does chat gpt purposely avoid rating you?

L

larpingricecel

Iron
Joined
Jan 21, 2026
Posts
4
Reputation
5
ive noticed that whenever I ask it to rate it chatgpt just avoids the question and gives me some surface level bullshit. any work arounds?
 
  • Ugh..
  • JFL
Reactions: Wexilarious and Deleted member 322969
dont even try with gpt, ive done it and it straight up lies and is too nice, probably because its coded to not diminish self esteem, but yeah dont use gpt for rating
 
  • +1
Reactions: Divineincel, rraymond, trashbinxoxo and 3 others
Short answer: yeah, kinda — and it’s mostly on purpose.
Longer answer (no corporate fluff, promise):
ChatGPT is trained to avoid directly judging real people, especially with stuff like attractiveness, worth, “rate me,” etc. Numeric ratings are blunt, easily hurtful, and often context-free, so the model defaults to safer, vague feedback. That’s why you get the “surface-level bullshit.”

Workarounds that actually work​

You don’t need to trick it — you just need to reframe the ask:
1. Ask for analysis, not a score
Instead of:
“Rate me 1–10”
Try:
“What features stand out about my appearance, and how are they usually perceived?”
or
“What would people likely find attractive or unattractive here?”
2. Make it hypothetical or third-person
“If a fictional character looked like this, how would they be perceived?”
“How would different groups probably rate someone with these traits?”
This sidesteps the “judging you” issue.
3. Ask for comparison, not judgment
“What kind of aesthetic does this fit into?”
“What celebrities or archetypes does this resemble?”
4. Ask for improvement advice
This one almost always works:
“What would maximize attractiveness here — grooming, style, posture, etc.?”
5. Be explicit about consent + intent
“I’m not looking for validation, just blunt feedback. You can be direct.”
That often flips the tone from HR-approved to actually useful.

Why it won’t just give a number​

A number:
  • adds zero information
  • locks in a value judgment
  • is easy to screenshot and spiral over
So the model avoids it by design.
If you want, paste exactly the kind of prompt you’re using and I can rewrite it into something that gets real, concrete feedback instead of the dodge.
 
  • +1
Reactions: rraymond and Zeuxx
I told it to rate me brutally based on overall attractiveness and it said im a 7/10-7.5/10 and when I asked it to rate me more it started to refuse
 
I told it to rate me brutally based on overall attractiveness and it said im a 7/10-7.5/10 and when I asked it to rate me more it started to refuse

That’s the rating Foids give too for “go away creep”
 
  • +1
Reactions: LTNUser and Deleted member 328836
Chat gpt said I’m 3

Brutal, but do you even know what that means?

30th percentile?
3/5?
3/10?
3/3?
2 standard deviations below average?
A complete vibe rating with no comparison?
 
  • +1
Reactions: rraymond
Brutal, but do you even know what that means?

30th percentile?
3/5?
3/10?
3/3?
2 standard deviations below average?
A complete vibe rating with no comparison?
Nah I’m actually like 2/10 u can see me tho I’ve posted
 
  • So Sad
Reactions: rraymond and LTNUser
  • +1
Reactions: rraymond
Yes
Jewpt sucks
 

Similar threads

bloodpact
Replies
6
Views
54
Nyrо
Nyrо
𝔻𝕠𝕨𝕟𝕨𝕒𝕣𝕕𝕘𝕣𝕠𝕨𝕥𝕙𝕔𝕖𝕝
Replies
3
Views
49
Azi.org
Azi.org
lockhead
Replies
3
Views
33
BeanCelll
BeanCelll
Nyrо
Replies
18
Views
129
jeremyy
jeremyy
alurmo
Replies
24
Views
176
Jason5000
Jason5000

Users who are viewing this thread

Back
Top