Kevin Cole, a data scientist at Pro Football Focus, joins me to talk football analytics. He describes his latest work in using Bayesian methods and player grades to evaluate NFL QBs. This allows him to assign a probability that Mitch Trubisky is better than Pat Mahomes and Deshaun Watson, no matter how small. We also discuss why QBs should not always minimize their interception rate. Finally, Kevin gives his thoughts on San Francisco, Green Bay, Cleveland and Baltimore.
Edward Egros, a professor SMU that works at the forefront of journalism and sports analytics, joins me to talk football. He explains his college football model and the impact of recruiting. Then we discuss SMU, Texas, Alabama and the Dallas Cowboys. Finally, he tells us about his sports analytics class, how to get NFL play by play data, and Florence Nightingale.
Whale Capper is an earthquake engineer who uses this technical expertise to predict the NFL. He tells us how he introduces variance into his model, and how this informs his betting. We talk about science vs art in his model, and how this pertains to the Miami Dolphins. He also tells us the team that he didn't like this preseason, and why he has changed his mind. We end with a great food discussion.
After six weeks of the season, we are starting to see early returns on college football teams. Which teams are living up to the preseason hype? Host Ed Feng looks at 3 teams: 1. the top 10 team that is no longer getting carried by its defense, 2. the team I had overrated this preseason that is looking good, but not in the way I expected, 3. the team in which everyone is talking about the QB, but that might not be its biggest problem.
Seth Walder of ESPN joins the show to discuss their excellent work on football analytics. He tells us how the Football Power Index (FPI) works for both college football and the NFL. In addition, he offers a college football and NFL team that FPI really likes. Then we get into ESPN's work with NFL Next Gen Stats which tracks the motion of all players. Seth tells us about pass rush win rate and the player on Tampa Bay tearing it up this season.
Rufus Peabody, a professional sports bettor, joins me to discuss how he uses analytics to make better bets. He talks about his Massey Peabody model for making college football and NFL predictions. We talk about teams like Wisconsin and Michigan in college football that have moved rapidly from their preseason projections. We also discuss Miami in the NFL, a most difficult team to quantify. Finally, Rufus talks about the data that goes into NFL player model.
Preston Johnson, a professional sports bettor and ESPN personality, joins me to talk college football and the NFL. He tells us about the two sets of numbers he uses in his football handicapping. This informs his positions on Michigan, Syracuse and SMU in college football and Denver and Green Bay in the NFL. Preston also talks about profitable versus unprofitable teasers in the NFL.
Dr. Eric Eager is a data scientist at Pro Football Focus pushing the frontiers of football analytics. He tells us whether pass rush or coverage matters more on pass defense. When the defense rushes the passer, Dr. Eager explains whether the offensive line or quarterback plays a bigger role in reducing that pressure. Then he explains the new player based model he uses to predict both college football and NFL games. Finally, we talk about the 2019 NFL season and the team that might become Jacksonville 3.0.
Aaron Schatz of Football Outsiders joins me for a wide ranging discussion on the NFL. He tells us about the most important frontiers in football analytics, which leads to a discussion of which defensive stats are least stable from year to year. Then we get into an overrated team, or Jacksonville 2.0. Finally, he tells us about an NFC North team that might surprise you.
The preseason is the time for high expectations in college football. However, not all teams live up to the hype. Using analytics and context, host Ed Feng identifies 3 overrated college football teams for 2019.
Host Ed Feng goes through 5 games based on predictions he usually saves for paying members of The Power Rank, his site for more accurate football predictions. These predictions allow him to dig deeper into the Clemson Tigers, the Wisconsin Badgers and the Kansas City Chiefs.
Colts QB Andrew Luck shocked the football world by retiring this weekend. To determine the effect of his absence on Indianapolis, host Ed Feng uses his NFL preseason rankings based on market win totals.
While Oklahoma and Texas dominate the headlines, Iowa State might be the most intriguing team in the Big 12. Host Ed Feng breaks down QB Brock Purdy and the pass defense by using a little known fact about the official statistics. He ends by ranking the top 3 Big 12 teams by The Power Rank preseason model.
Seattle had a solid 10-6 season in 2018 and made the playoffs. However, the Seahawks have a serious problem. Host Ed Feng looks at their offense based on research on how running the ball sets up play action.
It's hard enough to win a division title in college football. It gets even harder when your team has the most difficult schedule. Based on division win probabilities, host Ed Feng identifies 3 college football teams getting screwed by the injustice of schedule.
Despite going down to Ohio State as the favorite last season, Michigan lost a heart breaking 62-39 game. Host Ed Feng explores the hidden side of the game based on modern football analytics. Then he projects whether Michigan can finally catch Ohio State in 2019.
NFL teams run the ball more when they're up late in games. So it makes sense that 100 yard rushers on winning teams get lots of their yards late in games. Host Ed Feng dug into the data to test this hypothesis.
Why preview a Nebraska team that went 4-8 last year? First, they were better than their record in 2018, and looking at their season shows us the perils of small sample size. Second, Nebraska has interesting upside for 2019.
Predictive analytics is a never ending game of distinguishing skill from luck. Ed Feng determines which statistics reflect the most skill for NFL quarterbacks by looking at how each statistic has evolved over time. This ends with a discussion of Derek Carr and Jared Goff.
Over the next two weeks, Ed Feng brings you 10 episodes of spoken word episodes to get you ready for the 2019 football season.
Josh Hermsmeyer of FiveThirtyEight joins me for a wide ranging conversation on the NFL. He tells us why he considers himself an analyst more than a data scientist. Then we get into play action and whether running it more times reduces its effectiveness. Finally, we discuss the 2019 prospects for Oakland, Chicago and Arizona.
Bill Connelly of ESPN joins me for a wide ranging conversation on college football. He tells us which parts of his preseason model gets broken down into offense and defense, and which doesn't. We discuss whether any team can catch Alabama and Clemson in 2019. Finally, Bill breaks down the contenders in the Big Ten, Big 12 and Pac-12. Our apologies, as the audio quality is not has high as usual due to technical difficulties.
Jim Sannes, senior writer and analyst at numberFire and FanDuel, joins me to talk the upcoming NFL season. We discuss how he ranks NFL offensive lines from 1 to 32. Then he tells us about an offensive line that isn't ranked well but has a high upside and could power a Super Bowl contender. Jim ends with a quarterback that could have a remarkable 2019 NFL season.
Gadoon Kyrollos, who goes by Spanky on Twitter, joins the show. He's a professional sports bettor who was recently profiled on The Ringer. We discuss his approach to sports betting and how sports books treat sharp bettors like himself. He also gives us a hilarious criteria for whether you're doing well as a sports bettor.
Host Ed Feng discusses his new sports betting podcast called Covering the Spread.