What prevented slaves in Classical Greece from running off?

INTRO

First, it is important to note that Greek (and other) slaves performed many kinds of jobs and this fact alone could influence the likelihood of a slave running away:

The status of slaves, and the conditions under which they lived, depended in part on what kind of work they did....Some slaves received a formal education and training and managed to attain executive positions in business and industry...Slaves...could also attain positions of management and oversee the work of other slaves....

Source: Theodore M. Sylvester, Slavery Throughout History

At the other end of the spectrum,

Some of the hardest work for slaves was in the farm fields, but the worst possible fate for a slave was to be sent to the mines, where the hours were long, the work was backbreaking

Source: Sylvester

Thus generalizing about why they did or didn't run away is impossible, and there is also the character of the individual to consider - ancient sources refer to some slaves being meek while others were difficult to manage. It is also impossible to say what percentage of slaves ran away, but we do know that some did.

REASONS WHY SLAVES DIDN'T RUN

There are a variety of reasons why many slaves didn't run away:

1. Slaves acquired through conquest would, in some cases at least, have found it difficult to return home, either because their city was still under the control of the conquering power or because there was nothing / no one left there for them (e.g. Melos).

2. Some slaves were born into slavery, others were abandoned babies (resulting from the practice of infant exposure) who were found and brought up as slaves. In both cases, they had no 'home' to go to and were not citizens of any state (a major handicap in Classical Greece).

3. Others were sold into slavery as children – difficult to return to one’s parents under those circumstances. This was common for Thracians.

4. Some people became slaves due to extreme poverty - slavery at least usually meant food and a place to stay. In Athens, though, Solon (d. circa. 558 BC) made debt slavery of Athenian citizens illegal and had all enslaved Athenians released.

5. Fear of being caught. The comic poet Antiphanes' play Drapetagogos (The Runaway-catcher or The Catcher of Runaway Slaves) is evidence that some slaves clearly did run away but also, at the same time, that those who did could not expect their masters not to do something about it. The risks were considerable: being caught might well mean exchanging a relatively comfortable position for working in the silver mines, probably the worst fate for a slave.

6. It has been argued that some slaves became very close to their master or mistress and were generally content with their lot, or even almost part of the family:

Some have suggested (e.g. Westermann 1955: 18) that slaves failed to rebel there because they were relatively well treated and content.

Source: K. Bradley, P. Cartledge (eds.), The Cambridge History of Slavery

Note also

Euripedes' tragic character of Medea confided her deepest feelings with her nurse, who both advised and comforted her in her troubled times.

and

Tombstones of upstanding Athenian women often depict scenes of familiarity between the deceased and her slave companion.

7. It has also been argued that the wider dispersion of slaves in Athens (Mines of Laurion excepted) was a factor in slaves not rebelling:

Paul Cartledge (2001b), however, suggested that Athens differed in key ways from modern societies that experienced slave rebellions. Athens had a lower proportion of slaves (a third or less), and they were dispersed in relatively small groups with a relatively personal relationship to their masters.

Source: K. Bradley, P. Cartledge (eds.), The Cambridge History of Slavery

8. The slaves who would have had the most reason to run away would have been the ones who had the worst tasks. The prime example here would be slaves working in the Mines of Laurion, but they were under guard and sometimes (at least) chained (but see below for more on this).

9. A final factor is that research indicates that the large majority of slaves in Athens (at least) were probably either non-Greeks or else Greek women / children, for Greek men captured in wars between Greek states were usually put to death rather than enslaved. The relevance of this is that a young, fit Greek male would find it easier to escape than a 'barbarian' (less far to safety/home city) and a woman / child (on average, better able to evade slave hunters).

EVIDENCE OF RUNAWAY SLAVES

We know that some slaves did run away as this is referred to in a number of ancient sources. For example, in Xenophon's Memorabilia,

Socrates expresses surprise that people sometimes give more effort to hunting runaways (or looking after sick slaves) than cultivating friends who are much more useful.

Source: K. Bradley, P. Cartledge (eds.)

Also, Thucydides mentions that

Athens punished the town of Megara for (among other things) harbouring runaways (Thuc. 1.139–40)

Source: K. Bradley, P. Cartledge (eds.)

Further,

Several law-court speeches mention owners chasing escaped slaves (Ps.-Demosthenes 49.9, 53.6).3 Travelling after runaways could be a risky business, but these texts do not imply that it was unusual. There is some epigraphic evidence too (SEG iii 92.9–19).

Source: K. Bradley, P. Cartledge (eds.)

There is also a reference to slaves running away noted by R. Zelnick-Abramovitz in Not Wholly Free,

Socrates, speaking about good estate management, claims that there are households in which slaves are fettered and yet attempt to run away, whereas in other households, although they are without fetters, they are willing to work and remain

Finally, there was a major rebellion with thousands of slaves running way to nearby Decelea following an Athenian defeat at the hands of Sparta in 413 BC during the Peloponnesian War.

DISTINGUISHING SLAVES FROM CITIZENS

As noted in the question, slaves could easily be mistaken for citizens. On this, J.W. Roberts says:

The similarity of dress is not surprising in view of the known overlap of occupation: citizen and slave artisans worked at the same tasks for the same wages

Source: J. W. Roberts, City of Sokrates (2nd ed.)

It is also quite plausible that the favoured slave of a wealthy Athenian would be better dressed than some less well-off citizens. However, it is fair to assume that Athenian citizens would not be found in certain jobs - for example, down the mines.

There is evidence of confusion even among ancient historians concerning Argos in the aftermath of their defeat at Sepeia the hands of Sparta in 494 BC as to whether the men who defended Argos after the catastrophic defeat of the Argive army were slaves or local farmers.

According to Kostas Vlassopoulos, it is no surprise that, as

Slaves and freemen exercised the same professions; this overlap made it impossible to differentiate status solely on the basis of profession or living conditions. Thus, many slaves were in a position to take advantage of this blurring of identities to escape detection and create better conditions for themselves.

The only evidence I've found for any kind of 'marking' of a slave is this in J. W. Roberts:

A runaway slave who was recaptured could expect to be branded.

A NOTE ON SPARTA

IF one considers the Messenian helots as slaves (which many historians do not, preferring to call them serfs), there were a number of revolts against Spartan ownership of the land and the people. For the most part, though, the helots of Messenia were not 'runaways' - Messenia was, after all, their home.

However, the Athenians helped to establish the city of Naupaktos

as a refuge for liberated ex-helots during the great post-earthquake revolt of the 460s.

Source: Paul Cartledge, The Spartans

Subsequently, over the years, a number of helots did escape from Spartan-controlled Messenia to Naupaktos but, mostly, they stayed put. This was most likely because they considered the land theirs (so why should they move). Also, despite being slaves / serfs, they were left enough of their produce by their Spartan overlords to survive.

Other sources

S. Murnaghan, Women and Slaves in Classical Culture

Robert Osborne, Classical Greece 500 - 323 BC

R. A. Tomlinson, Argos and the Argolid

M. Gann & J. Willen, Five Thousand Years of Slavery