Lets be honest...In NL hold em...ALL hands are coin flip if you go all in preflop. Once all the chips go all in preflop...all the skill has now been taken out of the game on the flop, turn and river....there is no more betting startegy...no more bluffing strategy , etc. It is now all LUCK .
In other words, any hand preflop, can end up the winner on the river.
...Well, if you're thinking in terms of one hand wins/one hand loses, then yes, each hand has a 50/50 chance of either of those results. That's a given in all-in situations and with the exception of a chopped pot, a hand can only win or lose.
However, in poker, we're more interested in getting our money in with hands that have more equity preflop. Equity is simply the chance something has at winning. If we take a coin, each side has 50% equity because it can only land on one of two sides.
A poker hand with more preflop equity will win more in the long run. Simple as that. AA has more equity over JJ, so it will win a lot more money over a decent sample size.
The term coin flip , is usually regarded as meaning a 50-50 hand , but the reality is, poker is not like a heads or tails coin and if you play:
...Actually, like a coin, the term 50/50 or coinflip is referring to each hand's preflop equity, or chance they have at winning by the river.
So, your classic AKs vs QQ is a coinflip because both hands have a ~50% chance of winning by the river. QQ does have a slight edge, but I'd say any 2 hands falling that fall somewhere between 45%-55% in equity is a coinflip.
AK VS JJ
1000 times , I guarantee it wont work out to each hand winning exactly 50 % of the time...especially if you do this experiment on 1000 different days, a 1000 times each day.
...Well, no, it probably won't land at exactly 50/50, but over a large sample size, it'll be pretty close. In a perfect world where everything is exact, JJ would have slight advantage as far as wins and profits go since it does have a slight equity advantage. However, since poker isn't exact and w/ a large enough sample size, their results would end up pretty equal. Maybe not exactly 50/50, but 48/52 in favor of JJ.
27 vs AA...is still a coin flip hand, all in preflop ...its just that the AA is expected to win much more then the 27...BUT...because of the luck factor, variance in NL hold em poker , there is no guarantee that is gonna happen repeatedly.
...27 vs AA isn't a coinflip because AA has more equity. Also, even though variance is a thing in poker, it balances out over large sample sizes.
To give an example, lets say we agree to flip a coin 4x. If it lands on heads, I pay you $100 and if it lands on tails, you pay me $50. Sounds like a sweet deal, right? Now lets say it lands on tails 4x in a row and I pocket $200. Well, if it's 50/50, you would expect the coin to land 2x on head, 2x on tails, right? What happened?
Variance happened. Just because something is expected to happen doesn't mean it'll happen exactly. Flip a coin 10x and see it lands 5x heads/5x tails. It's possible, but it's also not uncommon for it land a bit off at 3x heads/7x tails or such. However, the more times you flip the coin, the better chance of heads vs tails becoming equal.
So, applying this to poker. Yes, AA will lose to 27 about 20% of the time. In fact, as long as hand has some equity, it can win. However, in the 27 vs AA situation, over a large sample size, AA will win more often and, thus, be more profitable. If we don't worry about results and focus more on decisions, we can see that getting our money with AA vs 27 will always be profitable in the long run.