I have read it all and one of the wise members @
Eternad already pointed out the obvious mistake in all the data provided by OP;
But you should input the winning percentage set on the game to determine the win/loss ratio if it will match to the outcome or not. 300K is indeed a good size for sample imho which can already use to check the math.
Look how he wasted the opportunity to get some sincere attention. Instead of specifically answering he used the word 'flawlessly'

.
I played the game flawlessly and therefore should have experienced the advertised 0,5% house edge.
Now let me tell you why knowing this information matters;Let's say you have a winning chance of 'one in a million' so you will lose on an average of around one million bets before winning one bet and your data will look like the following;
Total bets: 1 million
Lost: 1 million
Won: 0
and now you can mislead others by claiming it is impossible with a 1% house edge.