Accuracy

How often the network's forecasts for the next 24 hours turned out right, stream by stream. Every forecast is saved before the outcome is known and checked against the real price a day later. Misses stay in the record.

How to read these numbers

Right about the direction: The share of forecasts that correctly said whether the price would be higher or lower a day later.
Average error: How far the forecast was from the real price a day later, on average, as a percentage of that price. Lower is better.
If the price had not moved: The error you would get by simply assuming the price stays where it is.
Landed in the network's range: Neurons never agree exactly. The range is where the middle 80% of their forecasts fell. This is how often the real price ended up inside it.

Showing 85 to 86 of 86 · all time

YEC on SafeTradeSAFETRADE_YEC

Open forecast

All time · checked on 44 days · 1,009 forecasts

47.4%

right about the direction

Average error: 10.48%
If the price had not moved: 10.31%

Worse than assuming no change

36.5% landed in the network's range

Was it right more often when more of the network agreed?

50 to 60% agreed30.6% right · 229 forecasts
60 to 70% agreed48.3% right · 178 forecasts
70 to 80% agreed46.3% right · 188 forecasts
80 to 90% agreed54.2% right · 155 forecasts
Over 90% agreed58.3% right · 259 forecasts

ZEPH on SafeTradeSAFETRADE_ZEPH

Open forecast

All time · checked on 44 days · 1,009 forecasts

49.4%

right about the direction

Average error: 6.68%
If the price had not moved: 6.84%

Better than assuming no change

45.2% landed in the network's range

Was it right more often when more of the network agreed?

50 to 60% agreed36.6% right · 451 forecasts
60 to 70% agreed55.9% right · 270 forecasts
70 to 80% agreed53.2% right · 141 forecasts
80 to 90% agreed64.9% right · 57 forecasts
Over 90% agreed77.8% right · 90 forecasts