The Standings → GPT‑6.1 Sol (medium)
OpenAI
medium reasoning effort · via Codex CLI · judged by Claude Sonnet 5 · the setting shown on the standings; also run at xhigh63.1 high63.0
Results from run ‘first’: ten items, ten verdicts.
GPT‑6.1 Sol (medium) scores 66 on craft, 3 points above the field’s 63 and 27 points short of the authors themselves. Its shape differs from the field’s: relatively strong on restraint and folding the brief into the story, relatively weak on momentum and building scenes. On the mechanical checks it kept to the brief’s word range on 10 of 10 items and is unusually free of stock phrases.
The record: 10 items · 10 verdicts · 0 errors · 2,296 words on average · est. generation cost $0.75
Eight dimensions, each 0–100, averaged over ten items; the author’s real chapter was in the room every time.
Craft by dimension, 0–100 · judged by Sonnet 5 beside the author’s real chapter
The authors themselves average 93 on the five writing dimensions, the field 63, this model 66. Hover or focus a bar for the field and the authors on that dimension.
The judge’s one-sentence best quality and worst flaw for each chapter, written for a general reader.
Best The money argument is staged as a clear, well-paced scene, and Anne's proposal followed by Elizabeth's surprising support lands cleanly and is easy to follow.
Worst The writing is careful and controlled but a bit flat next to the real thing — it reads like competent imitation rather than prose with its own ironic bite.
Best The evasive small talk about phoning tomorrow and about George Vane is genuinely well built — both characters clearly want something they won't say, which is exactly the right note for these two.
Worst Several moments state feelings outright rather than letting the dialogue and action carry them, which flattens the guarded, oblique mood the source establishes.
Best The narrator's wife gets real personality through her own sharp observations, like noting that Ogilvy already seems to know what he'll find, which makes her more than a passive companion.
Worst The chapter breaks into many short, clipped paragraphs and quick exchanges that read more like a factual report than the flowing, cumulative sentences of the original book, so it doesn't fully sound like Wells.
Best The scene where the carriage stops for sheep and Mary climbs out onto the dark moor is well-paced and physically vivid, with good sensory detail (the wind, the mud, the sudden sense of smallness).
Worst The ending basically states out loud that Mary has decided to find out what the noise was, which reads like it's checking off the assignment rather than showing that decision happen naturally.
Best The Renfield scene is quietly and convincingly creepy, giving him his own fussy, logical madness ('Either the lady has accepted him, or she has not') instead of just having him say something spooky.
Worst The closing twist about a forger inventing news of Jonathan feels imported from a detective story rather than grown out of Stoker's slow-building dread, and it introduces a plot thread (Anton Weiss) that the rest of the chapter never earns.
Best The clues (the mismatched screw with silk thread, the latch-key contradiction) are planted cleanly and Poirot's refusal to explain them builds real curiosity without the narration over-explaining.
Worst Poirot doesn't sound like Poirot here — none of his French exclamations or theatrical flourishes come through, so he reads as a generic quiet detective rather than the vivid character from the books.
Best It delivers every required plot point cleanly and in clear order, so the reader never loses track of what is happening.
Worst Large stretches are carried by clipped, businesslike dialogue and a dry lawyer's letter, so it reads flatter and less alive than the warm, rambling voice of the original.
Best The swimming scene is admirably understated -- it shows Edna's small, real success without inflating it into a grand epiphany.
Worst Too much of the chapter is summarized rather than played out on the page, with flat lines like 'The evening passed without incident' standing in for scenes that should be dramatized.
Best The bit where Lord Emsworth accidentally quizzes Eve with his gardening questions instead of library ones is a clever, character-true piece of comic business.
Worst The jokes and dialogue feel flatter and more workmanlike than the original, missing the sparkle of wordplay the real author is known for.
Best It is lean and efficient, moving through the bank, the widow, the note, and the wire without a single wasted or padded moment.
Worst Most of the chapter is rapid back-and-forth questions and one-line answers, so it reads like a police transcript instead of a scene with people in a room.
Headline, compliance and craft for each chapter, 0–100.
| Item | Headline | Compliance | Craft | Words |
|---|---|---|---|---|
| Persuasion item 001 · after ch. 1 · Jane Austen | 75.8 | 100 | 76 | 2,445 |
| The Sun Also Rises item 021 · after ch. 3 · Ernest Hemingway | 74.2 | 100 | 74 | 2,124 |
| The War of the Worlds item 006 · after ch. 1 · H. G. Wells | 73.8 | 100 | 74 | 2,375 |
| The Secret Garden item 013 · after ch. 2 · Frances Hodgson Burnett | 73.2 | 100 | 73 | 2,243 |
| Dracula item 015 · after ch. 5 · Bram Stoker | 66.0 | 100 | 66 | 2,341 |
| The Mysterious Affair at Styles item 017 · after ch. 3 · Agatha Christie | 64.6 | 100 | 65 | 2,503 |
| The Moonstone item 011 · after ch. 8 · Wilkie Collins | 61.6 | 100 | 62 | 2,489 |
| The Awakening item 016 · after ch. 7 · Kate Chopin | 59.8 | 100 | 60 | 1,730 |
| Leave It to Psmith item 022 · after ch. 3 · P. G. Wodehouse | 57.0 | 100 | 57 | 2,461 |
| Red Harvest item 023 · after ch. 2 · Dashiell Hammett | 55.4 | 100 | 55 | 2,252 |
Words written per item, against each brief’s maximum.
Word count per item vs. brief maximum · run ‘first’
Ten of ten items landed inside their brief’s word range.
Each verdict names the chapter’s single worst flaw, with evidence. Quoted verbatim from run ‘first’.
The closing twist about a forger inventing news of Jonathan feels imported from a detective story rather than grown out of Stoker's slow-building dread, and it introduces a plot thread (Anton Weiss) that the rest of the chapter never earns.
The chapter breaks into many short, clipped paragraphs and quick exchanges that read more like a factual report than the flowing, cumulative sentences of the original book, so it doesn't fully sound like Wells.
Poirot doesn't sound like Poirot here — none of his French exclamations or theatrical flourishes come through, so he reads as a generic quiet detective rather than the vivid character from the books.
The ending basically states out loud that Mary has decided to find out what the noise was, which reads like it's checking off the assignment rather than showing that decision happen naturally.
Automated screening results across the ten outputs.