0)
this is still the best wg game concept out here, as it's an actual game.
1)
The biggest problem is still the mismatch between descriptions.
E.g.
>>
111.0 lbs +9.0 gained · 163cm · 18 yrs
● Denial — She hasn't acknowledged what's happening to her body — or has decided not to. The changes are visible. They are not discussed.
👗 Her clothes pull just slightly across the belly and hips. She adjusts them without comment.
Jaryn has striking, unconventional features — the product of origins she hasn't explained. Her face doesn't belong to any single tradition and is more interesting for it.
Jaryn's chest has grown first and fastest — fuller, heavier, straining at her clothes in a way that was not an issue when she arrived. Her arms have softened too, losing their previous definition.
Jaryn is powerfully built and lean with it — flat belly, dense arms, muscled thighs, the sharp physicality of an active fighter.
Jaryn's arms are genuinely soft now — upper arms that jiggle slightly when she moves, forearms rounded and warm, the whole arm heavier and fuller than it was. Sleeves that once slid easily over her arms now cling. Jaryn's bust is full and heavy — a weight she carries constantly, her clothes adjusted to accommodate them. They rest warm and prominent against her, impossible to ignore. Jaryn's hips have widened noticeably — she moves through doorways a little more carefully now, her silhouette wider than it was. Jaryn's thighs press together now — a warm contact that wasn't there when she arrived. She adjusts her stance slightly without seeming to notice she does it.
A soft layer has settled over Jaryn's hard stomach — warmth, a yielding quality that wasn't there before. The definition that was so crisp is now padded.
<<
This is BS. I recommend just rewriting the whole thing. I have no idea how it can be that wrong? Because e.g. Claude is actually pretty good at estimating sizes from photographs etc.
2) suggestion image wise
I asked chatgpt to draw her, lol. I think re: images ,the thing to do would be stick to a bunch of pre-defined heroines, have ~12+ possible stages, an image for each action (e.g. getting them drunk) and also perhaps images of the inn as it's getting improved.
GPT image is now really good, but it costs some. If you find someone willing you can do plenty even with local models now. OTOH, if you live in the west and don't eat out 2-3x times, instead roll your own sandwich, that'd pay for it too certainly. ~50 bucks is probably enough to get a 5 minute HD AI genned film done if using an open / commodity model.