Methodology: How BTC 15-Minute Up/Down, Win Rate and Net Wins Are Calculated

Public Methodology and Statistical Definitions

Positioning: a market-information and reference research tool, not investment advice

This page publicly explains 24K's data sources, time windows, direction rules, signal results, statistical methods, page markers and update mechanism, so users can understand, verify and trace them independently.

Core principles: transparent data, consistent definitions, settlement at close, traceable results.

1. Data Sources and Time Basis

24K's real-market reference data comes from public quotes of major exchanges, e.g. Binance. The actual data source, trading pair and fallback strategy are as disclosed on the product page at the time.

The system reads quotes in fixed absolute time windows and uses data validation, anomaly checks and multi-source fallback to minimise the impact of a single faulty source, network latency or brief glitches.

Underlying data, time windows, direction rules and statistical definitions are applied uniformly. Pages show times in the visitor's local time, and dates in the Research Center and Live Table follow the local date; this conversion never changes the underlying time windows, direction rules or historical results.

2. 15-Minute Periods and Direction

24K uses 15 minutes as one observation period. Each calendar day contains 96 periods of 15 minutes.

The final direction of each valid cycle is judged strictly by the result of the matching Binance candle:

Close above open: recorded as Up；
Close below open: recorded as Down。

A direction is only settled after the period closes. Confirmed historical results are never altered afterwards because of later price moves.

3. What "In Progress" Means

In progressmeans the current 15-minute period has not closed yet and its final direction is still pending.

A period in progress:

is not counted as a hit;
is not counted as a miss;
is not counted in net wins;
is not counted in the win rate;
is not counted in the Research Center or final historical statistics.

After a period closes, the system reads and confirms the actual result, then updates the statistics. This avoids drawing conclusions from an unfinished price path.

4. Teams and Start-Time Sections

4.1 Teams

Teams are labels 24K uses to group different time systems and signal sections. Currently:

A team is not a method in itself and does not represent any up/down conclusion. After choosing a team, you can view its start-time sections and method-group data.

4.2 Start-time sections

A start-time section is the point in time at which a method group starts producing signals. For example:

from 10:00: the method group starts producing signals from the 10:00 time window;
from 21:00: later signals may continue into the next calendar day.

For sections that cross midnight, the system joins the continuous time windows automatically and marks them with labels such as "next day".

5. Method Groups

A method group is a set of methods that produces direction calls independently. Different groups may form their own signals from different data dimensions or logic.

The page may currently show, among others:

Breakout, Volume, Divergence, Trend, Reversal, MA, Gap, Pattern, Swing, Momentum, Support, Resistance, Turnover, Rhythm, Formula.

Each method group outputs its direction independently and accumulates its own history, including:

hits;
misses;
net wins;
win rate;
settled bars;
days of sample coverage.

A group's history only reflects past results under fixed rules; it does not represent future results or actual trading returns.

6. Watchlists and Observation Sets

You can add method groups to a watchlist to build your own observation set. The page offers three independent watchlists:

Watchlist 1
Watchlist 2
Watchlist 3

A star next to a group name means it is in one of your watchlists. The number after a watchlist name is how many groups it contains.

Watchlists only change how you group and display data; they don't change the groups' data and are not a recommendation or rating by 24K. Adding or removing does not change:

hits;
misses;
net wins;
win rate;
Research Center results;
method-group consensus;
history records.

7. Hits and Misses

Hitsis the number of times a group's direction matched the real market direction of the period.

For example:

Group direction: Up;
Real market direction: Up;
Result: Hit.

If the directions differ, it is recorded as a miss. Only valid periods that have closed and been confirmed enter hit/miss statistics.

8. Net Wins

Net winsis the difference between hits and misses among settled samples.

Net wins = hits − misses

For example:

30 hits;
20 misses;
50 settled;
Net wins = 30 − 20 = +10。

Positive net wins mean more hits than misses in the sample; negative means more misses than hits.

Important: net wins are not a return rate and do not represent actual trading profit or loss.

9. Win Rate

Win rateis the share of settled samples in which the group's direction was correct.

Win rate = hits ÷ settled × 100%

For example:

30 hits;
50 settled;
Win rate = 30 ÷ 50 = 60%。

Periods in progress are excluded. Win rate is a historical ratio; it is neither a probability of future success nor an actual investment return.

When reading the win rate, also consider:

settled bars;
days covered;
the statistics time window;
the period during which the group was active.

Avoid drawing conclusions from a single percentage.

10. All-Group Consensus

All-group consensusshows how much multiple groups agree on the direction of the same period.

For example, with 15 groups in total:

10 call Up;
5 call Down;
The all-group consensus for the period is Up.

The consensus reflects the distribution of directions among groups; it is not a separate group nor a guarantee about the market. Even with consensus, the actual result can differ.

11. Settled

Settledmeans the 15-minute period has closed and its direction has been confirmed, so it formally enters historical statistics.

Settled data may be counted in:

hits;
misses;
net wins;
win rate;
the Research Center;
history review.

12. Sample Coverage and Research Center

Groups differ in daily signal counts, active hours and valid periods, so comparing counts or win rates alone does not fully reflect sample reliability.

24K also records the number of days covered. For example:

Covers 25 / 30 daysmeans the group effectively covered 25 days within a 30-day statistics period.

The Research Center can show:

hits;
misses;
net wins;
win rate;
settled bars;
days covered.

When viewing results, read the sample size and days covered together to avoid over-concluding from too few samples.

13. Page Colors and Result Markers

24K shows "direction" and "result" separately:

13.1 Direction colors

Green: Up, the group or signal points up;
Red: Down, the group or signal points down.

Green and red only indicate direction, not hit or miss.

13.2 Result markers in each view

Views may use different result symbols, but the underlying statistics are the same.

View 1

Green = Up;
Red = Down;
Circled = Hit;
Blank = Miss.

View 2

Green = Up;
Red = Down;
▲ / ▼ in circle = direction;
Outer ring = Hit;
Faded ✕ = Miss;
Finished bars are shown faded; bars not yet finished stay bright.

View 3

Green = Up;
Red = Down;
Filled circle = Hit;
Hollow circle = Miss.

The three views only change appearance; they never change direction, hits, misses, net wins, win rate or history.

The interface uses the following short legend:

Green = Up Red = Down Filled / circled / ringed = Hit Blank / hollow / ✕ = Miss

The actual legend changes with the current view to avoid confusion.

14. Live and History View

14.1 Live

"Live" is the time window the system is currently running and updating, with its latest status. The current unclosed 15-minute period is marked "In progress" and is not counted early.

After a period closes, the system confirms the actual direction and updates the data. Across midnight, live data continues in the same continuous window without interruption.

14.2 History view

"History" lets you view completed historical data and statistics. You can filter by date, time window or other conditions provided on the page:

each group's direction;
hits and misses;
net wins;
win rate;
settled bars;
days covered.

History and live data use the same rules. Confirmed results remain traceable and are never rewritten due to later market moves.

15. Statistics Workflow and Audit Trail

24K's basic statistics workflow:

Read public quotes → fixed time window → groups produce directions → period closes → real direction confirmed → hit or miss settled → statistics updated → history kept.

Core principles:

No conclusion before the close;
No polishing after confirmation;
Statistical definitions stay consistent;
Historical results keep verifiable, traceable records.

16. Positioning and Risk Disclosure

24K is a market-information and direction-reference research tool. It does not constitute investment, financial, legal or tax advice.

The data, signals, method groups, all-group consensus, watchlists, Research Center and historical statistics on this site:

are not advice to buy, sell, hold or trade;
are not a promise of gains or of avoiding losses;
do not mean future markets will follow historical results;
do not mean actual trades will achieve the same results;
do not include managed accounts, asset management or trade execution.

Crypto-asset prices can be highly volatile and past statistics do not indicate future performance. Make your own independent decisions and bear the risks and consequences of using this information.