problems with infantry vs cavalry

[Bertram]

I have earlier reported that there are some problems with the battle resolution regarding the infantry vs cavalry casualties. In several battles my cavalry did way to much damage, without gettiing hit themselves. The last turn I played against Squarian is a good example, which I would post here, to bring the problem to attention again.

We are playing under 1.02 beta 4.

The place is Erfurt, a wooded area. The Austrians are defending it with an infantry brigade. It consists of 2399 men (and 2 horses) and is commanded by comte de Maupeau (3-0-0). It consists of 16 elements.
They are attacked by a cavalry brigade of 2364 men (in 15 elements), commanded by Georg von Driesen (5-5-4). Von Driesen is already pretty famous on our side, in several battles his charges did tremendous damage to the enemy. He has 3 stars. His cavalry units are experienced, they have on average 4 stars (two have no stars, one has seven stars, three have 6 stars, the others are somewhere in between). There are no "hero units" though (units with absurd strengths due to a bug).
Battle consists of one round. Ranged hits give 9 hits on my side, 0 on the enemy units.
Assualt gives another 9 hits on my side and 134 on the enemy side. The enemy loses 13 infantry elements..... I can not see the remaining 3 enemy infantry elements after the battle, I suppose they are destroyed in pursuit.....
In numbers I lost 313 men, the enemy 2334.
Looking at the detailed combat report we can see that I started with 140 hearts/ 2051 cohesion, and ended with 124 / 1809. I inflicted 56/316 and suffered 14/154.
The enemy started with 142/1272 and ended with 45 /501. (inflicted and suffered is of course the reverse from mine).

The result is that I my cavalry charged an infantry force of equal strength, smashed their line, routed them and destroyed them, with minimal losses to themself. And this in wooded terrain and near a town. Though this could happen in this timeframe, it was very rare, and only happened when (or after) the infantry panicked. In the game it seems common.

I cannot look at the battle resolution routines (as far as I know) but it looks to me that
- the possibility of the charge going home is to large. If I remember correct Pocus stated somewhere that the engine checked if a charge was succesfull. It looks like it is always succesfull. If it isn't it is doing damage like it is.
- the damage to the cavalry is very low, even for a succesfull charge.
- if some chages are not succesfull, not only do they damage like they were succesfull, the cavalry also avoids damage like the charge was succesfull (an unsuccesfull charge should really hurt the cavalry).

The strange thing is that we complained at the start of the scenario about cavalry that got always wiped out - now it is the reverse. I got to say that this bunch of cavalrists is pretty experienced - maybe there is a value for their experience below which the charges are unsuccesfull, and above which they are always succesfull?

Anyway, this behaviour isnt very much 30-years war, it is more 100-years war

[Baris]

Hi Bertram,

I tested(patch3 beta5) with Gessler with only cavalry army. It was clear terrain but raining. As seen in the screenshot in the assault stage many elements of Austrian army lost. Prussian suffered only 29 hits while Austrians 134. Add to that many elements of infantry is lost. Maybe cavalry assault is a bit overpowerful against infantry.Even Prussians defeated. Also cavalry doesnt get too much losses or hits. In the 2nd. screenshot,detailed combat report there was 4 rounds of fighting. Many cavalry elements commited as it is clear terrain but only 2 units of Austrian army commited in each round of battle. While Prussian cavalry is very succesfull they decided to retreat.

Note: 5 NM lost for 500 men killed, while Austrians have 2400 combat losses.I thought I was succesfull.

[ATTACH]12149[/ATTACH][ATTACH]12150[/ATTACH]

[JacquesDeLalaing]

I've just quickly scanned through the post. One question arises: How have the troops been organised? I.e. have the cavalry units been formed into a brigade or have they been seperate units in the stack? Bertram mentions a cavalry "brigade", whereas in Baris' screenshot (detailed combat report) there are many (9 units) seperate cavalry units against only one infantry bigade (=1 unit).

I can explain to myself Baris' example to a certain extent: As far as I know, any "unit" (seperate unit/brigade) targets only one enemy "unit" (seperate unit/brigade) in ranged combat. So it is possible that the whole infantry brigade targeted only one seperate cavalry unit. 2 hits seem a bit low though...either the unit routed quickly or else it has been low in strength, or else the infantry showed a lack-luster performance (Piccolomini and Gaisruck are no heros...). Then, in close combat, all cavalry units hit the infantry brigade and virtually slaughtered it. We all know that cavalry inflicts much more hits in close combat than any infantry e.g. Austrian Dragoons inflict 10 hits and 10 cohesion loss everytime they successfully strike at the enemy! Also take into consideration that cavalry in the open can perform special cavalry charges, inflicting even more hits. Moreover, infantry looses some "attempts to shoot at the enemy" due to the rainy weather (range modifier -1), but then again, cavalry's protection is reduced by 1 in this weather too.

Usually such a tactic (swarming the enemy with seperate cavalry units) should be prevented by the high command deficit that such a stack usually suffers from. Though I don't know how a command-penalty affects the success-chance of close-combat-charges...In this case, Gessler with his cavalry-ability might have compensated the bad effects of the command-penalty?

But this is only my explanation... I'm not the developer of the game and can only hint.

Also I don't know how a cavalry-only-force can inflict ranged combat-hits.

[Baris]

Looked at the Austrian files(loaded) and looked at detailed combat report again that Gaistruck Brigade and a supply wagon are the only units commited to the battles in each round. So no cavalry from Austrian army commited it looks. I agree the cavalry targetted one unit. But even cavalry target only infantry, cavalry doesnt take many loses but it is deadly against infantry. so maybe it isbetter for units to be seperated rather than combined in birigade level because they can be a target for enemy. Where as only seperate units have much chance to survive. But downside is too much command penalty for seperate units. But as I remember combat penalty for Gessler was 25%. Did it have impact on charge of cavalry? and why didnt Austrian cavalry commited to battle it was clear but rainy terrain.

Maybe you should look at the save game.

[Bertram]

Jaques,
interesting point. With the cavalry doing so much damage when assaulting it could be a valid tactic (within the rules of the game, not in reality) to offer up one cavalry element to have the others charge home.
If it works that way, it should be stopped also.

My example points to a deeper problem though. I had one unit on both sides, so they should be firing at each other/ fighting each other.

And Baris his example (combined with my earlier remarks in the AAR and elsewhere) reinforces this point. In Baris his example the Austrian cavalry never got involved. This has been point of discussion before: often in mixed stacks either the cavalry or the infantry does not get involved at all. This leads to one sided losses, where either the cavalry or the infantry gets wiped out, leaving the other half of the force beaten, but without casualties.

It might be that the game checks at UNIT level if the charge was succesfull. If that is the case it would explain my result: either the charge is succes full, and all cavalry elements do triple (I think) damage and wipe out the infantry, or it isnt succesfull, and the cavalry takes triple (?) damage and gets wiped out. If experience and leader tactical efficiency is a modifier, it could fit the results show: at start of the game the (inexperienced) cavalry fails to charge, and gets wiped out, later, when more experienced, it succeeds, and the infantry gets wiped out. Main problems are then that:
- either the check is at unit level, and with a unit against unit battle, one side gets wiped out
- or the check is to sensitive to the experience level of cavalry and leader, and has the elemnts fail or succeeed "en masse".
- or both are the case

Another, and deeper, problem might be that the combat resolution algorithms try to solve all combat with the same routine. Like in AACW, where the naval battles don't work to well, and especially when a large number of transports gets engaged by an ironclad.
To get consistently a historically correct result, the routine to solve a cavalry vs infantry battle should be different depending on if that battle is part of a large scale engagement, with mixed forces on both sides, or if it is an engament between a single cavalry brigade against an infantry brigade. And if the infantry lugged a couple of guns with them, it should not only matter for the number of hits, but for the question if the cavalry would engage at all.
There should be a higher kind of algorithm which should first decide if there would be a battle at all, and if there should be a battle, what kind of parameters should apply.
Unfortunatly this is something that is a scale smaller then the game is depicting - and quite a bit of work. So I fear the best we can hope for is a tweak to the parameters, to get rid of the worst outcomes.

[Baris]

I didnt load the game again to see but in the battle report(not detailed one) it shows that Austrian started firing before Prussians. But again Astrians losing 2 line inf elements in range combat.

But very interesting you find out Bertram, engine maybe must check scale of battle before deciding which elements to engage and some parameters as you said for not to engage only cavalry to infantry after many many one type of engagements.

[JacquesDeLalaing]

to Baris:
Wait I missed that there were 12k Austrian cavalry present in the region (don't have the time to check the save right now). But you say that the cavalry has been in the same stack as the infantry? That’s really odd, because then it should indeed have been committed to the combat (frontage would have allowed it).

But, in case that the cavalry has been formed into a separate stack (with a defensive posture), the result seems plausible, because stacks with offensive command postures (i.e. in this case the infantry stack) are commited first. Also the amount of protection on the Austrian side (combat report: 177) is a hint that there was a seperate (cavalry?) stack with a defensive command posture present in the region.
Then again, I don't know exactly how the commitment of stacks with different command postures is carried out once stacks begin to rout or retreat. I.e. should the defensive cavalry stack have plunged into the combat once the offensive infantry-stack retreated?

to Bertram:
Well I guess it’s really a balancing issue then. However I do think that each element individually tests whether it charges and/or carries out a “special” cavalry charge (then the infantry gets a chance to form square…?). I don't know the factors which are taken into account for this, but I think that the result is to be explained by the very high quality of the cavalry force (command, discipline, experience, cohesion). The infantry might be unable to rout the "advancing" cavalry because of their high cohesion values (everytime an element gets hit in ranged combat, it makes a cohesion-check (absolute or % of it's max. cohesion?) in order to see whether it routs) and their high discipline-values (once an element has suffered hits, it undergoes discipline tests to see whether it is considered to be shaken during the following round). It’s all so complicated Let's wait for the answers of the competent guys!

[Baris]

[quote="JacquesDeLalaing"]to Baris:
Wait I missed that there were 12k Austrian cavalry present in the region (don't have the time to check the save right now). But you say that the cavalry has been in the same stack as the infantry? That’s really odd, because then it must have been committed to the combat (frontage would have allowed it). But, if the cavalry has been formed into a separate stack (with a defensive posture), then the result seems plausible, although I don't know exactly how the commitment of stacks with different command postures is carried out once stacks begin to rout or retreat.

Hi jacques, There was Austrian cavalry in the stack of the Austrian army. But not in the brigade that is targeted. İ looked at the save files for Austrians and didnt find Gaistruck Brigade in their stack after the battle. In the battle report it wasnt showing cavalry committed for Austrians. As in the battle report huge losses are from infantry elements. I thought that Gaistruck brigade is made of only infantry elements. The other brigades were made of only cavalry birigades and some only infantry brigades and they didnt committed to battle. In all 4 round only 1 brigade and supply wagon committed to the battle.

[Baris]

I load the game again as Prussians. And in the replay i saw that Gessler is moving towards Picolomini corps. When Gessler get close i saw Gaistruck Brigade seperating from Picolomini coloum so get caught by Gessler. I thought they were in the same stack after looking at the battle report. But after replay i saw that they were seperate stack But even they are seperate they should help each other as they are in the same region. But as you said maybe Picolomini was in defensive posture. I thought at first Gessler targetted Pİcolomini forces but only targetted Gaistruck brigade. So the battle report with Pİcolomini portrait as enemy was misleading

[JacquesDeLalaing]

Ah so at least I was able to explain this oddity. It seems that the Prussians have been very lucky. Had the Austrians organised their forces differently (different command postures), the Prussians would have been negatively surprised *g* - or else Gessler would simply have retreated prior to the battle ( guess the success-chance has something to do with the avergae evasion-value (+strat.-rating) of a force compared to the average patrol-value (+strat. rating) of the enemy, but that's a totally different question...).

However, in this case, Gessler only withdrew after 4 rounds of combat and after having routed Gaisrucks Brigade. My guess is that Gessler retreated because/before the defensive Austrian cavalry stack got committed? So, I could imagine that the Austrian infantry force routed, thus the defensive Austrian cavalry stack got "activated" and committed to the combat. Gessler got an opportunity to decide whether to retreat before this round and did so because the power-relation between the forces got shifted in Austrias favour by the committment of the cavalry-stack? But this is just a hint, and I guess we should see a "Piccolominis corps commited"-message or something like that in the message log?

As for the misleading leader-portraits: the leader shown here is always the C-in-C, the highest ranking general in the province, regardless of whether his stack was committed or not. It's the same for the troops displayed in the "overall combat report". It shows all troops present in the region, regardless of whether they/their stack actually took part in the battle or not.

[Baris]

Gessler withdrew because of Piccolini's portrait because there is no hint in combat reports that Pİccoloinis corps got commited in battle. Im suspicious that supply wagon that get commited must be from his stack not from Gaistruck.That was a secret message from Piccolini to Gessler not to be foolishly brave

I didnt pay attention to the messege log during the battle maybe there was a message about Piccolini's commitment.

[JacquesDeLalaing]

Well, it is not as gamey as one might think, because bigger units (bigger in terms of hits; e.g. brigades) are far more likely to be targeted than smaller units (e.g. seperate units). In this case, cavalry is not really weaker, but the enemy also takes more losses, which leads to a battle result that feels more realistic.

However, as in the case above, it can lead to some unbalanced results when the only targets available are small units, and when there is no enemy big unit at all.

But this is getting too much off topic. We still need to talk about the Original Poster's issue.

[Baris]

Yes exactly when I think about it. Not to be misleading from topic i editted my post.

[squarian]

I do not know about the mechanics. What I do know is some military history for the period. So far, I've seen two forces suffer disproportionate casualties against cav-only enemy forces.

In the first case a combined column of elite infantry supported by cav encountered an inferior body of cav - in reality, the cav would be driven off every time, with minimal losses on either side.

Second case, a bde of infantry in prepared positions (lvl 2 entrenchments) is attacked by cav - in reality, the result would either be a death-ride slaughter or (if the cav commander knew his business) a skirmish.

In reality, formed infantry facing a cav-only force was virtually invulnerable - the only reason not to prepare to receive cav was if one was threatened by arty or infantry. If those other arms were not present, then only an idiot would not form to receive.

Since cav commanders knew the odds, if the opposing infantry looked reasonably steady, only an idiot would commit his troopers to close combat, because the result was inevitable - lots of empty saddles and no glory. Instead, you might skirmish a bit, but any colonel of cavalry with a week's experience would know to draw out of range.

Anyone with a reasonable familiarity with this period won't need them, but if necessary I can cite half a dozen sources.

Napoleon's Campaigns apart, ROP is the first game using this engine to be set in a shock-cavalry environment - both the ACW and AWI saw little if any cav used in a shock role. So perhaps it's understandable that this part of the mechanics haven't been perfected quite yet.

However, there clearly is something wrong, and while I'm just about able to suspend disbelief while the entire Russian army starves in the snow and certain other glitches in the current release, I cannot manage the same trick with cavalry charging in broken country, absorbing several volleys, then arriving with sufficient cohesion and momentum not only to storm entrenchments, but to slaughter every last man defending them. It's a game-breaker, I'm afraid.

[Baris]

Success with only cavalry armies are not realistic at all indeed. Even without knowing history it is common sense that soldier riding a horse can not fire efficiently against a foot soldier who is also entrenched. In game also most cavalry units have rate of fire mostly maybe 2 and range is 1. While the foot soldiers has 3 or 4 range and rate of fire. Cavalry should lose more cohesion and health before charging the foot soldiers in close combat. In game what needs to be balanced as Bertram and Squarian pointed it out is owerpowered cavalry charges. Ironclad against trasport may be similar issue as said by Bertram.(Im havent played that much ACW)

[squarian]

Thanks, Baris - but I want to point out (directing these remarks in the spirit of constructive criticism primarily at the design team) that the dynamics of the battlefield in this period - horse and musket, whatever you want to call it - turned on the interaction of the various arms more than on the technical functions of any given arm.

In other words combined-arms tactics - the rock-paper-scissors dynamic - is what is missing here. Clearly the designers were aware that they were entering slightly new design country because this period involved cavalry in a shock role. So they introduced two new features - "cav charges" and "forming square". Well and good, but I'm afraid the interaction of the arms hasn't been reflected.

Having infantry check to form square vs. cav is one thing. But if a body of inf is facing cav and art, with no other support, it is doomed whether it forms square or not: if it forms it makes magnificent targets for the arty; if it doesn't form it will be ridden down by the cav. Likewise, inf alone facing a combo of cav and inf is in big trouble - the enemy inf will pin you to its front while the cav rolls up your flanks.

Conversely, inf need not worry about cav much if the cav is without support - most especially if the inf is dug in!!!

I suspect that's what going on here - we've got rules for cav shock effect, but no rule that says cav shock only works when cav operates in conjunction with other arms. Alone, cav can only skirmish or charge other cav and should recoil from enemy inf.

But let me repeat, this is fundamental to a wargame simulation of the period. If the dynamics of the 18th century battlefield aren't accurately reflected in this game, we might as well be playing Super-Mario. I dislike putting the case so strongly but here it is warranted - in my opinion, until this is fixed, ROP is unplayable.

[Baris]

Yes exactly, if the game evolved to a certain degree about adding special behaviours for cavalry and infantry.(I didnt practice acw or wia like ROP) than it must introduce more concepts like "combined armies" . Because it is never be logical to have all cavalry or all infantry brigades. As Jacques explained that units target same units in each combat round. Then with some probability and luck there will be all inf against all cavalry battle, if both brigades made of only cav or inf. We dont see how the battle occurred or units engaged of course.

Also commanders have some bonuses given as brigade level like militia man or cavalry man, so it is not practical to have mixed units in a brigade because of commander abilities. Maybe some bonuses can be applied to mixed stacks.

It is a good game and can be better ,we all here for constructive opinions of course

[Baris]

I was reading some historical articles about Prussian cavalry in the command of Friedrich Wilhelm von Seydlitz. As only cavalry army he had managed to win some crucial battles against the infantry of French while they were mobile. and they caused severe casualties to the infantry. After that Prussian infantry join the battle where the French units were already defeated.

But it looks like Prussian army's one of the most important units are cavalry. In some battles they are nearly as much as infantry in numbers. So it may not be so much bizarre about cavalry success.

But In Leuthen battle, Squares with a line inside them are cavalry. Red Boxes are Frederick's troops against Prince Charles of Lorraine. Circle in the middle of the picture is a rock. I couldn't find the first image but Prussian cavalry making a fake attack on enemy while the other cavalry and the rock hiding fast moving prussian infantry. On the other hand Charles's cavalry moving from left to right flank( behind the lines of infantry) to flank the prussian cavalry. Main Prussain army is attacking Charles of Lorraine from the flank so there is good coordination of units here(very Disiplined and fast march). And it is success for Frederick(reds) after that.

Combination of infantry and cavalry is a very big factor in the success leuthen battle.

[ATTACH]12229[/ATTACH]

[Pocus]

I tried reproducing the problem, but no luck. I would need a save with backup1 showing the problem.

[Bertram]

It has been some time... I'll look if I still have the saved games

Took some digging in my comp, but I had still the old game saved (Squarian and me restarted this turn, with my cavalry doing something else, as the result was to odd to keep).

This is about a fairly small battle, at Erfurt. 2300 Prussian cavalry attacked 2300 dug in Infantry (level 2 entrenched, in wooded terrain). The result is 300cavalrist casualties, and 2300 infantry casualties, all infantry units wiped out.

Results in the Cavalry Win.zip, previous turn in the Backup1.

[Bertram]

Added files to the previous post, but realised that it might not show up as new post- so this is just to point to those files .

[Pocus]

I replayed then analyzed the battle.

I would say that the problem is a combination of several factors, so overall it produced the effect you saw. To summarize, the losses done by the cuirassiers were mostly done on assault. They had at this time 90% chance (on average) of doing their attack, and when they did, they did 3 hits. On the other side, the infantry has 30% of doing an assault or counter assault and when they did, they did 1 hit.

So this led to a ratio of hits of 9 to 1, which is roughly what you see in the battle report.

Now, the engine is not out of whack, it is just a kind of snow balling effect of several factors:

1. Forest did not hamper much the cav, they only suffered 1 to discipline. This should be much higher I would say, to reduce charge and probability to 'assault' (= attack at close range)

2. The stats for damages are too high, 3 hits done per cuirassiers compared to 1 for infantry, I would say 2 is sufficient, as they can get (cherry topping ) a x1.5 coeff to that if they charge, and they charged often.

3. Experience. The experience bonus was significant, but normal considered their experience level which was very high (experience level 4 to 6!!). Here, I believe that as they fought since the start of the campaign, they raked such experience. But they did gain this experience because their damages done were too high. If the data for damages done was lower since the start, they would have gained less, so would have less boost from stats (snowballing effect).

Charge seems right. The dice used is a 20 sided dice, and the test is discipline + offensive bonus of the leader, so has they have 11 discipline and the leader is excellent (5 in offensive), you get in the extreme range of 16/20 chance of charging. This seems very high, but that's because the circumstances are exceptionnal too... The leader is exceptional, the base discipline is excellent, and the discipline from experience added 2 points. In normal situation you have less than 50% of doing a charge, for the cavalry ... if we now say that cav in forest loses 3 discipline, charge will be even rarer.
So charge is fine for me. Charge will x1.5 damages, so if we reduce to 2 damages done, we are good I believe.

So, I would say that:
a) difficult terrains should impact more discipline of cav units (-2 to -3 even -4 in mountain).
b) heavy cav should only do 2 damages, other cav only 1, because they can have a +50% bonus from charge.

That combined, will mean that for a new game, experience gained for cav will also be lower, so even less extra bonus.

[PhilThib]

Well, this could be done rather fast if you all agree, for next (present?) patch ?

[squarian]

a) difficult terrains should impact more discipline of cav units (-2 to -3 even -4 in mountain).
b) heavy cav should only do 2 damages, other cav only 1, because they can have a +50% bonus from charge.

That combined, will mean that for a new game, experience gained for cav will also be lower, so even less extra bonus.

In the first place, thank you very much Pocus for taking the trouble to examine this case so carefully. Not that it's unprecedented - you're a grand lot of fellows at AGEOD.

I'm still a little concerned that cav (AFAIK) is still going to be as likely to get the charge bonus against entrenched inf/art as against any troops in the open. But I don't know how it works "under the hood" - maybe entrenchments already have sufficient effects I'm not seeing on the surface?

But that apart, I say give these changes a try!

[Narwhal]

I'm still a little concerned that cav (AFAIK) is still going to be as likely to get the charge bonus against entrenched inf/art as against any troops in the open. But I don't know how it works "under the hood" - maybe entrenchments already have sufficient effects I'm not seeing on the surface?

But that apart, I say give these changes a try!

The change goes in the right direction. I like gradual change - if it does not work perfectly, it will still work BETTER, and we (well, Pocus) can change it more if needed.

Thank you for listening to your communauty that much, AGEOD (even though I am very recent in the communauty, I am already a die-hard fan trying to convince his friends to buy BoA

[Ebbingford]

Perhaps the cavalry charge bonus should also have the target infantry's cohesion/experience/discipline factored into it aswell?
Unbroken, steady infantry should be able to keep cavalry at bay and not sustain many casualties. On the other side of the coin though cavalry should not be charging steady formed infantry anyway, they just wouldn't do it.

[Charles]

The result of the battle against entrenched infantry does seem a little strange. I would personally be in favour of the changes suggested.

Unbroken, steady infantry should be able to keep cavalry at bay and not sustain many casualties. On the other side of the coin though cavalry should not be charging steady formed infantry anyway, they just wouldn't do it.

agreed. From what I have read, cavalry in the SYW would not charge unbroken infantry. All the successful charges occurred against infantry which was already breaking or broken, either caught in the flank or rear, with most of the casualties occurring during the pursuit.

It also appears forming squares was also rarely used. The usual tactic was to have the third line turn around and fire towards the rear to drive off enemy cavalry.

[Pocus]

Good suggestions indeed. Code changed, for the elements capable of charging, there are now 2 additional penalties possible:

-2 to the test roll for each trench level of the defender
-1 to the test roll for each 25% of cohesion of the defender (so a defender with 75% average cohesion will give a -3)

[Baris]

Excellent I congrulate you for your support. The game will be in a good balance.
Thanks to Bertram and Squarian also about their proposal and influence about improvement.

About the entrachment level I think it is 2 in ROP. So I think it will need to be much higher for the inf to be more effective against cavalry charges.
But the New cohesion rule is well thought and will balance it. Im eager to test it

[PhilThib]

DB changes done, will be in upcoming patch