The basic premise of RTP is that you will always return to the average – over a long period of spins. Take for example betting on the flip of a coin – it isn’t unreasonable to think of 5 tails in a row, followed by 2 heads and then another 3 tails. This obviously means in those 10 flips, it’s 80/20 Tails to Heads. However, after you’ve done 1,000,000 spins – that number is going to come down to 50/50 as firstly you are going to throw a lot of heads to balance that out, but given it’s a fair coin – you should theoretically have a 50% chance either way you bet.
So in the case of RTP it’s the same deal except it isn’t a 50/50 chance but rather a 90% chance of having a return. So if you do 100 spins you might get a return of like 500% of what you started with – like deposited £100, did £1 spins and now have £600, that’s a 500% RTP. But say your next session with £100 you get like 40% return over the same number of spins – you are not at an RTP of (500%+40%)/2 – 270% – I.e. the more spins you play – the quicker you will regress to the mean.
It’s basically the law of large numbers – if you perform an experiment or something with a known theoretical outcome enough times, you will end up with that theoretical outcome every time.
Hope this helps.