A Look Inside Orb 2.0
We are always looking to update and improve our model to try and make the Orb’s picks as accurate as possible. This unfortunately is not always a one-to-one ratio of work put in to results put out. However, after some development we are are eager to share our most exciting modeling update yet and want to give everyone an inside look at what we have been working on internally. The Orb’s algorithm has always been built on a multi-linear regression model. This allows us to make three types of picks (moneyline, spread and total) off of a single score prediction. The use cases for multi-linear regression are common, but it does have some limitations compared to other model types for decision making such as picks. For this reason, we have spent most of our recent development time creating and testing the Orb’s new logistic regression model and are beyond excited about its potential.
Since starting the Orb we have had the vision of using multiple models to confirm pick accuracy and confidence. Our current valuation of picks (low, medium, high) has been a good start, but is still derived from one model type. With the addition of our new logistic regression model, we will not only have a second stand alone predictor but also will be able to track how the two models perform on picks they agree on. Our 2.0 model is still new and we are testing its picks internally throughout the playoffs before we start incorporating them into the official picks we give out. Looking forward to next season this could change how our pick structure looks. As our readers know, we currently give out all picks across all games which is something we have been looking to move away from as such a large sample size naturally brings down accuracy. Using linear and logistic regression together, we hope to be able to give out less picks per week at a higher confidence and accuracy rate.
How did Orb 2.0 do over Wildcard weekend? Our current testing is only for spread picks but we will lift and shift this to include totals for the remainder of the playoffs. The log model went 4/5 on spreads so far, hitting on 49ers -9.5, Dolphins +13.5, Giants +3, Ravens +10 and missing on Chargers -2.5. The most interesting part to us is that the shared predictions made by both our linear and now logistic models went 2/2 (49ers and Ravens) with the third happening tonight on MNF. Both models predict Buccaneers +2.5 at home vs. the Cowboys so we will be eagerly watching to see if our early vision for next season can start 3/3.
If you have any questions about our new model or what this means for our readers going forward, please don’t hesitate to reach out. If you would like to see our new model’s picks the rest of the playoffs and how they perform compared to our current model just ask and we would be happy to send them privately.
- Team Orb Analytics
.
.
.
.
.
.
DISCLAIMER:
The information provided on this website is for informational purposes only. It is not intended to be gambling or financial advice, and should not be relied upon as such. We are not responsible for any actions or decisions taken by readers based on the information provided on this website.
The picks and predictions provided on this website are based on our own research and analysis, and are intended to be used for entertainment and informational purposes only. We do not guarantee the accuracy or completeness of the information provided, and we are not responsible for any losses or damages incurred as a result of using this information for gambling or other purposes.
By accessing and using this website, you acknowledge and agree to the terms of this disclaimer, and you assume all risks and liabilities associated with your use of the information provided on this website.