Skip to content
♣♦♥♠Poker Flows
Cut one free
Back

What actually makes a poker short work

Research notes behind Poker Flows. Everything in src/content/templates.ts and the beat engine in docs/clip-grammar.md traces back to something in here.

1. The format exists because of one structural accident

Poker is the only spectator game where the audience knows more than the players. The hole card graphic is the whole trick. A viewer who can see both hands watches a person walk into a trap in real time, and the dramatic irony does the retention work for free. No other sport gives you that.

This is why poker punches so far above its weight in short form. A soccer highlight needs the goal. A poker clip needs the decision, and a decision is legible in seven seconds if you frame it right.

Practical consequence: the reveal order is the edit. Not the footage quality, not the music. The single single most valuable choice in a poker short is when the viewer learns each piece of information. That is the thing an editor spends their hours on, and it is the thing this product automates.

2. Who set the conventions

Two distinct audiences fall out of this, and they need different edits:

PlayersCivilians
Knows what a squeeze isyesno
Wantsthe spot, the sizing, the readthe money, the face, the cooler
Hook that works"He called with this""$47,000 on one card"
Toleratesslow street by streetnothing over 2s without a beat
WhereYouTube Shorts, XTikTok, Reels

Most tools ship one edit. The difference between a clip that does 4k and one that does 400k is usually that it was cut for the wrong one of these two.

3. Anatomy of the clip

Short form retention data is consistent across genres: the first one or two seconds decide the swipe, a Short needs roughly 70 to 90% view through to escape its first test audience, and sustained attention wants a new visual beat every 1.5 to 3 seconds. Poker maps onto that skeleton unusually cleanly, because a hand already has beats, they're called streets.

0.0-1.5s   HOOK        The payoff frame, out of order. Never the setup.
1.5-4s     STAKES      Pot size, stack depth, who these people are. One line.
4-8s       PREFLOP     Hole cards revealed. This is the dramatic irony load.
8-18s      FLOP/TURN   Bet, sizing, the tank. Equity bar if it's a race.
18-30s     RIVER       The card. Hold on the hesitation. Do not cut early.
30-40s     RESOLUTION  Chips, reaction, the number.
last 1.5s  BUTTON      A line that closes it, or a loop back to frame 1.

The hook is the entire game. It is almost always stolen from the end of the hand and placed at the start: the all in shove, the winning card, the face at the moment of the call. Then the clip rewinds and earns it. Editors call this the cold open; this product calls it a hookSource and picks it automatically.

4. The five hooks that carry the genre

Derived from what the best performing poker shorts actually open on. These ship as templates in src/content/templates.ts.

  1. 1.The Cooler, two enormous hands collide, neither player can fold. Hook on the two hole card graphics side by side. Zero narration needed; the cards are the joke.
  2. 2.The Hero Call, someone calls a huge river bet with a marginal hand. Hook on the tank, then the call. The payoff is the bluff reveal, so hold the villain's cards until the very last beat.
  3. 3.The Bad Beat / Runner Runner, the mathematically doomed hand gets there. Hook on the outs graphic ("2 cards to win"). The suspense is arithmetic, so show the arithmetic.
  4. 4.The Bluff, hook on the villain's cards, immediately, out of order. The viewer knows it's air and watches the hero agonise. Pure dramatic irony, maximum load.
  5. 5.The Table Talk / Tilt, the hand is the setup, the human is the payoff. Hook on the line of dialogue. Needs face cam; dies without it.

A sixth exists but is a different product: the teaching clip ("what would you do here?"), which ends on a question and lives or dies on comment volume.

5. What editors actually spend the hours on

Ranked by time cost, from creators describing their own workflow:

  1. 1.Finding the hand. Scrubbing a six hour stream for the twelve hands worth cutting. The single biggest time sink and the most automatable.
  2. 2.Reveal timing. Frame accurate placement of the hole card graphic, the equity bar, the pot counter. Every graphic has a correct in point tied to a betting action.
  3. 3.Dead air removal. Live poker is mostly people thinking. A four minute hand becomes forty seconds almost entirely by cutting the tanks, except the one tank you keep, because that one is the tension.
  4. 4.Reframing 16:9 to 9:16. The table is wide, the phone is tall. Naive centre crop loses the board or the player. Needs subject tracking, or a stacked layout with the board on top and the player below.
  5. 5.Captions. Word level, high contrast, poker vocabulary spelled right. Generic ASR writes "check raise" as "Czech raise" and mangles "the nuts" into something worse.
  6. 6.Sound. A held beat under the tank, a drop on the river card. Cheap and effective.

Items 1 to 5 are the product. Item 6 is a v2 with a licensed library.

6. Where the footage comes from

These are three different ingest problems. Stream VODs already carry an overlay we can read; vlog footage needs the hand entered by hand. Poker Flows handles both, one by parsing, one by a hand entry form that takes about twenty seconds.

7. The competitive gap

The generic clippers, Opus Clip, Vizard, Klap, Choppity, CapCut Highlights, all do the same thing: transcribe, score the transcript for "virality", reframe, caption. Every one of them is transcript driven.

Poker's best moments have no dialogue. A player tanks for ninety seconds in total silence, shoves, and gets snap called. The transcript for the single most viewed beat in the hand is empty. Transcript driven scoring is structurally blind to it.

That is the entire wedge:

Generic clippers find the moment someone said something interesting. Poker Flows finds the moment someone did something interesting.

Poker Flows scores on game state, pot size relative to stack, equity swings between streets, all in events, showdown hand strength, bet sizing outliers, and uses the transcript only as a tiebreak for table talk. It also knows how to draw the graphics, because it knows the hand.

8. What this means for pricing

The market rate for a competent freelance poker editor runs roughly $50 to $150 per finished short, and a serious creator needs five to twenty a week. That is the number the product is priced against, not Opus Clip's $20/mo, which is not the alternative anyone is actually weighing.

At $100/mo unlimited, Poker Flows undercuts a single outsourced clip. That is the whole pitch, and the landing page says it in those terms.

Sources / reading

All of that, on one hand of yours

The first clip is free and it’s a real render, full quality, no watermark.

Cut one free