[Sun_Tzu]

The problem you are having, Hanzi, is that you are simply confused about how probability works.

If two outcomes are possible, that does not imply they are equally probable. Even if they are the only two outcomes possible, say either App will overtake Asc or it won't, without specifying a timeframe, the outcomes still have different weights, different probabilities. By identifying patterns and trends and identifying different factors contributions to these trends, we can assign them different predicting values. By calculating the risks involved in each assumption, we can asses the likelyhood of a default, and control for it.

The key here is large data sets. Where as it's impossible to say what will happen any given tick, any more than it's possible to say how long any given person will live, reality still tends to follow a bell-curve and regress towards the mean, allowing insurance companies to predict the average lifespan of their policyholders by controlling for risk factors, and allowing us to predict how an alliance will perform over a period of time.

In essence, you made a similar prediction, however you did not control for most variables, and you used only one point of data. Had you included multiple data points, you would have deduced that there was a discrepancy in your calculations, and you wouldn't have made such a foolish prediction as a specific tick.

This is all basic statistics.