Different confidence level in statistics Let's assume we have to through a ball in a  basketball ring a number of times. After throwing, we would like to understand the success rate and the uncertainty of the success.
The number of trials are = 100, 
number of time that I successfully through the ball in the ring is 75. Therefore the efficiency is 75%. 
Now often people use the phrase confidence level to interpret the  the uncertainty of the  efficiency (or any statistical study). 
What does it mean if I have to find the uncertainty of efficiency with the confidence level 68%?  What happens if we set the confidence level to 100%? 
 A: Your best estimate for the efficiency of the shooter of the basketball is 75%. However, this is only after observing 100 shooting attempts. What if the shooter were to shoot the ball 1,000 times? Or 10,000 times? Is the shooter's efficiency likely to still be 75%? Maybe, but it is likely to change at least slightly as we observe more shots. Our uncertainty about how this number will change is represented using a confidence interval.
What can we expect the shooter's ACTUAL efficiency to be over the course of an entire basketball career? We are uncertain. That is why we need to build a confidence interval, so we get some idea of how confident we are in the shooter's actual efficiency, beyond just what we observed in the 100 shots. 
A 68% confidence interval would mean that we build a confidence interval with a minimum value and a maximum value that has a 68% chance of capturing the shooter's true efficiency. In other words, if we had the shooter shoot 100 baskets over and over again, 68% of the time the shooter's efficiency would land within this confidence interval.
A 100% confidence interval would mean we build a confidence interval that has a 100% chance of capturing the shooter's true efficiency. Note: we can never achieve a meaningful 100% confidence interval. For us to be 100% certain our interval contains the shooter's true efficiency, we would have an interval of 0% to 100%.
