1. Recent team efficiency — how effectively each team has moved the ball and prevented scoring opportunities.
2. Offensive and defensive strengths — passing, rushing, scoring efficiency, and the quality of the opposing unit.
3. Matchup fit — how one team's style and strengths line up with the opponent's areas of weakness.
4. Home-field context — the venue and the usual advantages associated with playing at home.
5. Available outside context — information such as market expectations or weather, when it can add useful context.

The system weighs recent performance more heavily than older performance while still using earlier results when the current sample is small. This provides a balance between recognizing change and avoiding overreaction to one unusual game. In fact, the realism of the is so important that Coach C typically ignores Week 18 games of the previous season, when many teams rest their starters and/or alter their strategy as the playoffs loom ahead.

Raw yards and points are useful, but they do not tell the whole story. A short gain on an important third down can be more valuable than a longer gain when the outcome is already virtually decided.

For that reason, the system also uses situation-aware efficiency measures such as Expected Points Added (EPA). EPA estimates how a play changes the offense's expected scoring position after considering the game situation. It helps the analysis distinguish productive plays from statistics accumulated in less meaningful situations.

The prediction is then translated into a likely winner, an estimated margin or score context where applicable, and a confidence description. These outputs are estimates based on the information available at the time they are generated.

A team can be strong overall while facing a difficult opponent-specific matchup. The matchup view therefore considers the main position and unit groups separately, including:

• quarterback play and pass defense;
• running game and run defense;
• receivers and tight ends against the opposing coverage;
• offensive line play against the opposing front; and
• broader team efficiency and situational factors.

This approach is meant to answer not only "who is favored", but also "why the matchup may lean that way." The result can change when expected starters, injuries, recent form, or other relevant information changes.

Fantasy projections begin with a player's available historical production and recent performance. The estimate is then adjusted for factors such as:

• expected role and playing time;
• opponent strength;
• likely game flow;
• position-specific scoring patterns; and
• the player's current availability.

For positions where several players may contribute, the Coach compares eligible candidates rather than assuming that the highest depth-chart label is always the best statistical projection. Quarterback and kicker selections place more emphasis on the expected starter role.

Players with limited history are not automatically discarded. When a player does not have enough previous data, the system uses a reasonable position-specific baseline and updates it as new information becomes available. This helps keep projections useful for emerging players and other small-sample cases.

As you can imagine the projections are intended for comparison and planning. They are not guarantees of playing time, touches, targets, or fantasy points.

Availability data can change quickly. The app automatically removes a player from projection eligibility when the available status explicitly identifies the player as Out. A Doubtful designation is treated as uncertain rather than as an automatic exclusion, so the projection should be read with appropriate caution.

Depth charts, rosters, injury information, and the coach's gameday starter selections can all contribute to the expected-player decision. Coach C will update the prediction up to 30 minutes before game time so that coaching decisions and late injuries are incorporated into the final forecast.

Because source data can lag behind real-world news, users should review the latest team announcements before relying on a player projection.

Football data is published at different times and at different levels of detail. Early-season, injury, roster, and depth-chart information can be incomplete or delayed.

When information is missing, the system uses a conservative baseline or continues with the portions of the analysis that can be supported. A result based on less information should be treated as less certain than one based on a fuller and more current sample.

The site may also retain calculated results temporarily to keep the experience responsive. Refreshing the relevant prediction view ensures that newly available data and updated player selections are used for a new calculation.

Confidence describes how clearly the available signals separate the two teams or players. It is not a guarantee, and it should not be interpreted as a promise that an outcome will occur.

Confidence levels for the predictions refers to the picked game winner, not the score or fantasy projections. Anything over 50% suggests that Coach C thinks the game winner selected is more likely than not.

Confidence can be affected by:

• the amount and freshness of available data;
• the size of the matchup difference;
• uncertainty about player availability;
• unusual recent results; and
• information that has not yet reached the public data sources.

Last-minute inactive announcements, unexpected playing-time limits, weather changes, and coaching decisions can all make a previously reasonable prediction less useful.