Today we released a fairly significant change to our Win Predictor, ATS Predictor, and Over Under Predictor tools.
In their first incarnation, these tools were built to enable users to do three things:
1) Choose a specific set of games to predict, since some people only care about certain conferences or teams and because pick’em contests vary regarding what games are included each week
2) Customize betting lines for each game, again useful for spread-based pick’em games that freeze game point spreads early in the week, or for matching betting odds to those available to you at your sports book of choice
3) Create a data-driven predictive strategy based on stats of your choice to apply to all games, and then view the resulting picks
In short, we tried to do a little too much in one product, and today we moved to simplify the user experience.
First, we have completely removed item #3 from our Win, ATS, and Over Under Predictor tools. Now, you choose games to predict, pick a pre-existing predictive model to use, and click to generate picks. Much faster, much simpler, much easier.
Best of all, you no longer need to create your own customized prediction model to get picks. You can just use one of the predictive models already created by the Team Rankings nerds. But unlike getting predictions from our Picks page, in our Predictor tools you can alter the applicable betting lines of any game.
(Our picks pages just show predictions based on the most recent betting odds we get from our primary odds supplier.)
Second, we have built a separate tool, Model Builder, for more advanced users who are interested in creating and backtesting their own data-driven predictive strategies. This tool allows you to build a predictive model based on stats of your choosing, then backtest its performance over recent games, and save it to your Team Rankings user account.
Now, as soon as you save a custom model using Model Builder, bam — your newly saved model appears as one of your prediction model options in the Win, ATS, and Over Under Predictor tools.
We’ve still got a little cleanup work to do on the user interfaces, but the new versions of the predictor tools are released and functional. Take one for a spin:
NFL Win Predictor
NFL ATS Predictor
NFL Over Under Predictor
NFL Model Builder (for advanced users)
Side note to current tool users: If you used the prior versions of Win, ATS, or Over Under Predictor to save custom predictive strategies, the updated Predictor tools should automatically load your custom models into the new interface.
We’re extremely excited to announce the release of the first version of the Team Rankings Model Builder, a technology and application that we have been working on for quite some time.
Over the past weeks, months, even years, we’ve heard from people who use our sports prediction tools to create unbiased, data-driven algorithms for predicting games. Many of our users have enjoyed strong success with sports betting and in pick’em contests using their own customized Team Rankings prediction models.
However, the most common — and understandable — request we get is this: “If you guys could devise a way for me to not only create my own objective algorithms for predicting games, but also be able to quickly test how my strategies have done in the past, that would be huge.”
We’re happy to say that today, we’ve brought that idea to life. Instead of creating a predictive strategy and having to wait weeks to measure how it performs, now you can see how your theories have played out over the last 100 games — in seconds.
Not only that, but Model Builder shows you how your custom prediction strategy has done picking game winners, ATS (against the spread) winners, over under line winners, and money line winners. And it displays both percentage correct and betting unit based performance results.
We already have a number of enhancements on the list for Model Builder, but the core is in place. Check it out and let us know what you think. It’s powerful functionality, but we also hope we’ve made it easy and fun.
So stop following the herd and use Team Rankings to craft your own proven winning strategy!
As one of the final steps in rolling out our new user-driven NFL prediction tools and college football prediction tools, we decided that it would simplify the experience for users if we released several distinct products.
There are many different use cases for football predictions. Some people are in pick’em contests where they need to predict straight-up game winners, while others are in point spread based pools. Some are in pools in which they assign confidence points to every pick, while others aren’t.
On the betting side, some people only bet point spreads, others only totals, and some a little of both.
As we thought about how to design our latest tools (which allow you to build customized, data driven prediction systems and use those systems to predict the upcoming week’s games), we were having trouble, as is often the case, figuring out how to fit lots of information on one screen — and still have the tool be simple enough for a first time user to understand.
(That’s the ultimate irony of our site, by the way. With so much nerdy math going on behind the scenes, our biggest challenge is often figuring out how to create an interface that lets users take advantage of all that math power without overwhelming them with complexity and numbers.)
In the end, we remembered the old adage: if you try to serve too many masters, you’ll please none.
The result was that we released not one but three weekly prediction tools: Win Predictor, ATS (Against The point Spread) Predictor, and Over/Under Predictor. If you need to make picks, we hope it’s clear which tool you should use, based on the type of pick’em pool you’re in or type of bets you like to make.
Here are links to all the tools. Give them a whirl and let us know what you think:
NFL Win Predictor
NFL ATS Predictor
NFL Over/Under Predictor
College Football Win Predictor
College Football ATS Predictor
College Football Over/Under Predictor
You are currently browsing the archives for the College Football category.


