Detect volatility

I think I found my answer

https://www.mql5.com/en/forum/102326

I use ATR or standard deviation to measure market volatility. If you don't want to use any kind of indicators, consider using mathematical calculation.

Calculation

1. Calculate the SMA for Period n
2. Subtract the SMA value from step one from the Close for each of the past n Periods and square them
3. Sum the squares of the differences and divide by n
4. Calculate the square root of the result from step three

Standard deviation = Sqrt [(Sum the ((Close for each of the past n Periods – n Period SMA for current bar)^2))/ n]

  // calculate simple moving average
  double calculateSma(int period){
    double closePrice = 0;
    double sma;
    for(int i=1; i<period; i++){
      closePrice = closePrice + Close[i];
    }
    sma = (closePrice + Bid)/period;
    return sma;
  }
  // calculate standard deviation
  double calculateStd(int period){
    double sum = 0;
    double std = 0;
    for(int i=1; i<period; i++){
      sum = sum + (MathPow((Close[i] - calculateSma(period)), 2)); 
    }
    std = MathSqrt(sum/period);
    return std;
  }

The higher standard deviation means the higher volatility. You also can convert the standard deviation value to pips by multiple by Point.