Special Case Probability Suppose there exists an ideal system in which you have perfect data 100% c.i. This data tells you that the price of gasoline will move a distance (not displacement) of +/- 10% in the next 7 calendar days with a 70% probability.
On day 1 the price of gasoline is $1.00.
In the following days from 1-6 the price is in flux and arrives at precisely $1.00 on day 6. 
Your data tells you that gasoline will move a distance of +/- 3% in 1 day with 70% probability. 
Are both of these probabilities still valid? The 7 day and the 1 day? If yes, is there any edge that can be calculated when these special cases present themselves. Have the 7 day odds changed and is there less than a 70% chance of the +/- 10% move happening in one day? If so, please explain how/why to calculate these changes.
What type of statistical analysis is this? I don't even know what to research to learn more about this - this would be quite helpful for my future education in this area.
Please explain this at the level of basic statistics, as my education in statistics is not advanced in any way. Thank you. 
 A: It is simple enough. The first prediction was valid on the day it was made relevant to a day seven days in the future, and, it is not valid on day six. On day six, a new probability was calculated and the clock was reset. The second prediction is relevant to day six for predicting day seven prices. 
Imagine that a prediction is a box containing a Schrödinger's cat. Here we assign a probability of the cat in a box being alive of 70% at day seven and 30% of being dead. Now if you peek into the box at day six, you are no longer looking at a probability, you are then looking at a cat that is either dead or alive. If the cat is still alive, and you reseal the box then you have restarted the clock and the cat now has a better probability of being alive at day seven than it would have had. 
In order to calculate what the exact probabilities are, we need to know more information about how prices change from day to day, that is if the price changes are erratic the calculation would be different than if the price changes are more or less the same amount when they occur. For prices, like stock market values, the changes tend to be lognormal distributed. 
