Sports Betting & Pick'em Analysis Tools

View NFL packages and pricing
View college football packages and pricing

In the upcoming weeks we plan to roll out several new football related tools, features, and data. Whether you are a football bettor, handicapper, pick'em player, or all of the above, our goal is to help you make smarter and more efficient decisions. Here's how we plan to do it.

Predictive Model Builder

Creating an effective, long-term football picking strategy begins with a repeatable, unbiased, data-driven forecasting system. Although every football game has its "intangibles" that require more subjective handicapping, sophisticated math models provide the baseline expectations that serve as the foundation for subjective adjustment.

Our new predictive model builder will give you access to all the data and technology you need to build Ph.D.-worthy, data-driven predictive systems and use those models to forecast future games. It will take your personal "secret sauce" for predicting games (e.g. emphasizing pass defense and turnover ratio) and within seconds, generate predictions with an underlying level of research that would take months to perform manually.

Model Backtesting Features

Building data-driven sports prediction models is always fun, but the real question is, how do they perform? Some predictive strategies may have solid long-term results, while others could be red hot over the month; some models may be great for picking early season games, while other approaches work better in the playoffs.

To help answer these questions, we will launch our unique model backtesting features. As you create custom predictive models, with the click of a button you will be able to see how your systems did picking the last five years of historical games, via performance reports that illustrate wins, losses, and performance trends in a number of insightful ways. No more committing to a system without a full, trustworthy, five year performance history.

Team Rankings Math Models

In addition to enabling users to create their own predictive models, over the years we have developed a number of data-driven algorithms for predicting games that do not incorporate user input and interaction. Sometimes referred to as "black box" models, these proprietary algorithms involve us shoveling a ton of data into a computer, which then applies complex calculation methods and, usually within an hour or three, spits out prediction results. Our basketball simulation model, where we use possession-based efficiency statistics to "play out" games in a hypothetical fashion, is a good example of a black box model.

We will have several black box models available for football, along with their five year performance histories, and they can serve as a powerful cross-reference for predictions from other sources (or your own gut instinct).

Pick'em Analysis Tools

Everybody loves pick'ems; for some reason there's no better feeling than out-predicting your friends or office mates and taking their money at the end of the season to boot. However, many people don't have the time to commit to researching every single NFL or college Top 25 game every week. We plan to launch some new tools that enable users to quickly and easily apply a data-driven prediction strategy to their weekly pick'em games. We're looking into survivor pool analysis as well.

Smart Situational Screeners

Situational analysis, an approach used by many professional handicappers, can help unearth "angles" on games that change your expectations for one or both teams. For example, a math model might say that based on each team's statistical profile, the Patriots, coming off a bye week, have a 65% chance to beat the Dolphins away. Then you find out that away teams coming off a bye week are 2-300 in the last 10 years. (That's a joke, but you get the point...)

Situational and trend analyses can be valuable, but the most common risk is mistaking random trends for meaningful information. Our new situational screener will not only make it easy for users to identify relevant situations and trends, but also weigh in on how meaningful the results are likely to be.

Matchup Stats & Analysis

We believe that good matchup analysis finds the right balance between comprehensiveness, sophistication, and ease of use. Most matchup analysis on the web (and especially on sports radio/TV) is either so simple and vague that it's just plain silly ("Indy has Peyton at QB but Pittsburgh has the better defense, so it's going to be close"), or it's such an orgy of poorly organized stats and data that it's impossible to make sense of it.

Our new game matchup zones will give users access to all the information and stats they need to research a game like a pro, in a way that highlights key information and makes the entire research process more efficient.

Pick Tracking, Sharing, and Discussion

Team Rankings users are smart folks, and we're going to make it easier for our users to share their perspectives and learn from one another. Especially when it comes to predicting games, situations that are difficult if not impossible to model with hard data alone will always occur. What's going to happen the game after a star receiver gets injured -- total offensive meltdown, or will a rookie backup finally get the chance he needs to shine? How can we translate a "revenge factor" into expected additional points or fewer points allowed?

In addition to providing features to record and track your own picking prowess, we're going to make it easy and fun to interact with fellow users and explore these tough questions.

Mobile and Social Applications

Web-based features and applications that do everything mentioned above are just the start. We love Facebook and our mobile phones as much as many of our users do, and we plan on developing and releasing more apps for social and mobile platforms. We've just started with a few prototypes, Five Bets: Pro Baseball for Facebook and Odds: Pro Baseball for the iPhone and iPod Touch.