Every Rejection Is Priced Too

The whole cut lives in 9 files. A 2,714-byte review, an 8,912-byte cutlist, a 1,912-byte EDL, a 3,446-byte FCPXML, 2 title cards at 26,907 and 28,952 bytes, a publish package of 2 files and 1 finished reel at 4,895,300 bytes. The reel runs 14.6 seconds over 6 beats with 0 gaps. The delivery is 1080x1920 at 30 fps and -14.0 LUFS. Inside the cutlist sit 6 picks and 6 distinct sources from 2 projects, and not 1 of them was used before. Inside the review sit 18 candidates: the 6 that shipped at 6.20, 5.942, 8.942, 6.20, 8.942 and 5.942, and the 12 that did not. The 6 picks add to 42.166 and average 7.028. The 12 rejects add to 74.07 and average 6.1725. Eighteen candidates were priced before the timeline was closed, and every price was on disk before anyone watched the reel.

Sort the 12 rejects and take the best 6: 7.61, 7.61, 6.20, 6.20, 5.94 and 5.94. That is 39.50. The 6 that actually shipped score 42.166, a 2.666-point margin over the strongest alternative cut the folder could have offered. Now read it the other way. The single best reject sits at 7.61, and 4 of the 6 shipped picks sit below it at 6.20, 6.20, 5.942 and 5.942. The top of the drawer outranks the bottom of the timeline. What holds the shipped order is the top of it: 2 beats at 8.942, both clips from Sal_y_Miel, and no reject reaches either. A ranked drawer does not stop a weaker pick from shipping. It makes the weaker pick a decision instead of an accident.

Each reject carries the score of the beat it was competing for, and that is what makes the number useful. The hook drew 2 alternates at 5.94 and 5.94 against a pick of 6.20. The context beat drew 5.94 and 4.61 against 5.942. The proof beat drew a still at 7.61 and a clip at 5.94 against 8.942. The turn drew 6.20 and 6.20 against a pick of 6.20, the 1 tie in the folder. The payoff drew a still at 7.61 and a clip at 6.20 against 8.942. The release drew 5.94 and 5.94 against 5.942. Twelve rejects in total, 3 of them stills and 9 of them clips, drawn from 4 different productions. A score pinned to a beat is a replacement with a known price, and a price is what turns a leftover into a second option.

Run a search for the word reason across all 9 files and it returns 0 hits. The why column in the review reads tags[hook], 768x1364, score 6.20 - a tag, a dimension and a number. No sentence explains why b10 lost the hook, or why the still of b07 lost to the clip of b07. That absence is the design. Eighteen candidates, 18 numbers, 0 explanations, and the number is the part that survives a rewrite. A rejection with a number on it can be re-run against a better brief next quarter. A rejection with an opinion on it has to be re-argued by whoever still remembers it. Keep the number and the drawer keeps working.

9 files. 18 candidates. 14.6 seconds.

Every source runs 24 fps into a 30 fps delivery, so 6 of 6 clips need retiming before they can be used. Four of the 6 picks are 848x480 wide frames feeding a 1080x1920 vertical canvas, and 2 are 768x1364, which is already 9:16. The 6 chosen sources hold 43.9167 seconds, of which 14.6 ship, or 33.25 per cent. The music bed runs 206.160978 seconds, enough for 14.12 reels at this length, and the cut uses 1 of them. Three captions were written at 192 of 2,200 characters, 123 of 2,200 and 177 of 5,000. None of that is a defect of the drawer. It is the drawer's inventory, and the score on each reject is what says which frame is worth the crop and which frame is a crop with nothing left in it.

Here is why the discipline pays, on the record. Entrade built 200 wood-gasification machines across the UK and Europe, and 200 is a count somebody can walk and check. A EUR 75 million industrial group was restructured, and a counterparty can recount it from its own records without asking anybody. The book published on 8 April 2026 runs 393 pages, and anyone holding it can confirm that in 1 minute. Three claims, three numbers, all three checkable by a third party. Now put the method beside them: 9 files, 18 priced candidates, 14.6 seconds, 6 beats, 0 gaps, 42.166 against 74.07, and 4,895,300 bytes of finished reel. The probability that an unpriced drawer gets re-mined on taste, with a deadline and no time, is not 0. A priced drawer answers that before the question arrives. Every rejection is priced too. Keep more of yourself, for longer.