Holdem tournaments and all-ins.?

You are thinking about something that is about as irrelevant as could possibly be. If the ENTIRE tournament boiled down to you only going "all-in" these 3 times, and THAT would be the determining factor of whether or not you'd be the winner of the tournament, then yes - you'd have about a 34% chance of winning the tournament, regardless of when each of these hands take place.

This is SO FAR from ever being a possibility that it borders on ludicrous to even think about it.

About the only relevance I could possibly glean from your question would be to tell you this: People who can play tournament poker profitably, (in other words they WIN more money over time than they spend), consistently are able to do just what you are talking about. They consistently get their money in with an advantage.

Now if you ALWAYS got your money in the middle with an advantage, it would NOT matter what the odds of winning any particular tournament were! You WOULD make money over time, and you WOULD win some of these tournaments. THAT is all that matters! The very best tournament players in the world don't expect to win each tournament. They only know that they will make more money than they spend, and that they will win tournaments every now and then. Why? Because they consistently know how to get their money in with the best hand, AND they know how to get an opponent with a better hand OFF their hand!

I don't understand if your question is on poker theory or just math.

You have a 34% chance of surviving those 3 all ins, you can't possibly think the math changes based on timing, can you?

But when you say if the 3rd was 3 hours later, do your chances never go up? That doesn't make any sense either. The odds of winning those 3 all ins are 34% before anything starts. But if you happen to survive the 1st one... then it's over and done. The odds of surviving the next two are 49%. If you survive the 1st two shots, then your odds of getting by the 3rd one are 70%, whether it happens 3 minutes or 3 hours later.

This is related to ICM or bubble factors. Your stack size and opponents stack size compared to the prize pool indicates when you should be going all in and calling them. If you are a big stack you want to be shoving wider than usual vs medium size stacks because they have a high ICM/BF. If you have a medium stack vs big stack you want to be tighter than usual. Also it depends if there are any short stacks that are almost out and the prize pool moves up. There is a math formula but it is easier to just understand the patterns of different situations. Here is an extreme example of ICM, 3 players left in a satellite, 2 places get paid. Stack sizes are Player 1 - 10000, Player 2 - 1000, You 7000. Blinds 100/200. You get dealt AA, short stack folds and Player 1 goes all in. Call or Fold? You fold this.... obviously you have the best hand but if you use the math formula this has a negative return in the long run.... considering ICM/BF.

Read Kill Everyone and play around with SNGwizard.

There is no such thing as a 70% hand. the probability one hand wins is dependent on the other active hands.

If you are 70% likely to win the first allin, and 60% likely to win another allin, then the likelyhood of you winning both of these is .7*.6=.42=42%

You are correct in your understanding that this number continues to decrease. It is the nature of the tournament, and it is why/because the tournament will continuously loose players.

Two things to think about. if you go allin with 10,000 chips and are 50% likely to win, your ev (assumably) is 0 chips. if you then win, and go allin again for 50% your ev is then 0. Your probability of winning both is only 25%, but your ev is 0. You can look at this for 80% as well, the odds for aa vs kk. If you go all in with 10,000 chips your ev is 6,000 chips, and your likelihood of winning is 80% If you win and get it in again with 20,000 chips your ev will be 12,000 and another 80% for this allin. Combined, this makes an ev of 18,000 but a likelihood of winning of only 64% So you can see that as you play you should be gaining chips, despite the fact that the likelihood of winning continuously decreases.

PDQs answer can be applied in some live situations, and could be applied to tournament poker a decade ago, but online tournament strategy has evolved away from this.

You're thinking is actually pointing to a very fundamental concept exploited by some of the best players. Arnold Snyder discusses this line of thinking in his series of poker books. If you rail some of the best online players you'll notice that their results are highly polarized. They wither go bust early, or do very well. This is the result of their efforts to accumulate chips, and reduce the number of chances they have to flip for their stack. They're basically trying to put players in the situation you have described, without putting themselves in the situation. This can also be used in cash game play, if you put yourself in a situation where you are bankrolled correctly, and your opponent isn't. The same logic applies in that you drive the game to a flipping competition, under the assumption that your better funded to survive.

Keep thinking in the ways you are. One thing you will find in poker forums, is that a lot of information is bad, limited, and dated. Very rarely is the answer to the question of value, but the line of thinking used to generate the answer is.

I don't think you can compute the probability of winning the tournament that way. Each hand IS independent, so the nimber of all ins does not affect your overall tournaments chances.

If you usually cash in tourneys, but overall you are losing.... how does that work exactly?

In Australia, books and other material, are available that explain what you ask. I suggest trying to find

them (Amazon, maybe) and study them.

The knowledge gained may be well worth the cost.