Bitcoin price prediction FAQ

How is the Bitcoin price prediction calculated?

Up to 1,000 daily closing prices from Binance are converted to log returns and projected forward as a geometric Brownian motion. The trend is split into a market component (BTC's beta to Bitcoin) and a coin-specific component (alpha), each shrunk toward zero as a Bayesian posterior mean. The market trend is extrapolated assertively so that a coin's long-run direction reaches the forecast; measured against 24 coins this costs about 2.6% in one-year RMSE versus assuming no change, so the forecast is a trend extrapolation rather than a statistically validated prediction. Coins with under two years of history use a conservative prior instead, and the drift is capped. Technical indicators are not used to tilt it: backtested across 46 coins their 5-day directional accuracy was 49.8%.

Will BTC reach $100,000 (or any round number)?

The page answers that with two numbers: the chance BTC ever touches the level at any point before a date, and the chance it closes at or beyond it on that date. The first is much larger and is usually what people mean — historically Bitcoin touched a +58% level within 61.4% of one-year windows but closed above it in only 44.0%. Both come from a fat-tailed distribution calibrated so a stated 50% band really contains about 50% of outcomes.

Why not use moving averages, RSI and MACD to predict the price, like other sites do?

Because we implemented exactly that method and measured it. Across 25 coins with non-overlapping forward windows, a composite of moving averages (20/50/100/200), RSI(14) and MACD(12,26,9) predicted the 5-day direction 49.4% of the time — a coin flip. MACD alone was the weakest of the three at 49.5%. The indicators do appear significant until you account for the fact that every coin shares one market: they are really detecting "the market is trending", and once forward returns are made market-neutral the effect collapses (MACD's t-statistic falls from 2.15 to 0.33). Stating that a site uses technical indicators is a description of its method, not evidence that the method works.

Will Bitcoin go up?

This page does not answer that, and no honest model can from price data alone. For Bitcoin, as for nearly every cryptocurrency, the historical trend is not statistically distinguishable from random noise, so the long-run median projection sits close to the current price. What the model can tell you is how wide the plausible range of outcomes is at each horizon.

What will Bitcoin be worth in 1 year?

The 1-year row gives a single forecast price with the range containing half of all outcomes under BTC's measured volatility, plus the probability of gaining or losing 10%. The forecast is deliberately modest because Bitcoin's historical drift is not statistically distinguishable from zero. Separately, a historical scenario table shows the median outcome across every 1-year window BTC has actually lived through, along with how many of those windows were independent — a number worth checking before trusting any long-horizon figure.

Is this Bitcoin forecast investment advice?

No. These are statistical projections of a price distribution based only on past prices. They ignore news, regulation, liquidity and market structure, and they are not investment advice.