Saint Blackmoor

Saint and Sinner
Donor
Oct 26, 2017
6,030
18,448
Had to split due to technical issues. We are thinking about making major adjustments to the first part to combine both parts into one game.
I can safely say you won't do that. The cost-benefit ratio is unfavorable, and this game shows playability issues based on the questions asked.

Some factors are out of your hands.

Market saturation.

Sites like these foster a sense of entitlement in the more unintelligent members because it's all free, thus creating a petulant attitude in them. These are the foul-words-spitting, loud, and hateful ones that give this site its toxic reputation.

The review system here sucks. Reviews often damage a game's reputation before it can even get started. Many are from the reviewer's ignorance of what a proper is supposed to achieve or those who use the review system to hurt games they dislike.

Also, many sub-par games here cheapen the experience and disillusion the player to new games.

I wish you good luck. The art does look good, IMO.

I bet this gets deleted. :LOL:
 
Last edited:

hgffdh

Member
Feb 26, 2018
158
207
The review system here sucks. Reviews often damage a game's reputation before it can even get started. Many are based on the reviewer's ignorance of what a proper is supposed to achieve or those who use the review system to hurt games they dislike.
I can't speak for everyone but I've never ever read a single game's review. First, they aren't that visible and second, I don't care. Games are good if they have tags I like. 95% of the judgement the game receives is due to the tags. The rest is due to stuff that tags don't mention - is the MTL horrible? Is the pregnancy a mechanic? Is NTR avoidable? Etc. If above criteria are satisfied - tags are right and the rest is fine - I will play it, I don't give a shit about the rest. AI CGs are good, compared to some of the hand drawn art we had in some games. Sekira 2 art is certainly worse than 1st but a game with mediocre art is better than no game at all.

Personally though, for me bad art is a deal breaker in this case because Sekira 1 art was top tier and while the tags weren't to my tastes (no pregnancy at all! no rape! no violence or vore!) I still enjoyed it purely thanks to the great art. So mediocre art means I won't bother trying out the second part since it still has no tags I particularly like. But that's just me, for people who don't share my tastes AI CG game is still better than no game at all.
 
  • Like
Reactions: Evizzy89 and lemaro

torohror

Newbie
Nov 13, 2023
88
128
[X] Doubt.
I mean, it does look the part, but the original game (so-called "Part 1") also had CG with somewhat nonsensical parts, back in 2021. Given that AI-gen of this kind of CG was simply not available then, I would more readily attribute that to the artist not knowing how bothering to draw anatomy properly.
AI image generation exists since 2016. DALL-E is around since 2021. So, actually, Part 1 also could be AI-Generated.
 
  • Angry
Reactions: qwerty-1234567890

Arkanae

Newbie
May 29, 2018
81
157
AI image generation exists since 2016. DALL-E is around since 2021. So, actually, Part 1 also could be AI-Generated.
Not really, no. The actual timeline of model performance doesn't work.

The original DALL-E was crap at bodies, and generally pretty crap at anything not abstract or geometrically simple. DALL-E 2 wasn't released until Sep. 2022, and wasn't very good either. Midjourney, in 2022, was a bit more self-coherent, but still crap at bodies and detail accuracy. The first commercially released LLM model that was able to generate relatively "anatomy accurate" images was SD 1.5, out in Oct. 2022.

Even today, SD XL is probably the best commercial LLM for this kind of work, and it still requires lots of fine-tuning with post models to get to that level of quality. Fine-tuning that is nearly impossible to implement for a non-data-scientist without tools that didn't exist until mid 2024. And even a royal pain to set-up for a data-scientist used to doing image processing... (source: me. I did, before ComfyUI was a thing, and it made me swear all the way through).
 
Last edited:

EnterYourNameHere

Engaged Member
Mar 16, 2022
2,075
2,794
Reviews often damage a game's reputation before it can even get started.
Nobody will be happy with everything. Reviews works also bad in the other site. Negative reviews might destroy the reputation but positive do the same. Its not the first time I see reviews like
"very bad game, just a bugfeast, worst engine / pictures but cuz its your first game I rate it 5 star".
People will see a 5 star game, download and notice its actually just a bad game cuz the reviewer rate it very high just because its the first game.
 

Saint Blackmoor

Saint and Sinner
Donor
Oct 26, 2017
6,030
18,448
Nobody will be happy with everything. Reviews works also bad in the other site.
Many are based on the reviewer's ignorance of what a proper is supposed to achieve.
True, that was part of what I tried to say. ;)
Rule 8 was added not long ago to combat members' attempts to counteract messed-up reviews. The review system needs a big overhaul, but it won't happen. There's no benefit to the investors/owner/owners to do that.
Reporting reviews that don't meet the rules is the best solution.

I did try this game, didn't play the first. You definitely need to play the first.
 
Last edited:

defnotalt

Member
Jul 13, 2021
490
1,733
Reviews often damage a game's reputation before it can even get started.
Devil's advocate: people have the right to leave a bad review if the game is in a barely functional state/barely has any content. People used to sell games when they were finished, now you're lucky if you get to have fun in their first version. It doesn't matter if F95 users get to download it for free, because if it's here that means it's available somewhere else and thus has been provided as something to pay/support in that state.

Now, is it unfair that reviews from early versions are kept as the game progresses to an actually competent state? To some extent, yes. I'm pretty sure I heard the next site overhaul will have version-based reviews, though not sure how that'll work exactly.

Also, from anecdotal experience, I haven't really seen average review scores in this site being wildly different than the percieved quality of the game. Bad apples and retards exist, but I've found people interested in a game tend to self-regulate within the thread pretty well.

AI image generation exists since 2016. DALL-E is around since 2021. So, actually, Part 1 also could be AI-Generated.
Yeah no, Stable Diffusion was first released in 2022, and anything earlier than that was absolute dogshit :KEK: