Precog methods applied, label definitions, statistics are explained with the weekday posts.
Recent SPX precog forecasts
Latest week SPX bar chart:
Next-Session forecasts
By now there have been
482 forecasts with ITK (Implied via Telekinesis), by way of physical dice and Remote Influence. Actually, I think it is also considered dowsing by some.
445 forecasts with Divination, by way of Thoth Tarot deck.
206 forecasts with Generalized Psi Expectation (S*), by way of 1million random number generator iterations with Remote Influence.
Total number of half-day forecasts performed since inception: 583
244 Composite Forecasts by simple average, C^
211 Composite Forecasts by majority vote, C[M]
1-week forecasts
There’s been 130 1-week forecasts for each method applied, since 16th of Sept.
Total number of 1-week forecasts performed since inception: 130 x3= 390.
Implied Expected Values
*Since the aim of the forecasts is to produce +EV (Positive Expected Value) biases, I would NOT expect the average win rate to deviate far from 50%. You shouldn’t either, if your goal is to make money from trades.
Hedge Fund benchmark IEVs from HFR, for the same period
It’s been a rough period for most hedge fund benchmarks:
Precog 1-week forecast IEVs
Precog half-day forecast IEVs
The morning forecasts have done better than expected.
Precog next-day Open-to-Close forecast IEVs
*Since this week I have begun to do a performance weighted average for the S* forecasts, so I expect it to improve over time.
Probabilities of forecast performance due to chance
We can use the z-score to estimate the probability that the hypothetical trading performance had occurred by chance. We usually use Z-Score tables, but they only go up to about +/- 3. So in cases of bigger Z-scores, we have to use a Z-Score calculator:
All of the above is explained in more detail at www.fourmilab.ch
Rational analytics from me are available at Unconventional Trades.