Really weird, I've been able to get it to be quite consistent and replicate the outputs most others on X are reporting. Such as Obama's score of 75 out of 100. Also I've prompted it two ways. First time was, "what is my X score?" The other was "what is my user score?" Or "what is Obama's user score?"

It also gave me this breakdown of how the user score is calculated:

User scores on X are calculated based on a variety of metrics and qualitative factors, though the exact formula isn't officially disclosed. Here's a general breakdown based on posts found on X and general web insights:

Engagement Metrics: Likes, retweets, comments, and bookmarks contribute to a user's score. Each type of interaction might have different weights. For instance, likes might have a lower impact than retweets or comments.

Content Quality and Nature: The tone and content of posts are considered. Positive, constructive, or engaging content can boost a score, while negativity might reduce it. This aligns with X's recent algorithm changes focusing on "no negativity" and "unregretted user seconds."

Behavior Score: This could involve the user's behavior on the platform, including consistency, the frequency of posts, and how they interact with others. Positive interactions and a consistent presence might improve a score.

Violation Penalty: Violations of platform rules or community standards can negatively impact a user's score. This includes spam, misinformation, or any behavior leading to account actions like suspensions.

Follower Engagement: The engagement rate with followers, including how followers perceive and interact with the user's content, plays a role. A high engagement rate from a large follower base can significantly increase a score.

Unregretted User Seconds: This newer metric measures how long users spend engaging with content without regretting their time spent, which could be a factor in determining content quality and user score.

User Invites: Some platforms incentivize user growth by considering the number of invites or new users a person brings to the platform.

Here's a speculative formula, based on user discussions on X:

X_Score = (Engagement_Score + Follower_Score + Behavior_Score - Violation_Penalty) / Total_Possible_Score * 100

Where: Engagement_Score could be derived from likes, comments, retweets, etc., with each having different multipliers.

Follower_Score might reflect the quality and engagement of the follower base.

Behavior_Score could include consistency, frequency, and positive interactions.

Violation_Penalty would subtract points for any negative behavior or violations.

This explanation is based on community insights and speculative discussions rather than official documentation, as the exact scoring method isn't publicly detailed by X. Scores can range widely, from negative values to over 100, reflecting both the complexity and the subjective nature of how they're calculated.