Orb Analytics Week 5 Recap – Performance & Results
The Orb had mixed results in week 5. Our spread models had a winning weekend going 3-2, 60% against the spread while our moneyline models once again struggled going 4-4, 50%. We care about and put more emphasis on our models against the spread than on moneyline picks but the disparity between the two of them is now at the highest it has been since starting this project, which is fascinating since predicting outright winners should theoretically be easier than spread picks. The Orb is now hitting 62.5% against the spread but just 51.5% on moneyline picks this season. The moneyline models had a chance to have a neutral or even slight +unit weekend, needing just one of the 49ers or Seahawks to win at home as 6.5 or greater point favorites. Unfortunately for our models, both NFC West teams got upset in their own stadiums. However, while this weekend ultimately resulted in a -0.661 unit performance overall, we had a lot of positives and interesting data points to take away from it.
It paid to pay in week 5! Our models went perfect on the non-Sunday premium picks, going 2-0 on both spreads and moneyline picks taking the Falcons and Chiefs. The Orb extended its winning streak for premium picks to 5 in a row! This resulted in our week 5 premium picks being +3.022 units while the Orb struggled on Sunday going -3.683 units in the main slate of games. It is interesting to track how the models do in the stand-alone Thursday/Monday games compared to the full slate of Sunday games each week, but we are starting to see the importance of both halves to the weekly whole.
A fun trend we saw this week is that favorites are fighting back! They covered and won 71.4% of the games this week, ending a dominant run by the underdogs. While there was just a one-game difference between a winning and losing week against the spread for us in this one, it was cool to see that our models got ahead of the trend and predicted a good week from the favorites. Considering that underdogs were covering 60.4% of games in our training data from this season, I was surprised the models backed favorites to turn it around. Of the 5 spread picks they made in week 5, 4 of them were on the favorite to cover. Of these, they predicted 3/4 correctly, only missing on the Bills who went from favorites to underdogs in the time between giving out the model’s pick on Friday and kickoff on Sunday. The models missed on their only underdog spread and moneyline of the weekend, as once again the Panthers let us down. I had to make a promise to Andy Diamonds and am now making it to you publicly so I can’t go back on it. For the second season in a row, we will be removing any picks on the Carolina Panthers moving forward. One team can only hurt our season numbers so badly.
Here is how each pick performed in week 5:
Spread:
Falcons -1.5 (TNF - Premium) ✅
Ravens -2.5 ✅
Panthers +3.5 ❌
Bills -1.5 ❌
Chiefs -5.5 (MNF - Premium) ✅
All spread picks: 3-2, 60.0%, +0.727 units
Moneyline:
Falcons (TNF - Premium) ✅
Vikings ✅
Ravens ✅
Panthers ❌
Bills ❌
49ers ❌
Seahawks ❌
Chiefs (MNF - Premium) ✅
All moneyline picks: 4-4, 50.0%, -1.388 units
Overall -0.661 units.
Week 5 Model Performance:
Favorites covered 10/14 games (71.4%) and won 10/14 outright (71.4%).
Weekly spreads:
Weekly Moneyline:
Season-To-Date Model Performance:
Not including week 1, favorites have covered 29/62 games (46.8%) and won 35/62 (56.5%) outright this season.
Season spreads:
Season moneyline:
Official Orb Picks Season Results:
Season spread picks: 15-9, 62.5%, +4.6 units
Season moneyline picks: 17-16, 51.5%, -3.8 units
If you asked me how the Orb is performing this season, I would have to say overall well. While the moneyline performance is shocking, I do expect the models to turn things around and start chipping away at the spread deficit. Particularly if favorites continue to win at their expected rate moving forward. But I started the Orb mainly to pick spreads and tend to focus on that in my own feeling of model performance. While the season is still young and anything can happen, I am very pleased with how the models have performed against the spread. Anytime a 60% performance brings your overall spread accuracy down, you are in a good place. So hitting at a 62.5% rate so far against a 52.4% target feels great. The spread model is outperforming itself from last season at this stage. Last year after week 5 our spread models were +3.5 units compared to the +4.6 they are now. The difference in overall performance is that the moneyline models were +2 units at this point, rather than in the -3.8 unit deficit they are now. Last season week 5 was our worst performance overall, so the damage control of a -0.7 unit week feels better in perspective. We do hope that mirroring last season continues, as the Orb went on a 5-week +unit streak between weeks 6 and 10! But we don’t want to get ahead of ourselves and will continue to take things one week at a time, focusing on week 6.
Finally, I wanted to break down the overall 62.5% performance and see how the Orb is doing when predicting favorites vs. underdogs. Queue everyone’s favorite confusion matrix:
As a baseline to compare our results against, during the Orb season, so not including week 1, favorites are covering 46.8% of games and underdogs are covering 53.2% of the time. When the Orb has predicted the favorite to cover, it is hitting at 53.8%. This means our models are hitting 7% higher on favorites than they are actually covering. When the Orb has predicted the underdog, it is hitting 72.7% of the time. This is a 19.5% probability increase when taking an Orb underdog rather than picking one at random. So while the hit rate on favorites is lower, it is above the rate they are currently covering and more importantly is still above the 52.4% target on its own. Orb underdogs are performing significantly above the target so far this season and have been our most profitable picks.
If you are new to the page and want to see our modeling strategy and goals for the season, you can find them linked here. The Orb has a pick for Thursday’s game so our models look to try and extend their winning streak on premium picks to 6 in a row! Let’s see if they can start off week 6 on the right foot and set us up for a good performance Sunday.
- 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.