These data-filled websites will help you dominate your March Madness pool

College basketball is like a comet. It burns at the center of the national sports world for exactly three weeks, and then largely disappears until the next year. During this brief window American sports fans become obsessed with figuring out who is going to win March Madness games, often involving teams they’ve never watched play and know nothing about.

The old adage is that the more college basketball you watch, the worse your NCAA Tournament bracket will be. But in the information age, you can gain an edge. If you know where to look and how to parse the information, you can find all the data you need to make educated calls on your tournament bracket.

Here are seven websites that could help you dominate your bracket pool:

Bart Torvik

Bart Torvik’s T-Rank is such a staple within the college basketball community that the predictive team ranking metric is on the list of criteria that the NCAA Selection Committee uses to determine the field.

But Bart Torvik goes much deeper than just one number. It’s a hub for team and player statistics, with a deep repository of free data. You can look at advanced box scores from every Division I game; split up rankings by date range, opponent quality, venue, and more; and look at a vast array of different player stats. While the database isn’t as vast, there’s also a deep array of women’s basketball statistics on the website, including full T-Rank pages for each team.

Among my favorite features is the one that lets you look at which teams have the most similar profiles, historically, to this year’s teams. For instance, based on the base weights of four factors (shooting, turnovers, rebounding, and free throws), results, adjusted efficiency metrics, and play style metrics, the most similar team to Arizona this year (at least since 2008) is the 2007-08 Memphis team that lost to Kansas in the National Championship game.

Bart Torvik is what you make of it: a rich database of free information. 

KenPom

Just like Bart Torvik, KenPom is a flagship predictive metric that is part of the NCAA’s selection criteria. This one has become such a part of the college basketball conversation that you’ll often hear it mentioned on national television broadcasts, and it’s always the first place I go for information about a team or player. 

KenPom has many of the same basic features as Torvik, but with a more accessible user interface, a more robust history, and super clean visuals that make it easy to understand everything you’re seeing.

Without paying, you can access KenPom’s rankings, offensive and defensive efficiency, tempo, and strength-of-schedule stats dating all the way back to 2001-02. With a paid subscription ($24.95 per year) you can access detailed team and player pages, schedules, national leaderboards, game previews and box scores, and more.

One thing I love doing on KenPom this time of year is going to the “opponent tracker” tool and seeing which types of teams certain teams have struggled with throughout the year. For instance, Louisville went just 3-8 against teams with a top 50 defensive efficiency (20-2 against all other teams). Its first-round opponent, South Florida, has the 40th-ranked defense. You have to be smart to determine which of these statistics matter, how much they matter, and why, but it puts things into perspective in a digestible way.

Hoop Explorer

If you’re a little more of a hoops sicko, Hoop Explorer is a terrific resource for finding plenty of data that you can’t find on Torvik or KenPom. I’ve spent hours on Hoop Explorer throughout the season looking at the Team/Roster Analysis Tool, parsing through on-off statistics of four factors, rim rates, and much more for hundreds of different teams and players.

Looking to find out how a team has survived without a player who may be injured? Hoop Explorer’s on-off data can show you not only exactly what the difference is for a team when a player is on or off the floor, but also what changes in terms of how they generate offense and how opposing offenses generate offense.

Do you want to know which teams run the most high-low actions or generate shots through a drive-and-kick offense, and who is the most efficient at it? Hoop Explorer has that, and you can even analyze each team’s tendencies by lineup as well. Like Bart Torvik, it also contains data on women’s college basketball.

Haslametrics

Haslametrics is based on a flagship predictive metric that ranks each team in the country. It’s similar in premise to T-Rank, but it has a few additional interesting metrics. Each game is given a predicted outcome based on predictive metrics, and then each team is given a score for each game’s performance based on that prediction, both offense and defense.

Those scores help Haslametrics generate rankings for momentum and consistency. Momentum tells you which teams are playing the best relative to expectation over the last few weeks, while consistency tells you which teams tend to play the closest to their projections and previous projections.

EvanMiya

Evan Miyakawa’s Bayesian model gives a different perspective to player and team modeling than other metrics. Miyakawa rates each player in the country offensively and defensively, and also puts together a predictive team metric.

His statistics are commonly used and referenced by coaching staffs. You can access a lot of the data on EvanMiya for free, such as player and team ratings, and data on which teams go on the most 10-0 scoring runs, known as “kill shots.” Miyakawa found in studies that teams that go on more 10-0 scoring runs than their opponent in a given game win that game 82% of the time.

The website also has free data on which teams play the best against better teams, and which play worse against better teams.

With a basic $5-per-month or $30-per-year subscription, you can see lineup ratings, on-off splits, and game predictions. You can even export the data to CSV if you want to go through it yourself. With a premium subscription, which he’s offering for $22 for one month on a March Madness deal, you can look at player skill grades, keys to victory for every team, and more.

CBB Analytics

This one is for total sickos. CBB Analytics is mostly designed for coaches, but it has public-facing subscription packages that fans can use. It essentially has everything that all the other sites have, but put together in one place.

You can dive into more splits on CBB Analytics than on any other site, and view more detailed team, player, and conference pages. Its shot charts, both team and player, are wildly customizable and can give you insight into which types of shots teams and players take. You can customize a leaderboard of pretty much whatever you want for player stats or team stats and then make a graph of it.

I’ve been looking at shot volume and box score differentials a ton recently, and CBB Analytics features a dashboard that helps me figure out which teams generate the most opportunities to score compared to their opponents. Depending on your subscription level, you can view extremely customizable on-off splits, player pages, and more.

Everything that you can find on CBBAnalytics.com for Division I men’s basketball you can also find for Division I, II, and III men’s and women’s basketball.

Her Hoop Stats

Her Hoop Stats is a dedicated site for women’s basketball analytics and articles. Like KenPom and Torvik, it has a flagship predictive metric, known as Her Hoop Stats Rating. Most of the site’s analytics features are locked behind a paywall, including ranks and percentiles, résumé comparisons, and a vast suite of sortable and customizable data relating to both women’s college basketball and the WNBA.

A basic subscription costs $4.99 per month or $49.99 per year, but there are more expensive tiers that feature automated game predictions and expert betting picks.

---