Week 2 Newsletter and Picks

How fun was that? It was the first full weekend of meaningful football games, and I hope you guys enjoyed them as much as I did. With week 1 in the books, we now have a small amount of 2023 NFL data to analyze and overreact to! We always want the Orb to perform better, but this week's results help highlight our hypothesis that bad data = bad results. For the second year in a row, we saw that the previous season's advanced stats do not accurately represent the teams we see on the field to start the season. They may wear the same uniforms and have similar rosters, but these aren't your 2022 NFL teams. As we have mentioned many times before, we use the first few weeks of the season to load up the Orb with an appropriate amount of team data so that it feeds our models accurate representations of the teams. Ideally, this loading time should be as short as possible. Still, the inaccurate starts can be directly attributed to insufficient information as we continue to gather data. We hammer home this point a lot, but I want to be fully transparent as we work through the modeling process. All this has been to thank you in advance for your continued patience, and we hope to repay you with long-term, consistent accuracy. 

Let's break down what happened in week one. Using only 16 data points, here are the early trends we are seeing:

• 10/16 (62.5%) underdogs covered the spread. 8/16 (50%) won the game outright.

• Of the six favorites that covered the spread, only two were at home, meaning road favorites covered in 4/6 games (66.65%).

• Of the eight favorites that won outright, only half were at home. Homefield advantage did not exist in week 1.

• There were only two games in which the favorite won, and the underdog covered the spread. If you liked the underdog to cover, it was worth it to take them to win outright, and if you wanted a favorite to win outright, it would have been worth it to take them to cover the spread.  

• Models A and B tied for the best individual performances this week, going a lackluster 8/16 against spreads across all 16 games, while Model C struggled at just 6/16. 

We will continue to break down exciting trends we see throughout the season as we get more game data weekly. 

Turning the page to week 2, let's see how the data overreacts to week 1. In the same vein as our disclaimers about input data in our week 1 picks, week 2 is based solely on one week of team data. Again, the models have plenty of historical data, but the teams' metrics are based only on last weekend's performances. The input data sees Dallas as the #1 defense, Miami as the #1 offense, and the Giants as a team that scores 0 points per game (lol). Here are the Orb's picks for week 2 in which all three of our models agree:

Spread:

• Falcons -1.5

• Seahawks +4.5

• Titans +3

• Colts +1.5

• Cardinals +4.5

• Cowboys -8.5

• Patriots +3

Moneyline:

• Falcons

• Buccaneers

• Bills

• Chargers

• Ravens

• Cowboys

• Saints

• Browns

Even though the models got out to an expected cold start, the founders of the Orb somehow stumbled our way into a 7-1 record in our head-to-head picks, which you can find here. If you missed this blog, Andy Diamonds and I are doing a weekly picks competition where we give out two underdogs and two favorites we like against the spread. These picks are not the Orb's (although they may sometimes overlap) but are powered by the brains of two idiots. But after starting at 87.5%, we may be the Orb. 

Last bit of housekeeping to close out the week. As you may have noticed, this week's newsletter did not come out on Wednesday like previous ones. Last week was a gentle reminder of why we decided not to give out Thursday night picks. You spend all offseason developing three distinct machine learning models that all pick Kansas City only to have the Chiefs' best defensive player sit out because of a contract dispute, the best tight end in the league injure his knee days before the game, and their theoretical #1 wide receiver turn this play into 7 points for the Detroit Lions and swing the spread: 

But in all seriousness, we want to focus on the accuracy of our picks, and putting them out five days before kickoff on Sunday only allows for lines to move and other factors like injury news to come out midweek. So, moving forward, we will send these out closer to Sunday.

Let's have a fantastic week two! Data overreaction time!

- 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 readers take based on the information provided on this website.

The picks and predictions provided on this website are based on our 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 from using this information for gambling or other purposes.

By accessing and using this website, you acknowledge and agree to the terms of this disclaimer. You assume all risks and liabilities associated with using the information provided on this website.