How the DNA Score works

The DNA Score is one number from 0 to 100 for how an athlete projects in football. An athlete who plays more than one sport has a score for each. It is worked out again when a scout locks an evaluation and as its other inputs are updated, and every change is kept in the athlete's score history.

There is no DNA Score until a trained scout locks an evaluation. Paying never changes this score.

The score blends up to seven parts, each on its own 0 to 100 scale, by these weights:

• Scout gradeThe scout's grades for the position's traits.40%
• Coach reviewsWhat the athlete's own coaches say, adjusted for how each coach grades. It counts once two different coaches have reviewed the athlete.15%
• Position ELOThe player rating, built from the team's results and the athlete's production.15%
• CampTested numbers against sourced ranges for the position.15%
• ProductionSeason stats against a typical season at the position.10%
• Film qualityWhat kind of footage the film is: game film counts most.3%
• Recent formThis season's stats against last season's.2%

DNA Score = (weight × part, added up over the parts with data) ÷ (those parts' weights, added up)

Only the parts with data count. Dividing by the weights that are present scales them back up to 100%, so a missing part is never scored as zero: it simply doesn't count until there is something to measure.

An athlete with every part in place:

The same athlete early on, with a scout grade, Camp numbers and production but no coach reviews, rating or film record yet. The three parts present carry 65% of the weight between them, so each is scaled up by 100 ÷ 65: the scout grade's 40% becomes 61.5%.

A trained scout watches the film and grades each trait for the position from 1 to 10 against a fixed rubric. Scouts never see the athlete's name, school or photo.

Scout grade = average trait grade × 10

With more than one locked evaluation, the grades are averaged with each scout weighted by their Scout Authority Score (SAS), a 0 to 100 measure of their background and track record:

Scout weight = (SAS ÷ 50) ^ 1.5

Example. A scout with SAS 80 (weight 2.02) averages 8.0, and a newer scout with SAS 40 (weight 0.72) averages 7.0. The scout grade is (2.02 × 80 + 0.72 × 70) ÷ (2.02 + 0.72) = 77.4.

Each Virtual Camp number is compared with sourced ranges for the athlete's position at each level, and scores for the highest level whose range it reaches.

• Reaches D1 FBS100
• Reaches D1 FCS80
• Reaches JUCO65
• Reaches D260
• Reaches D3/NAIA40
• Below every level20

Camp = average of the test scores, with unverified numbers counted at 50% weight

A test with no sourced range for the position is left out rather than guessed. If no test has one, this part has no data yet.

Example. Two verified numbers reach D1 FCS (80) and D1 FBS (100); an unverified one reaches JUCO (65, counted at half weight). The part is (80 + 100 + 0.5 × 65) ÷ 2.5 = 85.

Unverified numbers count at half weight by default; the platform can change this.

The latest season's stats are added into one production number for the position and compared with a typical season at that position.

Production ratio = (verified production + unverified production × 0.25) ÷ typical production for the position

Production = production ratio × 50, at most 100

A typical season scores 50 and twice typical scores 100. A classification multiplier can adjust the ratio for the size of the school.

Example. Verified stats 40% above typical give a ratio of 1.4, and Production = 1.4 × 50 = 70.

Unverified stats count at a quarter by default; the platform can change this.

The player rating starts from the team's ELO, which moves with every game result against the strength of the opponent, and adds an adjustment for the athlete's own production.

Position ELO part = (player ELO − 1200) ÷ (1800 − 1200) × 100, between 0 and 100

Example. A player ELO of 1560 gives (1560 − 1200) ÷ 600 × 100 = 60.

The range mapped to 0 to 100 is 1200 to 1800 by default; the platform can change it.

Reviews from the athlete's own coaches in this sport, once they are linked to the athlete's profile. Each coach's scores are measured against every review that coach has written, so a generous grader doesn't lift a score and a tough one doesn't sink it. The result is a 0 to 100 part.

Each clip counts by the kind of footage it is, and the part is the average:

• Game film (pads)× 1
• Camp or 1v1× 0.7
• 7v7 / flag× 0.4
• Drill or workout× 0.2

Film quality = average clip weight × 100

Example. Two clips of game film (pads) and one of drill or workout: (1 + 1 + 0.2) ÷ 3 × 100 = 73.3.

How the latest verified season compares with the one before. It needs two verified seasons.

Recent form = 50 + 50 × (latest season's production ÷ previous season's − 1), between 0 and 100

Example. Production up 20% on last season gives 50 + 50 × 0.2 = 60. The same as last season gives 50.

Paying, or the plan a family is on. The athlete's name, school and photo, which scouts never see. And a missing part, which waits for data instead of counting as zero.

More on the whole process: DNA for families and how it works.