For X amount of draws, the frequency of each number being drawn were tablulated.

The highest frequent winning number was 34. It was a winning number 392 times

The lowest frequent winning number was 15 at 309 times.

Is this within the expected variance or should the winning number frequency be closer together?

If it outside the expected variance, should you eliminate the highest frequent numbers if you were to play the next draw or does it not make a difference?

On the face of it it doesnt seem unusual. I think you would need more info (such as total number of draws or where those numbers came in the ranking).

But those proportions could fit a normal distribution.

Consider this; what are the chances of each number in a lottery being picked the same number of times? VERY, VERY small (almost impossible)

Even the simple example of a die gives a chance of less than 2% that you could roll one of each number once in 6 throws. Whereas the chances of getting a 'lucky' number peaking with 3 occurences are about 12%. Randomness happens.