Frost Log — July 11, 2026 (Inaugural Edition)

Executive Summary

Feeds went live at 18:01 PT July 10 and within sixteen minutes we had nominations: Week-1 HOH Dee put up Mallory, Taylor, and Yash, with Mallory as the stated target (confirmed privately to Kamu). The veto field drew Barrett and Melody alongside the HOH and nominees; the comp began at 13:58 PT July 11 and feeds were still down at window close — outcome unknown, so everything past the ceremony is written in branches. The structural story of the window is not the nominations; it's the Angela–Barrett–Devens cluster, the only group observed doing organizational work (regrouping talk, covert-signal design at 11:16 PT). The tactical story is Dee's contradictory contingency planning — Melody named as backup replacement to Kamu, Barrett floated to Taylor — which is either strategic ambiguity or the first entry in a conflict-avoidant signature file. The unforced error of the window belongs to Mallory, who handed her full retaliation list (Dee, Kamu, Haley) to Melody, a veto player who had already privately flagged Mallory as untrustworthy under pressure. And the negative space belongs to Chuk: named repeatedly as socially central, zero logged conversations. Silence is data — with alternatives noted below.

The Cold Read

Read 1: Dee is running an appeasement HOH, and the house will figure it out before she does. Look at the shape of her window: nominations locked almost immediately after her full entry into the game, a "one nominee per social group" rationale offered to Taylor (the diplomatic HOH's favorite fiction — it's not personal, it's arithmetic), a shmooze circuit of all three nominees, a noncommittal "yeah, yeah" when Angela declared trust, and — the tell — two different replacement-nominee names given to two different audiences. Melody to Kamu, in the HOH room, in an actual planning conversation. Barrett to Taylor, in a reassurance context, to a nominee she was soothing. I'm treating the Kamu version as the real plan (Grade B) and the Barrett version as a comfort object (people rarely lie to their sounding board and tell the truth to the person they nominated). But here's the leak: Taylor now has the Barrett name, Taylor talks to Jason and Latrice, and Barrett is actively positioning himself as a "swing vote or sleeper." If the Barrett line travels — and Taylor's face is, per Latrice, LOUD — Dee has manufactured a Week 2 enemy out of a Week 1 pleasantry. This is the classic cost of tell-them-what-they-want-to-hear HOHing: every soothing lie is a scheduled detonation.

Read 2: The Angela–Barrett–Devens cluster is the only group playing Game 2. Everyone else this window was doing survival triage or vibes. These three were observed designing infrastructure — regroup discussions, discreet signaling systems. That's electorate-design behavior in week one. Caveat stamped in red: the "Cookout-style" framing and Drew's plus-one expansion come from the narrative summary only, no timestamped scene (Grade C at best), and invoking the Cookout comparison is doing them a huge favor they haven't earned — that alliance held because of mission-grade defection cost; this one, so far, is three people who like each other and a hand signal. Rated defection cost: low, repeated-game trust only. Watch whether they add binding sources or stay a vibe.

Read 3: Mallory is campaigning on the human channel and detonating on the game channel simultaneously. The human channel is working — Angela is in full protector mode, promised veto use, and reports the house is sympathizing rather than judging. That's real; crying campaigns win votes. But at 22:09 she told Melody her retaliation nomination list by name: Dee, Kamu, Haley. Melody had, fifteen hours later, told Mallory to her face she can't trust her to "keep it together." A player who distrusts you and holds your kill list is not a friend; she's a courier. I expect that list to reach Dee or Kamu within the week, at which point Mallory converts from "sympathetic pawn" to "confirmed threat if she stays" — the single worst ledger transition available to a nominee.

Read 4: What the house hasn't noticed — Chuk generated zero content. Week-1 safe, identified by the Ashley/Melody recon as tight with Haley and by extension Dee, and not one logged conversation in a 40-hour window. I ship the alternatives: (a) deliberate low-visibility positioning enabled by safety, (b) cameras chasing the nomination drama, (c) genuinely passive. No conclusion yet — but if the pattern holds another window, the ledger opens a position-without-visibility file, and those are the files that end 6–1.

Honorable mention: Ashley and Melody's 21:50 storage-room session was the first genuine transmission-map work of the season — they correctly triangulated Haley–Dee tightness and flagged Jason as an unmapped node. Meanwhile Jason, who spent the window telling Taylor they need to "teach Dee how the game works," doesn't register as a strategic actor to that bloc at all. Two live siloes already: information is not flowing across the Taylor/Jason axis and the Ashley/Melody axis. In week one, that's the whole map.

Detailed Timeline

July 10

July 11

Alliance Standings

Transmission-map framing. Declared alliances are a light overlay; edges and leak lines are the real chart. Everything here is week-one provisional.

1. Angela–Barrett–Devens (+Drew, unconfirmed) — "the Signals Cluster." Solidity: highest in the house, which in week one means "moderate." Only group with observed infrastructure work (11:16 PT). Defection-cost rating: low — repeated-game trust only, no mission, no exposure, no showmance binding it. Leak risk: Barrett is simultaneously telling Taylor he's a swing vote — he's running a second legible game, which is either good decoy discipline or a genuine hedge. Drew's membership is Grade C, secondhand only.

2. Dee–Kamu — the HOH axis. Functional strategy pairing, one week old, born of power. Kamu got the real replacement plan; nobody else did. Edge attributes: high fidelity, fast, strategy-class information. Durability beyond Dee's reign: unknown.

3. Taylor–Jason–Latrice — the support pod. Emotional-support edges with strategy content on the Taylor–Jason line. Latrice's "Rome and Jason have your back" claim implies a Taylor–Jason–Rome–Latrice voting bloc that has not been observed as such — Grade C.

4. Lyric–Rome — showmance, covert. Cover story pre-built. Leak coefficient on this secret: unknown, but Jason's flirt-adjacent needling (19:45) suggests the house is already sniffing.

5. Reported F2s: Chuk/Haley, Drew/Melody. Single-source, zero corroborating scenes. Logged, not credited. If Drew/Melody is real, note Melody is also the recon partner of Ashley and the courier holding Mallory's list — she'd be one of the best-positioned information nodes in the house.

Conspicuous non-relationships:

Player Status

Prediction Ledger

No prior predictions to grade — inaugural edition. Naive baselines now on the clock: "stated house target goes home," "noms stand," "random veto-pool member wins (1-in-6)."

# Class Prediction P Grade Basis
P1 comp-branch Veto outcome — no edge claimed. Baseline 1/6 per player; ~50% a nominee wins (3 of 6). Branches: (A) nominee wins → self-save → replacement ceremony; (B) Dee/Barrett/Melody wins → veto unused, noms stand (Dee/Kamu agreed plan) A: ~50% / B: ~50% B (branch structure), n/a (winner) Class-4 ceiling — no comp-type info, no point estimate offered
P2 nom If a replacement is needed (Branch A, Melody not the winner), Dee names Melody, not Barrett 60% B Dee pattern, twice observed: private plan to sounding board (Kamu) outranks reassurance line to nominee (Taylor)
P3 vote If Mallory is on the block at eviction, Mallory is evicted 65% B Dee/Kamu target alignment (Grade A convo) + no counter-coalition observed; capped low — week one, zero vote history, Angela sympathy channel live
P4 flip Mallory's retaliation list (Dee/Kamu/Haley) reaches Dee or Kamu by July 18 60% C Melody's stated distrust of Mallory (12:47) + Melody's active recon behavior; single courier path
P5 nom-adjacent The Barrett-replacement breadcrumb travels from Taylor to a third party by July 15 55% C Taylor is an open book ("face is LOUD" — one scene, but a loud one) + her three active support edges
P6 flip/social Lyric–Rome showmance becomes house-known (discussed by ≥2 non-members) by July 18 65% B Cover story already deployed = already leaking; Jason already probing the perimeter
P7 structure Angela/Barrett/Devens formalize (a name, or ≥5 members) within one week 55% C 11:16 infrastructure conversation; discounted for single-source "Cookout-style" framing

The Counter-Read

The lead read — Dee as appeasement HOH scattering contradictory promises — has a strong alternative: Dee isn't managing anyone; she's improvising in real time, and the contradictions are indecision, not diplomacy. Consider her actual constraints: she entered the game as a full player roughly seventy minutes before nominations were locked, as the 17th houseguest, with the thinnest social map in the house. Under that reading, "Melody to Kamu, Barrett to Taylor" isn't a soothing lie — it's two genuine snapshots of a plan that changed overnight, because she has no settled model of the house yet. The "one nominee per group" rationale isn't diplomatic fiction; it's the honest heuristic of someone nominating from a org chart because she hasn't met the org. If this is right, the signature file I opened ("soothes the listener in front of her") is premature on two data points, and the correct forecast is not "scheduled detonation when the lies cross" but "Dee's week-2 behavior will be uncorrelated with week-1 behavior once she has real information." The test that distinguishes the reads: watch the veto ceremony. If a replacement is needed and Dee agonizes visibly, canvasses widely, and lands wherever the last conversation pointed — that's improvisation. If she executes the Kamu plan cleanly while telling Taylor a different story right up to the ceremony — that's appeasement-as-strategy, and the file stands.

Retrospective

Nothing to grade — this is the inaugural edition. No prior predictions exist; no ex ante branch tables to preserve; no narrative leakage possible yet. Taran's daily read for this window hadn't landed in our coverage by press time, so no peer comparison this edition — that column opens the moment his coverage enters the record.

What this section will do going forward, stated now so it's binding: every prediction in today's ledger (P1–P7) is frozen as written, with its probability, class, and grade. Tomorrow's edition scores whatever resolves — starting with the veto outcome sitting behind the current feed blackout — on the process/outcome grid, against the naive baselines published above. If P2 hits because Dee flipped a coin, that's process-unknown/outcome-right and gets said plainly. The log grades itself first.

One housekeeping note for the record: the pre-window Rachel Reilly departure was a pre-planned production stunt — she was never a player, and her exit carries no equity implications for anyone in the house. It does not enter the twist-accounting ledger and will not be treated as evidence about production's posture toward actual players.


Coverage: July 10, 00:00 PT → July 11, 16:36 PT. Regime call: this is a Week-1 chaos regime — no stable structure, no vote history, and every behavioral read below rests on one or two observed scenes. Every confidence number below is ceiling-capped accordingly. Horizons are short on purpose. Claims and predictions carry evidence grades: A said privately and backed by behavior; B repeated behavior or multiple independent conversations; C a single statement, possibly performed for an audience; D inference from silence, body language, or the edit.