Articles & Features
How EA Sports Calculates Ratings for Madden and NCAA Football Games
Photo: Oct 28, 2023; Pasadena, California, USA; Colorado Buffaloes quarterback Shedeur Sanders (2) carries the ball against UCLA Bruins defensive lineman Gabriel Murphy (11) in the first half at Rose Bowl. Mandatory Credit: Kirby Lee-USA TODAY Sports

How EA Sports Calculates Ratings for Madden and NCAA Football Games

For sports video game enthusiasts, few series are as beloved as EA Sports' NCAA Football. These games have captivated fans for years, offering a nice blend of realism and fun. Central to this experience are the player ratings, which can turn the tide in a close game or validate the dominance of a powerhouse team.

The Birth and Evolution of Player Ratings

The concept of player ratings in sports video games dates back to the early 1980s. One of the pioneers in this field was "Micro League Baseball," released in 1984 for the Commodore 64. This game was among the first to use real player names and statistics, necessitating a system to differentiate player abilities. The solution? Numerical ratings that attempted to quantify a player's skills.

As technology advanced, so did the complexity of these rating systems. Early football games like "Tecmo Super Bowl" (1991) used a limited number of rating categories - often no more than seven per player. By the late 1990s, EA Sports' NCAA Football series had expanded to 14 attributes per player, while Madden NFL began using 17.

Today, the rating system used in modern Madden NFL games is exponentially more complex. Each player is evaluated across 43 different categories, with nearly 20 additional "trait" tags that control specific player behaviors. This evolution reflects not just technological advancements, but also a growing demand from players for more realistic and nuanced gameplay.

The Fundamental Categories

In current EA Sports football games, player ratings fall into several broad categories:

  • Physical Attributes: These include speed, strength, agility, acceleration, and jumping ability. These ratings are often derived from measurable physical performances, such as 40-yard dash times or bench press repetitions.
  • Position-Specific Skills: These vary by position but might include throwing accuracy for quarterbacks, route running for receivers, or tackling ability for defensive players. These ratings are typically based on a combination of statistical performance and subjective evaluation of a player's technique.
  • Mental Attributes: Ratings like awareness, play recognition, and football IQ fall into this category. These are some of the most challenging ratings to quantify, often relying heavily on expert opinion and analysis of a player's decision-making in game situations.
  • Special Abilities: These include ratings for specific skills like spectacular catches, break tackle ability, or pass rush moves. These ratings often correspond to memorable plays or recognized strengths of particular players.
  • Intangibles: Ratings for clutch performance, injury resistance, and stamina fall into this category. These attempt to capture aspects of a player's performance that aren't easily quantifiable but can significantly impact their effectiveness on the field.

The Importance of Granularity

The high number of rating categories serves a crucial purpose: player differentiation. With 43 different ratings, the game can create nuanced differences between players that might be similar in overall skill level. For example, two quarterbacks might have the same overall rating, but one could excel in short passing accuracy while the other has superior throwing power. This level of detail allows for more realistic gameplay and strategic depth, as users can tailor their play style to the specific strengths of their players.

Moreover, this granularity allows for more accurate representation of player development over time. As players improve or decline in specific areas, individual ratings can be adjusted to reflect these changes, providing a more dynamic and realistic simulation of player careers.

The Foundation: Official Statistics

At the core of the player rating process is a vast amount of statistical data. For NFL players, this includes official game statistics covering every aspect of performance: passing yards, completion percentage, rushing yards, tackles, interceptions, and much more. For college players entering the NFL draft, their collegiate statistics form the basis of their initial ratings.

However, raw statistics are just the starting point. The rating process goes far beyond simply translating stats into ratings. A player who leads the league in passing yards won't automatically have the highest quarterback rating. Context is key, and this is where more advanced data gathering techniques come into play.

Advanced Analytics

In recent years, the use of advanced analytics has become increasingly important in the rating process. Sources like Football Outsiders and Pro Football Focus provide in-depth analysis that goes beyond traditional stats. These might include metrics like:

  • DVOA (Defense-adjusted Value Over Average), which compares a player's performance to a baseline adjusted for situation and opponent
  • PFF grades, which are based on play-by-play analysis by football experts
  • Next Gen Stats, which use player tracking data to provide insights on things like separation for receivers or time to throw for quarterbacks

These advanced metrics help provide a more nuanced view of a player's performance, informing ratings in ways that raw stats cannot.

Physical Measurables

For rookies entering the league, NFL Combine and Pro Day results are crucial in establishing baseline ratings for physical attributes. These events provide standardized measurements for attributes like speed (40-yard dash), strength (bench press), agility (shuttle run), and jumping ability (vertical and broad jump).

The translation of these measurables into ratings isn't always one-to-one. For example, a 4.3-second 40-yard dash time might translate to a speed rating in the high 90s, but the exact rating might be adjusted based on how that speed translates to on-field performance.

Game Film Analysis

Perhaps the most time-consuming aspect of data gathering is the analysis of game film. This involves watching hours of footage to assess players' performance in real game situations. This analysis is crucial for rating aspects of performance that don't show up in the stat sheet, such as:

  • A lineman's blocking technique
  • A quarterback's decision-making under pressure
  • A defensive back's ability to read and react to plays

Film analysis also helps contextualize statistical performance. A quarterback's completion percentage might be low, but film might reveal that this is due to receiver drops rather than inaccurate passes.

Scouting Reports

Scouting reports, both public and private, provide another valuable source of information. These reports often contain detailed breakdowns of a player's strengths and weaknesses, as well as projections for their potential development. This information is particularly valuable for rating rookies and young players who haven't yet established a long track record in the NFL.

Continuous Data Collection

The data gathering process isn't a one-time event. Throughout the NFL season, performance data is continually collected and analyzed. This allows for regular updates to player ratings, reflecting changes in performance over time. These in-season updates help keep the game current and reflective of real-world NFL performances.

From Data to Ratings: The Conversion Process

Once the vast amount of data is collected, the next step is to convert this information into actual numerical ratings. This process involves a complex interplay of quantitative analysis and subjective judgment.

Quantitative Analysis

For many attributes, particularly physical ones, there's a relatively straightforward correlation between measurable data and ratings. For example:

  • 40-yard dash times can be directly converted into speed ratings
  • Bench press repetitions inform strength ratings
  • Completion percentage plays a significant role in accuracy ratings for quarterbacks

However, even these seemingly straightforward conversions often involve nuanced adjustments. A player's game speed, for instance, might not perfectly match their combine speed. The rating system needs to account for how physical attributes translate to on-field performance.

Comparative Analysis

Ratings are not assigned in a vacuum. A key part of the process is comparing players to their peers. This comparative analysis helps ensure that ratings are properly scaled across the entire player population. For example, a quarterback's arm strength rating isn't just based on how far they can throw, but how their throwing power compares to other quarterbacks in the league.

Subjective Evaluation

For many attributes, particularly those related to technique, awareness, and other less quantifiable aspects of performance, subjective evaluation plays a crucial role. This is where the expertise of the ratings team becomes paramount. They must interpret game film, scouting reports, and expert opinions to assign ratings for attributes like:

  • Route running for receivers
  • Block shedding for defensive linemen
  • Play recognition for linebackers

This subjective evaluation is often the most challenging and controversial aspect of the rating process.

Weighted Attributes and Overall Ratings

While individual attribute ratings are important, many users focus primarily on the overall rating for each player. This overall rating is calculated as a weighted average of the individual attributes, with the weights varying by position.

For example, a quarterback's overall rating might be calculated with the following weights:

  • Awareness: 20%
  • Throwing Power: 20%
  • Short Accuracy: 20%
  • Medium Accuracy: 15%
  • Deep Accuracy: 10%
  • Play Action: 10%
  • Throw on Run: 5%

These weights are carefully calibrated to ensure that the overall ratings accurately reflect each player's total value on the field.

The "Trait" System

In addition to numerical ratings, modern Madden games also use a system of "traits" - binary attributes that represent specific skills or tendencies. These might include things like "fights for extra yards" for running backs or "throws tight spiral" for quarterbacks. These traits add another layer of differentiation between players and can significantly impact how a player performs in certain situations.

Rookie Ratings

Rating rookies presents a unique challenge, as they have no NFL track record. For rookies, the process typically starts with a template based on their draft position and college performance. This template is then adjusted based on combine results, scouting reports, and expert projections. Rookie ratings are often more conservative, with room for rapid improvement as players prove themselves in the NFL.

The Balancing Act

One of the most challenging aspects of the player rating system is maintaining a balance between realism and gameplay enjoyment. While the goal is to create an accurate simulation of NFL football, the game must also be fun to play. This often requires careful tuning of how ratings translate into on-field performance.

For example, in real life, the difference between a good player and a great player might be subtle. In a video game, however, users expect to feel a significant difference when using a highly-rated player. This might require exaggerating the impact of high ratings to some degree.

Similarly, if lower-rated players performed as poorly in the game as their ratings might suggest, it could lead to frustrating gameplay. Therefore, the impact of low ratings often needs to be softened somewhat to maintain playability.

The "Dice Roll" System

To introduce an element of unpredictability and prevent the game from feeling too deterministic, Madden employs a "dice roll" system for player interactions. When two players interact (e.g., a receiver trying to catch a pass while covered by a defender), the game:

  1. Considers the relevant ratings of both players
  2. Applies a random factor (the "dice roll")
  3. Determines the outcome based on the combination of ratings and randomness

This system allows for occasional upsets and unexpected outcomes, mirroring the unpredictability of real football. It also prevents users from being able to guarantee success simply by using the highest-rated players.

Regular Updates

Player ratings are not static. Throughout the NFL season, ratings are updated regularly to reflect changes in player performance. These updates typically occur weekly, with hundreds of players potentially seeing rating adjustments based on their performance in the previous game.

These regular updates serve several purposes:

  • They keep the game current and reflective of real NFL performance
  • They provide an added layer of engagement for users, who can see their favorite players improve (or decline) over time
  • They allow the ratings team to correct any perceived inaccuracies in the initial ratings

Community Feedback and Controversies

The player rating system often generates significant discussion and sometimes controversy within the Madden community. Players, fans, and sometimes even NFL athletes themselves may disagree with certain ratings.

The ratings team pays attention to this feedback and may make adjustments in response to community consensus about certain ratings. However, they must also be careful not to overreact to short-term fluctuations in performance or public opinion.

Technological and Methodological Challenges

As the rating system has grown more complex, it has faced increasing technological and methodological challenges:

  • Data Management: Handling the vast amount of data required for the rating process requires sophisticated data management systems.
  • Consistency: Ensuring consistency in how ratings are assigned across thousands of players is an ongoing challenge.
  • Automation vs. Human Judgment: Balancing the use of automated systems for data processing with the need for human expertise in player evaluation.
  • Emerging Stats: As new forms of statistical analysis emerge in football, the rating system must evolve to incorporate these new metrics.

Future Developments

Looking ahead, we can expect the player rating system to continue evolving. Potential future developments might include:

  • More granular ratings to further differentiate players
  • Increased use of AI and machine learning in the rating process
  • Greater integration of real-time data to allow for even more frequent rating updates
  • Expansion of the trait system to capture even more specific player tendencies

As technology advances and our understanding of football analytics deepens, the process of translating real-world athletic performance into video game ratings will undoubtedly become even more sophisticated, further blurring the line between virtual and reality in sports gaming.