Week 6 was the worst week of the season so far for the model. The ML predictions were still positive, but the spread predictions were pretty bad. The record for the week was 6-8 on spread and 8-6 on ML.
At the end of the day, the followers were very kind on my post on instagram about the results this week. I expected some hate (and there was) but the comments I received were overwhelmingly positive. I think it is because this page is one of the few pages that is extremely transparent. I post the losses just the same as I post the wins. Obviously I have to market the wins much harder than I market the losses, but still.
I’ve said it before and I’ll say it again: I am not a handicapper and I am not consulting/advising anyone to make bets. I post statistical outputs of my machine learning model and if people want to use them they can, if they don’t they don’t have to!
I love when people message me that they won big with my picks, and it has happened a LOT. I’ll always celebrate a win with the people. However, I am not responsible for any of those wins because I didn’t tell anyone what to do with THEIR money. In the exact same way, I will take no responsibility for losses because I am not advising anyone to make any decisions with their money.
This is a data science blog, and it always has been. I will never tell anyone to make decisions with their money on sports betting because it is truly not a smart investment decision. If you go into the sports betting world with a mindset of “I’m going to get rich from this” you’ve already lost. If you want to really make returns on your money, go to the stock market! Bet sports because it’s fun and no other reason.
Overall, this week I was disappointed with the results from the model. I have made some tweaks to the model that I think will result in some improvement, but I am not sure how confident I am in the predictions this week coming off a tough week.
Especially now that I have gained hundreds of new followers over the past couple days. Let’s dive into Week 7.
This week the model is again making sweeping decisions on its picks. The first few weeks, the model was often choosing the underdog to cover but the favorite to win. This strategy actually proved to be beneficial for the records because it guaranteed a 50% win rate on each game. When the model starts picking underdogs to win outright and favorites to cover, it becomes a zero sum approach. Either all or nothing. Which is what it’s done again this week.
As a fan however, the model seems to be pretty close to what people are expecting to happen which a) makes sense and b) is somewhat nerve racking. When the model just agrees with the public, what is the value? There are definitely some surprises but the stats need to show something a bit more valuable than just public opinion as well.
There are a couple of surprises this week. Interestingly enough, the model has predicted two ties. The model believes that the Falcons-Lions game and the Seahawks-Cardinals game will both result in ties. Technically, the model is predicting ties because there is rounding that happens and there are slight advantages for each of those games, but its still the first time the model is showing two ties in the same week.
The biggest surprise game-wise is probably the Seahawks-Cardinals. The Seahawks are 5-0 and coming off a bye week, and the model believes they could lose to the Cardinals by less than a point. This is interesting because it might be showing that bye weeks are actually worse for a team. We shall see!
The other surprise is the prediction for the Panthers to outright win against the Saints. The Saints are -7, which means Vegas has them heavily favored to win. All season, the model has disliked the Saints and has loved the Panthers. For whatever reason, we will see whether or not the model truly understands these teams.
Unfortunately, the model has once again picked the Jets to cover this weekend. The number of times that the model has picked the Jets to cover and they didn’t is a lot. I don’t think they stand a chance against the Bills and, personally, I think the Bills are going to smoke them, but the stats say otherwise I guess we’ll have to find out!
Excited to see how the model performs this weekend but nervous for another bad week, let’s go model!
Everything I do is for free because of my passion for sports and data science. I always post the predictions on instagram the day before gameday, but if you’d like the advanced analytics and predictions for the games days before everyone else, please check out the Patreon page linked here. In the same notion of this being free, every donation is appreciated if you’d like to help keep operations running, click the yellow button at the top of the page for our PayPal.
Check out our YouTube channel with our first video of the inner workings of the model here.
Check out our merch shop here. And if youd like’ to buy NFL data, check here.
And as always, good luck bettors!