>>13964142This. Though you should understand what you're doing. Essentially you are creating two models and comparing them, the model that the coin is fair and the model of the true results you are seeing.
You then are checking to see if the results are reasonably the same or different. With results like yours, they'd be reasonably different and you'd have high confidence in their difference. If you had only done it 5 times, they'd be different, but you'd have less confidence in the results. This naturally makes sense when you think about the concept of variances and means (which is ultimately what you compare in these tests). They logically converge to a single value with enough data points (assuming all data is coming from the same system/model). You're comparing whether you have enough data and difference to reasonably decide that things are different. These things do NOT tell you how different things are though and if the difference is even meaningful for further analysis, just that there is a difference and quantifies a confidence on that difference.
