In the world of sports analytics, there is a constant focus on trying to predict the winner of games, win/loss totals and ultimately predict the Super Bowl Champion. What oddsmakers and data scientists have spent decades trying to decipher, is a formula that could deliver with some reliability a high percentage probability on the outcome of a game or season. I too have spent years toiling over formulas and combing through data, but today I believe I have stumbled upon the promise land. I present to you the RPC computer modeling algorithm. I will continue to tweak and work on this formula as time passes. However, I want to document the models, their predictions, and ultimately results. To paraphrase a wild eyed scientist… when this algorithm is finalized you’re gonna see some serious shit.
So what has the model actually predicted? I’m so glad you asked, let’s go year by year starting with the 2020 season.
This model was run prior to the 2020 playoffs. The BRIGHT GREEN line indicates the score on the RPC scale. The higher the number, the more likely the team is to win the Super Bowl. According to this model the most likely Super Bowl Champion was the Tampa Bay Buccaneers. This is the model I ran that started this entire project. As the playoffs went on I went back to my data and realized, “Oh my GOD, that’s either a nice coincidence or maybe I’m on to something.” I went into the lab, triple checked my numbers and not only did the model correctly predict the Super Bowl winner, but also predicted all four teams in the Conference Championship games. Additionally, 10 of the top 11 teams in our model all made the playoffs. This is where I would like to point that no part of model takes into accounts wins or losses, so to have 10 of the top 11 teams in the model make the playoffs, the top 4 teams be the last four remaining on Conference Championship weekend, AND the top team win the Super Bowl?…. yea’d I’d say we are F**KING on to something. Our model also went 10-2 in the playoffs as a stand alone modeling system. I will break down this data further in a bit. For now lets go back and look at what the SAME modeling system told us in 2019.
Above is the 2019 projection. I want to reiterate that none of the data uses wins or losses, and all data is compiled using regular season data. Again, the BRIGHT GREEN LINE points you to the predicted Super Bowl winner, in this case… The Kansas City Chiefs! 2/2 doesn’t sound too bad. Now here is a good time to explain the rest of the lines on the chart. The greyish/blue lines represent alternate versions or our computer model. Think of it like a spaghetti model for a hurricane chart. While these other lines have not shown the accuracy of the RPC model, they do show strong strengths that correspond to success in certain environments. You will note that 3 of the models on the 2019 chart show the 49ers at the top of the chart. There is some additional really fun data on this chart I will get into later, but trying to keep this as simple and straight forward as possible for now.
Now of course you are wondering, okay well what does the model actually use to compute probabilities? Fair question, and I’ll say as nice as possible, “none of your f**king business.” What I will say is that it takes into account all kinds of statistical categories (for example first down conversion percentage, turnovers, penalties, etc.). After which a value has been placed on each category based on the correlation of the value to being important for chances to win a Super Bowl. That’s as detailed as I would like to be at the moment as I have now wiped my main computer, put flash drives into safety deposit boxes, and written my entire code in cyphers so that only I can understand it. Sound crazy, perhaps. But if I’m anywhere close to as accurate as I think I am this information is priceless. Find out for yourself though, as the season goes on we will post updates of our model and you can follow along and find out how accurate our models are for yourself.
I will continue to update not only prior years of data as I continue to pull it (you have no idea how time consuming it is), and also post the 2021 data as the season progresses. I will say the data is beyond promising, 4 out of the last 5 years our top team in our model made the Super Bowl, with the top teams in 19 and 20 also crowned champions. We will definitely be laying heavy on our models midway through the season and towards the playoffs. Follow us on Twitter for updates and let us know your thoughts.
Week 6 Update
We are about a third of the way through the season and while we have a lot of football left to be played, I wanted to drop the first snapshot of what our model is looking like so far this season. See the leaders below:
Now I know what you’re thinking… what kind of ass backwards power rankings is this. Remember we don’t do power rankings, they are mostly subjective click-bait lists made to be polarizing and divisive. This list is a proven computer analytics model that doesn’t take into account wins and losses and doesn’t give a fuck about your feelings or mine. Per the computer… after 6 weeks, the Cowboys are the strongest team in the NFL. More to come soon.
Week 10 Update
A little more than halfway through the season and not too much has changed, the Cowboys remain on top, but the Patriots are ascending fast, while other teams have taken a tumble. The first place Titans have yet to make an appearance. That’s the great thing about the model is that it has no idea the records of anyone, so there is no bias. Stay tuned for more updates as the season progresses. Let us know on Twitter if you agree or not with the model.
Week 16 Update
As you can see our top teams have not changed, the computer models are still very high on Chiefs, Bucs, Cowboys, and Bills. The more we run these models and cross reference them with the odds that Vegas puts out, the more I am convinced that we are on to something huge. As I continue to mention, these algorithms have ZERO knowledge of wins and losses, playoff seeding etc. All it know is that there are a bunch of weighted statistical categories, that have proven to be predictors of Super Bowl winning teams with great accuracy. It weighs and calculates all of those categories for current teams and spits out a score that represents how high the computers confidence is in each team. With two weeks until the computer model is finalized, all signs are pointing at a Chiefs, Bucs rematch, but anything can happen. Stay tuned for our final update after Week 18.
WEEK 18 FINAL UPDATE
This is it, the moment we’ve all been waiting for. The numbers are in and the algorithm has spoken. The Bucs have slid into first place, which as great because at the time of this posting they were +710 odds to win the Super Bow. They are followed by the Cowboys at +1075, and the Chiefs at +410. Remember, we have successfully predicted the winner of the Super Bowl in back to back years. Now a greedy person might say to put the mortgage on the Buccaneers, but I did not come this far by ignoring math and tempting fate. I have extreme confidence in our top 4, and last year our top 4 was the final 4 teams remaining. You will notice the Chargers are on this list, because as I have pointed out several times, the computer has no information about wins and losses, no information about division standings, or playoff position. That’s the beauty, it spits out the odds of whether each team could win the Super Bowl at the present moment.
Use this information however you want. We just put everything on our top 3 teams, and will monitor the situation as the playoffs get underway. If you have always wanted to bet on the Super Bowl, now is the year. Not only do you have access to the results of a proven algorithm that slides the odds in your favor, but if you use our code “REGULATORS” at MyBookie.ag you can DOUBLE your first deposit (up to $1,000). What more can you ask for. We are putting our money EXACTLY where our mouth is. We bought Bucs, Cowboys, and Chiefs… and we bought heavy! Let us know if you are riding with us, or where you think we are wrong. Good luck, and Godspeed!