Deacon dashboard readings, May 3 – Jul 17, 2026 · 11 weekly windows
"By day of week" uses real dots read directly from the Deacon screenshots, dated off the label under each point (afternoon duplicate readings dropped, morning kept). But treat it as an early signal, not a finding: Sunday and Saturday are structurally over-represented because every weekly screenshot starts on Sunday and ends on Saturday, so those two days are the most likely to have a plotted dot. Wednesday and Monday, sitting mid-week, often fall in a gap between readings — sample sizes (n) are shown on every bar so you can see how thin some days are.
Covers the eight metrics with a value in every week: green speed, smoothness, trueness, air temp, precip, dew point, humidity, and wind. Air temp, dew point, humidity, and wind now come from an exact daily weather station record rather than the visually-read Deacon chart, so those four (and anything correlated with them) are on firmer ground than the turf metrics. Firmness and soil moisture are still left out — too few weeks (3–5 of 11) for a meaningful correlation.
This goes past one-to-one correlation: pick two driver metrics and an outcome, and see the outcome's average in each of the four combinations (both high, both low, and the two mixes), plus a combined-effect regression across both drivers at once.
Type in a weather scenario — temperature, rainfall, humidity, dew point, wind — and this fits a regression against the 11 weeks of history to estimate what a metric would likely read under those conditions. Check or uncheck predictors to see how the estimate and its reliability change.
With only 11 weeks of history, every predictor you add uses up one of a very small number of degrees of freedom — start with one or two predictors for a steadier estimate, and treat anything with 3+ checked as a rough sketch, not a forecast. This will get meaningfully more trustworthy as more weeks of data come in.
Firmness and soil moisture aren't being tracked consistently. Speed, smoothness, and trueness get measured almost every week; the two metrics that most directly answer "are the greens getting too soft or too hard" are the ones going uncollected. This is the biggest thing standing between you and a real firmness benchmark for RVCC — a data-collection habit issue before it's an agronomy issue, worth raising directly with Jim.
Smoothness and trueness move together strongly (r ≈ 0.80). Expected — both track ball-roll quality and both get knocked down by the same disruptive events (aeration, verticutting, heavy topdressing), then recover together.
Heat and rain together explain about 43% of green speed's week-to-week swing (R² ≈ 0.43), and rain matters more than heat. In the combined-effect regression, an extra inch of peak weekly precip is associated with roughly a 0.35 ft drop in green speed, versus about a 0.04 ft drop per extra degree of air temp — rain is nearly 9x more influential on speed than heat over this stretch. In the quadrant view: weeks that were both hot and rainy averaged the slowest green speeds, while hot-but-dry weeks stayed closer to normal — heat alone isn't the driver, heat combined with rain is.
Trueness is the metric most sensitive to rain (r ≈ -0.49) of the three ball-roll/speed metrics — a useful early-warning signal to watch in the days right after a storm, rather than waiting for the weekly average.
The mid-July week (Jul 12–17) is your best-documented week — the only one with all five metrics measured, including both firmness and soil moisture. Worth using as a template for what a fully-logged week looks like.
This analysis is limited by two gaps: some metrics aren't measured every week, and none of the weeks note what maintenance actually happened. Closing either would unlock real answers, not just better-looking charts.
Consistent firmness & soil moisture, by green. A real firmness benchmark per green instead of one club-wide number; the actual moisture-to-firmness curve for these greens (how many points of soil moisture it takes to move firmness from "suitable" into "firm"); a drought-stress threshold specific to RVCC; and a check on whether irrigation programming is delivering what's intended.
Dates for mowing height/frequency and rolling would let this tool tell a maintenance-driven dip from a weather-driven one — right now a bad smoothness/trueness week could be aeration or could be rain, with no way to separate them. Logged dates would also reveal actual recovery time after aeration/verticutting, rolling's isolated effect on speed and smoothness (separate from mowing changes), and whether a height-of-cut change produces a lasting gain or a short-term bump.
Product, date, and rate for fertilizer, PGR, and fungicide applications would show whether PGR timing matches the actual speed decay pattern rather than a calendar schedule, whether a firmness dip after a feed is a growth flush rather than weather, and whether fungicide timing lines up with the humidity/temperature conditions that predict disease pressure.
A weather-adjusted baseline. With consistent metrics and a maintenance log together, this could separate what the weather did to the greens from what the maintenance program did — which is really the question the committee is trying to answer. None of this needs new software: a shared log (date, practice, product/height/setting, which greens) alongside the existing Deacon exports would be enough to start answering these questions within a season.
The charts below use invented numbers to illustrate the kind of question this analysis could answer for RVCC once mowing, rolling, and application dates are logged alongside the Deacon readings. The shapes and values are constructed for illustration only — they are not based on RVCC's greens, staff, or performance in any way. The purpose is to show the board what becomes possible, not to report a result.
The honest version of this pitch: RVCC doesn't have this data yet. The real, factual case for building it is already in the Data wishlist and Top findings sections above — firmness and soil moisture are only measured 3–5 weeks out of 11, and no week has any record of what maintenance actually happened. That gap, not any invented chart, is the strongest reason to ask for consistent logging.