The Strength and Organization of the Persian Army at Plataea

[Sean Manning]

I was playing around with a copy of Herodotus and taking some notes when this leaped out and I had to write it up. Comments and criticism would be appreciated.

Readers of Herodotus' account of Xerxes' invasion of Greece are struck by a contradiction. On one hand he mentions the many races in the Persian army, including numerous half-armed skirmishers, but on the other hand his battle descriptions focus on troops from a few peoples. Most of these troops are sparabara, men armed with bows, spears, and sidearms and protected by tall wicker shields and some armour. Sparabara fought in formation and stood up to attacks rather than evading them, so they must be considered heavy infantry. The best-informed scholars have generally assumed that sparabara were the mainstay of the Persian army. But generally the balance between heavy and light troops is hard to measure, as is the balance between Medes and Persians and other troops. However, Herodotus gives us information that lets us estimate the proportions of different troops in the Persian army at Plataea in 479 BCE.

After his naval defeat at Salamis Xerxes King of Kings had left Greece in charge of Mardonius, with the best of the invading troops and those of his new Greek subjects. We do not know how strong this force was. We can throw out Herodotus' claim that there were 300,000 barbarian troops and perhaps 50,000 Medized Greeks in the imperial army. Among other reasons, the stockaded camp they built would fit no more than 120,000 men, some of whom would be noncombatants, and Herodotus gives no breakdown of these troops into individual contingents. However, Herodotus is better informed about the opposing Greek forces. Herodotus tells us that there were 38,700 free Greek hoplites, and 69,500 light armed. The second figure looks inflated, and the Greek light troops did not form part of their main battle line. However, the figures for hoplites in the coalition army is precise and plausable and based on the total strengths of twenty individual contingents. It may be somewhat optimistic, but each city had several neighbours ready to denounce it if it sent less men than it had agreed. We get the impression that the Persian line was not significantly longer than the Greek, and are told explicitly that the main Greek and Persian lines consisted of infantry only.

During the campaign of Plataeae both armies spent many days lined up opposite each other trying to lure their opponents to advance and fight on unfavourable ground. Herodotus tells us the strengths of the independent Greek contingents, and which units on each side faced each other. I have summed up these strengths. The Persians faced 11,500 independent Greek hoplites, the Medes 8,900 hoplites, the Bactrians 3,400 hoplites, the Indians 1,300 hoplites, the Saca 2,000 hoplites, and the Medized Greeks 11,600 hoplites. This is confirmed by Herodotus' statement (8.113) that the Persians were the most numerous contingent, with the Medes about as strong in numbers. He mentions that a few troops from other peoples are included in the above totals, but apparently their numbers were much smaller than any of the 'big five' contingents and some must have fought in front of the main line. These numbers need to be multiplied by some factor depending on the relative strengths of the Greek and Persian armies, but they tell us clearly that most of the non-Greek troops were Medes and Persians. The Bactrian, Saca, and Indian infantry fought as unshielded archers and should be considered light infantry. I see no reason to believe that the mix of troop types was much different in the original invading army, since many Persians went home along with most of the troops from the smaller contingents.

Unfortunately, we have no evidence for the strengths of different cavalry contingents. Herodotus does suggest that the only ethnic Persian cavalry present were the elite regiment of 1,000 horse. He also mentions a substantial force of Boeotian and Thessalian cavalry. I expect that Bactrian and Saca cavalry formed a larger propotion of the cavalry force than they did of the infantry since these regions produced many excellent horsemen. Based on other ancient armies, I would estimate that the Persian cavalry strength was about 10-20% of their infantry force. But we don't know how many infantry were in either army.

These numbers explain the emphasis on sparabara in Herodotus' accounts of battles in the Persian War. The unshielded archers and skirmishers who fill out his lists of peoples were only a small proportion of this army. And two thirds of the invading infantry were heavy infantry. The idea is not new (eg. Pierre Briant suggests that most of the satrapal contingents were very small or left behind in Asia), but I believe this argument is. This evidence also confirms that Persis could support several tens of thousands of soldiers.

Incidentally, Herodotus also says (8.113) that this was all the Medes, Saca, Indians, and Bactrians from the original invading army. People playing the game of guessing the strength of Xerxes' invasion force may want to consider this.

This essay is, of course, copyright Sean Manning 2008

[Gaius Julius Caesar]

OK, so what is the main theme of this? the actual numbers left with Mardonius?
What is the evidence that all the allied or satrapal forces were included in the stockaded camp site? Perhaps they were left outside, as a sign of their lower status?

Perhaps Herodotus focus on the heavier troop as the only ones worth detailed description as equal opponents of the hoplite classes makes it seem as though the proportions were heavy on their side? i always concluded that there were a wide variety of troops and the Greeks were in a position to win if they pulled together, the archers and cavalry being their biggest threat, especially when it came to positioning their troops, as in the need for water and the high ground, if i am remembering the correct engagement. You would need plentiful numbers of lighter troops to harass the enemy, to stop them from getting to water etc.

But interesting to see this recurring theme! 8)

[Giannis K. Hoplite]

The numbers of the Persian invasion are so messed up among ancient and modern scholars that we have heard every possible number,from the bigger to the smaller.And it is most difficult to estimate the plausible ones,due to so many factors.I have only questions to add.
First,we can give a higher limit of the total persian force from the size of the camp.Of course we reject Herodotus's numbers,but we can accept (?) the size of the camp. The only thing we can make out is that Herodotus(his sources) contradicts himself. Now,lets take the size of the camp for granted. 120.000 soldiers in it. In fact not soldiers,because as you said,a big proportion of them must have been non combatands. So,if the Greek numbers are correct,the Greek fighters were more numerous than the Persians.
Another strange thing is that we hear about units from the most distant areas of the Empire. I remind that Kyrus the younger was able to gather a vast army only from his Satrapy.And in much less time than the four years that Herodotus gives to Xerxes(actually he gives 10,but the previous 6 years he was busy in Babylon and Egypt).Herodotus speaks of the Persians as the less practical of people.They are so rich and numerous that they don't care about such small things as the cost.Why bring soldiers from India,for instance?
Another curious,to me, thing.The Persians knew the Greeks better than Herodotus describes.They had fough with them in Asia Minor,they had conquered them,they had ended their revolt,and they had made alliances with them. In 80 years,the Persian troops never dared successfully to attack a much smaller Greek hoplite force. Wouldn't it be insane to leave a relatively small (ok 120.000 is NOT small) army in unknown unfriendly land,with the fleet already defeated one time by the Greek? And in the end they were left to face an equal in numbers Greek army of much better quality.
I mean,nobody believes any more that the Persians were some dump who relied in the masses only to win a battle.This is only Greek views,of those who won them!Never forget that Herodotus's sources were the winners of these battles!This is good and bad in the same time.
And by the way,if the Medized Greeks were 50.000 as herodotus said,then the Greeks must have won another equal greek army PLUS the hundreds of thousands of Persians hock:
I had read somewhere that the numbers Herodotus uses for the Perisians must be always reduced to 1/10 due to a similarity of 100 and 1000 between Greek and Persian language.Unfortunately I cannot buck up this,or find the reference,but I suppose if it was that clear,we all should have considered it by now.
Khairete
Giannis

[Artabanos]

The reason why so many different people were recruited was because the Persians could not face Hoplite armies with tens of thousands of soldiers without full cavalry support. The Persians were horseman not infantry and there was probably no large heavy infantry contingent except the 10.000 Immortals.

Ethnic Persians were not numerous, those "Sparbara" with large wicker shields or better said pavise were either different people from all the empire or non-Aryan aboriginal people from the Iranian plateau. The Persians never allowed non-Aryans, or better said people not belonging to the nobility of the Iranic tribes, to enter their ranks. This only changed with Khosrow I's military reforms in the late Sasanian period 1000 years later. Herodotus and Xenophon were about right when they said that Persians never showed themselves without being mounted of their horses.

As for Plataea:

Xerxes army must have been indeed larger than the Greek because the cavalry could not be used like in Asia and their empire had not enough heavy infantry to mess with the Greek hoplites. Therefore they most likely took a larger number of their "multi-role" infantry supported by the Iranic cavalry mainstay.
After Xerxes left, he took with himself the Immortals and most likely a good part of the elite heavy cavalry plus other troops.
Mardonious had therefore less heavy infantry to face the Greek Hoplites and the task to hold Attica and repel Greek counter attacks.
His retreats to Thessaly clearly shows that he didn't believe that his infantry could hold the Greek Hoplites busy giving time for they main force, the Cavalry, to manoeuvre in the difficult terrain and outflank the Greeks.
His favourite battleground Thessaly was not taken by the Greeks and his mission was not all that dangerous as it seems because even after a defeat, the main force, the Iranic force could easily escape back to Asia. So he Xerxes knew that Mardonious and the Iranic force had a backdoor in the case the new possessions could not be hold.

In fact the Macedonian cavalry was probably the first Hellenic force that had a defence against the capability of Persian cavalry armies to attack and retreat at will without loosing larger numbers of troops. But even the small number of Macedonian cavalry in comparison with that of the Persians did let the Persians use their retreat tactics. This was necessary for success because forces like 40.000 Greek or Macedonian Hoplites are not only heavy and well trained but also larger in numbers than any heavy infantry force the Persians could ever field.

Conclusion: The Persians had access to a large number of troops, but since horsemanship and cavalry was an important part of their culture they only fielded one body guard contingent of heavy infantry, much smaller in number of the Hoplite armies the Greeks could field. But being a cavalry force was no handicap of the Persians, this gave them the capability to attack and retreat at will, reducing the numbers of casualties of Persian armies after lost battles.


It seems to me that the scholars ignore these points, wanting to believe the Hellenic point of view in which the Greeks defeated vastly superior "looking" Persian armies due to the latter’s "decadence"... But it must be realized that "ancient middle eastern/oriental armies" like that of the Babylonians and Assyrians were completely different than Iranic armies. The Persians were not the successors of the Assyrians in terms of military.

[Paralus]

The Persian numbers for Plataea are hopelessly confused and the true figure is likely irrecoverable. What can be said is that were the Persians 300,000 or more, we would likely be speaking Farsi today. I’d tend to agree that the hoplite levy given by Herodotus is near to correct (as near as memory allows). The numbers reflect that which will have been raised by each city contributing to the force.

The Greek “light armedâ€

[Sean Manning]

Well, ths has inspired more controversy than I expected. Some of it focusses on things I didn't know I was talking about. In particular, I make no claim about how strong the Persian army was except that it did not greatly outnumber the Greeks and that Herodotus' 350,000 is nonsense.

Paralus\\n[quote]The Persian numbers for Plataea are hopelessly confused and the true figure is likely irrecoverable. What can be said is that were the Persians 300,000 or more, we would likely be speaking Farsi today.

The Greek “light armedâ€

[Aryaman2]

Herodotus arrives to his number for Greek light infantry through "mathematical operations", he assigns 7 to each spartan and 1 more to each hoplite of the rest of cities. To sum up, it is a number that should be plainly dismissed.
As for the hoplites, it is very likely from an official list, but mind the round figures given for each force, it is not from a muster roll, it is probably the contingents the each Greek city promised to send. Reality, as always, would be a considerably lower number. The research of David Parrot on French armies in the 17th century shows that real strengthon average was about 60% of the official strength, that would put the Hoplites at about 25.000. Mind you, it is still a large army for premodern logistics.
The Persian army, with a strong cavalry force, would be probably even less numerous.

[Paralus]

Well, the evidence suggests there were only a few thousand Indians. It is surprising that he mentions them, since they were from a remote area and we have no evidence that they were especially good or bad troops. But bringing at least a few troops from every people in a big royal army seems to have been associated with the Persian ideology of "many lands, one king of kings." The evidence I have founds supports the idea that most of these contingents were small in number and little used in battle.

The most likely explanantion, propounded by Cawkwell, is that Herodotus' army list is a satrapal list or "list of nations" (as with Behistun for example) that dates to the early part of Xerxes' reign. This allows for Herodotus to assign a contingent for each. It also explains the missing Medes such as the Mardians and Cadusians as well as the lumping of ethnic contingents under commanders with absolutely no attachment to them: Artayktes, "regent of Sestos" who commanded the troops from the Balck Sea.

In all of the battles we hear of predominantly Iranian troops. It is more likely that they formed the bulk of the invasion forces. Even Herodotus, supposedly in possession of an "order of battle", can find no real place for the Egyptians (hitherto marines), Ethiopians, Mysians and Phrygians who he resorts to describing as "mixed up as well" in the line...somewhere.

As for Xerxes taking the bulk of the land forces with him on his "flight", most unlikely. He returned - reasonably continently - through a subdued satrapy with his guard. Even should we acknowledge Artabazus going with him with a large contingent, it returned for Plataea.

As for the hoplites, it is very likely from an official list, but mind the round figures given for each force, it is not from a muster roll, it is probably the contingents the each Greek city promised to send. Reality, as always, would be a considerably lower number. The research of David Parrot on French armies in the 17th century shows that real strengthon average was about 60% of the official strength, that would put the Hoplites at about 25.000.

Excellent point and well taken. As I wrote above, Herodotus’ numbers will have derived from the hoplite contingents to be mustered by each state. The interminable wait, if we are to take Herodotus at his word, allowed for those and others to trickle in. That, if true, is an indication of the somewhat “softâ€

[Sean Manning]

Well, the evidence suggests there were only a few thousand Indians. It is surprising that he mentions them, since they were from a remote area and we have no evidence that they were especially good or bad troops. But bringing at least a few troops from every people in a big royal army seems to have been associated with the Persian ideology of "many lands, one king of kings." The evidence I have founds supports the idea that most of these contingents were small in number and little used in battle.

The most likely explanantion, propounded by Cawkwell, is that Herodotus' army list is a satrapal list or "list of nations" (as with Behistun for example) that dates to the early part of Xerxes' reign. This allows for Herodotus to assign a contingent for each. It also explains the missing Medes such as the Mardians and Cadusians as well as the lumping of ethnic contingents under commanders with absolutely no attachment to them: Artayktes, "regent of Sestos" who commanded the troops from the Balck Sea.

I agree that Herodotus' 'list of contingents' may well come from some other source than a description of Xerxes' army. Whether every satrapy did send a small contingent is something we will never know for sure. But the Plataea list is quite different.

In all of the battles we hear of predominantly Iranian troops. It is more likely that they formed the bulk of the invasion forces. Even Herodotus, supposedly in possession of an "order of battle", can find no real place for the Egyptians (hitherto marines), Ethiopians, Mysians and Phrygians who he resorts to describing as "mixed up as well" in the line...somewhere.

Yes, and my evidence supports this. Four sevenths of the Persian line was West Iranians, two sevenths Greeks, and one seventh other contingents including Saca and Bactrians who we know were prominent in Persian armies. The only oddities are the Indians, the Egyptian marines, and the various skirmishers from the lands bordering the Aegean. I'm suspicious of the Egyptians and surprised by the Indians, but don't think any of these can be rejected.

As for Xerxes taking the bulk of the land forces with him on his "flight", most unlikely. He returned - reasonably continently - through a subdued satrapy with his guard. Even should we acknowledge Artabazus going with him with a large contingent, it returned for Plataea.

Again, I mostly agree.

As for the hoplites, it is very likely from an official list, but mind the round figures given for each force, it is not from a muster roll, it is probably the contingents the each Greek city promised to send. Reality, as always, would be a considerably lower number. The research of David Parrot on French armies in the 17th century shows that real strengthon average was about 60% of the official strength, that would put the Hoplites at about 25.000.

That is why I said it may be optimistic and need to be discounted by some fraction. I wouldn't discount is as far as 40%, however, because the situation was very different from a seventeenth-century war.

[Aryaman2]

That is why I said it may be optimistic and need to be discounted by some fraction. I wouldn't discount is as far as 40%, however, because the situation was very different from a seventeenth-century war.

It is just an average, in the 17th century there were also a good number of different situations and outcomes. However the larger part of the attrition is suffered in the first months after raising a force, so I think the Greek forces had enough time in campaign to suffer a significant attrition

[Tomenable]

The Persians faced 11,500 independent Greek hoplites, the Medes 8,900 hoplites, the Bactrians 3,400 hoplites, the Indians 1,300 hoplites, the Saca 2,000 hoplites, and the Medized Greeks 11,600 hoplites.

Interesting.

This would mean that the Medized Greeks faced 30% of the entire force of enemy hoplites (11,600 of 38,700). And according to Herodotus, the Medized Greek force numbered in total "perhaps 50,000". According to Herodotus, the entire free Greek army numbered 110,800 (38,700 hoplites - including 5,000 Spartiates -, 35,000 helots - 7 per each Spartiate -, 34,500 skirmishers of other hoplites, 800 Athenian archers and 1800 Thespians of uncertain equipment) - thus hoplites were roughly 35% of this army, according to Herodotus. If we assume a similar proportion of hoplites to other troops among Medizied Greeks, and accept their number as 50,000 - this would mean they had ca. 17,500 hoplites.

So the Persians deployed 17,500 of their own hoplites vs 11,600 enemy hoplites? Considering that Medized Greeks were probably the best heavy infantry that Persians had in this battle - this indicates a rather "generous" use of this force by the Persian commander. If I was him, knowing that I have numerical superiority in general, but smaller number of good quality troops, I would deploy my 17,500 hoplites facing maybe even 25,000 enemy hoplites - and just support my hoplites with huge number of other troops, perhaps skirmishers or archers, exploiting my overall numerical superiority.

Deploying 17,500 hoplites vs enemy 11,600 hoplites would be a waste - considering that it means we (Persians) have nothing of equal quality to deploy against the remaining 27,100 enemy hoplites. This would mean 60% of our (Persian) line is vulnerable to enemy attack in close-combat, while only 30% of our line is protected by troops of equal quality to enemy troops facing them. If my reasoning is correct.

This would indicate that either Persians made a terrible mistake, or that Herodotus gives an inflated number of Medized Greeks. Or - that proportion of non-hoplites to hoplites in Medized Greek forces was even higher than in free Greek forces (so even less than 1/3 Medized hoplite per each non-hoplite). This, however, is highly unlikely, considering that the bulk of light troops in "free" Greek army were helots - 50% of all light troops (7 per each Spartan). When not counting Spartans, the proporition of hoplites to light troops was rather 1:1 than 1:3. But if similar proportion was present in Medized Greek forces, it would mean they had even more - because nearly 24,000 - hoplites! Did Persians deploy 24,000 hoplites in a line which faced enemy line of just 11,600 hoplites?! Even more unlikely. The most probable conclusion is that Herodotus inflated the number of Medized Greeks and that his estimate of 50,000 should be decreased - perhaps at least by half - to get a more realistic number.

Of course another possibility is that Persians wanted to concentrate the best part of their infantry against relatively small (30%) section of enemy line - perhaps to achieve a numerical superiority without qualitative inferiority there, and as a consequence - to attack and win the battle there. In such case - if we believe Herodotus - they massed all of their hoplites along 30% of the battle front, achieving a more or less 2 to 1 numerical superiority just in hoplites (!) over the enemy army in that section of the front! But in such case - why they lost? And why they lost so badly (if - once again - we believe in what Herodotus wrote about casualties, which is even more hard to believe - if not impossible).

[Tomenable]

In 80 years,the Persian troops never dared successfully to attack a much smaller Greek hoplite force.

Really? So how they won their victories against Greek hoplites during the Ionian Revolt?:

http://en.wikipedia.org/wiki/Ionian_Revolt

[Macedon]

Wow... now this must be the epitome of a thread resurrection... Peter, I hope you noticed that the posts you are commenting on have been made about 5 years ago... On the other hand, the battle of Plataea always is a very interesting battle to disagree on.

[Sean Manning]

Dear Peter,

Doing arithmetic on Herodotus' figures for Xerxes' army can be fun, but I personally don't think this is a good approach, because its so unlikely that they have any systematic relationship to the truth, and if they do there is no way to find out. I think that the fleet strengths and Free Greek hoplite numbers are more likely to have some connection to the truth, because while very high they are less so, and because Herodotus would have found it easier to find evidence for them.

I also am not sure that Mardonius would have believed that his men were mostly inferior to the free Greeks. He probably thought back to the land fighting of the Ionian Revolt, about which we know very little. We aren't even certain that the Greeks fought in ranks and files by 480! Both sides behaved cautiously during the Plataea campaign, suggesting that they respected each other's fighting abilities.

[Tomenable]

Hi Macedon and Sean,

Dear Peter,

Doing arithmetic on Herodotus' figures for Xerxes' army can be fun, but I personally don't think this is a good approach, because its so unlikely that they have any systematic relationship to the truth, and if they do there is no way to find out. I think that the fleet strengths and Free Greek hoplite numbers are more likely to have some connection to the truth, because while very high they are less so, and because Herodotus would have found it easier to find evidence for them.

I also am not sure that Mardonius would have believed that his men were mostly inferior to the free Greeks. He probably thought back to the land fighting of the Ionian Revolt, about which we know very little. We aren't even certain that the Greeks fought in ranks and files by 480! Both sides behaved cautiously during the Plataea campaign, suggesting that they respected each other's fighting abilities.

Well sorry for resurrecting a very old thread but I think I had a good reason to do this (and it's better to resurrect old threads than creating new threads which are exactly the same, IMO). :wink: Regarding the credibility of numbers given by Herodotus - I found one article written by prof. Livio C. Stecchini which defends Herodotus and his credibility as a historian:

http://www.iranchamber.com/history/artic..._wars5.php

http://www.iranchamber.com/history/artic..._wars8.php

It is a very interesting article to read even if the author is wrong in his judgements. But he seems to have some good points. On the other hand he doesn't explain the casualty figures for both sides given by Herodotus (which are actually much more ridiculous than his strength figures).

There are also other possible explanations of numbers given by Herodotus, without accepting the theory that he was simply a liar and / or a Greek propagandist. For example, he might have given figures of so called "paper strength" - while real strength could be much smaller than "paper" one (especially at Plataea, considering that diseases were almost always taking heavy toll in all armies before 19th century). Another possibility is that Persians counted their "strength" in a smiliar way as later Ottoman Turks did.

I explained this on another forum - let me quote that post:

Why Turkish Medieval / Early Modern Era armies were so massive (compared to their enemies) according to sources from that time?

Well, if you count camp followers you get enormous strength of armies.

For example the Turkish army in the battle of Vienna in 1683 had 50,000 horse-drawn wagons with it (and we must count one coachman and a few camp followers for each horse wagon). So apart from 150,000 soldiers the Ottomans would have probably 150,000 - 200,000 camp followers (including coachmen) with them at Vienna - and 50,000 horse-drawn wagons (one wagon for each 3 soldiers). The same would apply to European armies of course.

And it seems that Turks actually had a habit of counting also camp followers and even horses and camels as "army strength" in their official military documents. I guess it had some propaganda impact on their enemies.

But they didn't count camp followers and horses when counting casualties, on the other hand.

For example Polish 17th century soldier Jan Ostrorog commented the discrepancy in numbers given on Turkish forces in the battle of Chocim in 1621. Turkish own counts said that their army at Chocim in 1621 was 300,000 Turks and 100,000 Tatars - in total 400,000. While Polish envoy who was also acting as a spy during his mission to Turkish camp, wrote that there were not more than 150,000 Turks and not more than 60,000 Tatars. And in fact the first number (300,000 + 100,000) most likely includes all men and horses and the second one (150,000 + 60,000) probably includes all men (including camp followers).

And here is how Jan Ostrorog explains these great differences in numbers:

"(...) Sir Zelenski [that Polish envoy] counted their forces using Polish method, which is considerably different than Turkish method - in the Turkish army they count separately each living creature, for example if they have a mounted knight and this knight has another horse, or a mule, or a camel - then they count all of this separately, and that's why they get so huge numbers of their armies (...)"

And Ostrorog concludes:

"(...) That's why when there will be a 10,000 strong Polish force, there will still be more actual soldiers among these 10,000, than in a Turkish force which is said to be 20,000 strong, or maybe even in a Turkish force which is said to be 30,000 strong (...)".

Another Polish envoy - Krzysztof Zbaraski - who visited Turkey few years later, wrote:

"Forces of the Turkish tyrant are bigger on paper than in actual armies, because when they count strength of their forces, they count both men and horses"

So maybe Persians also counted not just actual soldiers, not just soldiers, camp followers and other non-combatans (camp servants, baggage train, etc.) but even also animals (horses, baggage mules & donkeys, camels, etc.) as the "strength" of their armies? And maybe Herodotus indeed had access to Persian figures - but he didn't realize the fact, that they "counted camels as soldiers"? :mrgreen:

I also am not sure that Mardonius would have believed that his men were mostly inferior to the free Greeks. He probably thought back to the land fighting of the Ionian Revolt, about which we know very little.

The Ionian Revolt was surprisingly / relatively easily smashed by the Persians. So if anything I would say that experience from combats during the Ionian Revolt gave the Persians the overconfidence in their own strength and underestimation of capabilities of the enemy. But then we have battles like the Marathon. However, the Marathon could be considered by the Persians as just an "on-the-job accident" - especially considering that the battle was fought in a rather unfavourable conditions for the Persian army (a surprise charge by the Greeks down the hill against a probably not prepared and not deployed in a proper battle formation, Persian army - which probably had no cavalry in this battle, as it either already departed back on their ships, or was grazing their horses away from the battlefield).

Both sides behaved cautiously during the Plataea campaign, suggesting that they respected each other's fighting abilities.

... or that one side respected enemy fighting abilities, while the other side respected enemy numbers.

Question is which side respected what.