Hill Street Dog Park: Community Evidence Report

The observations support a review of how the park is configured, while preserving useful separated space. Across 576 observations from July 3 to September 25, 2026 (partial day), the main area accounted for 4,265 of 4,892 dog sightings (87.2%). It was occupied in 98.1% of snapshots. The small-dog area was occupied in 34.9%; in 371 observations (64.4%), it was empty while the main area had dogs.

Recent observations continue to show higher use of the small-dog area. The small-dog area's share was 12.8% across the full sample and 20.5% in the latest rolling four-week window. Its mean count rose from 0.5 dogs per snapshot in July to 1.1 in August to 2.4 in September to date (partial month). The report continues to treat the pattern as changing, rather than assuming early-summer use will continue unchanged.

Dog size does not fully explain area use, and counts do not establish safety. Observers recorded 788 small-dog sightings within the main area's total, compared with 627 sightings of dogs of any size in the separated area. Reported tension or conflict occurred in 39 snapshots (6.8% of 576 answered) and coincided with higher average dog counts. These observations do not establish that fencing caused conflict, that a given occupancy is unsafe, or that small dogs do not need separated space.

The next decision needs both better documentation and wider user input. On September 17, the City supplied a drawing labelled “As-Built” identifying different area figures (0.30/0.71 acres, ≈30%/70%) from the earlier preliminary drawing (0.26/0.80 acres, ≈25%/75%); its measurement method has not been independently verified. Depending on the assumed area and the observation period, calculated use per acre changes substantially — from roughly equal in the latest four weeks under a 20% assumption, to nearly 3× cumulatively under the as-built assumption. Confirmed dimensions, access and amenity assessment, the user survey, and the community vote should inform the City's review together. This interim report does not select a winning proposal.

All observation figures are calculated from the public observation data, captured September 25, 2026 at approximately 4:11 p.m. Eastern Time (20:11 UTC). The last included observation was September 25 at 10:45 a.m. Eastern Time; September 25 is a partial day, since this snapshot was captured before midnight. Voting remains scheduled to close September 27 at 11:59 p.m. Eastern Time. Ballot

This report brings together community observations, public records and the current consultation process to support a review of Hill Street Dog Park’s configuration. Its central question is how a limited neighbourhood park can provide useful separated space while meeting the needs of everyone using the main area.

The Hill Street Dog Park Community Coalition supports retaining a separated option for dogs and owners who need one. Counts help describe where dogs were observed; they do not establish why owners chose an area, the amount of space each dog needs, or the best final design. Those questions require user input and practical assessment.

This is an interim publication during community voting. The observation snapshot is dated, and no final voting or survey outcome is implied. The coalition is a volunteer community initiative independent of the City of Hamilton. Coalition purpose and process

An observation is one community-submitted snapshot of the two areas at a stated time. A dog sighting is a dog counted within a snapshot. Summing sightings can count the same dog again on another occasion; it does not measure unique dogs, visits or hours of use. An occupied snapshot has at least one dog in the relevant area.

Unless otherwise stated, this report uses the site's primary park-use sample: 425 current records submitted within 30 minutes, plus 151 historical records whose submission delay cannot be reliably reconstructed. Together these are 576 observations. A further 32 retrospective records remain valid but are included only in sensitivity comparisons.

The two maxima occurred in different snapshots. Small dogs recorded in the main area are already part of its total and are never added again. Percentages are rounded once from exact counts. Sources: observation data and published methodology.

The main area accounted for 87.2% of all dog sightings, with an average of 7.40 dogs per snapshot compared with 1.09 in the small-dog area. This is a cumulative description of the moments recorded over the study period.

The difference is not simply the result of a few very busy snapshots: the main area had more dogs in 533 of 576 observations (92.5%), the small-dog area had more in 29 (5.0%), and counts were equal in 14 (2.4%). Median counts were 7 and 0 respectively.

These figures make the distribution of usable space a reasonable subject for review. They do not imply that land should be allocated in exact proportion to recorded dog counts. Separation, avoidance of stressful interactions, accessibility and space to move can have value even at low occupancy. Data source

The main area was occupied in 565 observations (98.1%); the small-dog area in 201 (34.9%). Both were occupied in 194 snapshots. The main area alone was occupied in 371, the small-dog area alone in seven, and neither area in four.

Two similar percentages describe different things: the small-dog area was empty in 375 of 576 observations (65.1%), including the four when the entire park was empty. It was empty while the main area had dogs in 371 (64.4%).

This is evidence of unused separated space during many recorded moments alongside active use of the main area. It is not evidence that the enclosure was unused for two-thirds of all operating hours. A voluntary snapshot sample cannot determine the duration of either condition. Data source

The cumulative split conceals a material change. In the latest rolling 28-day window, 324 of 1,582 sightings (20.5%) were in the small-dog area, compared with 627 of 4,892 (12.8%) in the full sample. The recent window contains 149 observations and runs from August 28 to the September 25 capture; September 25 is a partial day.

The rolling window overlaps the monthly periods and the cumulative sample; these rows must not be added together. In that recent window, the small-dog area was empty in 76 of 149 snapshots (51.0%), compared with 65.1% overall.

The rising recorded use warrants continued observation through September 27. It could reflect changing attendance, different observation times or contributors, awareness of the enclosure, or several factors together. The data do not isolate an explanation. The recent result should receive equal consideration when assessing future space needs. Data source

Records cover 84 of 85 calendar days from July 3 through September 25, but that broad coverage does not mean continuous monitoring. July 4 has no included observation. The 5–8 p.m. period supplied 237 of 576 observations (41.1%); overnight coverage is sparse, with none from 2–4 a.m. or 6–7 a.m.

Across the recorded snapshots, the 11 a.m.–noon hour averaged 12.88 dogs across both areas (24 observations), and 6–7 p.m. averaged 11.62 (100 observations). These are observed hourly means, not forecasts of how busy any particular visit will be. At 11 a.m.–noon, the means were 11.50 main and 1.38 small; at 6–7 p.m., they were 10.23 and 1.39.

Weekdays and weekends had similar combined means: 8.55 dogs across 411 weekday snapshots and 8.34 across 165 weekend snapshots. Their small-area shares were 13.0% and 12.4% respectively. This broad comparison is descriptive; differences in which hours were observed can still matter.

The chart focuses on 7 a.m.–9 p.m., where each hourly bin has at least 12 observations. Continued collection should deliberately cover less-observed periods, quieter visits and both areas, without creating repeated submissions for the same moment. Data source

Observers recorded 788 small-dog sightings in the main area, occurring in 355 of 576 snapshots (61.6%). That is more than the 627 sightings of all dogs in the small-dog area. These are different groups: the second count includes dogs of any size and should not be labelled “small dogs.” In the latest 28-day window, the comparison reverses: 232 small-dog sightings were in the main area versus 324 dogs of any size in the separated area. That recent change is relevant alongside the cumulative finding.

The pattern supports asking owners what drives their area choice: available companions, space, shade, access, comfort, dog temperament or other reasons. Observation counts alone cannot identify those motivations or establish a preference among all small-dog owners.

The data also contain 38 large-dog sightings in the small-dog area, or 6.1% of that area's total. Physical occupancy includes those dogs. Their presence does not by itself establish a rule violation or an inappropriate use; posted eligibility and dogs' individual needs must be considered.

The September 15 edition flagged one historical record where the recorded small-dog count in the main area exceeded the main-area total. This edition's data-quality check found no such inconsistency in the current dataset. Data source

The same numeric thresholds provide a straightforward comparison:

Counts of at least 10 appeared on 64 of the 84 observed dates in the main area and five dates in the small-dog area. These are days with at least one qualifying snapshot, not the percentage of time spent at those occupancies. The busiest combined snapshot recorded 29 dogs (26 main and three small), on August 3 at 11:31 a.m. — unchanged as the all-time peak since the September 15 edition. The small-dog area’s own busiest recorded moment was different: September 16 at 7:04 p.m., with 15 dogs in that area alongside 13 in the main area (28 total).

The dashboard also uses count-based labels such as “light,” “moderate” and “busy.” Its numerical cutoffs differ by area. This report therefore uses explicit dog counts for cross-area comparisons. Neither those labels nor the thresholds above are independently established limits on safe occupancy. Data source · Dashboard definitions

Tension or conflict was reported in 39 of 576 observations (6.8%). The mean total dog count was 12.6 when conflict was reported and 8.2 when it was not, a difference of 4.4 dogs. Among the 37 reports with a location, 35 identified the main area, one the small-dog area and one both; two reports had no location.

The question records an observer's report of tension or conflict. It does not distinguish all severities, establish an injury or identify independently adjudicated incidents. Repeated observations could refer to overlapping activity. The available data also lack dog-hours of exposure, so these percentages cannot be interpreted as each dog's risk.

Higher dog counts create more opportunities for interaction and for observers to notice it. Dog characteristics, supervision, gate activity and other conditions may also matter. This association does not show that the divider caused conflict, or that a different fence line would prevent it. The sample starts after the June installation and supplies no comparable pre-installation baseline.

The practical implication is to assess entrances, circulation and usable space alongside observed occupancy and user experience. Do not turn the average occupancy in conflict reports into a “safe” density threshold. Data source

Observations tagged sunny/clear or cloudy/overcast continued to show more sightings in the main area. Mean total counts were 9.02 for sunny/clear (319 snapshots), 8.82 for cloudy/overcast (112), 6.00 for very hot (56) and 3.63 for rain (16). Their small-area sighting shares were 12.5%, 15.0%, 7.4% and 1.7% respectively.

Weather tags can overlap, so these sample counts are not mutually exclusive and must not be summed. Older records also use different wording for similar conditions. Very small categories, such as wind or smoke, cannot support strong comparisons. These observations span summer and most of September; they cannot establish winter use or isolate weather effects from time of day and contributor choices. Data source

The preliminary City-supplied drawing labels a total off-leash area of 1.06 acres, divided into 0.26 acres for the small-dog area and 0.80 acres for the main area. The labelled small-dog share is 24.5%. On September 17, 2026, the City supplied a further drawing labelled “As-Built,” identifying 0.30 acres (30%) for the small-dog area and 0.71 acres (70%) for the main area — a total of 1.01 acres, 0.05 acres less than the preliminary drawing’s total. Preliminary drawing · As-built drawing · Councillor Wilson’s reply

The public descriptions differ:

The coalition reports that the built divider sits farther east than the preliminary drawing shows. Its impression that the present enclosure is closer to one-third is not a survey measurement. The public record reviewed for this report does not establish independently confirmed as-built dimensions: the City has now stated as-built figures, supplied September 17, 2026, but their measurement method has not been independently verified by the coalition, and the 0.05-acre difference between the preliminary and as-built totals is unresolved. Any per-acre comparison therefore needs explicit assumptions; a range of scenarios is not a verified measurement range. Drawing concerns and follow-up

The site reports a September 30, 2025 creation date in the preliminary drawing’s PDF metadata, six days after the original petition began. That timing raises a records question. It does not establish when design work started, when approval occurred or that anyone acted improperly. Source context

The table holds recorded counts fixed and varies the small-dog area's assumed share of the total. Each entry is the ratio of mean dogs per acre in the main area to mean dogs per acre in the small-dog area. A ratio of 1 means equal observed use per acre under that assumption. The City's September 17 as-built figures are used as the primary assumption in this edition; the preliminary-drawing, one-third and 20% figures remain as labelled sensitivity scenarios, not verified measurements.

For example, using the as-built drawing's 0.71-acre main area and 0.30-acre small-dog area, the full-sample means become 10.45 and 3.58 dogs per acre per observation. Using the preliminary drawing's 0.80-acre main area and 0.26-acre small-dog area instead produces 9.26 and 4.19.

The cumulative main-area intensity is higher in all four scenarios. In the latest four-week sample, though, intensity is roughly equal per acre under the 20% assumption (0.97×) and only modestly higher under the as-built assumption (1.67×) — well below the cumulative 2.92×. This qualification materially changes how a single headline density ratio should be read.

These are arithmetic scenarios, not measured boundaries, confidence intervals, safe capacities or predictions of behaviour after a change. The as-built figures are the City's own stated numbers, not an independently verified survey. The actual area could differ from every description. Even equal dogs per acre would not prove that needs are equally served. Observation data · June 19 description · Preliminary drawing · As-built drawing

• September 24, 2025: The original petition sought dedicated small-dog space; it did not specify the final constructed configuration. Timeline and original petition link
• February 23, 2026: The Public Works Committee approved a motion for fencing and signage within the existing boundary, with a funding cap of C$25,000 including contingency and HST. It recorded 457 petition signatures as of January 7. Motion · Approval reported by Ward 1
• February 27, 2026: Ward 1 described a roughly one-third allocation and separate entrances. Newsletter
• June 5–19, 2026: A June 5 update scheduled construction for June 10; a June 19 update announced reopening and described a 20% allocation. Before · After
• July 12–28, 2026: The coalition’s records-request cover letter was dated July 12. The City’s July 28 acknowledgement recorded receipt on July 27, under FOI #26-256. Cover · Acknowledgement
• July 30–August 12, 2026: A drawing was supplied, followed by coalition requests for clarification on August 5, 10 and 12. Timeline
• August 13, 2026: A community meeting discussed the layout, evidence and alternatives. Meeting summary
• August 22, 2026: The user survey launched to examine reasons behind area choices. Survey · Launch record
• September 11, 2026: Community solution submissions closed. Submissions
• September 17, 2026: Councillor Wilson replied with a drawing labelled “As-Built,” identifying 0.30/0.71 acres (≈30%/70%), and said a buried gas line prompted the fence-alignment change from the preliminary drawing. Reply
• September 25, 2026: This report is republished with observations through today (partial day) and the City's September 17 reply. This report
• September 27, 2026: Voting is scheduled to close at 11:59 p.m. Eastern Time. Ballot

The ballot presents seven community-proposed options:

These are concepts for comparison, not construction plans or coalition endorsements of every feature. Their boundaries and proportions are approximate. City assessment would still be required.

Participants may support up to three options, without ranking them. Each selected option receives one support count. Published percentages should use valid ballots as the denominator and may sum above 100%. The process aims to identify three alternatives for City consideration; it is an informal preference poll, not a representative survey or a City referendum. Options and voting method

The API counts imply 626 submissions received: 617 validated public observations plus nine excluded under data-quality rules. A further nine records were removed by exact-timestamp deduplication (the same rule the live dashboard applies), leaving 608 unique records for analysis. Of those, 425 are current records inside the 30-minute rule, 151 are historical records and 32 are retrospective. Primary usage therefore contains 576 records, not all 608. The dashboard describes two of the nine quality exclusions as reports removed after someone on site disputed their accuracy; this report has not independently verified those moderation decisions.

The 30-minute threshold is the project's analysis rule, not a universal scientific standard. Retrospective records are not presumed false. Historical records are retained because their original submission delay cannot reliably be reconstructed, not because they are known to have met the threshold. Among the 457 current records with usable timing, the median recorded delay is two minutes.

The tracker allows submissions up to 24 hours after observation and asks contributors to wait 20 minutes between new observations. Those entry rules are separate from the 30-minute analytical cutoff. Tracker · Published methodology

This is a voluntary community sample. Contributors and dogs may recur, observation times are uneven, and the public data do not support a count of unique contributors or weighting by observer. Optional responses may have been normalized in older records; a zero in an exported field is not proof that every original question was actively answered. Conflict answer flags are present for all 576 included records, but that alone does not resolve historical questionnaire completeness.

The data check found no duplicate observation IDs, negative or invalid counts, missing observation times, future observations, or current timing classifications inconsistent with the supplied delay. It found no small-dogs-in-main-area count exceeding the corresponding main-area total (a single such inconsistency was flagged and noted in the September 15 edition; it is no longer present in the current dataset). These checks assess the exported records; they do not independently verify what happened at the park. Data source

Removing historical records or including later submissions changes the precise shares, but the main area remains the majority in each comparison. The 32 retrospective records considered alone have a substantially higher small-area share, 35.9% (93 of 259 sightings). Their different pattern is a reason to display sensitivity results, not a basis to dismiss those records as inaccurate.

Allowing delays up to 60 minutes while retaining historical records gives a small-area share of 13.1%; up to two hours gives 13.6%; all validated records give 14.0%. This robustness check addresses the timing rule only. It does not remove self-selection, repeated observations, changing attendance or seasonal limitations.

The same records contribute to several rows, so the samples are not independent groups and should not be added together. No population margin of error or significance test is asserted for this voluntary sample. Data source

A useful review should assess each option against the same criteria: safe separation where needed; usable running and play space; entrance capacity and double-gate arrangements; accessible movement; water, shade and seating; clarity of operating rules; maintenance; and cost, including possible reuse of fencing. Observed dog shares should inform that assessment without automatically becoming target land shares.

The next publication should preserve this dated snapshot, add observations through the voting period, explain ballot validation and counting, and report support for every option. It should state how ties among options affect which alternatives advance. Survey findings should identify the respondent groups represented and the number answering each question. Community process · Survey interpretation

The requested records should clarify the measured area and location of each enclosure, the approved construction drawing, and whether the completed installation was checked against that design. The community also needs to understand what consultation, usage evidence and needs assessment informed the allocation; which alternatives were considered; and how access, circulation and amenities were evaluated.

The absence of those answers from the public material reviewed here does not establish that no assessment or consultation occurred. It identifies information needed to evaluate the decision and future options. The published authorization is a funding ceiling; final expenditure should be confirmed separately. Records requested · Project motion

1. Continue observations through September 27. Record the counts actually present, including quiet periods and occasions when the small-dog area is busy. Follow the tracker waiting period and avoid repeated reports of the same moment. Submit an observation
2. Broaden participation. Invite users of both areas, occasional visitors and people who have stopped attending to explain their needs through the survey. Counts cannot substitute for those experiences. No survey results are asserted in this edition because no public result table was available in the material reviewed. User survey
3. Compare the proposals. Voting closes September 27 at 11:59 p.m. Eastern Time. Consider benefits, tradeoffs and feasibility across all seven options. View options and vote
4. Ask others to support the review. The petition expresses support for reconsideration and consultation. Its signatures serve a different purpose from observations or option votes and should be reported separately. Sign or share the petition
5. Publish a checked update. After voting closes and ballots are reviewed, issue a dated results addendum with every option's support, the counting rules, the final observation cutoff and available survey findings. Present the three selected alternatives to the City alongside unresolved measurement and design questions.

The aim is a park configuration that works well for different dogs and owners. The evidence supports asking better design questions and evaluating alternatives transparently; it does not supply a single answer to every user's needs.

Observation snapshot. Public observation API, retrieved September 25, 2026 at approximately 4:11:29 p.m. Eastern Time (20:11:29 UTC); observation dates July 3–September 25 (partial day). The report freezes that download. Every quantitative finding in this edition uses the downloaded snapshot. Snapshot content hash (SHA-256): fa90c352973f5399074de02088e0f0d8b4a40f6b8c1b9fb78ff9c346d20c48d1.

Calculation definitions. Sighting share = sum of an area's dog counts divided by the sum across both areas. Mean count = that area's summed counts divided by observations in the stated sample. Occupancy rate = snapshots meeting the stated condition divided by sample observations. Density ratio = (main mean ÷ assumed main acres) ÷ (small mean ÷ assumed small acres); the as-built basis converts the City's square-metre labels (1,232 m² / 2,868 m²) to acres at 1 acre = 4,046.8564224 m². The recent window is 28 × 24 hours ending at retrieval. Local dates and hours use America/Toronto. Quantiles use linear interpolation.

Method and records. Dashboard and methodology; tracker; timeline; media and source documents; preliminary reference drawing; City as-built drawing, September 17, 2026; Councillor Wilson's September 17 reply; Ward 1 newsletters of February 27, June 5, and June 19, 2026. Other primary documents are linked beside the relevant milestones.

Interpretive scope. This is a coalition publication based on publicly available records, not a City report, an independently administered census or a causal evaluation. Meeting recollections, public comments and coalition site observations do not become verified measurements by being included in a report. The September 9 media package remains a separate historical edition. No private survey responses, individual identities or ballot records are reproduced here.

Prepared for publication by the Hill Street Dog Park Community Coalition, Hamilton, Ontario. Interim edition 1.1 · September 25, 2026. Contact: [email protected].