Connection Analysis
Last revised: 2026-05-08 — see history. 2026-05-08 revisions: (1) Phase 1 cleanup expanded acronyms on first use (IPAC, KCNSC, PSFC, SBG-VSWIR, AMR, HIFI). (2) Neutrality re-pass replaced one verdict-style summary clause ("are within plausible base rates before controlling for other factors") with neutral framing of the unanswered base-rate question.
This document maps verifiable connections between the 11 cases. Connections are organized into three layers of decreasing evidentiary confidence. Every connection carries a confidence rating and evidence pointer. Connections are facts; whether they indicate causation is evaluated separately in the Hypothesis Evaluation.
Tight Layer -- Documented, Verifiable Overlaps
These connections are confirmed by primary or credible secondary sources.
Institutional Co-Employment
| Connection | Individuals | Evidence | Confidence |
|---|---|---|---|
| LANL (current or retired employees) | Chavez (retired 2017), Casias (active admin asst.) | NM DPS record (T1), ABQ Journal (T3) | High |
| JPL (current or recent employees) | Reza (Director, Materials Processing), Hicks (research scientist 1998-2022), Maiwald (principal researcher ~1999-2024) | LASD record (T1), Forest Lawn obituary (T1), Legacy.com obituary (T1) | High |
| Caltech system (Infrared Processing and Analysis Center (IPAC) is Caltech-operated) | Grillmair (IPAC, 1997-2026) | Caltech memorial (T1) | High -- but IPAC is institutionally distinct from JPL despite both being Caltech-managed |
What this means: Two institutional clusters exist. Three of 11 subjects worked at JPL; two at LANL. Grillmair is Caltech-adjacent but not JPL. JPL employs ~6,000 people and LANL ~14,000. Three deaths across JPL over a 2023-2025 window and two disappearances across LANL's current/former workforce of tens of thousands have not been formally evaluated against actuarial base rates; whether the observed counts exceed expectation is an open question. (updated 2026-05-08 — see history)
Geographic Clustering
| Cluster | Individuals | Geography | Dates |
|---|---|---|---|
| New Mexico | Chavez (Los Alamos), Casias (Taos County), Garcia (Albuquerque), McCasland (Albuquerque) | Within ~130 miles of each other in northern NM (LANL, Sandia, Kirtland AFB, Kansas City National Security Campus (KCNSC) NM operations) | May 2025 - Feb 2026 |
| LA County / Angeles NF | Reza (Angeles NF), Hicks (Sunland), Maiwald (Los Angeles), Grillmair (Llano) | Within ~60 miles in LA County area | July 2023 - Feb 2026 |
| Massachusetts | Loureiro (Brookline), Thomas (Wakefield) | ~15 miles apart in metro Boston | Dec 2025 |
Confidence: High that the geographic clustering exists. Moderate that it is meaningful beyond reflecting where defense-related institutions are located. New Mexico hosts LANL, Sandia, Kirtland AFB, and KCNSC. LA County hosts JPL, Caltech/IPAC, and major aerospace contractors. Massachusetts hosts MIT and numerous defense-adjacent research institutions. [T1/T3/T4 sources confirm all locations.]
Temporal Clustering
| Period | Cases | Note |
|---|---|---|
| June 2022 | Eskridge | Outlier -- 3 years before main cluster |
| July 2023 | Hicks | Pre-cluster death |
| July 2024 | Maiwald | Pre-cluster death |
| May-Aug 2025 | Chavez (May), Casias (June), Reza (June), Garcia (Aug) | Four disappearances in four months |
| Dec 2025 - Feb 2026 | Loureiro (Dec), Thomas (Dec), McCasland (Feb), Grillmair (Feb) | Four deaths/disappearances in three months |
Confidence: High that the temporal clustering is real -- 8 of 11 events occurred within a 10-month window (May 2025 to Feb 2026). Moderate that this is meaningful beyond media attention effects. The two earlier deaths (Hicks 2023, Maiwald 2024) were only linked to the cluster retroactively in April 2026. Eskridge (2022) was added even later. The question is whether an objective observer in real-time would have noticed a cluster, or whether post-hoc grouping created the appearance of one.
Documented Shared Programs or Projects
| Connection | Individuals | Evidence | Confidence |
|---|---|---|---|
| None documented | -- | No two subjects have been shown to work on the same specific program, project, or contract | High confidence in the absence |
This is significant. Despite institutional co-employment at JPL and LANL, no source has documented that any two subjects worked together, shared a lab, participated in the same program, or had a professional relationship. The one exception is a speculative claim about a Reza-McCasland professional connection (see Medium Layer below).
Medium Layer -- Research-Domain Adjacencies and Behavioral Parallels
These connections require inference beyond documented facts but are grounded in reported information.
Propulsion / Advanced Materials Domain
| Individual | Relevance | Confidence |
|---|---|---|
| Reza | Co-inventor of Mondaloy superalloy for rocket engine components; career at Aerojet Rocketdyne and JPL Materials Processing | High -- patent record (T1) |
| McCasland | Commanded AFRL (aerospace research including propulsion); career included space-based laser programs, GPS, and classified special programs | High -- official biography (T1) |
| Maiwald | JPL principal researcher on space instrumentation (Surface Biology and Geology — Visible to Shortwave Infrared (SBG-VSWIR); Advanced Microwave Radiometer (AMR); Heterodyne Instrument for the Far Infrared (HIFI) on Herschel); THz technology | Moderate -- instruments are adjacent to but not directly propulsion |
| Eskridge | Private gravity-modification research with NASA-retired father | Low -- private org, no peer-reviewed results, no government funding documented |
Assessment: Reza and McCasland are the strongest pair for domain overlap -- both worked in aerospace materials/propulsion at institutions connected to the AFRL ecosystem. WION (Indian media) flagged a "Mondaloy connection" between them. However, no source has documented that they knew each other or worked on the same program. McCasland's AFRL command (2011-2013) and Reza's Aerojet Rocketdyne career (1988-~2024) could have overlapped through Air Force contracting, but this is inference, not documented fact. Confidence in documented connection: Low. Confidence in domain adjacency: Moderate. [T4 -- WION; T7 -- independent media]
Fusion / Plasma Physics Domain
| Individual | Relevance | Confidence |
|---|---|---|
| Loureiro | Director, MIT Plasma Science and Fusion Center (PSFC); DOE/ARPA-E funded fusion research; SPARC tokamak partnership | High -- MIT News (T1) |
| Eskridge | Father (Richard Eskridge) was NASA MSFC plasma physicist | Low -- father's career, not Amy's; private org |
Assessment: Loureiro is the only subject with direct, confirmed work in fusion/plasma physics. His case has a named suspect with ballistic evidence, video confession, and documented personal motive. The fusion connection, while real, is not the explanatory factor for his death. Confidence that fusion research was a targeting motive: Very low. [T1 -- DOJ, MIT]
Nuclear Facility Workforce
| Individual | Relevance | Confidence |
|---|---|---|
| Chavez | Retired LANL employee; specific role unknown | Moderate -- LANL employment confirmed (T3), role unknown |
| Casias | LANL admin assistant; family disputes high clearance | Moderate -- LANL employment confirmed (T3), admin role |
| Garcia | Alleged KCNSC contractor; single anonymous source | Low -- T6 single source, unverified |
Assessment: Chavez and Casias share an employer (LANL) but had different roles and disappeared under different circumstances in different locations (Los Alamos vs. Taos County). Whether either had access to sensitive information is unknown for Chavez and actively disputed for Casias. Garcia's nuclear-facility connection depends entirely on an unverified T6 claim. Confidence in nuclear-facility targeting pattern: Low. [T1/T3 for LANL employment; T6 for KCNSC]
Pattern-of-Life Similarity: Left Without Belongings
| Individual | Items Left Behind | Source |
|---|---|---|
| Chavez | Wallet, keys, phone, cigarettes, car | T3 (Los Alamos Reporter) |
| Casias | Purse, wallet, cash, IDs, both phones, car | T3 (ABQ Journal, Santa Fe New Mexican) |
| Garcia | Phone, wallet, keys, car (carried firearm) | T4 (multiple outlets citing APD) |
| McCasland | Phone, prescription glasses, wearable devices (took wallet, revolver, boots, backpack) | T4 (multiple outlets citing BCSO) |
| Reza | Did not drive herself; unclear what she carried besides backpack with water | T4 (Yahoo/Men's Journal) |
| Thomas | Phone, wallet | T3 (Boston 25 News) |
Assessment: Leaving behind a phone and/or wallet is reported in 5-6 of 11 cases, which is striking on its face. However, this pattern is also common in cases of voluntary disappearance, suicidal crisis, and disorientation. McCasland is a partial exception -- he took his wallet and revolver but left his phone and glasses. Thomas had experienced severe personal grief days before. The behavioral pattern is real but not diagnostic of any particular cause. Confidence in pattern existence: High. Confidence in pattern significance beyond personal-crisis behavior: Low. [T3/T4 sources for each case]
Factory-Reset / Digital Anomalies
| Individual | Detail | Source | Confidence |
|---|---|---|---|
| Casias | Both phones (personal + work) factory-reset; found at home | T3 (family-sourced, not confirmed by NMSP) | Moderate -- single-family-source claim |
| Reza | Cell phone forensic data obtained but never released; SAR removed Facebook post about it | T4 (preserved via Reddit) | Moderate -- confirmed data was obtained; results unknown |
Assessment: These are the two most operationally anomalous digital details in the dataset. Phone-wiping before a disappearance is unusual in personal-crisis scenarios and more consistent with either counter-surveillance behavior or third-party action. However, the Casias phone-reset claim comes from a single family member and is unconfirmed by law enforcement. The Reza cell data suppression could reflect either investigative confidentiality or something more concerning. Confidence in anomaly: Moderate. Confidence in meaning: Low -- insufficient data. [T3 family source; T4/T7 Sentinel Briefing]
Corkboard Layer -- Speculative Connections
These connections are drawn from social media narratives, independent media, UAP-community theorizing, or inferences that go substantially beyond documented evidence. They are included for completeness and to make the speculative reasoning transparent, not because they are supported.
UAP / Disclosure Adjacency
| Individual | Claimed Connection | Evidence Quality | Confidence |
|---|---|---|---|
| McCasland | Tom DeLonge email to Podesta names him as adviser; wife confirms limited unpaid consulting; commanded AFRL at Wright-Patterson | Moderate -- email exists (T1 WikiLeaks); wife partially confirms (T4); DeLonge's characterization exceeds wife's | Connection to DeLonge: Confirmed. Depth of UAP knowledge: Speculated |
| Eskridge | Co-founded Institute for Exotic Science; claimed antigravity discoveries; father was NASA plasma physicist | Low -- self-reported claims on YouTube; no peer-reviewed results; father says "nothing suspicious" about death | Speculated |
| Reza | AFRL collaboration on Mondaloy in 1999; Wright-Patterson lineage | Very Low -- AFRL contract is routine aerospace work, not UAP-related | Speculated |
Assessment: McCasland is the only subject with a documented (not merely alleged) connection to the UAP disclosure community, and even that connection is contested in scope between DeLonge's characterization and his wife's. Eskridge made dramatic claims but produced no verified results. Extending a UAP connection to Reza (because she once worked with AFRL) or to others is guilt-by-institutional-association. Rep. Burlison's on-record statement that McCasland "has a lot of information" about UAP adds political weight but not independent evidence. Confidence in UAP pattern: Very low for the group; moderate for McCasland individually. [T1 WikiLeaks; T4 Newsweek, NewsNation; T7 independent media]
Foreign Intelligence Targeting Theory
See Foreign Intelligence for full evaluation. Summary: Former FBI officials (Swecker) have raised espionage as a possibility. Nuclear security experts (Rodgers/CSIS, Roecker/NTI) have pushed back, noting the strategic futility of targeting individual scientists in a workforce of thousands. No law enforcement or intelligence agency has publicly stated that foreign involvement is suspected in any specific case. Confidence: Very low based on current evidence. [T4 -- CBS News, Fox News, Newsweek]
"Silencing" Narratives
Social media and independent investigators (Greer, some Reddit communities) have proposed that scientists are being silenced to protect classified programs. This hypothesis (evaluated as H7 in the Hypothesis Evaluation) has no documented evidentiary support. None of the subjects are documented as whistleblowers, IGIC complainants, or participants in Congressional disclosure processes. Eskridge made pre-death claims of threats, but her father -- a retired NASA engineer who worked alongside her -- publicly dismissed conspiracy theories. Confidence: Very low. [T5/T6/T7 sources]
Find a Grave Anomaly (Reza)
A memorial for Reza was created on Find a Grave on June 26, 2025 (four days after disappearance), listing her death date as June 22 and describing a "green burial," while search operations were still active and no remains had been found. The memorial was removed in March 2026 after media reporting. Creator identified only as "lillian." This is genuinely anomalous but could reflect premature assumption of death by an acquaintance, memorial-site trolling, or something more concerning. Confidence in anomaly: High. Confidence in meaning: Low -- insufficient data. [T7 -- Sentinel Briefing]
Cross-Cutting Assessment
What the evidence supports:
- Geographic and institutional clustering is real but largely explainable by where defense/aerospace institutions are located.
- The temporal clustering in 2025-2026 is real and is the strongest objective feature of the dataset.
- Behavioral parallels (leaving without belongings) exist in multiple cases but are common in personal-crisis disappearances.
- Two cases (Grillmair, Loureiro) have named suspects with strong non-conspiracy explanations supported by ballistic evidence and, in one case, a video confession.
- One case (Thomas) involves a pharmaceutical researcher with no defense connection whose disappearance followed severe personal grief, with no foul play suspected by DA.
What the evidence does not support:
- No documented professional connection between any two subjects (no shared project, lab, or confirmed personal acquaintance).
- No common operational signature across cases (methods range from shooting to hiking disappearance to walking away from home).
- No law enforcement or intelligence agency has identified a connection between any cases.
- No physical evidence of third-party involvement has been publicly disclosed in any unsolved case.
Cases ranked by "fit" for a targeted pattern (strongest to weakest):
- McCasland -- Highest-value target by career and clearance level; unexplained disappearance; strong defense/SAP connections
- Reza -- High-value technical knowledge (superalloy patents); anomalous Find a Grave memorial; unexplained disappearance
- Chavez -- LANL retirement; unexplained disappearance; but role unknown and retired 8 years
- Casias -- LANL employment; phone-wiping anomaly; but admin role with disputed clearance level
- Garcia -- If KCNSC claim is true, moderate fit; but employment claim is unverified T6 single source
- Hicks -- JPL career on defense-adjacent missions; but cause of death undisclosed (not necessarily suspicious)
- Maiwald -- JPL career; cause of death undisclosed; but largely unclassified work
- Loureiro -- DOE-funded fusion director; but named suspect with video confession and personal motive
- Grillmair -- Named suspect; escalating local criminal; unclassified civilian research
- Thomas -- Pharmaceutical researcher; no defense connection; personal grief context; no foul play suspected
- Eskridge -- Private org; 3-year temporal outlier; ruled suicide; father rejects conspiracy