Four friends have 2, 4, 4 and 10 stickers. Pile all 20 stickers in the middle and deal them back out evenly: each friend gets 5. That 5 is the mean. Nobody had to own exactly 5 stickers to begin with; the mean is what each person would hold if the total were shared fairly.
The mean is also a balance point. Place the values on a number line: 2, 4, 4 and 10. The distances below 5 are 3, 1 and 1, which add to 5, and the one distance above is also 5. The two sides cancel, so a ruler carrying equal weights at those spots would balance at 5. Move the 10 out to 30 and the balance point slides right to 10, even though three of the four values never moved.
That sliding is what to watch for. Five friends get weekly allowances of $5, $5, $6, $8 and $76. The mean is 5100=$20, more than four of the five actually receive. The mean did its job, sharing the total evenly, but an even share no longer describes a typical friend. When one value sits far from the rest, the median (the middle value after sorting, here $6) is the better picture of typical; the mean stays the right tool when the question is about totals or fair shares.
Think of it like this. Tape an identical coin onto a ruler at each data value. The mean is the spot where the ruler balances on your fingertip. Slide one coin far to the right and the balance spot follows it, which is exactly how one extreme value pulls the mean away from the crowd.
Core idea. The mean is what every value would equal if you poured the whole sum together and split it evenly among all of them.