The Shield Problem(s)

[Zenicetus]

Of course, Agree 100%. We each have our own veiw on how things should be, and are free to implement them like that in our mods.

I was mearly trying to point out that the sheild fix only really disfavours HA as a unit class in general. Whilst not fixing it disadvantages 3 classes. Spear units, Sword & Sheild Units, and Cav with a sheild. It might give them slightly better rear charge resistance and Enfidle fire resistance than we'd like. But overall it fixes a much bigger set of problems much more cleanly and without creating as many more and is thus a better fix in general IMO.

Speaking as someone sitting on the sidelines (this is a very interesting thread)... comments like that make me very nervous about applying a "shield fix" to a full campaign game. At least, not without deep testing, in that context.

Someone mentioned up-thread that auto-calc is important because the AI uses it all the time, which is a crucial point. Every time you hit the "turn" button, the Ai is auto-calc'ing based on stats. I don't want to suddenly see the Mongols nerfed in the campaign game (just taking one possible example), because suddenly HA's are less effective.

Okay, back to lurking on this thread now.

[Foz]

I apologize ahead of time and mean no disrespect Blademun, but I have to rip your assumptions about the 3-quadrant system to shreds.

One of the very early tests I ran included 3 key elements. Those were the following:

-Testing a unit with 22 armour/0 skill/0 shield
-Testing same unit with 0 armour/22 skill/0 shield
-Testing same unit with 0 armour/0 skill/22 shield

It was one of the pivotal tests for me, because it showed that shield points did NOT do the same thing as other defense points. On the flip side of that coin, however, is a point pivotal to your discussion:

-The all armour point unit produced combat results nearly identical on average to the all skill point unit.

Since I can't seem to find it at the moment and it's possible I didn't actually post those numbers as they were part of other testing, I'll put the results that back it up right here:

Armored Swordsmen 22/0/0 vs. Dismounted Noble Knights 15 attack no ap

AS Kills/DNK Kills
50/30
40/29
55/29
40/25
45/43
-----
230/156

Armored Swordsmen 0/22/0 vs Dismounted Noble Knights 15 attack no ap

AS Kills/DNK Kills
39/17
60/48
60/59
38/23
34/19
-----
231/166

I'm also including a set I ran just now with lock_morale.  I set shield 1 to be sure skill would only affect the right side in case shield 0 makes it have 180 degree arc.

Dismounted Feudal Knights 13/8/1 vs. Dismounted Norman Knights 21/0/1

DFK Kills/DNK Kills
61/52
61/49
61/56
47/61
45/61
-----
275/279

The results immediately indicate your front-left and front-right thinking must be flawed. The first 2 sets show that an all armor unit performed the same in CC as an all skill unit. If skill only applied on the front-right, we would presumably see the all skill unit take a performance hit, as half the attacks against it should come from front-left and thus be against a 0 defense value.

The third set that I just ran tonight is much more carefully constructed. lock_morale removes the morale factor, making units fight to the last man. The DFK have 8 skill points where the DNK instead get an extra 8 armor points. I gave each unit shield 1 to force the skill to only affect right if it in fact becomes 180 degrees in the absence of a shield (this may be set already for each unit instead of relying on shield value to determine it). The result shows an even matchup, where we would expect DNK to be dominant if a front-left and front-right field were used, as half the attacks against the DFK would come from the front-left and thus have a 12 defense instead of their front-right 21. The DFK suffer no such performance hit because of having skill instead of armor.

What all this means is that there is a front rating which sums armor, skill, and shield, and into which the majority of attacks in frontal melee combat go. Skill therefore does not apply to front-right, but rather to front AND right, and likewise shield would go front and left. The way I picture it is with quadrants (that means 4 btw), one for each of the cardinal directions, with their borders at 45 degrees, 135 degrees, 225 degrees, and 315 degrees (i.e. diagonals, like NE would be on a compass). That representation may not be completely correct, but it is for certain at least that some field directly in front of the man gets all 3 values (armour, skill, shield) counted for it, and that field is wide enough that it accounts for the vast majority of melee attacks in frontal melee combat.

After understanding that, go back and look at the shield-to-armour fix, and you'll note that in all cases it makes the all-important front number the amount the devs would've assumed it to be, as well as the left number, while making the right and rear numbers too high by an amount equal to unit's shield amount. It's a VERY important distinction to make, as the screwy numbers don't affect typical front combat like you suggested.

Speaking as someone sitting on the sidelines (this is a very interesting thread)... comments like that make me very nervous about applying a "shield fix" to a full campaign game. At least, not without deep testing, in that context.

Someone mentioned up-thread that auto-calc is important because the AI uses it all the time, which is a crucial point. Every time you hit the "turn" button, the Ai is auto-calc'ing based on stats. I don't want to suddenly see the Mongols nerfed in the campaign game (just taking one possible example), because suddenly HA's are less effective.

Okay, back to lurking on this thread now.

That's exactly why I've been saying the armour fix is the way to go. All the discussion about HA's being less effective is only in the context of the battlefield. The shield-to-armor fix maintains the defense total as calculated by the game, and therefore doesn't affect the massive amount of auto-calcing the computer does! The mongol HAs will only be affected when actually in battle with you. And even then, the effect has been exaggerated: shield units will have at most 6 more missile defense to the right and rear than they used to (newly covered areas by the armor points), and the exact same amount they did on the left and front (where shield was already being applied). That at most 6 point difference is effectively halved if the archers in question have AP. The biggest thing is that HAs may be underperformers on the battlefield in melee now, since they might have been balanced stat-wise against the shield-bugged units, which would cause up to a 12-point defense swing in a lot of cases...

[lobo76]

Foz, using your fix, doesn't it make units with shields more vulnerable to missile from the front where the 'shield' defence is now zero.

[Zenicetus]

That's exactly why I've been saying the armour fix is the way to go. All the discussion about HA's being less effective is only in the context of the battlefield. The shield-to-armor fix maintains the defense total as calculated by the game, and therefore doesn't affect the massive amount of auto-calcing the computer does! The mongol HAs will only be affected when actually in battle with you. And even then, the effect has been exaggerated: shield units will have at most 6 more missile defense to the right and rear than they used to (newly covered areas by the armor points), and the exact same amount they did on the left and front (where shield was already being applied). That at most 6 point difference is effectively halved if the archers in question have AP. The biggest thing is that HAs may be underperformers on the battlefield in melee now, since they might have been balanced stat-wise against the shield-bugged units, which would cause up to a 12-point defense swing in a lot of cases...

Okay.... but what if I'm playing a faction like Turks, which is actually my current campaign (hence my interest in the thread)? I almost never auto-resolve battles, and while it's not their main strength, my HA's do sometimes enter melee when the arrows run out.

I suppose an argument could be made that vanilla HA's are overpowered, but it took me a long time to learn how to use those "overpowered" units.

[Tuidjy]

It seems that many of you are trying to change units to make them 'balanced'
in multiplayer. Speaking for myself, I abhor the idea of rock-paper-scisors
balancing applied to historical units. Heavy cavalry should devastate anything
but pikesmen. Horse archers should anihilate infantry. Shields should protect
only in an arc, and should help in melee.

I do not think that my wishes can be reconciled with those who wish to
balance multiplayer. I can only hope that someone (CA or a modder) will
strive for a "total realism" mod, as opposed to a "total balance" one.

[R'as al Ghul]

Hello,

I thought I drop in to compliment you on your findings. I've followed a different approach trying to find out how the stats actually influence combat. I've worked with zero stats etc. but didn't expect one value like the shield to be so completely bugged as it is.
I've just done some tests where I set the shield of ASergeants to zero and otherwise kept their stats as they are. I pitted 5 of them against 5 Mailed Knights (also with zeroed out shield).
Now those units actually work like the advisor tells you and like it's supposed to be. If the AS receive the charge in default formation they may win but take horrendous casualties. Before the fix they would've just been run down by the knights. Now they stand a chance and are actual anti-cav units.
What's even better, the Shiltrom works. It really does. When placed in shiltrom your AS won't take much losses from repeated charges and will eventually rout the knights.
If you still feel you want Spears to be stronger, change the shiltrom to shield_wall (unused feature) and the knights don't stand a chance.

Width of the shield_wall (unit in the back) compared to default:


Spear wall ready to receive charge:


Charge at shield_wall


Charge at Shiltrom


I was surprised that this simple fix makes it unnecessary to toy around with other values like collision mass or skeleton comp. factor.


R'as

[AussieGiant]

So this has become a hot topic.

What I find amazing is that no one from CA takes the time to jump on a thread like this and say something...anything.

Even something like;

"Guy's, great work, thanks for that and where going to get on it."

When is the next "update" scheduled?

[Point_Blank]

Is this 'Fix' working with armor upgrades? Having tried a number of battles with upgraded and non-upgraded units vs the same opponents, the upgraded units seem to do worse.

Maybe when a unit is upgraded it will not retain the higher armor value generated by the Fix?

Even unupgraded Armored Swordsmen now seem to be able to take om JHI quite well wiith the Fix...

Since the original problem is that the game subtracting the shield value in melee rather than adding it, setting the shield values to negative would in fact correct the melee values since subtracting a negative yields a positive. It would make shielded units more vulnerable to missile fire though.

[R'as al Ghul]

Since the original problem is that the game subtracting the shield value in melee rather than adding it, setting the shield values to negative would in fact correct the melee values since subtracting a negative yields a positive. It would make shielded units more vulnerable to missile fire though.

That doesn't work. Negatives are displayed as zeroes in the stats.

[Point_Blank]

That doesn't work. Negatives are displayed as zeroes in the stats.

Yeah I just tried it, would have been confusing anyway.

But I would still like to know whether armor upgrades are working properly in conjunction with the Fix.

[JeromeGrasdyke]

What I find amazing is that no one from CA takes the time to jump on a thread like this and say something...anything.

Even something like;

"Guy's, great work, thanks for that and where going to get on it."

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

[Carl]

The shield into armor fix nerfs all missile troops Carl... Even with foot archers you generally try to flank because enfilade fire is the best kill rate of any attack type in the game... Making it so that archers can't do any better when flanking/rear firing makes them basically worthless (Which explains why you think the Scots are so good) whereas they're the best killers in the game without that kind of fix.

Well just to get it out of the way at the start, my scots comments where based on a game where the only modification was a 2-hander fix and MY version of the shield fix which DID NOT get shield units to brace, and which DID NOT give them any extra armour. They're better with foz's fix I admit.

Second, I’ve said it before, and I’ll say it again. If your being competent and the enemy army isn’t bigger than your you WILL NOT get enfidle fire with foot archers, they can't get into position fast enough, and need guarding from cav. The result? You've got 2 small armies to beat as apposed to one, and while he's repositioning his archers you can advance on him with no missile fire losses. I can’t see any advantage to even trying against someone competent./ remember the AI is terrible at unit control, you have to think of what a good human player would do, not a stupid AI that can’t recognise when to take advantage.

Let me also point out, AGAIN, that you can easily create all kinds of unit layouts where you force the enemy to shoot through a unit whose shields are facing the HA if thy want to hit anyone in the back. Or fire in an arc instead. Either will cut the effects of enfidle fire ANYWAY, so your not getting as big an advantage as might first appear. It just lets you use sloppy formations as apposed to carefully thought out ones.

@ this post: Yep, I noticed the same thing, it's really shocking. From the sounds of it I’m not the only one finding Shiltrom 10X more effective than it used to be.

A little bitter, eh? Last time I checked, historically, a good HA army (mongol/Byzantine/Turk) was suppose to do just that to a all infantry army.

Nope, the only time I’ve ever fought against HA on the battlefield was those tests furthar up the thread. However, I have heard a LOT of reports of pure HA armies defeating mixed and pure Inf armies. That simply isn't BALANCED. It doesn’t matter if it's historical or not, this is a game and it must be fun to play, that requires BALANCE.

It is a shame that the shield does not contribute to melee fighting like it should. However, I still feel the idea that sheild values contribute negativly to defence in melee is still debateable. However, if it IS true, I would prefer to wait for a official patch.

Why? Well lets state the problem: We want our shields to have a effect in melee.

Well, of the possible solutions, the best two ways are to:

A: add sheild to the armour and 0 the shield value. This has the effect of nerfing the depth that CA took in simulating archery based combat. It also does not produce the above desired effect.

Why? Well lets talk about what happens when you add to armour.

There are three 'quadrants' of defense to each solder. The Front-Right, Front-Left, and Rear.

the Armour value adds to all three quadrants. The Skill value adds to the front Right, and the Shield value adds to the Front-left. So, Lets look at a vanilla DFK from a Quadrant perspective assuming the shield bug is true.

Front-Right: 7 Armour + 8 skill= 15 defense
Front-Left: 7 armour + -6 shield= 1 defense
Rear: Defense = 7.

Now add the Armour fix.

Front-Right: 14 armour + 8 Skill= 22 Defense
Front-Left: 14 armour + 0 shield.= 14
Rear: Defense = 14

This game is balanced under the assumption(not knowing of the shield bug) that DFKs are actually: 15, 13, 7 respectively, not 22, 14, 14. Thus, we get a unit that is overpowered in terms of the way this game was meant to be played. What -you- think is how it suppose to be doesn't matter as much. If you think Dfk's should be Zomg death knights, then great, mod your game...don't declare it a universal fix for everyone.


The other Fix:

B: Keep the shield value and add double the shield value to the skill value to offset the inverse effect sheild has on melee defense. This also does not produce the desired effect:

With this fix the DFK's Quadrants are

Front-Right: 7 Armour + 20 SKill(6*2+8)= 27 Defense(Zomg)
Front-Left: -6 Shield + 7 armour= 1 Defense
Rear: Defense = 7

So we get a Dfk who's Quadrants are 27,1,7. Crazy, but at least arrow fire will continue to work like its suppose too.

TO me, both of these 'fixes' are not really fixes at all. Assuming one used the Shield-to-armour fix..I think there is only way it could be done barring a official patch. You would have to go through every unit in the game and readjust their stats to balance accordingly with the new and improved shield units. With that goes hours upon hours of balance testing just like the designers went through when they made this game.

First your example are wrong, it should be 7+6 to the left. That's 13, 1 less than the value with the armour fix. defence skill also seems to be a 180 affect ATM, so it's only rear armour and right missile armour that’s borked.

Second, the shield isn't even working properly against MISSILE FIRE

RE READ THE ABOVE.

Got it, good.


Overall the fix actually makes the units have about the right defence values all the time from everything except the rear, and right if it's missile fire.

Okay.... but what if I'm playing a faction like Turks, which is actually my current campaign (hence my interest in the thread)? I almost never auto-resolve battles, and while it's not their main strength, my HA's do sometimes enter melee when the arrows run out.

Then you get smashed. As you should. Use a combined arms army so you have some infantry, or heavy melee cav to finish the job and you'll do fine. Byzantine has all those spear and shield units for a reason you know. As do the turks.

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

Thanks for the heads up.

[R'as al Ghul]

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things.

Thanks very much for posting but it's not true that there's not much to say.
What you just posted, an acknowledgement, means a lot to the community.
It's fine that we can have faith that you read the forums and take care of problems but it's even better to hear it. Seeing is believing. We don't want details, sometimes a simple "aha" or "mmmmh" or "ok" would be totally sufficient. Give us lifesigns now and then and we'll be as tame as kittens.
Of course, I realise that in the past some remarks have been taken as promises.
I see CA's problem there but there's always a compromise.

R'as

[AussieGiant]

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

LOL,

Mission accomplished, thanks Jerome for confirming that you guy's at the very least are aware of this issue.

I just needed a little positive reinforcement...I'm not too demanding...at least I don't think so.

[Slug For A Butt]

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

Sounds a little trite, but hey... I'm one of the TW series biggest fans so I hope it's true.

[Ulstan]

Lastly, I really don't like this fix mainly because it strengthens units that -should- be weak, namely spear milita and town milita. Spear milita should die like flys from Enfilade/rear fire, because Spear Milita are just peasants with spears.

Spear militia are currently too weak. They get dashed aside by cavalry as though they aren't there. They definitely need boosting, though more in their ability to stand against cav charges, not necessarily in their ability to absorb arrows from the rear.

A schiltron though, would be a fairly effective anti-archer formation in that suddenly, there isn't an exposed flank/rear to shoot at. Not quite as good as a testudo, but then, a schiltron isn't completely inept at melee combat either.

Also, don't just dismiss them as 'peasants with spears'. We have a separate unit of peasants to be the peasants. ;) That would be like calling longbowmen 'peasants with bows', as the french knights were happy to do until they got massacred :D

[dopp]

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

Ah, excellent. Hopefully it will be fixed soon.

[Carl]

however, because they lack the training and displine to properly brace themselfs for attacks by heavier forms of cavalry. Notice they don't plant their spear in the ground, and they only hold them WITH ONE HAND. That is why I feel they should largely get busted by heavy cav.

Two points here.

One, ALL spear units, (including Papal Guard, who are the best), use spears in EXACTLY the same way as Spear Militia. They ALL carry their spears in one hand, and NONE of them stick them in the ground.

Second, Spear Militia, ARE NOT peasants given spears and shields and shown how to use them, that’s Town Militia (who only get Light_Spears and Spear_Bonous 4), Spear Militia represent professional militia unit. Individuals given equipment just below the level of professional soldiers and the best training that can be given to Militia level units. Italian Spear Militia, (and similar top level Militia units), are better more because of better equipment, they actually have professional equipment instead of high quality Militia equipment.

NOTE: the above may or may not be historically accurate, but it IS how they are portrayed in game, and that’s what I really worry about as it shows designer intent IMHO.

My point on schiltroms is that it doesn't work that way in the game. The arrow trajectories in the game mean that half or more of your arrows are going over the guys facing you and striking the guys facing the other direction in the rear or sides...

This isn't actually as common as you'd think and comes with disadvantages. On small size at least, most arrows that would not hit someone with a shield will simply skim right over the unit. Larger unit sizes will probably result in the arrow hitting the shields as their more rank for it to fall into. Some will of course hit them in the back AND if you do really big angles on the arcing fire then YES you will get them hitting guys in the back, but your sacrificing a lot of hitting power to get that kind of angle, (it also won't work with Crossbow units at all, gunpowder being powerful enough to not care about the shield).

Don't get me wrong, the current shield fix does bork things, but in my experience, short range arrow volleys are too flat a trajectory to hit the unit without hitting a shield more often than the hit someone in the back. less flat trajectories introduce problems with reduced impact power. Their IS a perfect position I’m sure, but it's probably not easy to judge and get right.

Don't get me wrong, Schilstrom IS a LOT more resistant to missile fire ATM, but I don't think the difference will be quite as large as you think it will be either. In the end I think we should just wait and see.

To those worried about late era units getting wailed upon by early era units. I would point out that most good late era infantry is AP 2-handers and pikes. Even with the Shield immunity to AP back in, the shield units will be at a heavy disadvantage to them, so they won't be getting any weaker really. They also typically have much of the shield value dumped into armour anyway. This will make them a lot more resistant to enfidle fire, and hence HA anyway. (At least in the case of mounted units,. Late era Inf tends to be more vulnerable).

[Shahed]

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

Good to hear. Thanks.

[Jambo]

By the way, how do we know "skill" only works to the front and right? How do we know it's not cumulative with the shield?

[hrvojej]

If you still feel you want Spears to be stronger, change the shiltrom to spear_wall (unused feature) and the knights don't stand a chance.

R'as, did you mean shield_wall, or spear wall i.e. phalanx? I would love to be able to use shield_wall formation in game, but I think that the AI will not use it with his spearmen, which then imbalances things for the player. Have you tested it with the AI controlling the spearmen, is it able to employ shield_wall properly (or at all)? If not, do you, or anybody else for that matter, know how to get the AI to use a new formation?

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

Many thanks for posting this!

[Ulstan]

Hmmm looks like everyone who thought it was a bug was right, while Zhukov was wrong.

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Oh, we understand the difficulty in making a completely bug free product. I'm not interested in 'gotchas' or assigning blame to the developers, nor is anyone else, I think.

We just want assurance that what we are seeing actually is a bug, and that CA is aware of it, and working to correct it.

Icing on the cake would be an ETA on the fix :)

[Kraggenmor]

Hmmm looks like everyone who thought it was a bug was right, while Zhukov was wrong.

The opening shot at member who posted a completely valid point was totally unnecessary.

Especially considering that this

We just want assurance that what we are seeing actually is a bug

Was essentially Zhukov's point to start with: That just because you're seeing something you don't think is working as it should, doesn't necessarily mean its a bug.

[Carl]

@Kraggenmor:

I agree that was a cheap shot, but also, if you read my last reply to Zurkovs last post in this thread you'd see that i'd pointed out how it made fixed 2-handers work with origonal stats, (and thus in auto-calc), and also made late era sheild units competetive with all other late era units. This proved it was bug to me as it brought everything back into balance. The effects of getting working bracing, have only added to this extra balance where seeing.

[General Zhukov]

[which] was essentially Zhukov's point to start with: That just because you're seeing something you don't think is working as it should, doesn't necessarily mean its a bug.

Of course that is true Kragg, and nice of you to point it out. But, really, one can understand the compulsion that drives our hosts in this thread to improve the world, based on the evidence as they see it. Is not progress always driven by the unreasonable man? Nay, better to observe quietly as Carl and Foz4 work their mad schemes in Mod Manor, with Ulstan, ready, with ears perked, guarding the gate. When they have figured out exactly how shields should work, and exactly what balance the game should have, we will be able to eat the sweet fruits of their toil. And if they err in eradicating error? Then at least they have brought the issue to the attention of the TW gods, and for that we should be grateful at least.

[R'as al Ghul]

R'as, did you mean shield_wall, or spear wall i.e. phalanx? I would love to be able to use shield_wall formation in game, but I think that the AI will not use it with his spearmen, which then imbalances things for the player. Have you tested it with the AI controlling the spearmen, is it able to employ shield_wall properly (or at all)? If not, do you, or anybody else for that matter, know how to get the AI to use a new formation?

hrvojej, I meant shield_wall, sorry for the confusion. I'll edit my post.
Personally I just think that shield_wall looks better and is a bit more plausible. 75 men forming a shiltrom does look a bit funny in the battle.
I've edited the formation line where it says one or two of :"square, horde, schiltrom, shield_wall, phalanx, testudo, or wedge" can be used. So, now it's there I suppose that the AI will use it but to be honest I haven't seen the AI do a shiltrom formation. It may be difficult to test because it is rarely used.

[General Zhukov]

Considering the typically ambiguous wording of the guy from CA's statement, it's not clear whether CA considers this a major gameplay bug, a minor nuisance, or an embarrassing revelation of shortcuts they made to achieve the desired unit balance.

From a conversation a forum member had with someone from CA:

I also mentioned the shield bug to him(looks like CA changed the way shields worked and this was an unfortunate side effect)

However, was the unfortunate side effect known about before the game was released? The answer to that question may lead to which of the following idealized conversations are going on, or have gone on at developer HQ:

1. "Drat. They found our shield factor. Alright, redo the engine for the next patch so that the unit balance is roughly maintained, but so that the factor doesn't show up in homebrew testing."

2. "God's Nipples! I want that prog's head on a plate! This means that all shielded units have been underpowered in melee versus unshielded units and against units with smaller shields. Fix it for the next patch."

[Jambo]

I also like to know if the proposed shield fix, whereby the bonus is transferred from shield to armour, works with armour upgrades??

Some have stated that armour upgrades are kind of hardcoded and linked to how they appear on the battlefield. For instance, padded gives 4 armour regardless of the actual numbers displayed in the unit scroll. If this is the case then surely giving a massive armour boost is going to be completely negated as soon as the unit's armour level is upgraded.

[Foz]

Is this 'Fix' working with armor upgrades? Having tried a number of battles with upgraded and non-upgraded units vs the same opponents, the upgraded units seem to do worse.

Maybe when a unit is upgraded it will not retain the higher armor value generated by the Fix?

TBH, I'm not actually sure. I wasn't thinking about armor upgrades when I did it. If the upgrades add armour points to the unit, then for sure it should work. If they replace the unit's base armour amount with that of the given piece of armour (which I rather expect), then for sure they will mess up the fix. Anyone know for sure what happens? And what we might be able to do (if anything) to work around it if it's the latter case?

Great work guys, really good bug-spotting :) The reason that we don't usually comment is that - to be brutally frank - there just isn't much to say about these things. We try hard to prioritise core gameplay bugs, but it's a huge piece of software with many people working on it, and obviously a few things have slipped the net that shouldn't have.

Rest assured that we do read these forums, both here in the UK and in Australia, and that these things do get discussed internally and passed on for investigation to the folks in Oz who are dealing with the patches.

Understood, and appreciated. It even took me a while to spot this one, and I was running fairly large numbers of custom battles and playing with unit variables at the time. It's so much harder when you have to worry about the whole project at once, especially when the problem is something that only causes unintended behavior instead of compiler errors or game crashes.

If the fix is as easy as I believe it to be in the code, I think there's a very good chance we can expect them to get to it for the second patch. Data inversions, which is what I speculate to be the cause, are among the easiest fixes, often just requiring changing an incorrect minus to a plus, or vice versa. Yeah, amazing that it could be so simple, yet leave us no perfect way to mod a fix. Ahh, the power of real actual code...