How many people will be shot, injured or killed, in the United States between 1 July and 10 July 2026, according to Gun Violence Archive?

• My most likely prediction for the total number of people shot, injured, or killed between July 1 and July 10, 2026, is 1,215.00, with a 50% chance of falling between 1,095.00 and 1,345.00.
• Historical data for the first week of July shows a significant decline from 1,708 casualties in 2020 to 1,096 casualties in 2024, establishing a downward baseline for the holiday period.
• National crime trends in 2025 and early 2026 show a historic reduction in violence, with homicides in major cities dropping by 17.7% in the first quarter of 2026.
• The Independence Day holiday (July 4-5) remains a period of high volatility, with GVA historically recording more mass shootings on these two days than any other time of the year.
• The 10-day window (July 1-10) is expected to be approximately 15-20% lower than the equivalent period in 2024/2025 due to the sustained decline in overall firearm incidents.
• The resolution depends on the GVA snapshot as of August 1, 2026, which allows approximately three weeks for late-reported incidents and status updates to be verified and included.

Key figures

Historical context

The Gun Violence Archive (GVA) has tracked U.S. gun violence since 2013, consistently documenting a significant ‘holiday spike’ during the first week of July. Historically, July 4 and July 5 are the most violent days of the year for mass shootings. For example, in July 2023, there were 78 mass shootings in the month, with a heavy concentration in the first week, including high-profile incidents in Baltimore (28 injured) and Philadelphia (5 killed). The year 2020 represented a modern peak in holiday violence, with over 1,700 casualties in the first seven days of July, largely attributed to pandemic-related social stressors. Since 2021, however, a steady downward trend has emerged. By 2024, the July 1-7 casualty count had fallen to 1,096, the lowest in recent years. This aligns with broader FBI data showing that after the 2020-2021 surge, violent crime in the U.S. has been receding toward, and in some categories falling below, pre-pandemic levels. The 2026 forecast is set against the backdrop of 2025, which saw the largest one-year percentage drop in homicides since the 1930s.

Tailwinds

• Independence Day 2026 is the U.S. 250th anniversary (Semiquincentennial), likely leading to larger crowds, higher demand for fireworks (potentially masking gunfire), and more public gatherings than usual.
• Historical ‘Summer Spikes’: Despite Q1 declines, violence historically rises as temperatures increase; the 2026 summer could see a reversion to mean levels if prevention programs are stretched.
• The random nature of ‘Mass Shootings’: A single high-casualty event (e.g., 2022 Highland Park) can add 50+ casualties to the 10-day total instantly.
• Federal funding cuts to community violence intervention (CVI) programs reported in late 2025 could lead to a resurgence of violence in underserved urban areas by mid-2026.

Headwinds

• Sustained 2026 declines in major cities like Philadelphia (-54% in Q1) and New York City (-31.7% in Q1) could indicate a faster-than-expected nationwide drop.
• Enhanced ‘precision policing’ and federal violence intervention funding (CalVIP and others) reaching peak efficacy during the 2026 summer months.
• A relatively cool or rainy start to July 2026, as gun violence is historically correlated with high temperatures and outdoor gatherings.
• Potential under-reporting or delays in police data submission to the GVA by the August 1 snapshot deadline.

Detailed reasoning

My prediction for the total number of people shot, injured, or killed between July 1 and July 10, 2026, is based on an integration of historical seasonal baselines and strong evidence of a multi-year downward trend in U.S. gun violence.

Historical data from the Gun Violence Archive (GVA) for the first week of July (July 1–7) shows significant volatility: 1,177 casualties in 2019, 1,708 in 2020 (a pandemic-era peak), 1,581 in 2023, and a notable drop to 1,096 in 2024. To extrapolate these 7-day figures to the 10-day period required by the question (July 1–10), I add a baseline of three ‘ordinary’ summer days. Based on July 2025 data, where the daily average was approximately 113 casualties, adding three days to the 2024 baseline suggests a 10-day total of roughly 1,430 to 1,450 people.

However, the ‘Inside View’—current 2025 and 2026 data—strongly indicates that using a simple 2024 or 2025 baseline would overestimate the 2026 outcome. FBI preliminary data for 2025 showed an 18.1% drop in murders and a 9.3% drop in overall violent crime. More critically, GVA’s own full-year totals fell by approximately 17.8% between 2024 (48,424) and 2025 (39,791). This downward momentum has accelerated in early 2026. Reports from the Major Cities Chiefs Association (MCCA) indicate that homicides in 67 major U.S. agencies fell by 17.7% in the first quarter of 2026 compared to Q1 2025. Furthermore, firearm deaths in Q1 2026 reached their lowest level in 12 years.

In my reasoning process, I weighted the continued 2026 decline heavily (80-85% weight) but applied a slight ‘seasonality buffer.’ While annual homicides are dropping by nearly 18%, gun violence during the Independence Day holiday is often driven by specific events—mass shootings and celebratory gunfire—that may be less responsive to broad urban crime-reduction strategies. Since 2013, GVA has recorded a consistent spike on July 4 and 5 (averaging 6-7 mass shootings per day).

Therefore, my most likely estimate of 1,215 people represents a roughly 15-18% discount from the extrapolated 2024 baseline. The 50% High Density Interval (HDI) of 1,095 to 1,345 reflects the high probability that the 2026 decline will persist, while the wider 80% interval (985 to 1,465) accounts for the inherent randomness of mass casualty events that can significantly alter the aggregate total in a short 10-day window. I also factored in the GVA’s methodology; since the resolution uses a snapshot from August 1, 2026, any late-reported incidents or status updates (injuries turning into deaths) will be included, supporting a slightly higher central estimate than a ‘real-time’ count might suggest.

Key uncertainties

• The 250th Anniversary factor: It is unclear if the U.S. Semiquincentennial will lead to increased security that suppresses violence or larger ‘soft targets’ that increase casualty risks.
• Weather patterns: Severe heatwaves in early July 2026 would likely push the total toward the higher end of the HDI (1,345+).
• GVA Reporting Lag: The August 1 deadline is relatively firm; if major police departments experience reporting delays, the ‘official’ GVA count may be lower than the actual occurrences.
• Economic factors: If 2026 experiences a sharp economic downturn or further cuts to social services, the downward crime trend could reverse unexpectedly.

Conclusion

• The predicted casualty count represents a continued structural decline in U.S. firearm violence, potentially reaching the lowest holiday-period levels in over a decade.
• While the broader trend is downward, the high concentration of incidents around July 4 and July 5 makes this 10-day window highly volatile and sensitive to individual mass casualty events.
• Decision-makers should monitor the June 2026 crime data; if the downward trend stalls in early summer, the casualty count could shift toward the higher end of the 1,345-1,465 range.

Forecast and Reasoning

Forecast: Probability distribution:

• 1.00% chance of value below 780.0
• 5.00% chance of value below 910.0
• 10.00% chance of value below 1000.0
• 20.00% chance of value below 1100.0
• 30.00% chance of value below 1185.0
• 40.00% chance of value below 1235.0
• 50.00% chance of value below 1280.0
• 60.00% chance of value below 1335.0
• 70.00% chance of value below 1410.0
• 80.00% chance of value below 1505.0
• 90.00% chance of value below 1640.0
• 95.00% chance of value below 1795.0
• 99.00% chance of value below 2250.0

Forecast Reasoning

We need the GVA search-tool aggregate (Number Killed + Number Injured) for incident dates July 1–10, 2026, snapshot as-of Aug 1, 2026. Key drivers are (a) the underlying national daily volume of non-suicide shooting victims, (b) seasonality/holiday uplift around July 4, and (c) year-to-year randomness plus any GVA revision/backfill by the snapshot date.

Across the team, the main empirical anchor is the widely reported GVA “July 4 week” total for July 1–7, 2024: 1,096 (killed + wounded). A 10-day window should usually exceed that 7-day total by roughly “three more days” of near-baseline violence, but the correct level for 2026 depends on whether the post-2023 decline persists. Several analysts cite/assume a continued decline from 2024 to 2025 and into 2026, which would pull the 2026 July 1–10 total closer to the low-to-mid 1,000s, while still above a naive 10-day average because this window contains the Independence Day period.

I treat Forecaster 2’s distribution (median 1,500) as plausibly over-weighting holiday-week elevation from higher-violence years (e.g., 2023’s 1,581 for July 4 week) without fully discounting for the apparent 2024–2026 decline. Forecaster 1 looks somewhat low, likely under-weighting the holiday uplift and relying on at least one potentially mismatched “long weekend” figure. Forecasters 3 and 4 both center in the ~1,200–1,300 range while still allowing a substantial upper tail for an unusually violent holiday period.

Synthesis: I set the median at 1,280 (roughly consistent with “recent baseline 10-day total” plus a modest July 4 uplift), with moderately wide tails to reflect the large historical variability in early-July totals and the possibility of either an unusually quiet period, or a reversal/spike that pushes totals toward ~1,800–2,200.

An amateur forecast would often anchor on a simplistic 10×(annual daily average) (landing near ~1,000–1,200) and ignore the July 4 uplift, or else anchor on a single high year’s “July 4 week” and extrapolate linearly (often overshooting). My forecast explicitly blends (i) holiday-week empirical anchors across multiple years, (ii) the recent downward trend uncertainty, and (iii) the fact that the asked window is 10 days (not 7), producing a better-calibrated center with wider, more realistic tails than a single-number extrapolation.

Forecast rationale (numeric):

— Iteration 1 — Across the forecasts, the reasoning is broadly consistent:

• Baseline anchor: All models start from the 2026 year-to-date Gun Violence Archive pace of about 96.6 killed or injured per day.
• 10-day extrapolation: A simple unadjusted projection over 1–10 July gives a baseline near 966 people.
• Holiday/seasonal uplift: Each forecast then applies an upward adjustment for the Independence Day period, which historically sees higher gun violence.
  • One model uses roughly a 25–30% uplift.
  • Others use a larger 40–50% uplift based on July 1–10 historical patterns.
• Calendar effect: The fact that July 4, 2026 falls on a Saturday is treated as increasing the likelihood of a weekend/holiday spike.
• Historical comparison: One rationale explicitly cites recent July 1–10 totals in the 1,400–1,700 range, reinforcing an above-baseline expectation.

Consensus

The forecasts converge on a central estimate in the low-to-mid 1,000s, roughly 1,250–1,450 people, with a clear expectation that the period will exceed a straight 10-day baseline.

Main uncertainty sources

The models widen their ranges because of:

• Holiday volatility and weekend clustering
• Potential mass-casualty events
• Reporting completeness / delayed updates in the Gun Violence Archive
• General uncertainty in how closely historical July patterns will repeat in 2026

Key difference in emphasis

The main disagreement is not about direction, but about how large the July 4 uplift should be: one forecast is more conservative, while the others assume a stronger historical holiday effect.

— Iteration 2 — Overall, the forecasts converge on a midpoint around 1,500 people shot, injured, or killed in the U.S. between 1 July and 10 July 2026, with most of the variation coming from how strongly the models weight the July 4 holiday effect and how much they adjust for recent trends.

Main reasoning patterns

• Baseline GVA pace

  • All estimates start from the recent 2026 Gun Violence Archive pace, using late-May data to infer a daily national casualty rate.
  • That baseline is roughly 110–120 people per day, which implies about 1,100–1,200 casualties over 10 days before seasonal or holiday adjustments.

• Summer and holiday uplift

  • The dominant adjustment is early-July seasonality, especially the Independence Day weekend.
  • Models generally add a few hundred extra casualties above baseline, reflecting increased travel, gatherings, alcohol use, celebratory gunfire, and other holiday-related risk.
  • One forecast places extra emphasis on the fact that July 4 falls on a Saturday, which may concentrate activity across the weekend and raise the total.

• Historical July patterns

  • Historical GVA totals for July 1–10 are used as a cross-check, with one forecast citing typical totals in the 1,400–1,800 range.
  • This supports the idea that the period is usually above a simple baseline extrapolation.

Areas of agreement

• The expected total is well above 1,000 and most likely in the 1,400–1,600 range.
• The distribution is right-skewed, meaning rare high-end events can push totals upward.
• All forecasts treat the outcome as highly sensitive to short-window volatility, so wide uncertainty bands are appropriate.

Areas of disagreement

• How large the holiday surge will be:
  • Some forecasts assume a moderate bump, others a larger Independence Day spike.
• Whether recent trends should pull the estimate down:
  • One rationale notes a downward trend since the 2022 peak, limiting the upper estimate.
• Tail risk assumptions:
  • Upper bounds differ depending on how much weight is given to mass shootings or other unusually severe events, and lower bounds reflect possible reporting lag or temporary lulls.

Bottom line

The shared logic is: recent GVA daily rates provide the base, summer conditions and the July 4 weekend raise the total, and uncertainty is dominated by rare-event volatility. The net result is a forecast centered near 1,500 casualties with a broad but skewed range.

— Iteration 3 — Across the forecasts, the main expectation is that Gun Violence Archive counts for July 1–10, 2026 will land around the mid-to-high 1,000s, with a shared center of gravity near 1,700 people.

Key reasoning patterns

• Baseline rate extrapolation:
  All forecasts start from recent GVA patterns, especially the post-pandemic decline or a recent daily baseline in the 120–150 people/day range.
• Calendar effects matter a lot:
  The period includes July 4, which falls on a Saturday in 2026, so the models expect a weekend + holiday amplification. This is the single biggest upward adjustment.
• Seasonality in summer:
  July is treated as a higher-risk month than average, reinforcing the holiday-weekend bump.
• Definition/reporting uncertainty:
  The forecasts note that GVA figures reflect people shot, injured, or killed in incidents, not all firearm deaths, and that reporting completeness can shift the total.
• Tail risk is wide but not dominant:
  All models allow for unusually low or high outcomes due to a quiet holiday period, continued crime decline, or a major mass-casualty event.

Areas of consensus

• The forecasts strongly agree that the total will be well above a normal non-holiday 10-day period because of the July 4 weekend.
• They also agree that the distribution should be right-skewed, with a meaningful chance of an unusually high total.

Main differences

• The lower estimate leans more on a lower post-pandemic baseline and more caution about what GVA counts represent.
• The higher estimates put more weight on the holiday/weekend surge and a slightly stronger summer violence assumption.

Bottom line

The collective view is that recent declines provide the baseline, but July 4 landing on a Saturday should create a noticeable spike, making a total around 1,600–1,800 the most plausible range, with substantial uncertainty around mass-shooting risk and reporting effects.

Summary of Forecasting Rationales

The individual rationales broadly agree that forecasting gun violence for July 1–10, 2026, requires balancing a steep, long-term national decline in crime against several potent, short-term catalysts unique to the summer of 2026.

Macro Trends: Structural Decline in Gun Violence The foundational baseline across all rationales is the significant, multi-year deceleration of U.S. gun violence. Following pandemic-era peaks, casualty rates dropped substantially through the mid-2020s, with trends in early 2026 indicating that this downward trajectory is holding strong. Consequently, the baseline expectation for daily violence in 2026 is structurally lower than in previous years.

Calendar Effects: The Saturday Holiday Spike Despite the overall decline in violent crime, early July remains historically the most violent window of the year. In 2026, Independence Day falls on a Saturday. Forecasters unanimously note that when July 4th aligns with a weekend, it creates a “perfect storm.” Rather than diffusing across a workweek, routine weekend violence compounds with holiday celebrations, concentrating incidents into a highly volatile, high-casualty window.

Environmental and Event-Driven Accelerants Forecasters identify several severe upward pressures specific to July 2026 that will likely offset the broader macro-decline:

• Extreme Heat: Long-range meteorological models project a high likelihood of El Niño-driven heatwaves across the U.S. Criminological research consistently links unseasonable, extreme heat to measurable spikes in interpersonal conflict and shootings.
• Mass Gatherings: July 2026 marks the U.S. Semiquincentennial (America 250), which coincides with the FIFA World Cup and various planned protests. The resulting unprecedented density of crowds, parades, and block parties will severely strain law enforcement resources and heighten the statistical probability of mass casualty events.

Data Reporting Methodology Finally, the rationales address the mechanics of the Gun Violence Archive (GVA) and the resolution date. Because the question resolves based on an August 1 snapshot, forecasters agree that the approximately three-week buffer is more than sufficient for the GVA to log the vast majority of incidents. While a minuscule fraction of non-fatal or obscure police reports might miss this deadline, the snapshot is expected to reflect a highly mature and accurate count of the true holiday weekend casualties.

SUMMARY

Question: How many people will be shot, injured or killed, in the United States between 1 July and 10 July 2026, according to Gun Violence Archive? Final Prediction: Probability distribution:

• 10.00% chance of value below 1036.866667
• 20.00% chance of value below 1125.2
• 40.00% chance of value below 1250.2
• 60.00% chance of value below 1365.2
• 80.00% chance of value below 1531.866667
• 90.00% chance of value below 1686.866667

Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled

Report 1 Summary

Forecasts

Forecaster 1: Probability distribution:

• 10.00% chance of value below 1075.5
• 20.00% chance of value below 1170.5
• 40.00% chance of value below 1295.5
• 60.00% chance of value below 1405.5
• 80.00% chance of value below 1585.5
• 90.00% chance of value below 1760.5

Forecaster 2: Probability distribution:

• 10.00% chance of value below 1035.0
• 20.00% chance of value below 1115.0
• 40.00% chance of value below 1235.0
• 60.00% chance of value below 1350.0
• 80.00% chance of value below 1510.0
• 90.00% chance of value below 1660.0

Forecaster 3: Probability distribution:

• 10.00% chance of value below 1000.1
• 20.00% chance of value below 1090.1
• 40.00% chance of value below 1220.1
• 60.00% chance of value below 1340.1
• 80.00% chance of value below 1500.1
• 90.00% chance of value below 1640.1

Research Summary

The research compares a 2026 baseline (95.8 people killed or injured per day, based on 14,086 casualties through May 26, 2026) to historical patterns and a known July 4th holiday spike to project totals for July 1–10, 2026. A simple 10-day baseline gives ~958 people; using a higher recent-year reference (≈150/day) gives ~1,500 people. Accounting for documented Independence Day weekend spikes (city examples and nationwide GVA reports from 2025) shifts the central estimate upward, producing a most-likely range of roughly 1,200–2,400 total killed or injured: a lower bound ~1,200–1,400, a central estimate ~1,500–1,800, and an upper bound ~2,000–2,400. The brief notes GVA’s methodology (excludes suicides, continuously updated, sources include police/media/government) and highlights key uncertainties: continuation of 2026’s declining trend, the magnitude of any July 4th spike in 2026, and potential reporting lags in the GVA August 1 snapshot.

Websites/sources used (as named in the research):

• Gun Violence Archive: https://www.gunviolencearchive.org
• Kalshi: https://www.kalshi.com
• Polymarket: https://polymarket.com
• Metaculus: https://www.metaculus.com
• Manifold: https://manifold.markets

RESEARCH

Report 1 Research

Gun Violence Forecast Research Brief: July 1-10, 2026

Question Overview

You’re forecasting the total number of people killed or injured by gun violence in the United States from July 1-10, 2026 (a 10-day period), as measured by the Gun Violence Archive (GVA). This period crucially includes July 4th (Independence Day), which historically experiences elevated gun violence.

Current 2026 Baseline Data

Year-to-date 2026 statistics (as of late May):

• As of May 26, 2026: 5,048 killed + 9,038 injured = 14,086 total casualties over approximately 147 days [1]
• Daily average for 2026: 95.8 people killed or injured per day
• 10-day baseline projection: 95.8 × 10 = ~958 people

However, this baseline doesn’t account for the July 4th spike effect (discussed below).

Historical Base Rates and Reference Classes

Annual Trends (Non-Suicide Gun Violence)

2025 Full Year:

• 14,651 deaths (14% decrease from 2024’s ~17,000) [15]
• 40,000+ injuries [15]
• Total: ~54,651 casualties
• Daily average: ~150 people killed or injured per day
• Through Q3 2025: 11,197 deaths + 20,425 injuries = 31,622 total [19]

2024 Data:

• 15,364 gun homicides (excluding suicides) [20]
• Daily average: ~42 deaths per day
• Over 200 Americans visit emergency departments daily for nonfatal firearm injuries [25]

Key Trend: Gun violence has been declining. 2025 saw a 14% decrease from 2024, with deaths at the lowest level since 2015 [15][19]. The 2026 data through May suggests this downward trend may be continuing, with a daily rate of 95.8 (lower than 2025’s ~150).

July 4th Period - Critical Adjustment Factor

The July 1-10 period includes Independence Day, which is consistently associated with elevated gun violence:

July 4-7, 2025 (Philadelphia alone):

• 15 shooting incidents
• 48 victims total (killed + injured)
• 8 fatalities [12]

Independence Day Weekend 2025 (Nationwide):

• Gun Violence Archive recorded at least 22 mass shootings during July 4-7 period nationally [12]
• July 5 recorded the highest number of mass shootings in a single day [12]
• Multiple cities (Indianapolis, Chicago, Philadelphia, Brockton MA) reported deadly incidents [13]
• Historical sources note July is “one of the deadliest periods in the US” for gun violence [13]

Implication: The July 4th holiday period produces a significant spike above baseline daily averages. One city alone (Philadelphia) had 48 casualties over 4 days in 2025, suggesting the national 10-day total will be substantially higher than the simple 10-day baseline calculation.

Quantitative Estimates

Conservative Baseline Approach:

• 2026 daily average × 10 days: 95.8 × 10 = 958 people

Adjusted for July 4th Effect:

• If we use 2025’s daily average (150/day) as more representative of a typical period: 150 × 10 = 1,500 people
• Adding a July 4th spike (historically significant): 1,500-2,000+ people

Upper Bound Consideration:

• Using higher injury estimates (200 ED visits/day for injuries [25]) + 40 deaths/day: 240 × 10 = 2,400 people

Reference Class: Similar 10-Day Periods

2025 10-day equivalent (assuming 150/day average): ~1,500 casualties 2024 10-day equivalent (assuming higher rates): ~1,700+ casualties

The July 4th inclusion likely adds 100-300 additional casualties above a typical 10-day period, based on the spike patterns observed in historical data [12][13].

Gun Violence Archive (GVA) Methodology Notes

• GVA excludes suicides from daily counts (as of February 2024) [21]
• Data is continuously updated as incidents are verified [21]
• Sources: 7,500+ including police, media, and government agencies [21]
• The resolution counts will be aggregate totals displayed by GVA’s search tool as of August 1, 2026, 11:59 PM ET
• Mass shootings (4+ shot) accounted for only 2.8% of total shooting deaths in 2025 [19]

Prediction Markets

Search Results: I found extensive discussion of prediction market platforms (Kalshi, Polymarket, Metaculus, Manifold) and their growth to a multibillion-dollar industry [23][24], but no specific prediction markets were found forecasting this particular gun violence question.

The absence of active markets on this question means there’s no market-based probability data or liquidity metrics to reference for reliability assessment.

Most Relevant Forecast Range

Based on the research, the most likely range for July 1-10, 2026:

• Lower bound: ~1,200-1,400 (if 2026’s declining trend continues strongly)
• Central estimate: ~1,500-1,800 (accounting for 2026 baseline + July 4th spike)
• Upper bound: ~2,000-2,400 (if July 4th spike is particularly severe)

Key uncertainties:

1. Whether 2026’s declining trend (95.8/day through May) continues through July or reverts closer to 2025 levels
2. The magnitude of the July 4th spike effect in 2026 compared to historical patterns
3. Any lag in GVA reporting by the August 1 snapshot deadline

FORECASTS

R1: Forecaster 1 Reasoning

(a) Time left: Conditional on the question resolving normally and not being annulled, the outcome should be known at the GVA snapshot deadline: August 1, 2026, 11:59pm ET, about 64 days from today, May 29, 2026. The incident window starts in about 33 days and ends in about 43 days.

(b) Outside view: The best base rate is recent GVA killed+injured totals, adjusted for seasonality. GVA’s recent annual data imply roughly 109–112 killed or injured per day in 2025 and about 130 per day in 2024, with 2025 clearly below the pandemic-era peak. GVA’s own 10-year review lists 2025 at 14,719 deaths + 25,072 injuries = 39,791, while The Trace’s GVA-based 2025 snapshot gives 14,655 deaths + 26,238 injuries = 40,893. (gva-dev.sdev.site)

(c) Inside view: 2026 is running lower than recent years: The Trace reported 3,103 shooting deaths in Q1 2026, the lowest first-quarter figure in at least a dozen years, and noted that GVA can update incidents weeks or months later and does not track gun suicides in these figures. The research brief’s May 26 YTD rate of about 96 killed+injured per day is therefore a reasonable current baseline. (thetrace.org) The upward adjustment is that July 1–10 includes Independence Day: Northeastern’s James Alan Fox notes July 4 ranks as the most violent day of the year, and GVA data since 2013 showed 66 mass shootings on July 4 alone; 2025’s July 4 holiday weekend reportedly had at least 24 GVA-defined mass shootings. (news.northeastern.edu)

(d) Outcome if nothing changed: If I simply used the current 2026 YTD average from the brief, 95.8/day × 10 = about 958 people.

(e) Outcome if the current trend continued: If the 2026 decline continues but normal summer/holiday seasonality applies, I would expect roughly 1,250–1,400 people, centered near 1,330–1,350.

(f) Experts and markets: I found no liquid market specific to this question. Expert/analyst evidence points in two opposing directions: real-time GVA-based analysts see a continuing decline in shooting deaths in early 2026, while criminology/holiday evidence points to a July 4 spike. (thetrace.org)

(g) Unexpected low scenario: A low outcome could result if the 2026 decline accelerates into summer, several high-violence cities have unusually quiet holiday weekends, weather suppresses outdoor gatherings, and GVA reporting by August 1 misses or later adds a meaningful number of nonfatal injuries. That could plausibly land the total near 850–1,000.

(h) Unexpected high scenario: A high outcome could result from an unusually violent July 4 weekend plus one or more large mass-casualty shootings, especially in large cities, combined with fast GVA reporting before the snapshot. That could push the total above 1,900, with extreme tails above 2,400.

Rationale: I anchor on the recent GVA base rate rather than the research brief’s higher 2025 daily estimate: final 2025 GVA-based totals look closer to ~110/day, not 150/day. But July 1–10 is not an average 10-day period: it contains the July 4 holiday, an observed holiday Friday in 2026, and a weekend, all of which raise the expected count.

Starting from the current 2026 rate of about 960 people per ordinary 10 days, I apply a summer/July multiplier of roughly 1.20 and an additional July 4/holiday-window multiplier of roughly 1.10–1.15, giving a central estimate around 1,270–1,325 before allowing for GVA lag and incident reclassification. I add a small upward allowance for late reporting/reclassification by the August 1 snapshot, producing a median forecast of about 1,345 people.

My uncertainty is right-skewed: the lower tail is bounded by the very stable national daily volume of firearm injuries/deaths, while the upper tail can be moved by clustered mass shootings or a broad holiday surge. I therefore put relatively little probability below 800 or above 2,400, but not zero.

Percentile 0.1: 680.2
Percentile 1: 830.4
Percentile 5: 990.5
Percentile 10: 1,075.5
Percentile 20: 1,170.5
Percentile 30: 1,245.5
Percentile 40: 1,295.5
Percentile 50: 1,345.5
Percentile 60: 1,405.5
Percentile 70: 1,485.5
Percentile 80: 1,585.5
Percentile 90: 1,760.5
Percentile 95: 1,930.5
Percentile 99: 2,280.5
Percentile 99.9: 2,780.5

R1: Forecaster 2 Reasoning

Conditioning on normal GVA resolution / no annulment or technical failure.

I think the critic’s main correction is right: the 150/day 2025 baseline is too high. The better GVA-based 2025 total is about 14,655 deaths + 26,238 nonfatal injuries = 40,893, or about 112/day, and GVA explicitly does not track gun suicides in these counts. The Trace also notes GVA can update incidents weeks or months later, which matters for the August 1 snapshot. (thetrace.org)

I also agree that mass shootings should not drive the median. They matter for the upper tail, but most of the count comes from ordinary dispersed shootings. In Q1 2026, GVA recorded 3,103 shooting deaths, down almost 500 from Q1 2025, while mass shootings accounted for only 3.6% of GVA-tracked shooting deaths. (thetrace.org)

That said, I still would not move all the way down to a raw current-rate projection. The target window is seasonally and calendrically unfavorable: July 4, 2026 is a Saturday, July 3 is the observed federal holiday, and the 10-day window includes a Friday–Saturday–Sunday holiday cluster plus a second Friday. Historically, the July 4 week has been clearly elevated: GVA/CBS reported 1,096 shot killed or wounded from July 1–7, 2024, and 1,581 in the same week of 2023; 2024 was “only” about 15% above the first-half 2024 daily rate, but 2023 was far above its first-half pace. (cbsnews.com)

My updated model is: current 2026 underlying annualized GVA casualty rate around 100–105/day after adjusting the May YTD number for winter/spring seasonality and reporting lag, then a July/holiday/window multiplier of roughly 1.22–1.30. That puts the center around 1,270–1,340. I’m lowering my median from 1,345 to 1,290, mostly accepting the critic’s location critique, while keeping a fairly fat right tail for a hot holiday weekend or a major high-casualty shooting.

Updated prediction:
Percentile 0.1: 750
Percentile 1: 850
Percentile 5: 965
Percentile 10: 1035
Percentile 20: 1115
Percentile 30: 1180
Percentile 40: 1235
Percentile 50: 1290
Percentile 60: 1350
Percentile 70: 1420
Percentile 80: 1510
Percentile 90: 1660
Percentile 95: 1810
Percentile 99: 2130
Percentile 99.9: 2550

R1: Forecaster 3 Reasoning

Conditioning: I am forecasting conditional on the question not being annulled and on GVA’s standard source-verified killed + injured tally being the resolution source.

(a) Time left until known: Today is May 29, 2026. The incident window is July 1–10, 2026, and the resolution snapshot is August 1, 2026 at 11:59pm ET, about 64 days from now.

(b) Outside view: The best reference class is GVA killed + injured counts around the July 4 holiday. GVA’s 2025 year-in-review table implies about 39,791 source-verified non-suicide killed/injured in 2025, or about 109 per day, lower than 2024 and far below the 2020–2023 peak years. (gva-dev.sdev.site) The holiday week is consistently above baseline: CBS, citing GVA, reported 1,096 shooting victims from July 1–7, 2024, versus 1,581 in 2023, 1,708 in 2020, and 1,177 in 2019. (cbsnews.com)

(c) Inside view: 2026 appears lower than recent history. A May 26 report citing GVA put 2026 YTD at 5,048 killed + 9,038 injured = 14,086 through May 26, roughly 96 people per day. (news.mail.ru) The Trace also reported that Q1 2026 GVA shooting deaths were the lowest first-quarter total in a dozen years, though GVA can update records after incidents occur. (thetrace.org) This pushes me below the research assistant’s 1,500–1,800 central range; that range appears too anchored to an inflated 2025 injury estimate.

(d) Outcome if nothing changed: If the May 26 YTD average simply continued with no holiday/seasonality adjustment: about 960–970 people.

(e) Outcome if the current trend continued: Applying continued 2026 decline but adding July 4 seasonality gives a central estimate around 1,250–1,350 people.

(f) Experts and markets: I found no dedicated prediction market for this exact GVA question. Expert commentary points to continued declines: The Trace quoted Daniel Webster describing the recent decline as a set of “positive data points,” while warning that late-summer surges remain possible. (thetrace.org)

(g) Unexpected low scenario: Continued 2026 decline plus unusually cool/rainy holiday weather, no large mass-casualty shootings, and modest reporting lag by August 1 could yield a result below 900.

(h) Unexpected high scenario: A severe heat/holiday weekend spike, several large mass shootings or multi-victim urban incidents, and later GVA updates before August 1 could push the result above 1,800, with an extreme tail above 2,200.

Rationale:
I anchor on the recent GVA annual/day-rate base rate rather than the research brief’s higher 2025 estimate. The 2025 GVA source-verified killed + injured total looks near 40,000, not 55,000, and 2026 YTD is running lower still. However, July 1–10 is not an average 10-day period: it contains the highest-risk holiday, and GVA/CBS historical July 1–7 figures show a clear holiday multiplier. My median therefore sits well above the raw 2026 YTD 10-day run rate, but below 2023–2024 holiday-week levels after adjusting for the ongoing decline.

Percentile 0.1: 730.1
Percentile 1: 820.1
Percentile 5: 930.1
Percentile 10: 1000.1
Percentile 20: 1090.1
Percentile 30: 1160.1
Percentile 40: 1220.1
Percentile 50: 1280.1
Percentile 60: 1340.1
Percentile 70: 1410.1
Percentile 80: 1500.1
Percentile 90: 1640.1
Percentile 95: 1780.1
Percentile 99: 2060.1
Percentile 99.9: 2360.1

Forecast — Post 378

Modal: 996-1198 (30.0%) • partial_shift • 6m3s

Interpretation

Reading: strict

The resolution criteria are highly specific and mechanical: a precise date range (1–10 July 2026), a named data source (GVA), a fixed snapshot time (1 August 2026 11:59pm ET), and explicit rules for handling edge cases (non-numeric values = 0, aggregate totals override manual tallies, technical failure delays resolution). There is little room for charitable interpretation; the resolver simply reads the GVA aggregate at the designated snapshot.

Time bounds

• observation_window: GVA database snapshot at 11:59pm ET on 2026-08-01, covering incidents with Incident Date 2026-07-01 through 2026-07-10 inclusive
• resolution_date: 2026-08-01

Edge cases identified

1. Casualty status updates: an injury recorded during the window that is later reclassified as a death before the 1 August snapshot is included in the final sum — this could shift the count relative to an earlier read.
2. Reporting delays: incidents occurring 1–10 July but not yet entered into GVA by 1 August snapshot are excluded; this may undercount relative to eventual full reporting.
3. Independence Day (4 July) effect: historically elevated gun violence around this holiday inflates counts for this 10-day window relative to an average 10-day period.
4. Technical failure clause: if GVA returns zero results or is inaccessible at the snapshot time, resolution is delayed — the zero-result scenario is NOT treated as a zero resolution value.
5. Aggregate vs. manual tally: the GVA search interface aggregate totals take precedence over a manual sum of individual incident records if there is any discrepancy.
6. Timezone/incident dating: the Incident Date field is the sole arbiter; incidents near midnight boundaries could be attributed differently depending on local vs. reported timezone, but GVA’s field value is final.
7. The answer range (795–2405) is open on both ends, so extreme values outside this range are technically possible.

Research (3/3 variants, shared evidence pool)

Total evidence registered (shared pool): ?

Research Brief

Evidence confidence: medium

Scenario 1: Moderate decline from 2024 levels — total ~950–1,250 casualties [high evidence]

Conditions favoring

2026 YTD gun violence is tracking ~10–20% below 2025, which was itself well below 2024. If the ~19% YoY decline seen through April 2026 persists into the July 4 period and 2025’s July 4 pattern (Chicago down ~40–44% from 2024) generalizes nationally, a 2024 July 1–7 total of 1,096 reduced by ~15–20% yields roughly 875–930 for 7 days, or ~985–1,050 extrapolated to 10 days. The arrival-process projection (source 25) centers at ~1,250 using a ~125/day July rate, consistent with this range at the upper end.

Conditions against

The 2024 July 1–7 figure of 1,096 was already described as a five-year low; further decline of 15–20% may be optimistic if the rate of deceleration continues to slow (from 30%+ to 19%). A return toward prior holiday patterns or an unusually violent event could push totals higher.

Scenario 2: Sharp further decline — total ~700–950 casualties [medium evidence]

Conditions favoring

If 2026’s 10.3% YoY decline from 2025 is compounded on top of 2025’s own large decline from 2024 (2025 non-suicide casualties ~41,064 vs. 2024’s ~48,735, a ~16% drop), the July 4 2026 10-day total could fall to the low end of or below the answer range. Chicago’s 2025 July 4 was the least violent in at least 6 years; if a similar or larger drop holds nationally in 2026, totals could approach ~700–850.

Conditions against

The July 4 holiday consistently produces spikes well above baseline daily rates; even with a 20%+ decline, the baseline is high enough that 10 days including July 4 would likely exceed 700. No actual data for this period exists to confirm the magnitude of decline. Current 2026 YTD daily rate of ~50.7 non-holiday casualties/day means the holiday spike component would need to be quite modest to stay below ~800.

Scenario 3: Near-2024 levels or modest decline — total ~1,050–1,400 casualties [medium evidence]

Conditions favoring

If the rate of decline slows further or reverses near the holiday period — as the moderating trend (30%+ → 19% by April) suggests — July 4 2026 could resemble 2024 or 2025 holiday violence more than the steepest part of the 2026 decline. The arrival-process projection (source 25) p75 is ~1,518 and the mean is ~1,250; the upper portion of this scenario overlaps with that mean.

Conditions against

Every available data point shows 2026 running below 2025 and well below 2024. Reverting to near-2024 levels would require a break from the trend. The structural factors driving declines (policing, demographics, economic) show no signs of reversing as of the research date.

Scenario 4: Unusually high total — 1,500+ casualties [low evidence]

Conditions favoring

2023 July 1–7 saw 1,581 total victims (source 6), and 2021/2022 holiday weekends saw higher totals still (sources 4, 15). If there is a major mass casualty event, a reversal of the trend, or if holiday conditions are especially violent in 2026, the total could approach or exceed 1,500.

Conditions against

The sustained multi-year downward trend (sources 16, 18, 27) makes a reversion to 2021–2023 levels very unlikely absent a structural change. Q1 2026 data (source 13), mid-May 2026 GVA snapshot (source 31), and crime analyst projections (source 18) all point away from elevated totals. The 2026 decline rate, though moderating, is still clearly negative.

Key uncertainties

• The magnitude of the July 4 holiday spike in 2026 relative to baseline — the holiday effect is the dominant driver of the 10-day total, and its 2026 size is unknown
• Whether the moderating rate of decline (30%+ in Jan to 19% in April) will stabilize or reverse by July 2026
• Whether Chicago’s ~40–44% July 4 2025 decline from 2024 generalizes to the national GVA count for July 1–10
• The injury-to-death ratio in the July 2026 period (historically ~1.79–1.87, but mass shootings tend to have higher injury:death ratios)
• Potential for one or more large mass casualty events during the 10-day window, which could significantly shift the total
• GVA reporting completeness by the August 1 snapshot — reporting delays may undercount incidents occurring close to July 10

Forecast variants (3/3)

Variant 0: inside_view (inside_view_v1)

Inside View Causal Chain Analysis

Current state → mechanism → outcome:

The causal chain is: persistent structural decline in US gun violence (confirmed through May 2026) → moderating but continuing YoY decline into summer 2026 → July 4 holiday spike (the dominant driver of this 10-day window) → July 1–10 total.

Key causal anchors:

1. Baseline daily rate: 2026 YTD through ~May 29 shows ~50.7 combined casualties/day (source 31). This is the non-holiday rate. The holiday effect must lift this substantially for 10 days.

2. 2024 reference point: July 1–7, 2024 = 1,096 total (source 5), which is ~157/day for 7 days. Extending to 10 days with July 8-10 at slightly lower rates (post-holiday), the 10-day 2024 total was likely ~1,300-1,500.

3. 2026 trend: 10.3% below 2025 as of mid-May (source 18, 26). But the key point is: what was 2025 July 1-10? We don’t have this directly, but we know 2025 was ~16% below 2024 annually (sources 20, 21, 23). Chicago 2025 July 4 was down 40-44% from 2024 (source 22).

4. Arrival-process projection (relative_evidence_strength: high): Using ~125/day holiday-adjusted rate (2024’s ~157/day × ~80% for trend), mean = ~1,250, p5 = ~597, p95 = ~1,903 (source 25). This is the strongest quantitative anchor.

5. Moderating decline: The YoY decline rate slowed from 30%+ in January to 19% in April (source 27), suggesting the steepest part is past. By July 2026, the YoY decline may be ~15-20%.

Scenario weighting:

• “Moderate decline” scenario (~950-1,250): tagged high evidence strength. The arrival projection mean of ~1,250 falls here. The 2026 YoY decline applied to an estimated 2025 July 1-10 range of ~850-1,100 (itself ~30% below 2024’s 10-day ~1,400) yields ~700-950. But the moderating decline rate makes ~950-1,250 more likely. I weight this heavily.

• “Sharp further decline” (~700-950): tagged medium. Possible if Chicago-style 40-44% drops generalize, but that was 2024→2025; the 2025→2026 drop should be smaller (~10-20%). Low end of bin_0 and bin_1.

• “Near-2024 levels” (~1,050-1,400): tagged medium. The arrival projection p75 of ~1,518 and the extended 10-day extrapolation from 2024’s 7-day data support this range. The moderating decline rate makes this plausible.

• “Unusually high” (1,500+): tagged low. Against the strong trend. Only a major mass casualty event could push here.

My distribution:

The high-evidence scenario centers on ~950-1,250. The arrival projection mean of ~1,250 falls in 996-1198 (bin_2) and partially in 1198-1399 (bin_3). Given:

• The 10-day window extends 3 days past July 7, capturing more post-holiday violence
• The 2024 July 1-7 was 1,096 (7 days), so July 1-10 for 2024 would be ~1,400-1,500 range
• 2026 is ~15-20% below 2025, which was ~15% below 2024
• Compound effect: 2026 July 1-10 ≈ 2024 July 1-10 × 0.85 × 0.80 ≈ 1,400-1,500 × 0.68 ≈ ~950-1,020
• But arrival projection (125/day × 10 = 1,250) suggests a somewhat higher anchor

Central mass should cluster around 996-1399 range (bins 2-3), with meaningful probability on both sides.

Distribution:

• Below 795 (bin_0): ~7% — requires very sharp declines beyond evidence
• 795-996 (bin_1): ~18% — sharp decline scenario, plausible
• 996-1198 (bin_2): ~28% — high-evidence scenario core, near arrival mean
• 1198-1399 (bin_3): ~22% — high-evidence scenario upper / near-2024 scenario
• 1399-1600 (bin_4): ~12% — 10-day extrapolation from 2024 7-day data with modest decline
• 1600-1801 (bin_5): ~6% — approaching 2023-level territory
• 1801-2003 (bin_6): ~3% — arrival p90-p95 territory
• 2003-2204 (bin_7): ~2% — very elevated, tail
• 2204-2405 (bin_8): ~1% — extreme outlier
• Above 2405 (bin_9): ~1% — extremely unlikely given trend

Variant 1: outside_view (outside_view_v1)

Reference class: GVA-reported totals for short July holiday windows (recent years’ July 1–7/1–10 windows) and the arrival-process projection computed for July 1–10, 2026. Base rate: the arrival-process projection (treated as strong quantitative evidence in the brief) centers at ~1,250 total casualties for July 1–10 with p5≈597 and p95≈1,903 (source 25). Adjustment: I downweight very-high outcomes (>2405) and upweight the central bins slightly to reflect the strong, multiple-source evidence of a continuing national decline in 2026 (GVA YTD down ~10.3% vs 2025 per source 18; Q1 2026 deaths lowest in 12+ years per source 13), and because the arrival-process projection already internalizes a holiday spike. Final: probability mass concentrated around the projection’s center (bins covering ~996–1399) with a long but limited tail to higher totals and non-negligible chance of lower totals given the observed YoY decline. Weighting decisions informed by the brief’s scenario evidence-strength tags: the “Moderate decline” scenario is labeled high-evidence (so bins 2–4 receive the largest share), the two medium-evidence scenarios (sharp further decline; near-2024/modest decline) receive moderate mass (bins 0–2 and 3–5), and the low-evidence high-total scenario gets a small tail probability (bins 6–9). Specific citations from the brief: 2024 July 1–7 = 1,096 (source 5); arrival-process projection mean=~1,250, p5~597, p95~1,903 (source 25); 2026 YTD decline ~10.3% vs 2025 (source 18); Q1 2026 deaths = 3,103, lowest Q1 in 12+ years (source 13); 2026 YTD counts ~5,152 deaths + 9,233 injuries ≈ 14,385 (~50.7/day) (source 31). Reading: strict — I treat the GVA ‘Incident Date’ and the 1 August 2026 23:59 ET snapshot as authoritative per the resolution rules and did not broaden inclusion beyond those criteria. Calibration note (outside view): I anchor on the arrival-process projection (the most relevant quantitative base rate) and adjust only modestly to reflect the brief’s additional evidence (ongoing decline, holiday spike uncertainty, and scenario evidence strengths). The resulting distribution places most mass in bins 2–4 (combined 0.75) consistent with the projection’s central range while leaving ~0.25 total mass in the lower tail (0.28 in bins 0–1) and ~0.12 in the upper tail (bins 5–9) to capture reporting uncertainty and the (low) chance of unusual mass-casualty events.

Variant 2: contrarian_v2 (contrarian_v2)

Adversarial reading

The implicit consensus in the brief is the high-evidence moderate-decline scenario (~950–1,250 casualties) anchored on the arrival-process projection (mean ~1,250) and the 10.3% YoY drop persisting into the holiday window. This places the bulk of probability mass in bins 1–3.

Counter-evidence found (or absence)

The brief’s own evidence shows the YoY decline rate already moderating (30%+ in Jan to 19% in April), Chicago’s July 4 2025 drop was an extreme local outlier, and the arrival projection uses a holiday-adjusted 125/day rate that may over-weight the downward trend if the summer uptick or mass-shooting component reasserts even modestly. No data yet exists for the actual 2026 holiday period, leaving the 10.3% national extrapolation as the weakest link.

Distribution implications

This pushes materially more mass into bin_0 and the lower half of bin_1 than a naive centering on the 1,250 mean would suggest, while trimming the upper tail (bins 4+) relative to the arrival p75/p95. The high-evidence moderate-decline scenario still receives the largest single share, but the contrarian weighting widens leftward to reflect the possibility that continued deceleration plus a muted holiday spike produces totals below 995.

Consolidate

Strategy: logit_mean • Drafts: 3/3

Reviewers (3/3)

anchoring → shift_lower (low magnitude, medium confidence) anchoring_v1

The draft places p50 (and peak mass) in 996–1198 (bin_2), anchored primarily on the arrival-process projection mean of ~1,250 and the 2024 July 1–7 figure of 1,096. However, the arrival projection’s ~125/day rate was derived by applying only a ~20% discount to the 2024 July holiday rate of ~157/day — yet the evidence shows 2024→2025 was already a ~16% annual decline and 2025→2026 is tracking ~10.3% below 2025 YTD, implying a compound ~24–25% discount off 2024 is more defensible than 20%. More critically, the comparison base is the 2024 July 1–7 window (7 days = 1,096), not a 10-day window — the forecaster appears to have treated the 7-day rate as representative of the full 10-day period without adequately accounting for the fact that July 8–10 are post-holiday days running at much lower rates (closer to the ~50.7/day YTD baseline). Extrapolating 157/day for all 10 days is optimistic relative to a model where only July 2–6 are elevated and the remaining days revert toward baseline. A more neutral read centering on ~50.7/day × 7 non-holiday days + holiday spike days at ~200/day × 3 days yields roughly 900–1,050, which shifts the modal bin toward 795–996 (bin_1) or the lower end of bin_2 rather than the center of bin_2.

Flagged concerns

• Holiday-rate extrapolation anchored optimistically high: The arrival projection’s ~125/day derives from applying only a ~20% discount to the 2024 July 1–7 rate of ~157/day. But the 2024→2025 decline (~16%) plus 2025→2026 decline (~10%) compounds to ~24–25%, not ~20%. Using 125/day for all 10 days rather than a declining post-holiday rate is an optimistic anchor.
• 7-day vs. 10-day base confusion: The primary reference point (2024 July 1–7 = 1,096) covers 7 days. The forecaster extrapolates this at ~157/day for 10 days, but July 8–10 are post-holiday days closer to the baseline ~50.7/day. The realistic July 1–10, 2024 total was likely ~1,200–1,300, not ~1,570, making the 2026 central estimate lower than the draft’s peak bin.
• Modal bin placed at arrival-process mean rather than conditional median: The arrival-process projection mean (~1,250) falls in 1198–1399 (bin_3), yet the draft places peak mass in 996–1198 (bin_2) which is one bin lower. This is actually directionally correct but the rationale inconsistently references both bins 2–3 as the modal scenario, creating ambiguity about where the true modal expectation lies.

ceiling → shift_lower (low magnitude, medium confidence) ceiling_v1

The primary structural constraint the forecaster partially addressed but did not fully integrate is the holiday window multiplier ceiling. The brief states the 2026 YTD non-holiday daily rate is ~50.7 casualties/day (source 31). For a 10-day window to reach 2003–2204 (bin_7) or higher, the average daily rate over July 1–10 would need to be ~200–220/day — roughly 4x the current non-holiday baseline. The 2024 July 1–7 achieved ~157/day (source 5), itself described as a “five-year low.” With 2026 running ~10–20% below 2025 (which was ~16% below 2024), a 4x holiday multiplier on the current baseline is implausible, implying an effective structural ceiling near ~1,600–1,800. The forecaster assigns ~4.8% to bins 7–9 (above 2003), which overstates the plausible upper tail. Conversely, the lower floor is also constrained: even if July 4 barely spikes at all, 10 days × 50.7/day = 507 casualties, so below 795 (bin_0) requires the holiday period to achieve barely above the non-holiday non-summer baseline — plausible if the 2026 decline is sharper than estimated, but the brief notes the rate is moderating, making extreme lower outcomes less likely than the ~9.6% assigned to bin_0.

Flagged concerns

• Upper tail structural ceiling: The 2026 non-holiday daily rate is ~50.7/day (source 31). For outcomes in bins 7–9 (>2003 casualties over 10 days), the window average would need to exceed ~200/day — approximately 4x the current baseline and ~28% above the best comparable (2024 July 1–7 at ~157/day, itself a five-year low). With 2026 running 10–20% below 2025 (source 18, 26), a ~200+/day average over 10 days is structurally implausible. The ~4.8% combined mass on bins 7–9 is excessive given this ceiling.
• Holiday multiplier magnitude: The 10-day window’s July 8–10 days are POST-holiday and should pull the average below the July 4–5 spike. Source 9 confirms July 4 and July 5 are the top mass shooting days historically. The 2024 July 1–7 (7-day) was 1,096; extending to 10 days adds three post-holiday days at likely ~80–100/day (sub-holiday rates), implying a 10-day 2024 total of roughly 1,336–1,396. Applying 15–20% 2026 decline yields ~1,069–1,187 for 2026 July 1–10. The draft’s central mass allocation is broadly consistent, but the derivation should reflect this ceiling on the 10-day-extension component.
• Lower floor structural constraint: Even with no holiday spike, 10 days × 50.7/day = 507 casualties. Bin_0 (below 795) requires July 1–10 to average below ~79.5/day — roughly a 57% increase over baseline but still well below the holiday multiplier. While possible if 2026 declines accelerate, the brief notes the rate is moderating (source 27), making sub-795 outcomes require both a continued acceleration AND a muted holiday. The ~9.6% mass in bin_0 may be slightly overstated.

math → shift_higher (medium magnitude, high confidence) math_v1

The forecast’s probability distribution sums to 1.0007 (within acceptable rounding), but shows a critical internal inconsistency: the rationale emphasizes a “high-evidence scenario” centered on ~950–1,250 (bins 2–3, “996–1198” and “1198–1399”), yet allocates only 0.50 combined probability (0.288 + 0.219 = 0.507) to those bins. Simultaneously, 0.214 goes to 795–996 (bin_1), which the inside view explicitly tags as the “sharp further decline scenario” with only “medium” evidence strength. The distribution shape contradicts the stated evidence hierarchy: it’s weighted too heavily leftward (bins 0–1: 0.311) relative to a scenario tagged “high evidence” that should receive the largest share.

Flagged concerns

• Evidence-to-probability mismatch: Inside view rationale states the ‘Moderate decline’ scenario (~950–1,250) is tagged ‘high’ evidence strength and should receive heavy weighting. Yet bins 2–3 (which span this range) receive only ~0.51 combined mass. Bins 0–1 (sharp further decline, ‘medium’ evidence) receive 0.311. This reverses the stated evidence hierarchy. The arrival-process projection mean of ~1,250 (which the brief calls ‘strong quantitative evidence’) falls in bin_2; yet bin_2 receives only 0.288, less than bin_1’s 0.214.
• Centroid vs. stated reference point: The forecast cites the arrival-process projection mean of ~1,250 as a key anchor (‘strongest quantitative anchor’), which falls squarely in 996–1,198 (bin_2), yet the forecast allocates mode probability to bins 1 and 3, not bin_2. A distribution centered on the projection’s mean should peak in bin_2; instead, bins 1 and 3 split the mode roughly equally (0.214 and 0.219), leaving bin_2 slightly lower at 0.288.
• Internal inconsistency in tail treatment: The inside view assigns ~7% to below 795 (bin_0), calling it the ‘sharp decline’ tail requiring ‘very sharp declines beyond evidence.’ Yet the inside view also allocates ~18% to 795–996 (bin_1), which is the ‘sharp decline scenario’ scenario proper (medium evidence). The outside view cites ‘non-negligible chance of lower totals given observed YoY decline’ and places 0.28 in bins 0–1 combined. The inside and outside views are not fully reconciled; one could reasonably place 15–20% in bins 0–1, but then bins 2–4 must rise correspondingly, and the arrival projection’s mean (~1,250) must get more weight.
• Stated vs. computed weights: Inside view explicitly writes: ‘Central mass should cluster around 996–1399 range (bins 2–3), with meaningful probability on both sides.’ Bins 2–3 combined = 0.507, which is less than half the total mass. Bins 0–1 receive 0.311 and bins 4–9 receive 0.182, so the ‘side’ probabilities (0.493 combined) are actually larger than the stated central mass. This contradicts the written rationale.

Reconcile: partial_shift

Reviews consumed: 3

The draft places peak mass at 0.288 (bin_2) but with substantial leftward weight in 0.214 (bin_1) driven by the contrarian variant. Reviewers split on direction: anchoring reviewer flagged the 7-day→10-day extrapolation issue and suggested a slight downward shift (peak stays at bin_2 at 0.28, bin_1 at 0.24); ceiling reviewer agreed the upper tail (bins 7-9 ~4.8%) is too thick given the structural ceiling (~200/day implausible) and trimmed the upper tail while raising bin_3; math reviewer flagged the strongest internal inconsistency — the rationale labels the moderate-decline scenario “high evidence” with arrival-mean ~1,250 (which falls in bin_2/lower bin_3), yet bin_1 received nearly as much mass as bin_2, contradicting the stated evidence hierarchy. I accept the math reviewer’s core fix (concentrate more mass at bins 2–3 to match the high-evidence anchor) and the ceiling reviewer’s tail trim, but I do NOT shift as high as the math reviewer because the anchoring reviewer’s 7-vs-10-day point is valid — July 8–10 are post-holiday days running closer to baseline, which legitimately pulls the central estimate toward the lower end of the moderate-decline range. Net: partial shift — peak stays in bin_2 but strengthened, bin_1 reduced from 0.21 to 0.17, bin_3 raised slightly, upper tail (bins 7–9) trimmed per ceiling reviewer.

Rejected critiques

• [math] Suggested shift_higher with magnitude medium, placing 0.32 on bin_2 and 0.25 on bin_3 while reducing bin_1 to 0.12.

  • Why rejected: Partially accepted — I strengthened bin_2 to 0.30 and raised bin_3 to 0.23, but did not cut bin_1 as aggressively (kept at 0.17 vs. math’s 0.12) because the anchoring reviewer’s 7-day vs 10-day extrapolation concern is valid: July 8–10 are post-holiday days running closer to the ~50.7/day baseline, which legitimately pulls some mass leftward toward bin_1. Fully adopting the math reviewer’s distribution would underweight this downward pressure.

• [anchoring] Modal bin ambiguity between bins 2 and 3 in the rationale.

  • Why rejected: Acknowledged as low severity; resolved by placing clear peak at 0.30 (bin_2) with 0.23 (bin_3) as second mass — consistent with arrival-projection mean of ~1,250 sitting on the bin_2/bin_3 boundary and the 7-vs-10-day adjustment pulling slightly leftward.

Final Distribution

Evidence Sources (36)

1. Gun Violence By The Numbers in 2025 - The Trace — Gun Violence Archive’s 2025 annual data shows over 40,000 shootings occurred in the US in 2025, averaging more than 110 people shot per day.
2. July Fourth Gun Violence Plummets to a Five-Year Low - The Trace — A July Fourth 2024 weekend saw gun violence plummet to a five-year low, with 245 people injured or killed according to GVA, per The Trace.
3. Past Summary Ledgers - Gun Violence Archive — The Gun Violence Archive reported 2024 Total GV Deaths (all causes) = 41,062 and 2024 Total Injuries in their annual ledger.
4. Gun Violence Wreaks Havoc on July Fourth Weekend - Everytown — EveryTown reported that one July Fourth weekend saw over 500 shootings resulting in at least 180 deaths and over 525 gun injuries across the US.
5. Gun violence over July 4 week dropped in 2024, but still above 2019 levels - CBS News — From July 1 to July 7, 2024, GVA recorded 340 people killed and 756 wounded, for a total of 1,096 shooting victims across the U.S. (excluding suicides).
6. Gun violence over July 4 week dropped in 2024, but still above 2019 levels - CBS News — In 2023, the July 4 week (July 1–7) saw 451 people shot and killed and another 1,130 wounded, for a total of 1,581 victims (excluding suicides), per GVA.
7. 4th of July mass shootings kill 20, injure 126 - ABC News — ABC News reported that in July 2023, 20 people were killed and 126 injured in 22 mass shootings between 5pm ET July 4 Friday and 5am ET July 9 Wednesday (mass shootings only, not all gun violence).
8. Gun violence over July 4 week dropped in 2024, but still above 2019 levels - CBS News — GVA recorded 20 mass shootings over July 4 week in 2024, down from 28 in 2023. The week still saw a surge compared to other weeks throughout the year.
9. Why shootings and violence increase in the summer months - PBS News — Independence Day (July 4) topped the list with 58 mass shootings over the last 10 years — closely followed by July 5, according to GVA data through 2024.
10. Gun Violence Archive — GVA’s home page (as of May 2026) shows year-to-date deaths DOWN 866 and injuries DOWN 1,901, with mass shootings down to pre-COVID levels — indicating 2026 gun violence is tracking lower than recent years.
11. Gun Violence Archive — GVA’s homepage snippet states: “Report of first four months [of 2025]: YTD deaths DOWN 866, injuries DOWN 1,901. Mass Shootings down to pre-COVID levels.” This suggests a continuing downward trend in 2025.
12. USAFacts gun-related deaths monthly data — Gun violence deaths in the US dropped 8.2% in Jan-Sep 2025 compared to the same period in 2024, representing a continuing multi-year downward trend from pandemic-era highs.
13. The Data on Gun Violence in America, Q1 2026 - The Trace — GVA data for Q1 2026 shows only 3,103 shooting deaths, the lowest Q1 figure in the last dozen years — a decline of nearly 500 shooting deaths from Q1 2025, and vastly below Q1 2021-2023 figures that exceeded 4,600 deaths.
14. Past Summary Ledgers - Gun Violence Archive — GVA 2025 past summary ledger shows: Total GV Deaths (all causes) 38,840; Homicide/Murder/Unintentional/DGU 14,750; Suicide 24,090; Total Number of Injuries (non-suicide shooting injuries) for 2025 full year.
15. More than 220 shot and killed in U.S. gun violence over July 4 holiday weekend - CBS News — In the July 4 weekend of 2022 (Friday-Monday), GVA reported at least 220 killed and ~570 wounded, for a total of ~790 casualties. In the same holiday weekend of 2021, more than 180 were killed and 516 injured (total ~696+).
16. Past Summary Ledgers - Gun Violence Archive — GVA annual totals (homicide/murder/unintentional/DGU + injuries, excludes suicides): 2021: 21,463 deaths + 40,696 injuries = 62,159; 2022: 20,589 deaths + 38,712 injuries = 59,301; 2023: 19,256 deaths + 36,620 injuries = 55,876; 2024: 16,904 deaths + 31,831 injuries = 48,735.
17. Crime Trends in U.S. Cities: Mid-Year 2025 Update - Council on Criminal Justice — CCJ Mid-Year 2025 data: gun assaults in major US cities were 21% lower in H1 2025 compared to H1 2024, and homicides were 17% lower — reflecting a broad continuation of the downward trend into 2025.
18. 2026 Will Likely Have The Lowest US Murder Rate Ever Recorded - Jeff Asher — As of mid-May 2026, GVA data showed a 10.3% drop in gun violence in 2026 relative to 2025 year-to-date; murder is down ~19.1% through April in a 30-city sample. 2026 is on track to set the lowest US murder rate ever recorded, per crime analyst Jeff Asher (May 25, 2026).
19. Derived from GVA Q1 2026 data via The Trace — GVA Q1 2026: 3,103 shooting deaths in 91 days = ~34 deaths/day. With the injury ratio typically ~2.2 injuries per death (based on 2024 data: 340 deaths + 756 wounded = 1,096 victims in 7 days, injury:death ratio ~2.22), total shot per day nationally in Q1 2026 ≈ 34 × 3.22 ≈ 110/day (but 2026 is trending ~10% below 2025 overall).
20. Gun Violence Archive reports that in 2025, over 40,000 shootings occurred in the US — GVA 2025 annual final totals: 14,615 killed and 26,195 injured (excluding suicides), totaling approximately 40,810 shooting casualties in 2025.
21. Past Summary Ledgers - Gun Violence Archive — GVA 2025 full year: Total non-suicide injuries = 26,320; homicide/murder/unintentional/DGU deaths = 14,744. Combined killed+injured total = ~41,064 non-suicide casualties for 2025, implying ~112.5 per day on average.
22. Chicago’s July Fourth Weekend Shootings Plummeted - The Trace — Chicago’s July 4 weekend 2025 saw 44-60 shot (various sources disagree slightly), with 5-11 killed — a 40-44% decline from 2024’s holiday weekend of 105 shot; Chicago’s least violent July 4 weekend in at least 6 years. The national trend mirrors this decline.
23. Past Summary Ledgers - Gun Violence Archive — GVA 2024 annual ledger: Total GV Deaths ALL causes 41,060; Homicide/Murder/Unintentional/DGU 16,904; Suicide 24,156; Total Non-Suicide Injuries 31,831; Mass Shootings 502. Combined non-suicide killed + injured = ~48,735 (~133/day).
24. Past Summary Ledgers - Gun Violence Archive — GVA 2022 annual ledger: Total GV Deaths ALL causes 47,627; Homicide/Murder/Unintentional/DGU 20,589; Suicide 27,038; Total Non-Suicide Injuries 38,712; Mass Shootings 645. Combined non-suicide killed + injured = ~59,301 (~162/day).
25. Arrival forecast model based on GVA historical rates — Arrival forecast for July 1-10, 2026 gun violence casualties: Using ~125/day rate (based on 2024 July 1-7 rate of ~157/day adjusted ~20% down for the 2026 downward trend), with overdispersion=0.1: mean=1,250, median=1,250, p5=597, p25=982, p75=1,518, p90=1,759, p95=1,903.
26. 2026 Will Likely Have The Lowest US Murder Rate Ever Recorded (Jeff Asher / Jeff-alytics) — As of mid-May 2026, the Gun Violence Archive showed a 10.3% drop in overall gun violence in 2026 relative to 2025, consistent with the “10% reduction vs. 2025” benchmark used in forecasting.
27. 2026 Will Likely Have The Lowest US Murder Rate Ever Recorded – Jeff Asher (Jeff-alytics) — The downward trend in GVA deaths was still ongoing as of mid-May 2026, with GVA showing a 10.3% year-over-year drop as of that date; the rate of decline appeared to be gradually slowing — murder in a 30-city sample was down 30%+ through January, 22.3% through February, 20.8% through March, and 19.1% through April 2026.
28. Derived from Jeff Asher / Jeff-alytics (May 25, 2026) and The Trace Q1 2026 report — Applying the 10.3% GVA YoY drop (as of mid-May 2026) to the Q1 2026 rate of ~34 deaths/day implies an approximate current daily rate of roughly 30–31 GVA-tracked shooting deaths per day in Q2 2026 (before seasonal summer uptick).
29. untitled — No actual shooting data for July 4–6, 2026 was available as of the research date, confirming the July 4, 2026 holiday weekend has not yet occurred and no early count can be used as a resolution anchor.
30. Gun Violence in America: What the 2026 Data Shows and Why Solutions Remain Elusive – ZenNews24 — As of around May 12, 2026, there had been more than 180 GVA-defined mass shootings (4+ shot) in the US in 2026, maintaining a pace of more than one per day consistent with recent years.
31. Gun Violence Archive – GVA 2026 Homepage — As of approximately May 27–29, 2026, the Gun Violence Archive reported 5,152 homicide/murder/unintentional/DGU deaths and 9,233 total injuries year-to-date in 2026, implying an injury-to-death ratio of approximately 1.79 (9,233 ÷ 5,152) for the first ~5 months of 2026.
32. Gun Violence By The Numbers in 2025 / Q4 2025 – The Trace — For full-year 2025, GVA recorded 14,651–14,655 shooting deaths and 26,237–26,238 nonfatal injuries (excluding suicides), yielding an injury-to-death ratio of approximately 1.79 (26,238 ÷ 14,655). GVA does not track suicides in either figure.
33. GVA Past Summary Ledgers / The Trace 2025 roundup — For full-year 2024, GVA recorded approximately 16,993 shooting deaths (non-suicide) and 31,810 shooting injuries, yielding an injury-to-death ratio of approximately 1.87 (31,810 ÷ 16,993).
34. Gun Violence Archive 2026 Homepage – Published May 29, 2026 — The GVA homepage (published May 29, 2026) notes that “suicide estimates from CDC extrapolations have been removed until we can get a more accurate, timely number,” confirming that both the 2026 deaths (5,152) and injuries (9,233) figures exclude suicides — consistent with prior years’ methodology.
35. By the Numbers: Gun Violence in Q1 2026 – The Trace — In Q1 2026, GVA recorded 94 mass shootings — higher than Q1 2025’s 69, but below the 2021–2023 Q1 totals which all exceeded 100. This indicates mass shootings in 2026 are not at a reduced mix relative to 2025 that would shift the injury ratio downward.
36. Gun Violence Archive – 2026 First Four Months Report (May 1, 2026) — GVA’s first-four-months 2026 report (released May 1, 2026) stated YTD deaths are DOWN 866 and injuries DOWN 1,901 compared to the same period in 2025, meaning the proportional decline in injuries (-1,901) is larger than the proportional decline in deaths (-866) in absolute terms, which tends to push the injury-to-death ratio slightly lower in 2026 vs. 2025.

wooley_rhino • run 206da147 • $3.11 • 6m3s

Extracted Percentile Values: {0.1: 620, 1.0: 720, 5.0: 840, 10.0: 920, 15.0: 970, 20.0: 1010, 25.0: 1040, 30.0: 1070, 35.0: 1095, 40.0: 1120, 45.0: 1145, 46.0: 1150, 47.0: 1156, 48.0: 1162, 49.0: 1168, 50.0: 1175, 51.0: 1182, 52.0: 1190, 53.0: 1198, 54.0: 1207, 55.0: 1215, 60.0: 1255, 65.0: 1300, 70.0: 1350, 75.0: 1410, 80.0: 1480, 85.0: 1560, 90.0: 1660, 95.0: 1820, 99.0: 2100, 99.9: 2500}

Model’s Answer:

Forecasting Gun Violence for July 1-10, 2026

Understanding the Question

I need to forecast the total number of people shot (killed + injured) in the US between July 1-10, 2026, according to Gun Violence Archive (GVA). This is a 10-day period that includes:

• Independence Day (July 4, which falls on a Saturday in 2026)
• One full weekend (July 4-5)
• A mix of weekdays

Key Data Analysis

Current 2026 Baseline: According to GVA data from May 2026, year-to-date figures show 5,153 deaths and 9,236 injuries = 14,389 total people shot through late May. This works out to approximately 96-97 people shot per day on average in 2026.

Historical July 4 Data:

• July 4, 2024: 245 people injured/killed (single day) - this was 20% below 2023 and 46% below 2020 peak
• July 4, 2021 weekend (72 hours): ~189 killed and 500+ injured ≈ 700+ people in 3 days
• July 4 is historically the most violent day, with July 5 ranking second

Recent Trends: Gun violence has declined significantly from the 2020-2021 pandemic peak. Major cities saw substantial drops in 2024-2025, with 2025 potentially setting record-low murder rates. 2026 YTD data continues showing lower deaths/injuries than prior comparable periods.

Constructing the Estimate

Base calculation (non-holiday 10 days at 2026 average): 97 people/day × 10 days = 970 people

July adjustments:

• Summer months historically run higher than average
• July 4 holiday effect (typically 2-2.5x normal)
• Weekend effect (July 4-5)
• July 3 Friday pre-holiday elevation

Day-by-day rough estimate:

• July 1 (Wed): ~100
• July 2 (Thu): ~105
• July 3 (Fri): ~130 (pre-holiday)
• July 4 (Sat): ~220-240 (holiday peak)
• July 5 (Sun): ~160-180 (post-holiday weekend)
• July 6-10 (Mon-Fri): ~105-115 each

Total estimate: ~1,100-1,200 people

Key Assumptions and Failure Points

Assumption 1: Gun violence continues declining in 2026

• Probability of significant reversal: ~15%
• If reversed, could push toward 1,400-1,700

Assumption 2: GVA methodology remains consistent

• Probability of consistency: ~95%
• Changes could shift count either direction

Assumption 3: No unusually large mass casualty event during this period

• Historically rare but would spike totals
• Probability of major outlier event: ~5%

Outcome Scenarios

Low Outcome (below 900): Would require sharper decline than observed, unusually quiet July 4 period, or reduced outdoor activity (unlikely without major event). Probability: ~5-8%

High Outcome (above 1,600): Would require reversal of declining trend, multiple large mass shootings, or unusual conditions elevating violence. Probability: ~8-12%

Final Distribution

Given the declining trend evidenced in 2026 YTD data, but accounting for the elevated July 4 period and inherent uncertainty, I’ll center my distribution around 1,100-1,200 with meaningful tails in both directions.

The default range (795-2405) appears reasonable and captures most realistic scenarios, with the lower bound reflecting possible sharp continued declines and the upper bound reflecting possible trend reversals or major events.

Percentile .1: 620 Percentile 1: 720 Percentile 5: 840 Percentile 10: 920 Percentile 15: 970 Percentile 20: 1010 Percentile 25: 1040 Percentile 30: 1070 Percentile 35: 1095 Percentile 40: 1120 Percentile 45: 1145 Percentile 46: 1150 Percentile 47: 1156 Percentile 48: 1162 Percentile 49: 1168 Percentile 50: 1175 Percentile 51: 1182 Percentile 52: 1190 Percentile 53: 1198 Percentile 54: 1207 Percentile 55: 1215 Percentile 60: 1255 Percentile 65: 1300 Percentile 70: 1350 Percentile 75: 1410 Percentile 80: 1480 Percentile 85: 1560 Percentile 90: 1660 Percentile 95: 1820 Percentile 99: 2100 Percentile 99.9: 2500

The national annual casualty baseline constrains short-run ten-day totals to a narrow envelope.

Early-July is seasonally elevated but still represents a modest share of yearly violence rather than a structural break.

Year-to-date counts through late spring sit meaningfully below the prior-year pace, pulling the expected ten-day sum downward.

The Independence Day long weekend concentrates social activity and raises daily shooting rates, producing a measurable uplift over ordinary days.

A small number of large multi-victim events would create a pronounced right tail and materially increase the ten-day total.

Post-event reporting and late data updates between July 10 and August 1 will add cases to the ledger but are unlikely to overhaul the central estimate.

Regional clustering, policing shifts, and localized outbreaks are the principal sensitivities that can swing the result by several hundred people.

Balancing the lower baseline, the holiday uplift, and reporting lags yields a central estimate near 1,240 people killed or injured, with substantial right‑hand uncertainty.

Forecast: 1,181 people shot, injured, or killed (July 1–10, 2026)

Rationale

• 2026 baseline is well-established at ~97 casualties/day through late May, representing a 13–14% decline from 2025 and continuing a multi-year downward trend since 2021. This anchors the non-holiday portion of the forecast.

• July 4th holiday creates a structural spike, historically ranging from 1.17× to 1.50× the annual daily average. The 2024 multiplier (1.17×) was unusually low; the longer-term average is ~1.30–1.40×. Weighting recent suppression against historical norms yields an expected multiplier around 1.25–1.30× for the July 1–7 period.

• July 1–7 central estimate is ~880 casualties, derived either by scaling 2024’s observed 1,096 (adjusted for the 2026 daily rate decline) or by applying a 1.25–1.30× multiplier to the 2026 baseline across 7 days.

• July 8–10 reverts toward baseline at ~100–110 casualties/day (slightly elevated due to post-holiday weekends and summer seasonality), adding ~310 to the total.

• GVA retroactive updates over ~3 weeks between the July 10 window close and the August 1 snapshot add an estimated 3–5% to initially reported counts, contributing ~40–60 casualties.

• Right-skewed tail risk from potential mass casualty events (July 4–5 historically see the year’s highest mass-shooting concentrations) pushes the 90th–99th percentiles substantially higher, but the base case remains anchored to trend and historical holiday patterns.