See 80 bars of a real historical Binance chart with the coin and dates hidden, guess higher or lower 10 bars later, and compare your streak with luck.
| # | Coin | TF | Cut date | +10 bars | Your answer |
|---|---|---|---|---|---|
| No rounds yet. | |||||
Unfamiliar term? → Crypto & Stock Glossary
For information only, not investment advice. Data comes from exchanges' public APIs and may be delayed or interrupted. Charts: TradingView Lightweight Charts™
Can you call the next move from the chart alone? This quiz picks a real historical Binance chart of a major coin such as Bitcoin or Ethereum at random and shows 80 bars with the coin name and dates hidden. Guess whether the close 10 bars later will be higher or lower, and the hidden bars appear one by one before the coin, the dates, the actual move and the RSI and moving-average readings at that moment are revealed. Your streak and accuracy sit next to a guessing baseline, so you can tell whether your chart reading beats luck.
One of 22 major Binance spot USDT pairs (BTC, ETH, BNB, SOL, XRP, DOGE, ADA, AVAX, LINK, LTC, DOT, TRX, BCH, ATOM, NEAR, UNI, ETC, XLM, FIL, AAVE, HBAR, SUI) and either the 4H or 1D timeframe are picked at random. The cut bar is drawn uniformly from 206 bars after the listing day (120 bars for indicators + 80 on screen + 6 spare) up to the latest point at which the 10th answer bar closed at least 30 days ago. The chart shows the 80 bars up to the cut; the price axis is (price ÷ cut close − 1) × 100 and volume is shown as a multiple of its 80-bar average. The answer is move = (close of the 10th bar after the cut ÷ cut close − 1) × 100. Up/Down: above 0 is Higher, below 0 is Lower, and exactly 0 counts either way. 3 buckets: −5% or less, above −5% and below +5%, +5% or more. The RSI shown on reveal is the 14-period Wilder RSI (70 or above overbought, 30 or below oversold) and the moving averages are 20- and 60-bar simple averages, all computed only from data up to the cut. Windows with a bar missing because of exchange maintenance in the 80 + 10 bars, two or more bars with no trades or no price movement, or an 80-bar high-low range under 0.5% are skipped.
The top bar sets the answer type (Up / Down or 3 buckets ±5%), the timeframe (Mixed, 4H or 1D) and the MA20 button. Mixed picks 4H or 1D at random for each round. If a question is still unanswered and you switch the timeframe to 4H or 1D, a new question with that timeframe replaces it at once; switching the answer type keeps the same chart and only swaps the answer buttons. MA20 shows or hides the 20-bar simple moving average; its values are computed even while hidden, so turning it on shows them immediately. These three choices are kept in the address and in the browser, so your next visit opens the same way; the current round's code is kept only in the address. Accuracy and the guessing baseline are tracked separately per answer type, so switching type also switches the score panel to that type's numbers.
The title reads '???/USDT · timeframe' to hide the coin, with the round number beside it. The chart contains only the 80 bars up to the cut bar; the 10 answer bars are not in the chart data at all, and their space is left empty on the right. The price axis is (price ÷ cut close − 1) × 100 in %, with a dashed line at the cut close, 0%. Volume is a multiple of the 80-bar average on screen. The time axis and crosshair show only how many bars from the cut bar (−79 … 0) instead of dates, and a '?' marks the cut bar. When the 80-bar range is under 6%, the price axis shows two decimals. Everything is real Binance history; converting prices to % only stops you from recognising the coin by its price level. This is not a live chart but a stretch that ended at least 30 days ago, so nothing moves on its own.
The question under the chart asks whether the close 10 bars later will be above or below the last bar's close (0%), or which bucket it will land in, and states how many hours or days those 10 bars are. Up / Down has two buttons; 3 buckets has three: +5% or more, −5% to +5%, and −5% or less. Keys 1, 2 and 3 also work. Grading compares only the cut close and the close of the 10th bar after it. However far price swings in between does not matter, and if the change is exactly zero, both Up and Down count as correct. In 3 buckets, −5% or less and +5% or more are the outer buckets. While you decide, the next round is fetched in the background, so moving on is quick. Once you pick, the answer cannot be changed.
After you answer, the hidden 10 bars enter the chart one at a time about every 0.15 seconds (all at once with reduced motion), then the time axis switches to real dates and the title shows the coin. The cut bar is labelled and the 10th bar carries the actual change. The result panel shows correct or wrong with your streak, your answer and the truth, the dates of the cut bar and the 10th bar (UTC date for daily, this device's time for 4H), the change to the 10th close with the actual prices in USDT, and the highest and lowest move during the 10 bars. The sentence below states, as is, whether RSI(14) at the cut bar was at or above 70, at or below 30, or in between, how far the close sat above or below the 20-bar moving average, whether the 20-bar line was above or below the 60-bar line, and the change over the previous 80 bars. All of it uses data up to the cut bar only.
The four boxes are streak (best streak underneath), accuracy (correct/played), guessing baseline and all-time accuracy. The guessing baseline is the highest accuracy you would have reached by always picking one answer on the rounds played so far, computed from 5 rounds on (rounds with a zero change count as correct for any answer). In Up / Down mode, after 10 rounds the note shows in percentage points the range that pure 50% guessing commonly lands in (about 1.96 × √(0.25 ÷ rounds)); in 3-bucket mode it reminds you that always picking the most frequent bucket scores that bucket's share. Streak, accuracy and baseline cover this visit only and reset on reload, while the best streak and all-time accuracy are stored in this browser. Whether your accuracy stays above the baseline is what separates reading the chart from luck.
'Share score' turns the current answer type's accuracy and either your streak (when it is 2 or more) or your best streak into a sentence and sends it through the device's share sheet, or copies it with the link to the clipboard if there is none. With no rounds played yet it sends a sentence introducing the quiz. 'Share this chart' creates a link that opens the current chart. The round code in the address combines the cut-bar number, coin position and timeframe into a short base-36 string, so your friend cannot see the coin name before solving it. Reloading a round you already answered in the same tab gives a new round instead, so you cannot pile up all-time score while knowing the answer. If a shared window can no longer be used, a random round is served instead. If copying fails, a short note asks you to copy the address bar link yourself.
Rounds from this visit pile up newest first, up to 50, with number, coin, timeframe, cut date, change after 10 bars and your answer marked right or wrong. After a reveal, tapping a number shows that round's chart (with the 10 answer bars) and result panel again, and the row you are viewing is highlighted. While a new question is open the numbers are locked so you cannot peek. The table is useful for looking back at which chart shapes you tend to miss. For example, reopening rounds that were oversold or had the 20-bar line under the 60-bar line and comparing what happened 10 bars later gives you a feel for how common readings played out on these rounds. With only a few dozen rounds, though, chance plays a big part, so avoid turning it into a general rule. The table clears on reload.
Start with Up / Down and Mixed, and compare the RSI and moving-average sentence on the reveal with what actually happened to build intuition. Past 10 rounds, weigh your accuracy against the guessing baseline and the chance range, and from around 30 rounds see whether the gap to the baseline holds. Once comfortable, switch to daily 3 buckets to practise judging size as well as direction, and toggle MA20 to see how much a single line changes your call. Reopen missed rounds from the Rounds played table. This quiz works on past windows, so there is no need to keep it open, and it does not update live. To test chart reading further on real data, try a strategy backtester, an indicator scanner and a live coin chart alongside it.
They are real historical bars from Binance's public API. Only the coin, the timeframe and the window are chosen at random; no price is generated or altered. To hide the coin, prices are shown as % from the cut close and volume as a multiple of its average.
The close of the last visible bar (the cut bar) is compared with the close of the 10th bar after it. What happens in between does not matter, only the close of that 10th bar. If they are exactly equal either answer counts, and the 3 buckets are −5% or less, between −5% and +5%, and +5% or more.
The best accuracy you would have had on the same rounds by always giving one answer. If 60% of your rounds went up, pressing Higher every time scores 60%. Only accuracy that stays above this baseline suggests you are reading the chart rather than riding the market's drift.
All 10 answer bars must be complete, and recent prices you remember would make it too easy. So only windows whose 10th bar closed at least 30 days ago are used. Windows right after listing, where there are too few bars, and windows with bars missing because of exchange maintenance are skipped too.
Your best streak and all-time accuracy are kept only in this browser (localStorage) and never sent to a server. The current visit's streak, accuracy and round list start over when you reload. Other devices and private windows keep separate records.
For information only, not investment advice. Data comes from exchanges' public APIs and may be delayed or interrupted.