SWARM Stat theory

Example:

Below is an example stat of an algorithm pool stat:

{
  "Live": 0.00000000433038925,
  "Minute": 0.000000004330389,
  "Minute_5": 0.000000004330389,
  "Minute_15": 0.000000004332081,
  "Hour": 0.000000004332081,
  "Hour_4": 0.000000004332081,
  "Day": 0.000000004332081,
  "Custom": 0.000000004332081,
  "Hashrate": 27845.0,
  "Hashrate_Periods": 2,
  "Deviation": -0.15,
  "Deviation_Periods": 2,
  "Values": [
    0.0000000043329265,
    0.00000000433038925
  ]
}

Definitions:

Live

The current/last price documented.

Minute

The one minute moving average.

Minute_5

5 minute moving average.

Minute_15

15 minute moving average.

Hour

1 hour moving average.

Hour_4

4 hour moving average.

Day

Daily moving average.

Custom

Custom can be set by user. It is x number of periods moving average, where x is the value set through arguments. Essentially it created a moving average from x values.

Hashrate

This is a rolling exponential moving average of pool hashrate. It is used with -Volume argument, to calculate bonus/penalty of the pool based on how much hashrate it has.

Hashrate_Periods

The total number of periods that the moving average is calculating for hashrate value. This value is running counter for calculating the moving average itself.

Deviation

This is a moving average of % deviation (in decimal form between pool's 24 hour profit vs. its actual returns. In the example, the estimate the pool has been predicted was 15% higher over the past 24 hours than what the pool made. If you were using the -historical_bias parameter, incoming prices would be reduced by 15%. This value is capped by what -historical_bias value is. The exception is if the price is 100% (essentially 0 return past 24 hours), then it retains that value. If you had a -historical_bias 20, this value should never exceed 20%, so it allows you to define how heavy you wish the bias to be. Nicehash and whalesburg operate differently, not using direct market exchange, and therefor are not penalized, as their returns are usually very close to their predictions.

Deviation_Periods

This value is the counter for calculating the rolling moving average for deviation value.

Values

These are actually the estimate values SWARM pulled. I added this feature, as it allows you to investigate both pool pricing, as well as ensure SWARM isn't miscalculating stats.