In the Classical Greek phalanx which was eight deep, with the Greek spear being about eight feet long, "only ranks 1-2 (or possibly 1-3) actually fought" (Wheeler 206). The depth gained from further ranks added:
...momentum in attack and in defence renders the line more impenetrable. The front ranks receive psychological support from the rear ranks, whose mass checks any hesitation to engage and deters desertion. Depth also ensures immediate replacements for front-rankers killed or wounded (Arr. Tact. 12.4). Not least, depth, imposing in itself, exerts psychological pressure on an opponent (Wheeler 206).
I'm sorry to just quote a bunch of text at you, but Wheeler and the CHGRW are very fine sources on this matter. A flaired user in Greek warfare with a bit more time (I'm a bit pressed atm) will be able to give more detail, but suffice to say while all ranks are in combat in a psychic sense and very much suffering from the 'psychological pressure' of combat, the Classical Greek phalanx only had its front 2 or 3 at most ranks fighting at any time. Casualties however would be quickly replaced by the rear ranks. Further, once one side broke, the phalanxes scattered and individual duals might occur. So the course of the battle and the casualties suffered by one side or the other would drastically effect the odds of a rear-rank hoplite 'seeing' combat. I should emphasise again that the presence in the phalanx of a rear-rank hoplite, only a few metres from the killing and death of ancient battle, I think qualifies nonetheless as 'seeing' battle.
Regarding Marathon in particular,* the Greek centre was formed at only 4 deep. So 2-3 of their ranks would have been actively fighting, until they broke at which point they would have been trying to back off or flee to the best of their ability and fighting individually if necessary. So if you were in the centre of the Athenian line at Marathon, you almost certainly fought. On the flanks, where the Athenians were allegedly very successful and apparently quite quickly, we could probably assume they took relatively few casualties and so the front 2-3 ranks must have done basically all of the fighting. If we are to follow this perhaps overly simple model I have suggested, the "odds" of your fighting at Marathon are quite easy to broadly calculate depending on whether you were in the centre or the wings and what rank you were standing in.
Source: Wheeler's chapter on classical Greek "Land Battle" from the Cambridge History of Greek and Roman Warfare, 2007.
*I haven't read the account of Marathon lately so please correct me if I've mangled this broad outline of events.
(1/2)
I'll have to go through a bit of background to answer your question, so bear with me :)
The battle of Marathon
Our sources for Marathon are bad. We really only have one source - Herodotos 6.109-117 - and it's clear that the story was tainted by two generations of Athenian self-aggrandising before Herodotos wrote it down. A lot of details are unclear. There is a lot of focus on heroic deeds of individuals, like in the Iliad, but we're missing some essential practical details. For example, we have pretty much no idea how many Persians were actually involved in the battle; all we get from Herodotos is that 6,400 of them were killed, but this number is somewhat suspicious on the grounds that it can be derived from the number of reported Greek dead at the battle (192) simply by multiplying by 100 and dividing by 3. Even if we accept the figure, though, it doesn't really tell us anything about the total number of Persians who fought. We also don't know where the armies were drawn up, how wide their frontage was, what the Persian cavalry did during the battle, and so on.
As a result of the vagueness of the account, scholars have been arguing for centuries over what exactly happened at Marathon. There's not likely to be any consensus simply because there is little to no hard evidence to back up any theory. The facts generally agreed on are these:
The Athenian army went into battle with a weak centre; the Plataians on the left wing and the polemarch Kallimachos on the right were in a deeper formation.
The Greek army ran into battle, allegedly over a distance of a full mile; Herodotos says they were the first Greek army ever to do so.
When the lines met, the Persian centre, which included the Persians themselves and the Sakae (possibly mounted), broke through the thin Greek centre. On the wings, however, where the Persians had deployed unknown subject peoples, the Athenians and Plataians both routed their opponents.
At this point the victorious Greeks on the wings let their routed enemies escape and turned inward, attacking the victorious Persian centre from both flanks. The Persians broke and fled to their ships, pursued all the way by the Greeks.
There's a lot here for historians to bicker over. It's clear that the Persian archers couldn't do much to affect the outcome because of the quick Greek charge, but why didn't the Persian cavalry do something to tip the balance? Where were they even deployed? How did the Greeks manage to run a mile in full armour? How did they preserve anything like a formation over that distance? How did the Greek wings manage to hit the Persian centre in the flanks when it was already victorious and presumably pursuing the routed Greek centre? Was the double envelopment a planned manoeuvre, or was it ordered by Kallimachos in a moment of inspiration, or did it happen by accident? There are many theories, but little that can be proved.
Since there is so little we know for sure, the battle has triggered some very fundamental questions about the nature of Greek warfare in 490 BC. Can we really assume that this battle was fought in the same way as later battles about which we know more? This is very relevant to the question you've asked about whether the men in the rear of the phalanx would see combat. Specifically: do concepts like "phalanx" and "rear ranks" even apply to what happened at Marathon?
The phalanx at Marathon
It's generally assumed that the Greeks won at Marathon (and later at Plataia and Mykale) because of their dense, well-organised phalanx formation. However, there are a couple of problems with this assumption.
First is the observation that Herodotos never uses the word "phalanx". As a tactical term, used in the sense we use it, its first appearance is in Xenophon, who wrote nearly a century later. Perhaps this was merely a matter of personal preference, but perhaps its absence in Herodotos (and in Thucydides) is a result of the fact that the Greeks did not talk about their formations in such terms. Either they didn't have a name for what they were doing, or they simply weren't doing it yet.
Second, if the Greeks were deployed in a phalanx, Herodotos should be able to tell us how many ranks deep that phalanx was. But he doesn't. He only says broadly that the centre was thin and the wings were deeper. In fact, he never tells us how deep any formation actually was. Now, we could of course argue that he just wasn't interested in those details - but it's also possible to argue that he didn't know how deep the formation was and nobody could tell him, because it wasn't yet organised in such a way. The earliest source that tells us about the depth of ranks of an infantry formation dates to the 420s BC, when even Herodotos himself was already dead, let alone the men he wrote about.
Third, Herodotos never says the Greeks were deployed with deep wings and a shallow centre. In the relevant passage (6.111.3) the verb he uses twice is ἐγίνετο, "it happened". There is no indication that anyone made a decision to draw up the phalanx in a certain way. It's been suggested, therefore, that the shallow centre and deep wings were an accident - probably the result of men aiming towards the wings in the course of the long charge, to make sure that the line was wide enough not to be outflanked. This should not have happened if the formation was a carefully organised phalanx.
Fourth is the long charge itself. Later sources make it clear that a phalanx typically charged into battle. It was a great way for men to overcome their fear, and it had the added advantage of crossing enemy missile troops' kill zone as quickly as possible. This is often assumed to have been the reason the Greeks charged at Marathon, too (though Krentz plausibly argued that it was actually to rush into battle before the Persian cavalry could deploy). But the thing is that Herodotos specifically notes that this was the first time a Greek army ran into the fray. If they were a phalanx like any later phalanx, this wouldn't have been the case; charging was typical. There are only 2 possible explanations: either they were not a phalanx but a less organised pack of warriors, or (as Krentz thinks) they were in fact the first phalanx. They were experimenting with a new approach to heavy infantry combat, and it nearly ended in disaster when their centre became too weak.
Finally, a mile-long charge, or even a mile-long fast advance, would have torn up beyond recognition any formation that may have originally existed. This is exactly what Herodotos seems to imply when he says that the wings "happened" to be deep and the centre "happened" to be thin and the battle line "happened" to be as wide as that of the Persians. We really shouldn't overestimate the degree to which the Athenians and Plataians were able to form an organised front by the time they reached their enemies.
It's important to bear in mind that Marathon is in fact the earliest Greek battle of which a description survives. We can't compare it to other contemporary battles because we know even less about those. We don't know exactly what sort of world the Greek charge at Marathon changed forever, but if it did, as Herodotos specifically tells us, we should start from the assumption that Greek warfare in 490 BC was not like the Greek warfare we know from later times.
(continued below)