at the last two year. Please refer to this webpage, with links to the right outlining the statistical analysis for all figures.

Analysis shows that we have made a 5.4% ROI from 291 bets as previously stated since the player based model started. But interestingly, most of those profits occurred with the Away side. 11% ROI was made from 179 betting on the away team, as opposed to a 2% loss for the home team.

When looking at the probability, a small loss was made betting on teams with probs < 50%, whilst a 6.5% ROI profit was made on our calculated favourites. When looking at analysis by odds, we see an interesting picture. All up a 7.2% loss was made betting on teams with odds less than 1.90, whilst an incredible 21% ROI was made betting on the underdogs. Analysis by overlay showed an increasing profit with increasing overlay which is a great sign of an accurate model.

Analysis by round showed that early stages are very good betting opportunities, with an 11.8% ROI made in the first 20 rounds, whilst the last 10 rounds provided an 8.6% loss. Please note however, that these involve very small sample sizes and not too much should be written into it. However it is interesting analysis and one might want to change their strategies based on this information.

Probably the greatest analysis from above is how the away team and underdogs seem to have gone with the model in the past two seasons. Does the model just pick a good dog, or is it biased to the away team? Well one way to work this out is to test how the model has gone blind betting every game to win $1000.

As shown in the link above, blind betting the home team over the last 2 years has resulted in a 5.8% loss as compared to only a 2.8% loss blind betting the away team. A small difference, but more interestly, blind betting the favourite resulted in a 7.8% loss whilst blind betting the underdog results in a 7.4% profit. One could have easily just bet every dog the last two years and been way on top.

Blind betting the Underdog when away was even better, resulting in a 9.7% advantage.

So is this a long term trend, or just a recent adjustment, or do we not have enough data to come to any concrete conclusions?

Well I decided to look back on all data (that I have) since 2005. Analysis shows here that blind betting the home team resulted in a 2.8% loss, whilst the away team had a 6.7% loss. Blind betting the fav had a 5% loss, whilst blind betting the underdog only a 1% loss. Blind betting the underdog if playing away had a 3.8% loss.
The difference between the losses made blind betting the favourite and the underdog is significant, and it tells me that the underdogs are underrated, maybe moreso recently. Has home ground advantage diminished over the years? Possibly, and teams become more professional in their strategies.

So what should we conclude about the model and best form of attack in 2016? Well in total, we should hopefully be able to produce another great profit, and you must note that in the last two years more bets are suggested on the underdog than the favourite. So any bias in the odds is already accounted for with the model. Will the underdog bias continue? Most likely, but just betting the underdog could be a recipe for disaster as the market adjusts.

To subscribe to Sportpunter’s model in 2016 the click here for all the information.

Related Posts with Thumbnails
This entry was posted in Model, Sport Models. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.