Orb Analytics Week 12 Recap – Thank You Lamar Jackson!

and

Nov 27, 2024

The Orb avoided disaster striking twice in a row, albeit in a losing effort. A losing week is a losing week, but coming off of a season-worst performance in Week 11, our models saved themselves Monday Night from repeating the 1-4 stat line against the spread. Which days of the NFL week that the Orb hits or misses picks is as out of control as anything else in this random and rigged game that we play. But missing all 3 picks on Sunday is at the very least embarrassing for us internally. Thankfully an awful Sunday performance was bookended by wins on Thursday and Monday Night. In a similar vein of when a pick hits or misses, there is technically no difference between hitting/missing on a favorite or an underdog against the spread, but an unimportant positive that I am taking away from Week 12 is that the models were correct on both sides of the standalone non-Sunday games. Rather than just giving out two favorites or two underdogs and they both happen to hit, the Orb managed to predict the underdog Browns plus the points and the favored Ravens laying 2.5. Again, more of a moral victory than anything.

The Orb’s performance over the last three weeks has been anything but positive. It is probably the coldest spread model out there in that timeframe. This now puts our season-long goal in jeopardy of hitting and changes my mindset about this season. Unlike last year, where the question toward the end of the regular season was just “how much can we win”, now it’s simply “can we win”. Meaning can we finish the season above our Northstar goal of 52.4% against the spread.

One thing I highlighted in our ‘Welcome to the Season’ post is that we specifically don’t promise any level of results due to the difficult and intentionally unpredictable nature of what we are trying to do. As I mentioned then, the only thing that we can guarantee is that the Orb will miss picks along the way. Losing picks, losing weeks, and even losing seasons are statistically inevitable, and it's our job to work to try and limit them as much as possible.

One of the most common questions I, or anyone who runs a model, get is at what point do you tweak the model if results start going south. Like pretty much everything about the world of data, the answer is subjective and there isn’t one hard and fast rule. It comes down to each person running the data. If it is finally time to change the model beyond adding a new week’s worth of training data, the question then becomes how big of a change. Mid-season it is tougher, and riskier, to make fundamental alterations than it is in the offseason when you have more time to backtest. But in our promise of transparency, in light of recent results, I am making a slight update to our models.

In last week’s newsletter, I shared an analysis that showed the relationship between redzone performance and overall cover rate against the spread. While these numbers are baked into advanced stats, up until now the Orb models haven’t specifically been using redzone conversion rates as predictors. Moving forward I am adding them to the list of predictors we have already included, separated out from other advanced stats. Backtesting the models against the last three weeks with the new predictor shows early promising results, certainly better than how the picks played out in real-time. This is a great example of the safe world of testing historically vs. live upcoming events. And hopefully turns into an example of making a positive in-season data decision!

Here is how each pick from the Orb performed in a losing (but digestible!) Week 12:

Spreads:

Browns +3.5 (TNF - Premium) ✅
Rams +2.5 ❌
Cardinals -1.5 ❌
Raiders +6.5 ❌
Ravens -2.5 (MNF - Premium) ✅

All spreads: 2-3, 40%, -1.182 units

Moneyline:

Vikings ✅
Buccaneers ✅
Texans ❌
Commanders ❌
Lions ✅
Broncos ✅
Cardinals ❌
Rams ❌
Ravens (MNF - Premium) ✅

All moneyline, 5-4, 55.6%, -1.796 units

Overall a -2.978 unit performance.

Week 12 Model Performance:

Favorites covered 7/13 games (53.8%) and won 9/13 outright (69.2%).

Weekly spreads:

Weekly Moneyline:

Season-To-Date Model Performance:

Not including week 1, favorites have covered 86/163 games (52.8%) and won 110/163 (67.5%) outright this season.

Season spreads:

Season moneyline:

Official Orb Picks Season Results:

Season spread picks: 32-27, 54.2%, +2.1 units

Season moneyline picks: 61-30, 67.0%, -0.5 units

Confusion Matrix:

The NFL Thanksgiving schedule means we have 4 games before Sunday rolls around. So paid readers, be on the lookout for our yearly ‘Turkey Day Special’ with all the official picks and model leans that come out of Orb for the Thanksgiving and Black Friday matchups.

In the spirit of the holiday, Andy Diamonds and I are beyond thankful for you the Orb Faithful! Thank you from the bottom of our hearts for reading our work and supporting the page. By the nature of the beast, performances will go up and down so especially thanks for sticking with us through a bumpy patch in the season.

For all who celebrate, have a wonderful Thanksgiving! Hopefully it is filled with family, food, and lots of NFL football!

- 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.