Song Evaluation Scorecard

What this is

The scorecard above lets you rate any rap song across five dimensions — Flow, Lyricism, Delivery, Production, and Originality — on a 1–10 scale, and returns a weighted total out of 100. Originality is weighted ×1.1 because it is the one dimension nobody can template their way into. Production is weighted ×0.9 because a genuinely great song survives a rough mix, but no amount of mixing rescues a mid song.

This is the same framework SongBattle judges use as a starting point when evaluating submissions, and the same framework we recommend artists use before sending a song to press, curators, or contests. It is deliberately harsh. The scores are calibrated so that a 7 is genuinely good, a 5 is average, and a 9 is rare. If everything you make scores 8+ from yourself, you are not scoring — you are hoping.

Below the tool, this page breaks down what each dimension actually means, what a 3, 6, and 9 sound like in practice, and how to run the scorecard on your own songs, on peers' songs, and as an A&R exercise.

When to use it

Run the scorecard on a song after you have lived with it long enough that you can play it without wincing, but before you commit money to a rollout — before mastering, before press outreach, before pitching to curators, before entering a competition. It is also the right tool for A/B'ing two candidate songs for lead single: score both, get three peers to score both, average, and let the numbers argue.

The five dimensions

Flow

Flow is the relationship between your syllables and the beat's grid — how your rhythms lock, syncopate, and breathe against the pocket. It is the first thing a trained ear notices and the first thing that dates a record.

• 3/10: Bar structure is inconsistent. Cadence resets randomly between lines. You can hear the artist counting. Rhymes land on obvious downbeats without variation across the whole song.
• 6/10: Bars land on-beat with basic pocket control. There is one primary cadence and the artist stays inside it. Nothing offensive; nothing memorable. A listener could not hum the flow back.
• 9/10: Multiple cadences deployed with intent — a switch on the second verse, a triplet inside a straight-eighth pattern, a bar that lands a hair behind the pocket for emphasis. A listener could clap the rhythm of the hook after one play.

Lyricism

Lyricism is what the words actually say and how they say it — the specificity, the wordplay, the images, the compression. Not vocabulary. Not "big words." Precision.

• 3/10: Filler bars. Punchlines that telegraph. Rhymes chosen because they rhyme, not because they mean anything. Nothing you could quote back.
• 6/10: Competent writing. Some clean images, a few working setups. But the verses are interchangeable — you could swap the first and third and nothing would break.
• 9/10: Every bar is doing work. Setups pay off two lines later. Concrete nouns instead of abstractions. At least one line that a listener will send to a friend as a screenshot.

Delivery

Delivery is the voice — tone, texture, energy, conviction. Two artists can rap the same bars and one sounds like a demo while the other sounds like a record. That difference is delivery.

• 3/10: Monotone across the song. No dynamic range between verses and hook. The vocalist sounds like they are reading. Emotion does not match the subject matter.
• 6/10: Present, on-mic, in tune with the beat's energy. But the same energy runs from bar one to the end. No moment where the voice pulls the listener in closer.
• 9/10: Dynamic control — moments of restraint that make the loud moments hit harder, texture that shifts with the subject, and a distinctive vocal signature you would recognize in one bar on the radio.

Production

Production is the beat, the arrangement, the mix, and the way the vocal sits inside the whole. Weighted at ×0.9 because it is the most fixable dimension — and the one artists most often over-index on at the expense of the song itself.

• 3/10: Vocals fight the beat for space. Muddy low end. Hi-hats clipping. Arrangement is loop-loop-loop for three minutes. Sounds worse on a phone speaker than in the studio.
• 6/10: Clean mix, professional loudness, no obvious mistakes. Arrangement has an intro, verse, hook, verse, hook, bridge, out. Nothing distracting, nothing distinctive.
• 9/10: The beat has its own melodic hook. The arrangement earns each section. The mix translates from car to earbuds to laptop. Small production choices — a drop-out before the hook, a reversed hat, a texture that only appears on verse two — reward re-listens.

Originality

Originality is the "who else could have made this?" test. Weighted ×1.1 because everything else on this scorecard can be learned. This cannot. It is a combination of point of view, sonic identity, and the willingness to make choices that other artists would talk themselves out of.

• 3/10: Sounds like ten other current songs. Interchangeable references, interchangeable cadence, interchangeable ad-libs. You could tell the artist "make it more like X" and they would.
• 6/10: There is a recognizable voice or angle, but it is one you have heard before — a well-executed version of something familiar rather than a new thing.
• 9/10: The first fifteen seconds tell you who the artist is. A perspective that feels lived-in rather than performed. A choice — sonic or lyrical or structural — that most artists would not have made.

How to run this on your own song

1. Play the finished mix twice through, back to back, on the worst speaker you own.
2. Score it yourself. Be your harshest critic. If you find yourself giving 8s across the board, throw out your scores and start again.
3. Send the file to three trusted outside ears — a producer, a rapper who is honest with you, and one person outside music. Ask them to score it independently, no discussion.
4. Average the four scorecards. Discuss only after everyone has scored.
5. If any single dimension averages below 6 across all four scorers, fix that dimension or scrap the song. Two dimensions below 6 and the song is not release-ready.

The biggest failure mode is scoring alone and calling it done. Your own ear is the least reliable ear in the room. You have been living with the song for weeks, so you hear it as it is supposed to sound, not as it actually sounds.

How to run this on someone else's song

The same rubric works for playlisting decisions, A&R practice, and competition judging. Score without looking at the artist's name, socials, or streaming numbers first — score the audio. Then look at the context. If the audio scores 70+ and the artist has 400 monthly listeners, that is a lead worth following up on. If the audio scores 55 and the artist has 40,000 monthly listeners, that is a marketing story, not a song story.

Running this on ten songs a week for a month will retune your ear in ways that reading about music criticism cannot.

The blank template

SONG EVALUATION SCORECARD
Song: ____________________________
Artist: ____________________________
Scorer: ____________________________  Date: __________

FLOW              /10   Notes:
LYRICISM          /10   Notes:
DELIVERY          /10   Notes:
PRODUCTION        /10   Notes:
ORIGINALITY       /10   Notes:

Weighted total:
(Flow × 1.0) + (Lyricism × 1.0) + (Delivery × 1.0)
+ (Production × 0.9) + (Originality × 1.1) = _____ / 100

Overall verdict (circle one):
  RELEASE READY (75+)
  NEEDS ONE FIX (65–74)
  NEEDS TWO FIXES (55–64)
  RETURN TO DEMO (< 55)

One-line takeaway: ____________________________________

Completed example

Song: "Free Radio"
Artist: K. Vaughn
Scorer: Curator panel (avg. of 4 independent scores)  Date: 2026-05-11

FLOW              7 /10   Notes: Confident pocket in verse 1, second verse
                          switches to a triplet pattern that pays off. Hook
                          cadence is a hair repetitive by the third pass.

LYRICISM          8 /10   Notes: Concrete imagery throughout. Two lines that
                          function as quotables. One filler couplet in v2.

DELIVERY          6 /10   Notes: On-mic and present. Same energy from bar 1
                          to end. Would benefit from one moment of restraint
                          before the last hook.

PRODUCTION        7 /10   Notes: Clean translation across speakers. Beat has
                          a melodic hook of its own. Arrangement plays it a
                          little safe — no drop-outs, no bridge risk.

ORIGINALITY       8 /10   Notes: The perspective is specific and the sonic
                          palette (dry drums, mid-forward vocal, no reverb
                          on the ad-libs) reads as a deliberate choice.

Weighted total:
(7 × 1.0) + (8 × 1.0) + (6 × 1.0) + (7 × 0.9) + (8 × 1.1)
= 7 + 8 + 6 + 6.3 + 8.8
= 36.1 out of a possible 50 raw → scaled to 72 / 100

Overall verdict: NEEDS ONE FIX (65–74)

One-line takeaway: Strong writing and identity — spend a day on vocal
dynamics before this goes to curators.

Related SongBattle reading

• How to Judge a Rap Song
• Flow vs. Delivery vs. Cadence
• Why Some Songs Grow on You
• How to Tell if an Artist Has Star Potential
• The Six-Week Single Rollout