Using confidence scores and AI to predict the Baltimore Ravens wins and losses in 2024 NFL season
Weighted confidence scores, AI model picks, and Vegas over/under in the 2024-25 NFL season.
The Baltimore Ravens enter the 2024-25 NFL season with most of the same key players after a disappointing playoff exit to Kansas City last year. Biggest questions are if the offensive changes in the second year with Todd Monken, how Mark Andrews will look returning from injury, Zay Flowers development in his second season, and how many yards Derrick Henry still has in the tank. I think Henry is going to perform very well when he ends up one-on-one in the open field because Lamar demands so much attention from the LBs during any play action.
The Ravens return an excellent defense with Marlon Humphrey, Kyle Hamilton, and Roquan Smith. The big changes this year are the switch to Zach Orr replacing Mike Macdonald at Defensive Coordinator and the loss of LB Patrick Queen. Additionally, Justin Madubuike, David Ojabo, and Odafe Oweh need to be more consistent in generating pressure which has been an uncharacteristic weakness in recent seasons. Getting back to the Ravens' tradition of stout defensives will be essential for the team's overall success and a playoff run that doesn’t end in Arrowhead.
My Confidence Scores: 13-4
ChatGPT4o: 11-6
Vegas Over/Under: 11.5 (Over: +132, Under: -162)
The big question is if the Ravens are stacked to finally make it past the Chiefs for a Super Bowl run, or if last season’s 13-4 record was the peak from this current core group of Ravens. Luckily these questions will be answered early with a road primetime match-up in Kansas City to open the NFL season.
Confidence Scores
Instead of a simple W or L, I decided to use a 5-point scale to project each game:
1 (sure win)
2 (likely win)
3 (toss up)
4 (likely loss)
5 (sure loss)
Each point tier gets multiplied by the likelihood of a win to predict the total number of wins (e.g., scores of 1 get multiplied by 1.0, number of games with scores of 2 get multiplied by .75 win and .25 loss. toss up games are .5 win and .5 loss).
My confidence scores project a 13-4 record which is identical to how the Ravens finished last year.
AI Model
ChatGPT4o: Predicting the win-loss record for the Baltimore Ravens involves analyzing several factors such as team performance, player stats, injuries, and historical data against their opponents.
@Kansas City (Thu, Sep 5, 8:20 PM) - Loss: Kansas City is a strong team, especially at home.
vs Las Vegas (Sun, Sep 15, 1:00 PM) - Win: Baltimore should be able to handle Las Vegas at home.
@Dallas (Sun, Sep 22, 4:25 PM) - Win: Dallas is a competitive team, but Baltimore's defense could be the key.
vs Buffalo (Sun, Sep 29, 8:20 PM) - Loss: Buffalo is a top-tier team with a strong defense.
@Cincinnati (Sun, Oct 6, 1:00 PM) - Win: Baltimore tends to perform well against Cincinnati.
vs Washington (Sun, Oct 13, 1:00 PM) - Win: Washington is an inconsistent team.
@Tampa Bay (Mon, Oct 21, 8:15 PM) - Win: Tampa Bay has challenges without Tom Brady.
@Cleveland (Sun, Oct 27, 1:00 PM) - Loss: Divisional games are tough, and Cleveland has a strong roster.
vs Denver (Sun, Nov 3, 1:00 PM) - Win: Denver struggles on the road.
vs Cincinnati (Thu, Nov 7, 8:15 PM) - Win: Home advantage should help.
@Pittsburgh (Sun, Nov 17, 1:00 PM) - Loss: Pittsburgh is a tough opponent, especially at home.
@Los Angeles (Mon, Nov 25, 8:15 PM) - Loss: Los Angeles is strong at home.
vs Philadelphia (Sun, Dec 1, 4:25 PM) - Loss: Philadelphia is a solid team with balanced offense and defense.
BYE WEEK - Rest and recovery.
@New York (Sun, Dec 15, 1:00 PM) - Win: New York is in a rebuilding phase.
vs Pittsburgh (Sat, Dec 21, 4:30 PM) - Win: Home advantage in a key divisional matchup.
@Houston (Wed, Dec 25, 4:30 PM) - Win: Houston is a weaker team.
vs Cleveland (Fri, Jan 5, TBD) - Loss: Cleveland is a tough divisional opponent.
Predicted Record: 11-6
I asked ChatGPT to explain the criteria for predicting the games:
Here is an explanation of the criteria used to predict wins and losses for the Baltimore Ravens' schedule:
Opponent Strength:
Kansas City: Predicted loss due to Kansas City being a top-tier team, especially strong at home.
Buffalo: Predicted loss because Buffalo has a strong defense and balanced team.
Philadelphia: Predicted loss due to their solid and balanced team.
Home vs. Away:
Home Games: Generally, teams have a better chance of winning home games due to familiar surroundings and fan support.
Wins predicted against Las Vegas, Washington, Denver, Cincinnati, Pittsburgh (Week 16).
Away Games: Away games are typically harder to win due to travel and lack of home support.
Losses predicted at Kansas City, Cleveland, Pittsburgh, Los Angeles.
Divisional Matchups:
Divisional games are often unpredictable but usually highly competitive.
Wins predicted against Cincinnati (home) and Pittsburgh (home).
Losses predicted against Pittsburgh (away) and Cleveland (away).
Team Performance and Trends:
Consideration of the overall performance trends of the opposing teams and their recent seasons.
Special Situations:
Christmas Game: Special attention to performance on significant dates, which might affect player morale and focus.
This analysis incorporates a combination of home-field advantage, the strength of the opponent, historical performance, and divisional rivalry dynamics. The predictions aim to balance optimism and realism based on available data.
Vegas Odds
FanDuel likes the Ravens this year to win their division, beat the Chiefs in the playoffs, and get to the Super Bowl:
Ravens Over/Under 11.5 Wins
Over: +132
Under: -162
Odds to Make the Playoffs: -265
Odds to Win the AFC North: +140 (1st)
Odds to Win the AFC Championship: +490 (2nd)
Super Bowl Odds: +950 (3rd)
BetMGM has a less bullish outlook:
The Ravens’ season win total is set at 11.5 games.
Oddsmakers expect the under is the more likely outcome. Currently, Baltimore is -160 to win 11 or fewer games, an implied probability of 61.5%. Meanwhile, the odds for them to finish with 11 or more wins is +130, an implied probability of 43.4%.
Fox Sports is identical to FanDuel:
Over/under 11.5
Over +132
Under -162
David Riedman is the creator of the K-12 School Shooting Database, Chief Data Officer at a global risk management firm, and a tenure-track professor. Listen to my recent interviews on Freakonomics Radio, New England Journal of Medicine, and my article on CNN about AI and school security.