Sprint 8 — vW8 — Catch Up, Then Read the Volume
Every team finishes a gold page that passes the live test. Then every team adds the one thing a price-only page cannot tell you: how much conviction is behind the move.
Today, Monday 2 Nov: live test run sheet
About 15 minutes per team, using the 10-slide template, with the steps in this exact order. Each team has its own colour. The checker’s R4 is outlined in bold: in Practical A, the checker is always the R4 of a different team.
Team 1 · Team 2 · Team 3 · Team 4 ·
Practical A (in this order)
Slot 1Team 1 presents checked by Team 3 · R4
- 1Team 1 · R1 presents the team’s Step A status (slide 3).
- 2Team 1 · R6 starts the data script and the local server live. The bar time must be from today.
- 3Team 1 · R5 opens
index.html, then the gold page, and clicks through all seven tiles. - 4Team 3 · R4 (checker) names a timeframe and compares it with Investing.com live, on screen.
- 5Team 3 · R4 (checker) draws at random, on screen, one indicator row and one presenting member. That member explains the row.
- 6Team 1 · R3 explains the volume label: why it says what it says.
- 7Team 1 · R4 explains the biggest mismatch in
verify.md. - 8Team 1 · R2 presents the four weekly calls.
- 9Team 1 · R1 closes with teamwork evidence and the team’s first fix for Sprint 9.
After class: Team 1 · R4 commits docs/live-test.md today. Team 3 · R4 (checker)'s team commits docs/peer-check-team-1.md: the eight checks and the volume panel, each pass or fail, with one line of reason.
Slot 2Team 2 presents checked by Team 1 · R4
- 1Team 2 · R1 presents the team’s Step A status (slide 3).
- 2Team 2 · R6 starts the data script and the local server live. The bar time must be from today.
- 3Team 2 · R5 opens
index.html, then the gold page, and clicks through all seven tiles. - 4Team 1 · R4 (checker) names a timeframe and compares it with Investing.com live, on screen.
- 5Team 1 · R4 (checker) draws at random, on screen, one indicator row and one presenting member. That member explains the row.
- 6Team 2 · R3 explains the volume label: why it says what it says.
- 7Team 2 · R4 explains the biggest mismatch in
verify.md. - 8Team 2 · R2 presents the four weekly calls.
- 9Team 2 · R1 closes with teamwork evidence and the team’s first fix for Sprint 9.
After class: Team 2 · R4 commits docs/live-test.md today. Team 1 · R4 (checker)'s team commits docs/peer-check-team-2.md: the eight checks and the volume panel, each pass or fail, with one line of reason.
Slot 3Team 3 presents checked by Team 2 · R4
- 1Team 3 · R1 presents the team’s Step A status (slide 3).
- 2Team 3 · R6 starts the data script and the local server live. The bar time must be from today.
- 3Team 3 · R5 opens
index.html, then the gold page, and clicks through all seven tiles. - 4Team 2 · R4 (checker) names a timeframe and compares it with Investing.com live, on screen.
- 5Team 2 · R4 (checker) draws at random, on screen, one indicator row and one presenting member. That member explains the row.
- 6Team 3 · R3 explains the volume label: why it says what it says.
- 7Team 3 · R4 explains the biggest mismatch in
verify.md. - 8Team 3 · R2 presents the four weekly calls.
- 9Team 3 · R1 closes with teamwork evidence and the team’s first fix for Sprint 9.
After class: Team 3 · R4 commits docs/live-test.md today. Team 2 · R4 (checker)'s team commits docs/peer-check-team-3.md: the eight checks and the volume panel, each pass or fail, with one line of reason.
Practical B
Team 4Team 4 presents checked by Team 4 · R4
- 1Team 4 · R1 presents the team’s Step A status (slide 3).
- 2Team 4 · R6 starts the data script and the local server live. The bar time must be from today.
- 3Team 4 · R5 opens
index.html, then the gold page, and clicks through all seven tiles. - 4Team 4 · R4 (checker) names a timeframe and compares it with Investing.com live, on screen.
- 5Team 4 · R4 (checker) draws at random, on screen, one indicator row and one presenting member. That member explains the row.
- 6Team 4 · R3 explains the volume label: why it says what it says.
- 7Team 4 · R4 explains the biggest mismatch in
verify.md. - 8Team 4 · R2 presents the four weekly calls.
- 9Team 4 · R1 closes with teamwork evidence and the team’s first fix for Sprint 9.
After class: Team 4 · R4 commits docs/live-test.md today: the eight checks and the volume panel, each pass or fail, with one line of evidence, plus what the team fixes first.
Fair checking counts (Practical A). A checker who passes something that clearly fails, or fails something that clearly passes, counts against the checking team.
This sprint has two steps, in order. Step A — finish Sprint 7: your
ui/gold.html passes all five Sprint 7 rules. Step B — add volume: a Price & Volume panel on the same page (section 3). Teams that passed Monday’s live test start at Step B on day one. Teams that did not pass do Step A first; a reference engine is available to help (section 2). Nobody is left behind, and nobody gets credit for Step B while Step A still fails.Why. By Saturday, the Sprint 7 repos showed most teams close but not finished: needles pointing the wrong way, a Summary built from only some of the votes, missing team pages, files appearing on Saturday night. This sprint gives everyone the time to close those gaps — and the Wednesday checkpoint is now checked automatically from GitHub, so the work has to happen during the week, not at the end.
1. Where each team starts
Every team starts at Step A and checks itself against the eight checks in section 2. A team that passes all eight at Monday 26 Oct's live test starts at Step B straight away.
2. Step A — finish Sprint 7 (catch-up)
The target is unchanged: the Sprint 7 brief, sections 2–7, and its five pass-or-fail rules. Check yourself against this list before Wednesday:
| # | Check | How it is tested on 2 Nov |
|---|---|---|
| 1 | Seven tiles, each showing its own Summary rating | R5 clicks every tile live |
| 2 | Summary gauge = all 23 votes (11 indicators + 12 moving averages; ATR does not vote) | R3 adds up the Buy / Neutral / Sell counts on screen |
| 3 | Needle points at the score: −1 far left (Strong Sell), +1 far right (Strong Buy) | R4 shows needle and pill agree |
| 4 | Price, change and % change from the previous daily close | R4 compares with Investing.com live |
| 5 | Bar time = time of the last completed bar, in SGT (not the time the file was written) | R6 explains what the bar time means |
| 6 | Every indicator row opens its working; every moving average shows Buy/Sell | A row drawn at random is opened and explained |
| 7 | docs/verify.md: 30 Min, Hourly, Daily at the same bar time, with screenshots, mismatches explained | R4 shows the biggest mismatch |
| 8 | ui/index.html is the team dashboard and links to ui/gold.html | R5 opens index.html first |
Reference engine (optional). Teams may start from the course’s reference indicator engine (
tech_summary.py, released on the course site on Monday). Conditions: credit it at the top of your README; the page, the data script and verify.md are still yours; and on 2 Nov every member explains one indicator from it live. Copying it without understanding it fails in front of the class.No public hosting — this includes AI suggestions. Your page runs from your laptop through the local server (
python -m http.server), with the data script running live. Do not put it on Netlify, Vercel, GitHub Pages or any other hosting site, even if ChatGPT or another AI tool recommends it — an AI only knows this rule if you give it this brief. The only workflow allowed in your repo is .github/workflows/predict.yml. Why: a hosted page shows a copy that cannot be matched to your repo or proven live, and it puts your names on a public website. Teams will be asked on 2 Nov: “Where does this page run, and why?”3. Step B — add a Price & Volume panel
Every rating on your page is built from price. Volume tells you whether real money is behind the move. Add this panel to ui/gold.html, directly under the bar time, so it changes with the selected tile. This is what it must look like:
| Part | Exactly what it shows |
|---|---|
| Candles | The last 60 completed bars of the selected timeframe (green up, red down) |
| Volume bars | Under the candles, one bar per candle, same colour as its candle |
| Average line | Dashed line: the average volume of the 20 bars before each bar |
| Time axis | Times in SGT along the bottom (dates for 4 Hours and Daily) |
| Relative volume | last completed bar’s volume ÷ average of the 20 bars before it, shown as e.g. “1.62×” |
| Volume label | One of four, by this rule (same for every team): Confirmed — relative volume ≥ 1.5 and the last bar moved in the Summary’s direction Warning — relative volume ≥ 1.5 and the last bar moved against the Summary Thin — relative volume < 0.7 (low conviction) Normal — anything else |
- Data: Yahoo already gives you volume in the same download (
Volumecolumn). For 4 Hours and Daily, add up the volume of the 1-hour bars you combine. - Write it into
data.jsonper timeframe: the 60 bars (time, open, high, low, close, volume), the 20-bar average, relative volume and label. The page only readsdata.json. - Gotcha: the last bar of the week (Friday night) or a bar still forming has partial volume. Use completed bars only, and say so on the page.
- Verify: add one line per timeframe to
verify.md: your volume for one bar vs the volume shown for the same bar on Investing.com’s gold chart.
Stretch (optional, for teams that finish early): a chart of how the Summary score moved over the last 60 bars, drawn as a smooth line across shaded Strong Sell → Strong Buy bands.
4. How the team works this week (all teams)
1Working agreement by Tuesday 23:59 on FigJam: how you communicate (Discord), how decisions are made, what happens when someone goes quiet. Every member’s name on it.
2Sprint board: every member’s name on at least one card with a Definition of Done and a link to its commit.
3Stand-ups: three lines in your Discord channel (done / next / blocked) on at least three days.
4Everyone commits under their own GitHub account. No member makes more than 40% of the team’s commits this sprint.
5Branches, then merge: work goes on a branch and is merged into
main by R6 — not pushed straight to main.6Wednesday 28 Oct 12:00 checkpoint, checked automatically: Step A teams — the eight checks in section 2 at least half done; Step B teams — the volume panel showing on one timeframe.
5. Roles
R1Working agreement, sprint board, stand-up log; lock the weekly calls in
predictions/vW8.json by Wednesday 23:59; Wednesday check-in.R2Your weekly call; then the price line and change (check 4).
R3Indicators and their working (checks 2, 6); the volume label rule.
R4
verify.md, including the new volume lines. You are the auditor: in Practical A you check another team in rotation (1→2, 2→3, 3→1); in Team 4 you check your own team. You also commit your team’s docs/live-test.md.R5
gold.html: needles, tiles, and the Price & Volume panel.R6
dashboard_data.py: the 60-bar series, volume, 4 Hours and Daily volume; branches and merges.6. Files and deadlines
ui/index.html R1/R5 · team dashboard, links to gold.html ui/gold.html R5 · Step A checks + Price & Volume panel ui/data.json R6 · adds series + volume per timeframe engine/dashboard_data.py R6 docs/verify.md R4 · Saturday 31 Oct 23:59 predictions/vW8.json R1 · Wednesday 28 Oct 23:59 checkin/wednesday.md R1 · Wednesday 28 Oct 12:00 FigJam working agreement R1 · Tuesday 27 Oct 23:59
7. Monday 2 Nov: the team-run live test
Colour-coded run sheet for today, at the top of this page ↑
Each team runs its own live test in front of its practical class, about 15 minutes, following the same steps in the same order. Every role has a job.
| Practical | Who checks | How |
|---|---|---|
| Practical A: Teams 1, 2, 3 | Another team, in rotation: Team 1 checks Team 2, Team 2 checks Team 3, Team 3 checks Team 1 | The checking team’s R4 does steps 4 and 5 below, and the checking team commits docs/peer-check-team-N.md (N = the team it checked): each of the eight checks and the volume panel, pass or fail, one line of reason. |
| Team 4 | Team 4 itself | Team 4’s own R4 is the auditor and does steps 4 and 5. |
Every team uses the same 10-slide presentation template (download the template, PowerPoint), in the same order as the steps below. Replace every [bracketed] placeholder; do not add, remove or reorder slides. Commit the finished deck as
docs/sprint8-presentation.pdf by Monday 2 Nov 09:00.- R1 presents the team’s Step A status (slide 3).
- R6 starts the data script and the local server live. The bar time must be from that day.
- R5 opens
index.html, then the gold page, and clicks through all seven tiles. - The checker’s R4 (the checking team in Practical A; Team 4’s own R4) names a timeframe and compares it with Investing.com live on screen.
- The checker’s R4 draws at random, on screen, one indicator row and one presenting member (for example with a random-number generator); that member explains the row. Teams using the reference engine: every member explains one indicator.
- R3 explains the volume label: why it says what it says.
- R4 explains the biggest mismatch in
verify.md. - R2 presents the four weekly calls.
- R1 closes with teamwork evidence and the first fix for Sprint 9.
- The presenting team’s R4 commits
docs/live-test.mdthe same day: each of the eight checks and the volume panel, pass or fail, with one line of evidence, plus what the team fixes first. In Practical A, the checking team also commits itsdocs/peer-check-team-N.md.
Fair checking counts (Practical A). A peer check that passes something that clearly fails, or fails something that clearly passes, counts against the checking team.