Kickoff Scanner v.15.05

Kickoff Scanner v.14.07

CLOUD SYNC

DATA

ADMIN

DANGER

⚠ Some symbols aren't mapped to a sector.

📋 Watchlist — fetch data automatically ▼

Watchlist ⓘ 0 tickers

🔎 Kickoff candidates

Loading…

🌡 Warming candidates

Tap 🔄 to load.

⚙ Fetch data manually — only needed for a ticker the cloud has not seen yet

Period

Interval

Force refresh (re-fetch all, ignore freshness) Data merges with existing — re-runs are safe

📁 Or upload CSV files (manual) ▼

📂

Tap the button below to select CSV files
(or drag & drop on desktop)

Format: Date, Open, High, Low, Close, Volume · Min 20 rows

Results merge with existing history

⚙️ Sector mapping (advanced — usually automatic) ▼

Quick map: paste ticker(s) below — several at once is fine — pick the sector by name, tap ➕ Map. It writes the lines and saves in one step. The list uses TICKER → SECTOR_ETF per line; edit any row to override. To enable sector RS, add the matching sector ETF(s) to your watchlist and fetch them.

💾 Sector memory: Every mapping you save is remembered permanently. If you delete a ticker and re-add it later, or reset overrides, your mapping comes back automatically. Latest write wins per ticker.

For rows showing ??: paste the ticker in the quick-map row, pick its sector name, tap ➕ Map. Each ticker only needs this once — mapped forever.

🔔 Scheduled scan & notifications (worker v5) ▼

After the close, the worker reads everyone's watchlist straight from the cloud, scans them plus the Yahoo gainers screener, and pushes any kickoffs to Discord — no manual sync needed. It checks the kickoff candle only. Fresh screener finds appear in the 🔎 Candidates panel beside your watchlist — tap + on any to add it to your own watchlist.

the worker rejects /evening and /sync without a matching key. --> 🔑 Sync key — must match the worker's SYNC_KEY environment variable

📡 Latest scheduled scan ▼

What the after-close scan found across everyone's watchlists + market discovery. (P) means that stock's volume hadn't settled yet when the scan ran.

Loading…

Scan Results

Tap chip adds:

🧭 Where the trade stands

📔 My trades

🌡 Before the kickoff (watch only)

WHERE THE KICKOFF FIRED (information)

SUPERTREND (40,1)

DMA (10,50,10)

PRICE & MOVING AVERAGES (as on the card)

YOURS

CHART SLOPE (120 bars)

OPTIONS FLOW (research log, not tested yet)

PIVOT SETUP (the Discord pivot, on your cards)

PRE-EARNINGS (next 14 days)

FLOOR (3+ separated touches)

MAJOR SWING LOW (120 bars)

Compare a number

🎓 How to use this app — the complete guide

Ten minutes here explains the whole plan. Every result in this guide comes from tests on this app's own data.

0 · THE PLAN — start here

After 4:00 PM ET, the cloud updates stock_bars with the latest available Yahoo data.
Cards show ⏳ Provisional until Yahoo volume settlement is final.
When the Yahoo worker declares the day settled, the brain updates stock_bars again and cards show ✅ Final for that date.

The plan, in three rules:

Buy every kickoff. No filtering by grade, score, RS, sector, location or candle shape.
Sell into the first continuation kickoff. If that candle is climactic (≥10% on ≥2× volume), it is the best ending measured — 89% and 94% win across the two samples, ~+15–16% median, with only ~+2% more available in the following 10 days.
No continuation within 40 days → exit anyway.

Kickoff alerts now require:
• +5% or more
• volume at least 1.3× average
• green candle
• close in upper half
• prior decline at least -15%
• calm-after-decline gate:
– post-trough range ≤ 14%
– max daily swing ≤ 8%

What it delivers: 69% / +5.7% and 66% / +5.5% median across the two datasets, against 55% / +3.2% and 59% / +3.6% for simply holding 40 days — in roughly half the time in trade (~18 days vs 40). Realistic next-open fills cost about half a point: 67% / +5.4% and 64% / +4.7%.

There is no stop-loss, deliberately. Two were tested and both made things worse. A kickoff-low stop took 69% → 51%. A base-low stop made large losses more frequent (16% → 21%). About a quarter of trades dip below the kickoff low and recover — that is the retest of the accumulation zone, a normal part of the structure. Breaking that low is genuinely bad news (those trades win ~31% vs 96%), but by the time it happens you are usually near the worst price of the trade, so selling there locks in the bottom.

The one early read you do get — the week-one test. Did the stock's intraday high exceed the kickoff high at any point in the first 5 days? It is a one-time exam: pass once and you are permanently in the strong cohort (73–74% win); never pass and you are in the weak one (36–43%). Pulling back later does not undo a pass. The HOLDING banner reports exactly this.

One context signal, replicated: a kickoff in a stock more than 25% below its 52-week high reaches the continuation exit 82–84% of the time; one within 10% of the high reaches it only 39–71%. This changes how reliably the planned ending arrives, not the profit per trade.

What was tested and rejected — do not filter on these. Quality grades (all land 61–74%; marginal was best in one sample and worst in the other) · PRIME gating (keeps 28% of signals for the same win rate) · climactic entries (62% vs 68% — flips) · sector RS (flat either way) · stock RS (a drag in one sample only) · MA20 cushion and location (no gate; extended >20% was the best band, thin mildly worse) · candle wicks, body and close position (nothing; a large upper wick did fine) · pre-volume surging vs fading (the old 71%/37% split did not transfer — fading was equal or slightly better here) · moving averages and near-52w-high.

Why none of it works, structurally. The engine has already filtered hard before any grading happens: to be crowned at all, a day has to clear the kickoff algorithm's full set of structural and volume conditions. By then the ugly candidates are gone, and grading within the survivors measures noise around a filter that has already done its job.

Where the kickoff sits in the cycle. Decline → accumulation → the kickoff is the ignition out of accumulation → expansion (2–3 weeks) → crescendo. It is not the start of a smooth trend leg: five candidate "trend start" candles were tested and all returned roughly a coin flip afterwards. The expansion phase is the only part of the cycle with a measurable edge, and the continuation kickoff is its crescendo — which is why selling into it works.

Honest limits. One trade in three loses and one in six loses 10%+. A 66% win rate means losing streaks of 4–5 will happen. Both datasets are this watchlist — trend-biased names, with the top 5 tickers carrying 29–37% of total return; kickoffs on random stocks likely do worse. The recent stretch was weaker (2025: 61%, 2026: 63%, with fatter tails). Two overlapping periods, one engine, no commissions modelled. Nothing here is financial advice.

READING A CARD — what to act on, and what is only information

The plan is box 0: buy every kickoff, and sell at the first continuation kickoff or on day 40. A card tells you where each stock is in that plan.

1 · The banner at the top — the only part you act on.

▲ KICKOFF TODAY — ENTRY — the buy. When your bot is on, it buys at the next open.
◷ HOLDING — DAY 6 — you are in the trade: wait for the exit. It also tells you if the price went above the kickoff high in the first 5 days: ✓ confirmed (those won 73–74%) or ✕ never cleared (36–43%). Either way, the plan stays the same.
◆ EXIT — the first continuation kickoff fired today: sell. CLIMACTIC (up 10% or more on 2× the volume or more) is the best ending measured.
⊘ STALLED — 40 days and no continuation: sell. That is the time exit.
✓ LEG COMPLETE — the exit already fired. Anything after it is a new trade.

The strip above the banner shows the same thing in three steps: WARMING → KICKOFF → SELL. A card with no banner is not in a trade.

2 · Information only — read these, but do not pick trades with them.

Stk RS+ / RS− / neutral — how much more (or less) the stock went up than SPY over the last 21 trading days. +2% or more is RS+, −2% or less is RS−, anything in between is neutral. Tested as a filter: no help.
📊 XLK RS+ (XLK is the stock's sector fund) — the same sum for that sector fund against SPY. Tested as a filter: no help.
Position, in the card's 💓 Kickoff candle panel — how far below its 52-week high the kickoff fired: low is 30% or more below, mid is 15–30% below, high is within 15%. Over 5 years the three won almost the same: 62.0%, 62.2% and 62.9%.
deep pullback / near 52w high on the banner — a kickoff far below its high reached the planned exit more often (82–84%) than one near its high (39–71%). It changes how often the exit comes, not how much a trade makes.
The candle panel's badge and grade (PRIMARY, BEAR-TRAP, EXTENDED-BASE · textbook, standard, explosive, marginal) — they describe the candle. No grade changed the result.
⛏ base depth — how far the stock fell before the kickoff, and how often bases that deep reached their exit. Information only, not a reason to skip a trade.
✅ / ⚠ calm — how quiet the stock was before the kickoff, with the numbers.
The action badge (NO SETUP, STILL GOOD and so on) — only on a card with no banner. An older timing label, kept to read, not to act on.

3 · Your positions — each shows its gain or loss at the last close.

📔 +3.2% in the top corner — you hold it in your trade log. Tap it for the log.
🤖 +3.2% next to the price — your bot holds it. Tap it to open the bot. dry in the pill means a dry-run position, with no real shares.

4 · Watch, do not buy.

🌡 WARMING — a deep fall that has gone quiet. About 3 in 4 fire a kickoff within 30 days, but buying before the kickoff did worse than waiting for it.
📉 RSI(2) (Discord only) — a buy alert when RSI(2) closes under 10 above the 200-day average, and a sell alert when it closes over 70. Liquid stocks only. It posts to its own Discord channel, not to the app, and the bot does not buy it.

Not measured here: an earnings report during the hold can move the price a lot either way. The 📅 earnings words show reports in the next 14 days.

1 · What the app does

It watches the stocks on your watchlist and finds kickoffs: a day when a stock that has fallen at least 15% jumps 5% or more on heavy trading, after a calm spell (box 0 has the exact rules). Each kickoff starts one trade. The trade ends at the stock's next kickoff, called the continuation, or after 40 trading days (about 8 weeks) if none comes. Every card shows where its stock is in that trade, so there is nothing to work out yourself. The numbers in this guide come from two tests on past prices: 13 months and 5 years.

2 · Each evening, after the close

After 4:00 PM ET the cloud loads the day's prices by itself. A card shows ⏳ Provisional until the day's volume is final, then ✅ Final.
The scan runs by itself. 📡 Latest scheduled scan shows what it found, across everyone's watchlists plus new finds. (P) means that stock's volume was not final yet when the scan ran.
Sort by 🧭 Journey phase. ◆ EXIT cards come first, then ▲ ENTRY, ◷ HOLDING, ⊘ STALLED and ✓ LEG COMPLETE. ◆ EXIT and ⊘ STALLED mean sell — the STALLED cards come after the HOLDING ones, so read past those. ▲ ENTRY means buy, and ◷ HOLDING needs nothing. The "Reading a card" box above explains each one.
If you use the bot, its plan for the next morning is ready at about 5:30 PM ET, and you can veto any of its orders until 9:20 AM ET.
Log the trades you make yourself in 📓 Trade Log. A stock you hold there shows 📔 and its gain or loss on its card.

3 · Buying

Buy every ▲ ENTRY at the next morning's open. Do not skip the ones that look weaker. Many ways of picking were tested — candle grades, strength against the market or the sector, distance from the 52-week high, candle shape — and none did better than taking them all.
Expect a little less than the headline. Buying at the next open instead of the kickoff's close gave 67% winners instead of 69%, and +5.4% instead of +5.7% for the middle trade (half the trades did better, half did worse). In the 5-year test: 64% and +4.7%, instead of 66% and +5.5%. These are the numbers to expect.
Put the same amount of money in each trade — that is how the tests counted, and how the bot buys. There is no stop-loss (section 4), and one trade in six lost 10% or more, so choose an amount you can watch fall without wanting to break the plan. Many trades will be open at the same time.
Not buys: 🌡 WARMING, shown so you can watch it, and the 📉 RSI(2) alerts in Discord. A near-miss in the scan list is not a buy either, unless its own card shows ▲ ENTRY.

4 · Holding — no stop-loss

While the banner says ◷ HOLDING, do nothing. There are only two ways out (section 5).
There is no stop-loss, on purpose. A stop-loss is an order that sells by itself when the price falls to a set level. Two were tested, and both did worse. A stop at the kickoff day's low cut the winners from 69% to 51%. A stop at the bottom of the base — the quiet stretch before the kickoff — made big losses more common, not less: from 16% to 21% of trades.
A dip below the kickoff day's low is normal. About one trade in four goes below it and then recovers. Trades that break it do win less often (about 31%, against 96% for the rest), but by then the price is usually near its lowest point, so selling there would lock in the loss.
The week-one test. Did the price go above the kickoff day's high at any time in the first 5 days? Those trades won 73–74%; the ones that never did won 36–43%. The banner shows ✓ confirmed or ✕ never cleared. It tells you which trades are likely to work, but it does not change the plan.
Do not sell early because of a red day, a scary-looking candle or a dip. In older tests these were no better than a coin flip at calling the end of a move.
Earnings reports can move a price a lot in one day, either way. The plan has no earnings rule, because it was never tested. All signals → 📅 Reports within 14 days shows which of your stocks report soon, so you can decide for yourself.

5 · Selling — two ways out

◆ EXIT — the stock's next kickoff (the continuation) fired today: sell at the next morning's open. It marks the peak of the strong part of the move, which is why selling into it works.
CLIMACTIC on the exit banner — that day was up 10% or more on at least twice the usual volume. It is the best ending measured: about 9 trades in 10 made money (89% and 94% in the two tests), about +15–16% for the middle trade, and only about 2% more was there to gain in the 10 days after. Take it.
⊘ STALLED — 40 trading days and no continuation: sell anyway, and do not wait for one.
✓ LEG COMPLETE — the exit already fired. Anything after it is a new trade.
Why not just hold? Holding every trade for 40 days won less often (55% and 59%, against 69% and 66% for the plan) and made less (+3.2% and +3.6% for the middle trade, against +5.7% and +5.5%). With the plan, a trade lasts about 18 trading days on average.

6 · The bot

🤖 Bot, in the ⋯ More menu, runs the plan for you on an Alpaca paper account — practice money, never real money — with one change: it never sells below your buy price unless you confirm it. Dry run plans every evening and records what would have happened at the real opening price, without sending any order. Paper sends the orders. Off does nothing.

Three tabs: Trading (the mode, the plan and your open positions), Results and Settings. The window opens on Trading. A search finds a stock on all three.
Its day: at 4:05 PM ET Discord gets a list of your open positions at the close, the ones down your flag % or more first (15% unless you change it in Settings). At about 5:30 PM ET the list comes again at the final prices, and then the plan for the next morning, in the bot panel and in Discord. At 9:00 AM ET a reminder comes. At 9:10 it checks the early prices for anything that needs your OK. At 9:20 it sends every order you did not veto, and at 9:40 it records what filled and cancels the rest.
Veto any order in the bot panel before 9:20 AM ET.
It buys every kickoff on your watchlist that it does not already hold, with the same money for each, up to your limits. The order has a top price: the kickoff's close plus your gap limit (3% unless you change it). If the price stays above it until 9:40, nothing is bought — unless you confirm the buy, which sends it at any price.
Add: kickoffs over your limits are skipped. Veto a buy to free a slot, then tap Add on a skipped one to buy it instead, before 9:20 AM ET. Nothing goes over your limits: Undo on a vetoed buy also needs a free slot.
It sells at the first continuation that closes above your buy price, or after day 40. A sale has a bottom price — your buy price — so the bot never sells at a loss by itself. After day 40, a stock still below your buy price is kept, and the bot asks you every evening until you confirm the sale or the price is back above your buy price.
Sell at next open sells one stock by hand, at your buy price or better. Hands off means the bot never sells that stock — you manage it yourself.
Sell at break-even is offered on any stock below your buy price. The flag (down 15% or more, unless you change it in Settings) points out the worst ones. From the next 9:20 AM send, a sell order at your buy price works all day, every day, until it sells — it never sells lower. It is in the plan each evening, so you can veto a day. A continuation above your buy price still sells it at the open. Cancel break-even stops it.
Stocks you buy yourself in the Alpaca account are taken on in the evening and run the same way from the next day. A stock it manages that is missing from your watchlist is added to it.
Results: what is open now and what was closed, the profit and loss in all, by stock, by how long trades were held and by how they were sold, and every closed trade, a month at a time (tap a month to open it). A paper account, a dry run and a real account are never mixed.
Settings: the mode (Off, Dry run or Paper) and your Alpaca connection, money per trade, the most trades open at once, the most new buys per day, your gap limit, how far down a position is flagged (15% unless you change it), and your own Discord channel for its messages.

7 · What to expect — honestly

About two trades in three make money when bought at the next open (67% and 64% in the two tests), and the middle trade made +5.4% and +4.7%. One trade in three loses, and one in six loses 10% or more.
Losing streaks of 4 or 5 trades will happen, even when everything works as tested. Judge the plan over many trades, not over one week.
Results lean on a few stocks: in the tests, the 5 best stocks made 29–37% of the total return.
The recent stretch was weaker: 61% in 2025 and 63% in 2026, with bigger swings both ways.
The tests have limits. The stocks were picked for watchlists, which lean towards stocks that trend, so kickoffs on random stocks would likely do worse. Trading costs were not included. None of this is financial advice.

8 · Words used in the app

Kickoff — the buy day: a jump of 5% or more on heavy trading after a fall and a calm spell. Box 0 has the exact rules.
Continuation — the stock's next kickoff after you bought. The plan sells there.
Climactic — a day up 10% or more on at least twice the usual volume.
Candle — one day on the chart: where the price opened, its high, its low and where it closed.
Base — the quiet stretch after the fall, before the kickoff.
Stop-loss — an order that sells by itself when the price falls to a set level. The plan has none.
The two tests — the plan replayed on past prices: 13 months (about 270 trades) and 5 years (about 380 trades, on mostly different stocks).
The middle trade — line all the trades up from worst to best: half did better than the one in the middle, and half did worse.
Volume 2× — twice the stock's usual trading, compared with its average over the last 20 days.
Day 40 · D+5 — trading days since the kickoff. D+0 is the kickoff day itself.
Week-one test — whether the price went above the kickoff day's high in the first 5 days.
52-week high · low / mid / high position — the highest price of the past year, and how far below it the kickoff fired: low is 30% or more below, mid is 15–30% below, high is within 15%.
SPY — a fund that holds 500 large US companies. It stands in for "the market".
RS — how much more (or less) a stock went up than SPY over the last 21 trading days.
Sector fund (XLK, XLV …) — a fund that holds one industry's stocks, such as technology or health care.
MA20 · MA200 — the average closing price over the last 20 or 200 trading days.
PRIMARY · BEAR-TRAP · EXTENDED-BASE — the three shapes the engine finds: a kickoff after a fall and a quiet base; one that first dipped below a price where buyers had stepped in before, then came back; and one that breaks out of a long sideways stretch. They describe the chart.
textbook · standard · explosive · marginal — grades of the kickoff candle, from its size, its volume and where it closed. None changed the result.
⏳ Provisional · ✅ Final — whether the day's volume figures have settled.
💰 Fundamentals, in an open card — sales, profit, cash, debt and price against earnings, from a free data source, scored 1–5. Information only: never tested as a way to pick trades.

What we tested, and what we found. Every result behind the plan, with the size of the test where it is known. What worked is kept as evidence. What failed is kept so it is not tried again. How to trade the plan is in the user guide.

✅ The plan's own results

Two tests, replayed one day at a time, so each trade knew only what was known on its own day: 13 months (June 2025 to July 2026, about 270 trades) and 5 years (a second set of 131 mostly different stocks, about 380 trades). The two periods overlap.

Buy every kickoff, sell at the first continuation or after day 40: 69% won, +5.7% for the middle trade (13 months) · 66% won, +5.5% (5 years).
Just holding 40 days: 55%, +3.2% · 59%, +3.6% — worse, and in more than twice the time (40 days against about 18).
Buying at the next open instead of the kickoff's close: 67%, +5.4% · 64%, +4.7% — the realistic numbers.
A climactic continuation as the exit (up 10% or more on 2× volume or more): 89% · 94% won, about +15–16% for the middle trade, with only about 2% more to gain in the next 10 days. The best ending measured.
Losses: one trade in three lost, and one in six lost 10% or more. At a 66% win rate, losing streaks of 4–5 should be expected.
Where the profit came from: the 5 best stocks carried 29–37% of the total return.
By year: 61% in 2025 and 63% in 2026, weaker than the full period, with bigger swings.

Both sets are stocks picked for watchlists, which lean towards stocks that trend, and trading costs are left out. Kickoffs on random stocks would likely do worse.

🛑 Stop-losses and early exits — all did worse

A stop-loss at the kickoff day's low: winners fell from 69% to 51%.
A stop-loss at the bottom of the base: big losses became more common, not less — from 16% to 21% of trades.
Why stops fail here: about one trade in four dips below the kickoff day's low and then recovers. Trades that break it do win less (about 31%, against 96% for those that never do), but the break usually comes near the lowest price of the trade, so a stop sells the bottom.

An older test points the same way (396 kickoffs, July 2026, a different exit with a stop at the kickoff low):

Raising the stop to the buy price once a trade was up 10% cut the winners (48% → 44%) and nearly doubled the trades that were up 10% and still finished red (4% → 7.8%).
Selling on a close below the 20-day average lowered the average result per trade by 2.3 percentage points, because it sold healthy dips in the biggest winners.
Red candles, scary wicks, two or three down days and heavy-volume down days were no better than a coin flip at calling the top.

❌ Ways of picking trades that did not help

None of these picked better trades — which is why the plan buys every kickoff. Most were tried as filters in the plan's two tests; the rest come from other studies, and say so. Several had looked good in older tests with a different exit; both results are shown.

Candle grades (textbook, standard, explosive, marginal): all won 61–74%, and the best grade in one test was the worst in the other.
PRIME candles (+6–10% on 1.5–2.5× volume): buying only these kept 28% of the trades for the same win rate. Older test: 63% against 52%, on 1,900+ kickoffs.
Climactic kickoffs as a buy: 62% against 69% for all trades in the 13-month test, but the 5-year test went the other way. Not a reason to skip one, or to favour one. Older test: 45%, with a loss for the middle trade.
Sector strength: flat either way (68% and 66% for one group, 67% and 67% for the other). The older claim that it was the strongest filter rested on only 32 kickoffs.
Stock strength against SPY: stocks weaker than SPY did a little worse, in one test only. Older test: inside PRIME candles, strength added under half a point — nothing.
Distance above the 20-day average: no cut-off worked. More than 20% above was even the best band (76% and 70%), and just above the average was slightly worse. Older test: every band won 51–54%.
Moving averages: no filter worked. Older test: above the 200-day average won 59%, against 52% below it.
Distance from the 52-week high, in a separate, larger 5-year file (3,449 kickoffs, the same exit, about 62% won overall): low position 62.0%, mid 62.2%, high 62.9%. Older test: near the high won 54%, below average, once the kickoff day itself was left out of the high.
Candle wicks, body and close: nothing. A large upper wick did fine. Older test: a small upper wick looked better (61% against 51%), but as a filter it cost 40% of the trades for a gain that could be chance.
Volume in the days before the kickoff: rising or fading made no difference (fading was equal or slightly better in both tests). Older test, with a different exit: the biggest split ever found, 71% against 37% on 833 kickoffs — it did not carry over.
The medal, the golden setup and the 0–10 score were built from the filters above, and are gone. Older test: the score ranked trades slightly worse than chance (0.484, where 0.5 is a coin flip).
OBV (a running total of volume on up days minus down days). Older test (v12, 1,365 cases): 51% won when it disagreed with the price, 52% when it agreed — no difference. Removed.
The incubation chip (🥚, a tight range before the kickoff): removed in v12.3 after failing an older test.
The demand day (a quiet +6% day near the low, before a kickoff): retired in October 2026 (app v14.81, brain v14.33) as a failed signal. Its tracking rows were deleted.

Why none of them work: to be called a kickoff at all, a day must already pass a long list of price and volume rules. The weak candidates are gone before any grading starts, so grading the ones left mostly measures noise.

⏳ Before the kickoff — warming and waiting

🌡 Warming (435 lows on 229 stocks, July 2026). After a real fall, a fresh 60-day low turned into a kickoff 49% of the time. Only two things raised that: a deeper fall (down 37% or more: 73% fired; down less than 18%: 21%) and a daily range that was shrinking (61% against 34%). Together: 74% fired within 30 days, about 8 days before the kickoff in the middle case, and it held in both halves of the sample.

But no early buy beat waiting, in any test. Buying a warming stock meant sitting through a dip of about 8% (middle case) for a head start of about 5%. The kickoff then fired below that price 32% of the time, and never fired 26% of the time. Waiting for a dip under the low and buying the recovery was worse (a 12% dip, and only 42% fired). The price broke below the "low" 77% of the time even when it worked — only the kickoff shows that the selling has stopped.

How long after the low — an older study (6,767 lows). A kickoff came within 10 trading days only 15% of the time, within 20 days 27%, within 60 days 56%, and never 16%. The middle wait was 35 trading days, about 7 weeks. A kickoff soon after the low did no better than a late one (the win rates were within one point). One bucket — within 5 days of the low — won 68% against 57% for all kickoffs in that study, but it was one of five buckets tried, so it may be chance.

📉 RSI(2) (tested October 2026): over 5 years of our stocks, buying when RSI(2) closed under 10 above the 200-day average and selling when it closed over 70 won 64% of the time and made +0.8% a trade over about 4 days. Random stocks bought and sold on the same days did about as well, so it is posted to Discord to watch, not bought by the bot. It took the place of the capitulation bounce.

🔁 During the trade — what the first days tell you

The week-one test: if the price went above the kickoff day's high at any time in the first 5 days, the trade won 73–74%; if it never did, 36–43%. Found in both tests. A later pullback does not undo a pass.
The midpoint (13-month test only): above the middle of the kickoff candle in the first 5 days, 79% won; below it, 57%.
A kickoff far below its high reaches the planned exit more often: more than 25% below its 52-week high, the continuation came 82–84% of the time in both tests; within 10% of the high, only 39% (5 years) and 71% (13 months). It changes how often the exit arrives, not the profit per trade.
The cycle: a fall, a quiet base, the kickoff, a rise of 2–3 weeks, then the continuation — often the strongest day. Five other candles were tried as "the start of the trend", and each did no better than a coin flip after it.

🗂 Older results, not yet re-tested on the plan

From earlier studies, with exits different from today's plan. They were never re-run with today's exit, so they are ideas to re-test, not rules.

Volume after the kickoff (bought on day 3): when the 3 days after kept at least 85% of the kickoff day's volume, 65% won; when it fell below 60%, 53% won. A small effect that could still be chance.
Buying late (1,464 kickoffs): with the kickoff low unbroken, buying on any of the first 8 days won about as often as day 0 (54–56%); it faded after that and was gone by about day 13.
Broke the low, then hung on a moving average: the worst state measured (43%, on 188 cases), worse even than a stock that crashed straight through (54%).
A weak market: kickoffs fired while SPY was RISK-OFF won more often than in a strong market (58% against 51%).
Kickoffs come in waves: sector funds rise 1% or more on 18% of days, yet 48% of kickoffs land on those days.

📒 How the tests were run

One day at a time, no hindsight. Each kickoff was found on its own day using only the prices known then, and the trade was followed forward from there.
"Won" means the trade was sold above its buy price. "The middle trade" is the one in the middle when all trades are lined up from worst to best.
Checked where possible. The main results held in both tests (whose periods overlap) or in both halves of one. A result from one test only names that test, and one that could easily be chance says so.
Limits. Stocks that stopped trading are missing from the data, so results lean a little high. The stocks were picked for watchlists and lean towards stocks that trend. Trading costs and taxes are left out.
Live record. The brain also logs every new kickoff and how it ends, as it happens — a record with no hindsight at all.

▶📊 Sector Strength

📡

No scans yet — upload CSV files to begin