Research on battle map - post test results here

[Kenchi_Sulla]

Moderator edit: in the (perhaps temporary) absence of a Ludus Magna for M2TW, let's use this thread to post the results of all custom battle tests about units.

If you are not posting test results, please keep any commentary or questions to a minimum. This is not a thread to debate or to report on your solo games - it's for research results.

If you do research on battles and want to start your own thread for some reason, that's ok. But let's sticky this thread for those who want to use it.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++

I see so much speculation on these forums, so many small grievances. I think it's time to test the game. Instead of complaining about game mechanics let's try to figure out how they work.

First test: AI passive bug.

I think this one doesn't really exist, or it's a lot smaller then we think. It seems to be some kind of decision the game makes. I'll explain.

Ran a test with one unit of spearman and one unit of knights (A.I. on attack, me on defend)

Faced my spears towards the cavalry and stood still ("braced"). The enemy cavalry approached me from the front and halted about 40 yards out. They wouldn't move at all - until I decided to move my spears forward. The A.I charged and slaughtered my spears. It waited for a chance to strike - when the spears are moving they aren't bracing and when they aren't braced they don't seem to receive any kind of defensive bonus.

Repeated the test, but now turned the spears 180 degrees. Again the A.I. waited for this moment and struck home.

The confusing part of this routine is that the A.I. isn't predictable in this. Sometimes it waited for the right time, sometimes it just struck home. This could be indented, but it might be a small error.

Now, unit depth:

This time I used a unit of late era pikes vs english knights. First run I put the unit of pikes in deep formation (10 men front, 6 deep). The A.I. charged into the pikes (3 deep 13 wide), loosing their front horses, except the ones on the flanks. Those instantly killed my general and flanked my pikes on both sides. Killed approx half the knights, but lost all in the end.

Second test I put the pikes 3 deep, 20 wide. The A.I. charged directly into the pikes, loosing the front rank instantly. After a short melee, they pulled out in the following way:

edit: The picture didn't show correctly. The cavalry pulled out on a 45 degree angle (finding the flank of my pikes on the next charge)

After this they recharged and pulled out again. In the end I beat the cavalry unit but lost 35 pikemen.

I didn't test swordinfantry yet but I believe I would be able to conclude the following:

Deep formation might absorb a lot of the charge. Sacrifices about 10% of your men but force the A.I. in melee. I expect the cavalry to withdraw after a short time and recharge in.

A wide formation will probably get you killed since the A.I. has a larger "hit zone". It will simply hit more soldiers in your troop who will be instantly killed without killing anything. This sudden loss, loosing battle and being outnumbered will rout your soldiers before the momentum of the charge is zeroed.

Moving your troops and receiving a charge will hurt your force even more (Running towards horses that are charging you doesn't feel like a smart thing to do, so I like this feature)

Can someone confirm my tests?

-------------------------------------------

feel free to ad your tests here.

[Tamur]

Great thread Kenchi... I'm still waiting for the local store to get the game (fogged in airplanes are currently my worst enemy) so I can't confirm anything you've tested. However, you can be sure as soon as I get it, I'll be spending most of the first month or so testing and retesting systematically. Would be great to see others take this approach!

[Kenchi_Sulla]

Did the swordsmen test. It was as I expected

The A.I. behaved in a smart way. I put 3 units of armoured swords up vs 2 templar knights. the 3 swords were in a 10 rank deep formation, without spacing. Basicly it was a mass of soldiers of 18 wide, 10 deep.

The A.I. refused to charge this formation of men. I had to shuffle them around to get the A.I. to attack. This is not a bug, it's "smart" behaviour.

While not fully braced, the knights charged my soldiers. I lost most soldiers in the front rank (also because the A.I. didn't have much room to charge it wasn't at it's full potential). I was winning the melee and the A.I. pulled out. It again refused to charge back (my men were bracing). After some shuffling and exposing the A.I. recharged and beat my swords (there were 30 knights left in the last charge). Vs fresh full, deeply ranked, swords the knights couldn't win. vs fresh, 80% strength swords in a messed up formation the A.I. won easily while being heavily outnumbered.

Interesting - and a great improvement.

[TinCow]

The passive AI bug definitely exists. Not only did CA announce that it existed before the game was even released (prompting the 0 day patch development), I have personally witnessed it several times. It is manifested by an attacking AI army marching up to your lines, then simply halting and doing nothing until you engage. This IS a bug, because theoretically you can sit there and do nothing and win by timer.

[Kenchi_Sulla]

It's not engaging that triggers A.I. action. It is the movement of your troops. You could call it a bug and perhaps it wasn't intented to work like this but there is some benefit for the A.I. to behave like this (except when you work with timer).

[Sir Robin]

I thought I read, somewhere, that the Passive AI bug actually occured mostly when you had superior missile units. Basically the AI would just stand there until your archers run out of ammo and then attack instead of attacking immediately or something like that.

I am disappointed that the so-called Zero Day patch won't be out for at least two weeks. While I appreciate that CA is trying to fix as many bugs as possible as quickly as possible, I would prefer they release the patch that fixes large issues like the Passive AI bug now and release additional patches later regarding newly discovered or more difficult problems.

[Kenchi_Sulla]

Try to use this thread as a test thread - not a bug discussion

ps

I didn't have archer units so there goes your theory

[Orda Khan]

I didn't have archer units so there goes your theory

The AI does indeed become passive when facing archer superiority.
In one battle test, I charged and killed AI unprotected crossbow units and after that the AI was clueless. With superior infantry and with many good cav, it decided to do nothing. I expected it to attack my infantry, which it would surely have defeated and it had plenty of cav to rear my engaged units and cover my own cav. Even when I marched up nose to nose with them they did nothing

......Orda

[TheImp]

The only flaw in this behavior is with archers. Cause if u put your archers behind a good close formation of knights, the cavalry won't charge as u don't move.

So then, u have all the leasure to fire arrows at them and surely kill everyone. I have experienced it a few times in Azincourt.

[Brighdaasa]

Passive AI:
not exactly testing but i used this tactic in the campaign like almost every battle:

example of army composition:
ai: 1 or 2 missile units on the 1st line, 2 or 3 spear or militia in the 2nd line, 1 or 2 cav in the 3rd line (typical rebel army)

me: 2 heavy cav and a generals bodyguard (sometimes i bring other troops, even archers, but making sure none of those actually touch an enemy)

situation: i'm the attacking force, or the enemy is sallying, and i let them deploy outside the gates first

tactic:
Rush my cav to the enemy archers, on a ratio 1 to 1, in the standard ai army formation i can charge those archers full frontal. The archers always start running behind their infantry too late so i catch them just before the infantry and mow them down, making sure no single horse touches the infantry (distance can be as close as 1 feet, as long as my cav don't engage them). The ai infantry just stands there and does nothing, watching their archers get chopped up.
Then i pull back a little and send each cav around either side until they're positioned behind the enemy general or his cavalry, still no reaction from the ai (my bodyguard unit still is in front of the enemy army, maybe that's why they don't react to the cav going behind their backs, althoug at least the ai army's 3rd line, the cav, should start to face me).
I charge my cav home, slaughtering the enemy cav from the rear since they didn't face me. Sometimes the infantry now reacts, but not always, after which i pull back, line up a charge and hit the infantry in the back.

This tactic works _every_ time in my campaign battles, and feels kinda cheesy, so i eventually stopped exploiting the passive ai.
The rushing the front line archers works every time , and the rest of the army will idly stand by as long as no units engage anything but the 1st row of archers.

The ai clearly also fails to react when you move horses or horse archers behind his lines, except if it has ha's itself, which it starts using right after you start the battle, so i guess that breaks the passive ai routine for the whole army cause it's already in skirmish mode or something i guess.

[JonB]

In the interest of finding out about the anti-cav ability of my Danish units I decided to run some tests. I have posted the results here because they show off some odd behaviour on the part of the AI.

I ran a bunch of custom battles, in each case I was the attacker with the computer 'defending' settings of Large units/Very hard/Grassy plain. I used 1 unit if English Knights to charge the single infantry unit. I formed the cav into a 3 deep formation and single clicked on the target charging it directly from the front, after the charge was completed (this is not so easy to tell and I left the cav in longer if they were not getting killed) I withdrew them, rinsed and repeated until 1 unit routs. The units had no upgrades.

1st up
Danish Obudshaer. These have long halberds and formed a spear wall 3 men deep.
Results: Obudshaers routed after 3rd charge with 20 men remaining, 11 English knights remaining.

2nd
Danish Sword staff militia. These guys look pretty much like the Obudshaers, they also have heavy armour, long pikes and form a spear wall.
Result: Sword staff militia routed after 3rd charge with 22 men left, 15 English knights survived.


3rd
Danish Chivalric Knights. Unlike the defensive spear wall types the knights counter charged the horses...
Result: Chivalric knights routed inside 2 seconds with only 9 men remaining, 44 English knights survived.

This picture is repeated with the Norse axemen and Dismounted Huscarls as well (and presumably other non-spear infantry)

4th
Venetian Armoured Sergeants. I decided to skip the spear militia and try these instead (the Danish get axemen instead of these guys). These decided to form up into a schiltrom, I played nice and let them finish.
Result: Armoured Sergeants routed after second charge 7 sergeants remaining, 54 English Knights survived.

Not sure how good the schiltrom is supposed to be but surely it should be better than this.

5th
Spanish Terico pikemen. These long pikemen for a narrow spear wall formation 8 deep, very like the phalanxes from RTW.
Results: English Knights routed after 4th charge with 1 man remaining, 47 Terico pikemen survived.

The deep formation of the Terico pikes looked to be key, the impact of the horses charge on the formation was minimal whereas practically all of the first rank of chargers died instantly.

[econ21]

In his thread on "mechanics unite", beefeater raised some issues about researching battle mechanics which I am moving here:

Cavalry charges

From the forums it seems that quite a number of us have this problem. Cavalry are, apparently, well placed to charge. When the order is issued to do so (seemingly regardless of whether one single- or double- clicks on their victims), they either lower lances and then raise them again at the very last moment, robbing the charge of its impact, or they simply charge in with swords drawn. This latter looks quite impressive but again, you lose the shock factor of a heavy cavalry charge.

Since the AI seems quite capable of charging properly, this is either a player-specific bug or, more likely, I am doing something wrong. I'll start testing tonight to see if I can find what causes this but in the meantime, has anyone identified the factors affecting whether or not a cavalry charge is properly made (note I'm not asking about how it affects its target, merely about how to get them to couch lances and have a go). Some ideas I had were: distance you start the charge from, single/double clicking on target, whether or not the cavalry are in clear formation before the charge, the angle of the charge, and the speed they're moving just before impact.

Battle Morale

Back in MTW you could increase the chance of routing an enemy line, if their morale was weak AND if they were engaged in combat, by running cavalry around behind them. VI and BI made clear the factors affecting unit morale on the tooltip (I liked this feature, a lot). What do we know about factors that scare the troops, and what have we discerned from testing?

And Spendius replies:

Cavalry charges:

I tried in skirmish on an open plain, 1 heavy knight vs 1 archers:

- double clicking causes the 'sword charge'
- single clicking makes the knights walk towards target, then trotting, then finally galloping while lowering the lance. Looks awesome. however, as soon as the first 2 / 3 knights make it (sending archers flying), the rest of them stop/raise lance, then goes to normal melee without charge.
This is caused, I think, by an arrow volley that kills/slows down the first rank of the knights, disrupting the formation. I had the same issue in triangle formation.
The same situation occurs after I try to disengage: just double clicking outside the melee make the knights disengage but end up not facing the enemy: turning back to initiate a new charge leaves a bad formation.

I'll have to try again with 1 knights vs 1 peasants.

I want to find a workaround to use knights properly, at least on open field for charging the back of a melee, but this needs to be fixed (they should not stop because one of them gets engaged)

[econ21]

ObsOt has done some tests on which are the best infantry here:

https://forums.totalwar.org/vb/showthread.php?t=72302

And there is a nice test (with screenshots) of gendarmes vs spears by RabidGibbon here:

https://forums.totalwar.org/vb/showp...76&postcount=9

[RomoR]

I have also being making some tests on fighting in loose formation, melee against melee, and have an initial impression that it can be beneficial for the player to fight this way in some cases.

I am not trying to make any crazy battle stratergies and am not a mad pro-loose guy but am rather trying to make the most out of my men in different situations, just thought you should know.

1-1 encounters always, always seem to benefit the loose formation. You don't seems to get much of a moral penalty. for fighting this way and get the extra flanking bonus and when fighting general units get to kill the general very early on, even if you don't want to change your line to loose, spread it out more so you wrap around the enemy unit, Moral penalty seems very low on this.
I have already mentioned the cavalry problems against loose, this I am always doing now, it is always better to change to loose against cavalry.

In bigger unit encounters I still cant say if its better or not.
I have tried it out in SP custom battles, historical battles (4 times) and in MP (3 times) and don't seems to get more or less of an advantage or disadvantage unless fighting against heavy cavalry armies or heavy archer armies.
In any case I think a player should be getting more of an moral penalty than what we are getting now. It doesnt seem natural for me to win the agincourt historical battle twice after changing my whole army to loose (minus the archers, only when needed to fight hand to hand) with very similar end results has in winning the "normal way" with the same tactics.

[Rurik the Chieftain]

RomoR, those results sound...discouraging. If loose formation is better against charges, archers, and infantry, then why go tight formation at all? I sure hope this isn't a fundamental game mess up and I'm not seeing something here.

[Bullethead]

1-1 encounters always, always seem to benefit the loose formation. You don't seems to get much of a moral penalty. for fighting this way and get the extra flanking bonus and when fighting general units get to kill the general very early on, even if you don't want to change your line to loose, spread it out more so you wrap around the enemy unit, Moral penalty seems very low on this.
I have already mentioned the cavalry problems against loose, this I am always doing now, it is always better to change to loose against cavalry.

Geez, this sounds like a complete and total game-buster, an absolute abandonment of any semblance of realistic combat at the tactical level .

I won't have M2TW myself until near the end of the month so I can't verify these results. However, I wish somebody would take a good look. I'm not doubting RomoR, but it would still be nice to see further investigation of this problem. If this turns out to be as bad as RomoR indicates, then this is a SERIOUS problem and should be moved up to #1 on the fix-it list.

[BigTex]

After doing alot of testing to try and figure out if horse armor was giving later units the upper hand or if the 2 handed lance was doing something.

Going on 10 rounds each, all on grassy plains. I used demi lancers as the target, having neither armored horse and using 2 handed lances/no shield. I was on huge unit scaling. During this I found some interesting bugs. First of all the AI has been given way to much credit in its ability to charge its cavalry. Flat out the ai pretty much sucks at charging anything but infantry. They also have a bug of reforming their lines right before the charge. They usually bunch up in a ball, almost as if their turning their formation on its tail.

Out of 10 rounds of me commanding the Templar's vs the Demi Lancers, the Demi lancer won 0 of them. There were cases where the Demi Lancers reduced the Templar's numbers to 20 or a little below, but on average the Templar's were only reduced to around 40.

Playing the demi lancers myself vs templar's, the Demi Lancers won 3 rounds. Out of those 3 they were always left with under 20. In all 3 cases the Demi Lancers had the more momentous charge.

Out of 10 rounds of me commanding Lancers vs Demi Lancers. The Demi Lancers won 0 rounds. The Lancers walked away with exception of 1 time with over 60 men still remaining.

Playing the Demi Lancers vs Lancers. The Demi Lancers won 0 rounds. Though the lancers were on average at 40 men. There were a few rounds were they did get reduced to 20 though, but again the Demi Lancers had the more momentous charge.

Now I also did 10/10 rounds of Gothic Knights vs Templar Knights. Each and every time the Gothic Knights won, regardless of who commanded. Each and every time they walked away with less then 50% casualties.

I think I can say that the 2 handed lances arent playing a factor in the incredible results of the Later fully armored cavalry. Without having access to the export_unit file, and baring that there could be new hidden stats. I'd have to say that armor on horses seems to be very important in M2TW.

[grinningman]

I did some testing on the best formation for infantry to receive a cavalry charge.


I used a single unit of Berber spearmen (75) to receive the charge, the AI had control of a unit of mailed knights (40). I tried 4 formations for the spearmen

1. Tight 4 rank formation so the width of the spearmen front rank was similar to the width of the cavalry's front rank.
   
2. Loose 4 rank formation.
   
3. Schiltron.
   
4. Tight 4 column (file?) formation (19 ranks), so the front rank was much narrower than the cavalry's front rank.

My impressions from looking at what happens during the cavalry charge: The majority of casualties inflicted by the cavalry happen during the charge. During the charge, it looks like every target unit inside a small radius around each cavalry gets attacked with the charge attack value, i.e. they have a kind of 'killing zone' that knocks out any infantry within a certain range of a knight, so one knight can kill many infantry pretty much instantaneously during a charge.

Results: In all cases, the knights killed most of the spearmen and routed the rest. I did 5 test runs for each case, and the spearmen were never in guard mode so as many as possible would engage in combat:

1. 50-60 spearmen killed in the initial charge. The remaining spearmen killed ~14 knights before being routed.
   
2. 30-40 spearmen killed in the initial charge. The remaining spearmen killed ~25 knights before being routed. In one case, the knights pulled out and re-charged - in this case the spearmen killed 29 knights.
   
3. 40-50 spearmen killed in initial charge. The remaining spearmen killed ~28 knights before being routed. There was quite a large spread in the number of knights killed each test - I did 8 tests for this case, the results were 6,20,23,29,31,32,34,34. This was due to a large spread in the number of spearmen killed during the charge.
   
4. 20-30 spearmen killed in the initial charge. The remaining spearmen killed ~30 knights before being routed, and in 4 out of 5 cases, forced the knights to withdraw and re-charge.

These results make sense when you think about my impressions of the charge effect above - the more closely packed the infantry are, the more casualties they'll take during the charge. The case where the spearmen's front rank is the same width as the attacking cavalry's is the worst case, and the cavalry will inflict the most possible kills during this charge.

It looks like the best way to receive a cavalry charge is with as few ranks as possible (or a schiltron), so the impact of the charge can be taken with the smallest number of spearmen possible inside the 'killing zone' of the cavalry.

[Bullethead]

I did some testing on the best formation for infantry to receive a cavalry charge.

Thanks for doing this. Your results are definitely food for thought.

[quote[I used a single unit of Berber spearmen (75) to receive the charge, the AI had control of a unit of mailed knights (40). I tried 4 formations for the spearmen[/quote]

Did you do anything with the spearmen after the knights made contact? Specifically, did you tighten the loose formation back up to try to engulf the knights in melee from all sides? I'd really hate it if this ends up proving to be the best cav defense tactic, but I'm worried. Loose infantry formations should not stop a charging cav unit in good formation--the cav should go right on through. But if the loose formation does stop the cav, preserves the most infantrymen from destruction, and allows the dispersed troops out beyond the cav's flanks to wrap in after the impact, then there's a serious realism problem.

These results make sense when you think about my impressions of the charge effect above - the more closely packed the infantry are, the more casualties they'll take during the charge. The case where the spearmen's front rank is the same width as the attacking cavalry's is the worst case, and the cavalry will inflict the most possible kills during this charge.

IMHO, this doesn't make sense. Infantry in good formation historically was able to withstand cav charges fairly easily and with little damage. As long as the infantry's discipline held, and their ranks stayed tight while presenting the horses with a mass of long pointy things, the cav would bounce off or break up and go around them, from Hastings to Waterloo. So you'd think a unit of spearmen in good formation, with a frontage at least equal to the cav's, would be in good shape. But your results show this is the worst possible option. Hmmmm.....

OK, you can make the case that the cav in this situation has a greater reach due to its lances being longer than the spears. Also, cheap spearmen's light shields ain't much good against anything, anyway, especially not a lance with cav momentum behind it. So maybe cheap spearmen are just doomed from the start. But what about other situations, such as better infantry defense and/or longer reach?

Is there any unit in M2TW that can form a shield wall formation like in BI? If so, does a shield wall help the infantry withstand a charge better, even if the infantrymen have swords/axes instead of spears?

And what about pikes? Do pikes (even unarmored) do a better job against cav than spears?

And finally, does cav treat melee infantry differently than missle infantry? I've heard that missile infantry won't stop charging cav--the cav just runs through them. Is that true?

Thanks in advance.

[RomoR]

Got the 1.1 patch and so decided to re-test the cavalry charges on melee units (mostly spears) here is an example of one of those series of tests:

conditions:
Unit Scale-Large
Map- grassy flats
Whether- clear
Me with 1 unit of militia spearmen
Ai with 1 unit of Hobilares
After choosing desired formation I press start and dont give any commands to my unit.

Results.

Spears in default tight formation guard mode off. lost all five rounds
Spears in default tight formation guard mode on. lost all five rounds
Spears in Schiltrom formation guard mode off. lost 4 round won 1
Spears in 7 rank tight formation guard mode off. lost 4 rounds won 1

Spears in 2 rank loose formation guard mode off. won all 5 rounds.

So it still appears that the best formation to be in for receiving a cavalry charge with melee or spear units (not pikes) is in a 2 rank loose formation, I'm not even going to talk about the fact that a more expensive spear unit looses against Hobilars! ho I did...

On I side note my Dismounted Portuguese Knights are still Pants/Broken

Edit: Unit Scale info