System Logic: The Dial v1.1

I. Input Controls & Persistence

The Dial functions as a direct execution environment. We do not use predictive algorithms; we execute the specific parameters defined by your console coordinates.

Immediate Recalibration: Any adjustment to a knob or slider triggers an asynchronous re-calculation of your timeline.
Persistence: Settings are stored in your Operator Profile and persist across sessions.

II. Categorical Filters

Content selection begins with categorical inclusion/exclusion.

Strict Filtering: If categories are selected, the algorithm pulls only posts tagged within that set.
Uncategorized Fallback: Posts without tags are hidden when active filters exist, but visible in "Global Stream" mode (no categories selected).

III. Quantitative Ranges

We replace opaque algorithms with User-Rated Metadata. Also, binary "Likes" were replaced with high-resolution rates. Every post maintains a floating-point average for four core parameters on a \([-5, +5]\) scale. These averages are displayed on the top of each post card. You can select a range for each of the four parameters, as follows:

Approval: Weighted sentiment from Strong Dislike (-5.0) to Absolute Love (+5.0).
Politics: Perceived leaning from Far Left (-5.0) to Far Right (+5.o). Posts labeled as apolitical by the publisher will not have political rates.
Humor: Scale from Serious (-5.0) to Satirical (+5.0).
Veracity: Verification score from Completelely Untrue (-5.0) to Absolutely Truthful (+5.0).

IV. The Weighting Engine (Standard)

To prevent "brigading" (coordinated bias), the weight of a political leaning vote, for instance (\(W_p\)) is inversely proportional to the political distance between the voter and the post. This is scaled by the user's Political Leaning Rate of that post and the Self Identification of that same user. An additional variable (\(P\)) is set to zero in case the post is labeled as "apolitical".

W_p = 11 - (P \times |SelfIdentification - PoliticalLeaningRate|)

The final average for Approval, Politics, and Humor is calculated as:

{Average_{a,p,h}} = \frac{\sum (\text{Rate} \times W_{a,p,h})}{\sum W_{a,p,h}}

V. Veracity & Credibility Logic

Veracity uses a dual-weight system that rewards cross-partisan consensus and historical user credibility. The credibility is built over time, based on the veracity rate and number of posts of a user.

Political Weight (\(W_{v,p}\))

W_{v,p} = 6 + P \times (5 - |SelfIdentification - PoliticalLeaningRate|) \times \text{Polarity}

*Polarity is (1) for negative ratings and (-1) for positive ratings.

Credibility Weight (\(W_{v,c}\))

W_{v,c} = 1 + \text{Credibility}

Credibility Calculation

Credibility is calculated based on users (U) having at least 3 rated posts for veracity. A \(longevity\) index is determined, as a means to have credibility reach its maximum potential only after a good number of veracity rates (\(N_{veracityrates}\)) of posts. The variable \(z\) is set to 100. Then, the \(credibility_u\) is calculated, based on the user's veracity averages of published posts (\(\overline{V_u}\)), the average difference of the rates this user gave to other posts' averages (\(\overline{DV_u}\)) and the \(longevity_u\).

z = 100$$ $$x_u = N_{veracityrates,u}$$ $$longevity_u = \frac{(x_u + 1)}{(x_u + z)}$$ $$credibility_u = (\overline{V_u} * 3 + \overline{DV_u} / 4) * longevity_u

VI. Priority Sorting

The Following knob controls the priority for how other users are connected to you. It functions as a filter, but with a different concept, as it is not determined by range, but by connection proximity to you. The results will change immediately as you follow/unfollow other users.

9-10: Only users you follow. The algotithm selects only posts published by this set of users.
7-9: Users followed by the users you follow. The algorithm creates a new set of users, based on this criteria. If a user you follow changes who they are following, that takes effect immediately as well.
5-7: All of the above + users who liked the posts from users you follow. The algorithm looks only at approval rates greater than zero for that selection.
3-5: All of the above + users who liked the posts from users followed by who you follow. Same as above, for a larger set of users.
0-3: No restrictions. The algorithm by-passes this filter and all posts from the criteria match are selected

For the other knobs, Location (\(S_L\)), Engagement (\(S_E\)), and Time (\(S_T\)) are normalized to a range of 0 to 10 based on the posts filtered by the ranges selected. All four sorting parameteres are calculated and assigned to each post. The sum of these parameteres is then used to establish the sorting order of the timeline.

S_L = (2.0 \times \text{SameCountry}) \times (1.5 \times \text{SameState} + 1) \times (1.5 \times \text{SameCity} + 1) + 1.25$$ $$S_E = (1.0 \times \text{Approvals}) + (10.0 \times \text{Comments}) + (0.1 \times \text{Followers})$$ $$S_T = (\text{CurrentTime} - \text{PostTime})

CORE_SPEC_v1.1