Best missile unit (non gunpowder)

[katank]

Not necessarily true. If you place archers out front like xbows, their trajectory becomes noticeably flatter as well and can easily hit units behind it.

Besides, faster rate of fire -> more kills in typical battles -> more valor -> sniping machines

[Reapz]

Alatien

Thanks for posting test data. You had the result of Crossbowmen doing so well because of this little detail

The results below show the final numbers in units (starting 60) after depleting ammo by both sides

As I'm sure you will have observed in your tests the longbowmen fire faster. If you look at the numbers at the time the longbowmen exhaust arrows as Sinan said the longbowmen own crossbowmen. You are letting your longbowmen stand and take missile fire after they are depleted and that's when casualties accrue and flip the stats the other way.

If you play that way in campaign (let your units get riddled to allow the slower firing opponents to catch up) then you can expect the results to be as you posted. If you use longbowmen until they exhaust arrows and then move on to new tactics they will handily beat the slower crossbowmen.

I think the foz 4 was saying that

If you want to apply this to your missle vs. missle battles, the thing to do is stop both units when either is out of ammo, as then they've been active for the same amount of time. This is also the most realistic (as others have pointed out) because archers with no more ammo do not wait around to be slaughtered by crossbows unless they are commanded stupidly.

[katank]

I personally compare missiles 1 on 1 by having the unit that finishes firing first charge at the other and try to melee. Whoever wins combined shoot/fight is overall victor.

[Shahed]

Admittedly, longbows can get more volleys off in a certain time than x-bows, but I find, due to the trajectories involved, each x-bow volley is considerably more damaging than an archer volley - with the flat trajectory, x-bows are more likely to hit other units on the way/after as well.

This is where horse archers really shine. They can move fast enough to fire from the flanks. From where missiles which miss their original target can go along the line and hit someone else.

Reapz thanks for the refresher. Ofc that is what the problem with these tests is, the ammo-less longbows are just standing there.

[Blademun]

While many people stick up for longbows, I would have to say that Genonese Xbows do get a major advantage: Numbers. 8 Xbows vs 4-5 Retinues price for price, Genonese Xbows deal more damage per minute.

I think though that Xbows really shine when used in conjunction with pikes. Make a line of Pikemen milita, and then a line of Genonese Milita. Make the pikes 3 ranks deep and the archer 2 ranks deep. Place the Archers partialy inside the pikes, toward the front. Seperate, these two units are not so strong. TOgether they cover each others weaknesses perfectly. The Xbows act like human sheilds, their huge pavises intercepting enemy arrows for the vulnerable pikes. The pikes provide cavalry charge stopping power and offensive punch in cqc, while the Xbows are more defensive.

Heres a Example of the formation:



This is just with regular Pavise Xbows. Genonese provide even better results.

Typically, behind this I will put cheaper archers, like archer peasants unless I've got loads of cash. I like HRE for this best simply because I can mix in a couple Zwei handers and Gothic Knights. Milanese have great xbows but they seem to be lacking everywhere else.

[Reapz]

Blademun

I think it is a fallacious argument. Unit for unit Retinue Longbowmen beat Genoese Crossbowmen and inflict more damage per minute on targets. If you have to use 8 units in your stack to match 4 or 5 units in mine I am just going to use my extra slots to recruit other units (probably heavy cavalry) that add to the power of the stack. You don't win the fight because you built a cheaper stack.

Pavises do not protect any troops other than the crossbowmen carrying them. As long as you disable fire at will and manually target the longbowmen at the pikemen (or whoever is hiding behind the crossbows), they will be effectively targeted. I have repeatedly killed hundreds of pikemen or spearmen with single units of longbowmen this way. They are very powerful missile troops. Check out this set of pics in the amusing screenshots thread and look at the battle stats in the last pic - 5 units of longbowmen killed 1700 Danish troops in that battle and they were basic longbowmen (plus one or two experience), not Retinues.

I don't know how anybody can doubt the power of longbows if you use them to best advantage. Against an array as you describe with spears or pikes hidden behind crossbowmen I will ignore their missile troops in front and shoot over them to kill the pikes. I know I can kill their spears or pikes faster than they can kill mine with crossbows. I can then wipe out the crossbowmen, left at full strength, with cavalry.

[katank]

Also note that I don't need retinue longbows to get the 8 AP missile attack. I can easily settle for yeomen archers which are a helluva lot cheaper.

[Skott]

Crossbow men need to be out front moreso than Longbow men IMO. In RTW you could put archers behind the front line and just wipe out a enemy from safety but in MTW2 you have to put most archers up front to get any kind of good killing impact. Here they are subject to attacks.

However in battles, especially in defensive fights, I can put Longbow men behind main lines and get better results with them than I could with putting Crossbow men behind the main lines.

There are pros and cons between the two types of bowmen. But like mentioned earlier the person who uses the combined arms the most effecient will come out the winner. Wether it be defensive or offensive.

[Skott]

Double Post...Sorry

[Blademun]

Pavises do not protect any troops other than the crossbowmen carrying them.

I disagree. The Xbows DO protect the Pikes. Arrows are actual physical entities in this game. If something intercepts them before they reach their target, they cease to exist. Xbows protect the Pikes by physically intercepting arrows that come at low trajectorys at the pikes. That happens to be the most accurate trajectory and best one to defend against. The archers arn't just infront of the pike formation, they are inside it, just 1 line out ahead of the pikes. When you target the pikes, the arrows hit both units. If the arrows come from a lower trajectory, they will mostly hit the Xbows.



Of course, to put things in context, If I knew I was playing against the english, I wouldn't try to outshoot him. I would work on using cavalry and flanking manuvers (to avoid stakes) , and Serpentines if availible.

[Reapz]

Blademun

You are assuming that the animations drive the outcome of the battle. That is not the case. The game engine decides the outcome and the animations give some graphical representation to what the engine is doing but calculations about units' casualties aren't driven by animations. You can't intuit what is happening from watching animations as sometimes thay just don't represent what the game engine is doing.

[Foz]

Pavises do not protect any troops other than the crossbowmen carrying them.

I disagree. The Xbows DO protect the Pikes. Arrows are actual physical entities in this game. If something intercepts them before they reach their target, they cease to exist. Xbows protect the Pikes by physically intercepting arrows that come at low trajectorys at the pikes. That happens to be the most accurate trajectory and best one to defend against. The archers arn't just infront of the pike formation, they are inside it, just 1 line out ahead of the pikes. When you target the pikes, the arrows hit both units. If the arrows come from a lower trajectory, they will mostly hit the Xbows.

There's no real way to eye this up, and I really doubt either of you has a real leg to stand on concerning evidence. If so, post away, as I've heard nothing with any substance behind it yet. If not, then if you want anyone to actually know what's going on, figure out a way to test it empirically and post the results. If the arrows really ARE physical entities then I imagine you'll notice the pikes take loads less damage with pavise X-bows parked in front as described vs the pikes just standing there naked. If they're not physical entities or for some reason the game ignores the X-bows since the pikes are targeted, then I'm guessing the pikes continue to get blasted as usual.

You'll also want to test separately to determine if this interferes at all with the firing of the X-bows, as it's possible the non-front ranks will have their fire interfered with by the pikes near them, and this almost certainly would negate the benefit of blocking some incoming missiles from hitting your pikemen.

[Reapz]

foz

it is really a much more general issue - does mingling units allow a weaker unit to capture the armor bonus of the stronger one - protect it in some way? I have never seen anything posted by developers or anybody else to suggest this happens. It has not been my experience in game either.

Again I think people are misunderstanding the way the game engine works if they think combat outcomes are contingent on the graphically depicted events, animations, etc that play out during battles. The engine calculates the results. The display depicts the combat but the results of unit combat don't wait for graphics to play out to see what happens. Remember there is an autoresolve function that calculates battle outcomes with no graphics. Of course it matters where you position units, formation, etc, but the actual animations do not drive the outcome. If arrows can't be seen to strike a target or swords aren't swinging enough that does not mean that the game engine isn't going to calculate a succesful attack. Bottom line - animations are eye candy but not driving the engine's calculations of combat, just depicting what the engine is deciding, and sometimes not accurately depicting it.

[Spark]

You must be joking?

Custom battle, genoese crossbowmen vs. scottish pike militia and scottish highland nobles.

AI designates the pike militia as the general unit, places it behind the nobles. Both stacks march forward, with the nobles in front, towards your crossbowmen. Turn fire at will off, target pike militia. By the time they get close, the nobles will have suffered more casualties than the pike militia, despite the nobles' 9 defense compared to the pike militia's 1. It's clear as day that arrows do indeed behave as they should, as opposed to entirely based on some calculation. The pike/crossbow spearwall does work, the pavise will absorb a lot of the arrows.

[geala]

Of course animations mean a lot in M2TW. It's far more than mere optic. You should take a look to the modding sections. But I think that's not the actual problem.

Missiles are calculated with a certain trajectory. If someone is between the missile unit and the aim he will (at least partly) be affected. If he is after the unit aimed at he will be affected if the bolt or arrow does not hit someone in that unit.

But I don't know wether it is a good idea to put two units together in one place because there should be some modifiers to create a negative affect for such piled units.

[Reapz]

You must be joking?

Custom battle, genoese crossbowmen vs. scottish pike militia and scottish highland nobles.

AI designates the pike militia as the general unit, places it behind the nobles. Both stacks march forward, with the nobles in front, towards your crossbowmen. Turn fire at will off, target pike militia. By the time they get close, the nobles will have suffered more casualties than the pike militia, despite the nobles' 9 defense compared to the pike militia's 1. It's clear as day that arrows do indeed behave as they should, as opposed to entirely based on some calculation. The pike/crossbow spearwall does work, the pavise will absorb a lot of the arrows.

I don't follow this at all. Where are the pavises in a combination of scottish pike militia and scottish highland nobles?

If your point is that the troops nearby a unit targeted with missile fire also get depleted - that is not in dispute. If your point is that the casualty rate is lower in units posted further away than units nearer to the archers firing - that is not in dispute. What is being argued is a scenario involving a shielded unit occupying the same space as a non shielded unit - here is Blademun's post:

I think though that Xbows really shine when used in conjunction with pikes. Make a line of Pikemen milita, and then a line of Genonese Milita. Make the pikes 3 ranks deep and the archer 2 ranks deep. Place the Archers partialy inside the pikes, toward the front. Seperate, these two units are not so strong. TOgether they cover each others weaknesses perfectly. The Xbows act like human sheilds, their huge pavises intercepting enemy arrows for the vulnerable pikes

This is your impression:

By the time they get close, the nobles will have suffered more casualties than the pike militia, despite the nobles' 9 defense compared to the pike militia's 1.

Show me test numbers and then we can debate the reasons but I don't see any proof of this other than your impression.

It's clear as day that arrows do indeed behave as they should, as opposed to entirely based on some calculation.

Um, what do you think the game program does in computing combat outcomes other than calculations? The point is what type of calculation dictates the result? It is one independent of unit animations. The game engine does not compute the trajectory of every arrow before it can compute the outcome of combat. It computes the combat outcome while showing graphics of missiles flying.

[Musashi]

That's where you're wrong... the game does in fact appear to track projectile flights, and only units physically struck will be killed... If you put another unit in between the firer and it's target, particularly for a flat trajectory weapon, the unit in between will be killed and almost none of the other unit will.

[PaulTa]

I'd be really curious to hear something on this from the CA devs. From what I've heard, and this does make sense, calculating the kills based on the individual animations would be a huge drain on the computer. Having graphics that represent calculations, on the other hand, seems quite a bit more plausible.

I don't have any tests to back this up per say, but think about a unit of crossbowmen firing on some knights of the mounted sort. Some of those mounted knights will get pelted with ten or fifteen crossbow bolts before actually dying, while some take one or two and hit the ground. If the kills were calculated based on animations, I would expect the guy that sucks up fifteen crossbow bolts to be long past dead, not ready for more.

[Musashi]

No, because each "hit" is only a chance to score a kill. Much like in an RPG you roll to see if you caused damage after a successful hit.

More heavily armored units have a much higher chance to survive than less armored characters... Particularly against low end archers who don't have AP missiles. If you're talking about General's Bodyguards, they all have 2 hit points, so you have to actually hit them twice, and roll a successful wound twice to drop one...

[Silvershade]

My feeling is that dismounted Dvor are not only the best bowmen in the game but also the best infantry of any sort, fully fledged heavy infantry that have a ranged attack better than most archers makes these a fantastic unit.

[Tuidjy]

My feeling is that dismounted Dvor are not only the best bowmen in the game but also the best infantry of any sort, fully fledged heavy infantry that have a ranged attack better than most archers makes these a fantastic unit.

They are good, and they may be the best, but that is by no means certain.
The French have the Scottish longbowmen, and the adventurers. Both of
these have a nasty missile attack, and the longbowmen have amazed me
in hand to hand combat. Strangely enough, the adventurers don't seem
much worse statwise, but are not so overwhelmingly good at close quarters.

It's strange that the French, who early on lack any fancy missile units, have
three amazingly good ranged specialists: a longbowman, a crossbowman, and
a horse archer.

[Temujin]

Um, what do you think the game program does in computing combat outcomes other than calculations? The point is what type of calculation dictates the result? It is one independent of unit animations.

This is clearly wrong.

For instance, units using two-handed axes are currently bugged; some missing animations are causing them to NOT attack mounted enemies in melee, and they consequently die with 0 casualties inflicted in this situation. If the engine calculated the result and tried to approximate that in graphics, a couple of missing frames would only look odd, not affect the result. Yet the result is most definitely affected.

[Reapz]

If you think the game engine relies on graphical portrayal of individual troops within units, the trajectory of their weapons, the character of the strike of their individual weapon on individual opposing troops (not a whole unit but a single animated fighting man) to determine combat outcome that is wrong. The computing power required to run that kind of calculation isn't sitting on your desktop or mine. It is a nice idea but simply doesn't happen. Ask a dev or any game programmer.

The TW series game engine can do the whole thing with no graphics (autoresolve). THe TW series early games had flat 2D sprite animation for combat. You think the combat engine was transfromed when the game went 3D on the battlefield to wait for battleground physics for each of 10,000 animated men to be calculated before it gets a result?

[Zenicetus]

That's where you're wrong... the game does in fact appear to track projectile flights, and only units physically struck will be killed...

I think you have cause and effect backwards here. The game doesn't have to calculate individual arrow trajectories and then "find" a result and display it. If it did, the frame rate would crawl.... and it would be even worse with unit size set to large or huge.

If it works like other games of this type, then what's happening (roughly speaking) is that the combat algorithm is based on unit match-ups, not individual soldiers and missiles. When "unit A" fires arrows at "unit B", it matches up the stats and modifiers -- base attack rating, armor, defense, unit facing direction, terrain, experience, etc. -- and it calculates how many soldiers will die every time "unit A" fires an arrow volley at "unit "B". If the result is "3 soldiers will die", then the game shows a volley of arrows landing on unit B, with 3 of the normal soldier animations picked at random and replaced by the "stagger and drop to the ground, dead" animation.

It doesn't have to calculate an arrow trajectory to do this. If it did, then aiming in the game would be much more critical and results would be far more unpredictable. It would mean that a solider would be hit, or not hit, depending on whether he walked a few feet one side or the other of an incoming arrow.

We're not running supercomputers here, that can calculate results at the level of the individual soldier and arrow, with thousands of soldiers and arrows on the screen at the same time.

If you put another unit in between the firer and it's target, particularly for a flat trajectory weapon, the unit in between will be killed and almost none of the other unit will.

Sure, but the algorithm can still do that by working on the level of the whole unit, not the individual soldier or arrow.

[PaulTa]

I'd have to agree with Zenicetus on this one. One way is much more economic in terms of computer power, while the other seems like it would be a huge drain on your common PC. If minimum specs were anything less than a few gigs and a 2.8 intel proc, then I couldn't see it calculating the results based on the projectiles themselves instead of having the graphic reflect what the computer determined.

[Blademun]

Well, I went on and did those Impirical tests that were requested.

Setup: Enemy was one unit of Peasant archers, to rule out AP effects and other 'special abilities some archers get. I was HRE with a standard pike militia and Pavise Xbow. I let the archers fire their arrows till they ran out, then recorded the number of men left BEFORE the archers charged.

Pike/Xbow Formation: 49/33 respectively. Total Losses: 53

Xbows alone(pikes far away): 13. Total Losses: 47

Pikes Alone: 0 left (they went down to 6 then routed). Total Losses: 75


There you have it. While the total losses between the two were high, the damage was spread out between the two. This would suggest, that at least to some extent, flight trajectories are calculated and any units in the way get hit. Missiles are physical entities.

One unit of archers can only be selected to attack one unit, yet it did damage to two. THere is no possible way this could be possible judging by this:

If it works like other games of this type, then what's happening (roughly speaking) is that the combat algorithm is based on unit match-ups, not individual soldiers and missiles. When "unit A" fires arrows at "unit B", it matches up the stats and modifiers -- base attack rating, armor, defense, unit facing direction, terrain, experience, etc. -- and it calculates how many soldiers will die every time "unit A" fires an arrow volley at "unit "B". If the result is "3 soldiers will die", then the game shows a volley of arrows landing on unit B, with 3 of the normal soldier animations picked at random and replaced by the "stagger and drop to the ground, dead" animation.

This is NOT Age of Empires, by any means. Its a far more advanced game then that. Arrows and all other projectiles in this game have physical presence.

[Foz]

This is clearly wrong.

For instance, units using two-handed axes are currently bugged; some missing animations are causing them to NOT attack mounted enemies in melee, and they consequently die with 0 casualties inflicted in this situation. If the engine calculated the result and tried to approximate that in graphics, a couple of missing frames would only look odd, not affect the result. Yet the result is most definitely affected.

Yes, the result is affected, which also seems to indicate what blademun has seen - that attacks are caused by entities interacting. I think the mechanics of this need explanation though for people who may not be familiar with what is and is not possible.

Some people probably imagine that the game engine can somehow tell when an arrow or hand-held weapon on screen impacts a man, a shield, or whatever. Well yes... and no. In reality it has nothing to do with what's on screen. Identifying an arrow in a graphical scene by visually searching for the arrow and calculating whether or not it is currently intersecting anything that can be recognized visually as a man is probably far outside the scope of this game.

What IS possible, however, is that the game calculations for kills and the kill animations are directly tied to each other. From the way things seem to behave, I'm guessing the game tracks arrows and hand-held weapons (for melee perhaps the entire man) as separate entities, which are then checked to see if they have hit a man or an obstacle, using the coordinates of each thing in the scene... and this is where I think people get fuzzy. The calculations are based not on the picture you see, but on the numerical representation of the scene in coordinates, from which the scene you view is ultimately rendered. In short, the visual display on-screen doesn't dictate in any way what is going on, but rather is derived from what is going on at the calculation level. The game doesn't need to know that an arrow on your screen has gone through a man in the enemy unit - it simply calculates the position of the arrow object it is tracking, finds out it is inside the space defined as man #173 at the current moment, and so he has been hit and some piece of code that deals with arrow impacts executes to do something about it (and presumably determine if the man dies, and that the arrow must cease to exist). So the displayed picture doesn't matter at all, except to note that the game is probably displaying everything in the scene directly from the coordinate-based model of the scene that it uses to calculate everything.

As for hit detection, it seems likely that it's based on the intersection of objects as I suggested for arrows, in the case of melee too. This would illustrate why the bugged 2-handers get no kills: their weapons never intersect enemies when they attempt to swing, therefore the hit-code never triggers, and no kills result. I must conclude that all evidence points to the game keeping the entire battle as a physics-driven scene with interacting entities, as a simpler resolution of battle (at least none I've heard or know of) can not account for units with missing animations failing to hit anything, or arrows impacting objects that are in their path aside from their target.

[Reapz]

Blademun

If you look at some of the archery tests I did that are posted here you will see that the results vary markedly from test to test with exactly the same conditions - basically restart the same test and get different numbers. One test result doesn't tell you much as you can get a result you wont see even if you run another five tests.

Two units in the same location under missile fire will both sustain casualties. This does not negate what Zenicetus said.

You wrote "Arrows and all other projectiles in this game have physical presence." They are graphical representations. The developers could write code to show animations of the archers drinking beer instead of firing when you right click order the attack, and still have the target unit die when they are "fired" on. The animated men onscreen do not have to fire their arrows and hit their individual targets. This has been shown in other tests, that animations sometimes do not match the combat calculations from the game engine.

[Zenicetus]

This is NOT Age of Empires, by any means. Its a far more advanced game then that. Arrows and all other projectiles in this game have physical presence.

I'm not sure this has been proven by your test. I would be more convinced if people here reported that their frame rates dropped in half, when they doubled the number of soldiers and projectiles onscreen with the army size setting in Options. And that's not happening, as far as I can tell.

Different army size settings may cause a slight reduction in frame rate (depending on your video card) due to increased 3D rendering load, but not the kind of drop you'd expect by forcing the CPU to double or triple the number of individual projectile calculations it's (supposedly) doing. The fact that we can double or triple the number of soldiers onscreen with fairly minimal impact on frame rate, suggests to me that battle calculations and results are based on units (which don't change with army size setting), and not individual soldier/arrow calculations.

[Foz]

I'm not sure this has been proven by your test. I would be more convinced if people here reported that their frame rates dropped in half, when they doubled the number of soldiers and projectiles onscreen with the army size setting in Options. And that's not happening, as far as I can tell.

You are assuming many very shaky things by suggesting that double unit size should cause 1/2 frame rate:

1. That the battle calculations, not the rendering of the scene, is the limiting factor in your frame rate. I, on the other hand, would suggest that any bottleneck comes from something to do with the graphics processing as opposed to the battle calcs. It seems to me that some simple mathematical computations for moving an extra man and arrow around in a battle physics engine are completely miniscule compared to the thousands of computations that must be done to draw and dynamically light a single extra man and arrow into the scene. I don't think you could imagine enough detail into the battle mechanics to make the calculations even remotely compare to the ones done to render the scene.

2. That nothing else in the scene affects frame rate except the units. Look at the field of battle we're talking about here. The game draws a hugely complex scenery, including detailed blades of grass and numerous highly-detailed buildings often. As they account for a huge percent of the scene a lot of the time, it's misguided to suggest that doubling the unit size should halve the frame rate, since the unit drawing clearly can't account for 100 percent of the scene rendering time in the first place. If it's for instance 50% of the drawing time (not trying to guess the amount here, it's just as an example), then doubling it makes 150% of the original scene draw time, and your frame rate would be 2/3 original. Also, this isn't accounting for any non-unit-based calcs or routines the game does during a battle. There obviously are some.

3. That thousands of object computations should horribly lag your system. Consider chess engines for a moment. The Chessmaster engine can permute the game of chess 9 moves deep in less than a second (as soon as I get the window open to look at the best line of play it has found, it's 9 deep). That's 667,438 different board positions brought about by every possible 9-move sequence that's legal to play. And not only has it figured out each position, but it's run in-depth calculations on each one to determine whether the board favors continued play by white or by black, and to exactly what extent, AND it has compared all eventualities to find the best one for me. Stop and think about that. Can you even begin to think about 667,438 different chess board positions, and the implications of each position, in your head? Of course not. Try it for a few thousand men on a field of battle firing arrows at each other, though. Easier? It was for me, and I bet for you as well, and that means the computer can easily do it by comparison.