[Cavalry] Do they need a speed fix in the EDU? Update: Done

[QuintusSertorius]

I'm fairly happy with the performance of infantry javelineers after tweaking their animation and range in my submod, they do as they are supposed to now. But I wonder if javelin cavalry are labouring under the same problems? I just played a battle using Hippakontistai, who were soundly beaten in a javelin duel by a unit of Libyan Skirmishers. Only a small proportion of the horsemen would throw a javelin, where the infantry were volleying away nicely.

I believe the issue is primarily range, which was set (for all the right reasons) quite low. These were pretty realistic figures, but they don't work with the skirmishing ranges, causing javelineers to forever hesitate on the point of throwing, or only a few of them to complete a cast. At present all javelin cavalry have a range of either 45m or 55m; I wonder about levelling them all at 60m. That's intentionally shorter than the 70m range I've given all infantry javelineers; while you are throwing from an elevated position on horseback, you can't throw with your whole body as you can on foot.

Is there an issue here that needs addressing?

[QuintusSertorius]

I've made the change; it wasn't savegame compatible for me, but might be for other people.

[torongill]

I think an increased range may be a good idea, but something like 60m or so.

However, skirmisher cavalry would be skirmishing against cavalry or heavy infantry. Against actual skirmishers they would probably charge them. Talking historically, of course.
While it's true that skirmisher on horseback won't be able to use all his body strength to throw the javelin, we shouldn't forget that his javelin would already be travelling with the speed of the running horse. I'd say that range should be okay, ammo count should be higher(javelins can be carried in saddle quivers), accuracy should suck(a moving horse is not a stable throwing platform) and the javelins should be heavier.

But really the skirmisher cavalry shouldn't be recruited for battlefield prowess alone. It should be useful as a cheap substitute for melee pursuing cavalry(used to run down routing enemies), it should increase viewing distance of the army(through a hidden trait maybe?) and should lower the rate of starvation of the army in enemy territory. This is what it was actually used for in Antiquity.

[QuintusSertorius]

I did indeed go with 60m. They use a different weapon to infantry javelineers (cav_javelin45 rather than prec_javelin80 as I changed the infantry ones to), though I didn't check the projectiles file to see what its characteristics are.

The "cheap pursuit cavalry" was about all they were good for with non-functional javelins. Now at least they can skirmish too.

Not sure they can impact campaign properties, though, even if I agree with you that it was part of their role. Another one would be to obscure your own force from the enemy through screening.

[QuintusSertorius]

I'd appreciate some input on this, which is not entirely resolved yet. I've had feedback, and while javelin cavalry now throw their missiles (result!), they don't actually do much damage (fail...). I had a look at the descr_projectile.txt and found the culprit: cav_javelin45 has an accuracy_vs_units of 0.25. Bear in mind the lower this number is, the deadlier, and for infantry javelineers using prec_javelin80, it's 0.08 (about three times as deadly).

The nearest available cavalry type in there is the cav_heavy_javelin45, but this has an accuracy_vs_unit of a massive 0.045 - almost twice as deadly as infantry javelins and about six times as much as the present value, which seems overpowered. Can I just use the prec_javelin80, or will that do weird things when wielded by a cavalry unit? What I'm trying to avoid is editing anything besides the EDU, because we start getting into all sorts of complications, and make it harder for people to use it.

I note cav_javelin45 also has a damage of 0, where it's 2 for prec_javelin80 and more for the pilum. Though arrows and sling stones have damage 0. No idea if this makes any difference.


On a different point, I was talking to Ibrahim about movement speeds over on The Org, and he agreed that at present they're a bit strange. The relative speed difference between infantry and cavalry is too small; either infantry are too fast, or cavalry are too slow. The move_speed_mod attribute is responsible for speed, 1.0 is 120 steps per minute. Currently values range from about 0.5 for the slowest infantry (like phalangitai) to 1.03 for the fastest cavalry. But a lot of infantry is in the 0.8 range (and 0.98 for light infantry - the same speed as heavy cavalry!), so not that much slower than cavalry.

I was thinking about tweaking this to make cavalry (especially light cavalry) faster. Does anyone have any views on an appropriate value for this? I was thinking something like 1.8 for light cavalry, 1.5 for medium cavalry and 1.2 for heavy cavalry.

[torongill]

I'd appreciate some input on this, which is not entirely resolved yet. I've had feedback, and while javelin cavalry now throw their missiles (result!), they don't actually do much damage (fail...). I had a look at the descr_projectile.txt and found the culprit: cav_javelin45 has an accuracy_vs_units of 0.25. Bear in mind the lower this number is, the deadlier, and for infantry javelineers using prec_javelin80, it's 0.08 (about three times as deadly).

The nearest available cavalry type in there is the cav_heavy_javelin45, but this has an accuracy_vs_unit of a massive 0.045 - almost twice as deadly as infantry javelins and about six times as much as the present value, which seems overpowered. Can I just use the prec_javelin80, or will that do weird things when wielded by a cavalry unit? What I'm trying to avoid is editing anything besides the EDU, because we start getting into all sorts of complications, and make it harder for people to use it.

I note cav_javelin45 also has a damage of 0, where it's 2 for prec_javelin80 and more for the pilum. Though arrows and sling stones have damage 0. No idea if this makes any difference.


On a different point, I was talking to Ibrahim about movement speeds over on The Org, and he agreed that at present they're a bit strange. The relative speed difference between infantry and cavalry is too small; either infantry are too fast, or cavalry are too slow. The move_speed_mod attribute is responsible for speed, 1.0 is 120 steps per minute. Currently values range from about 0.5 for the slowest infantry (like phalangitai) to 1.03 for the fastest cavalry. But a lot of infantry is in the 0.8 range (and 0.98 for light infantry - the same speed as heavy cavalry!), so not that much slower than cavalry.

I was thinking about tweaking this to make cavalry (especially light cavalry) faster. Does anyone have any views on an appropriate value for this? I was thinking something like 1.8 for light cavalry, 1.5 for medium cavalry and 1.2 for heavy cavalry.

I can't really help you with the testing because I have my own EDU that has been tweaked too much.

In any case here are some points:

1. Accuracy for cavalry javelins should be lower than that of normal javelins - but it should be higher than the accuracy of arrows and slingshot. It's ridiculous to have slingshot and arrows that shoot for hundreds of meters, with all associated wobbling and wind interference, and get better accuracy than a javelin.
2. IIRC Cavalry and infantry have different default speeds. Cavalry should still be genuinely faster, much faster than infantry. And light cavalry should be considerably faster and have more endurance than heavy cavalry. The same way skirmishers must be able to break away from medium and heavy infantry. In this aspect, one big problem is the fact that the battlefields are small and often the two forces start an arrow shot away. This makes it impossible to use the skirmishers, both infantry and cavalry, to their best effect.
3. Coming back to the javelins, the damage you saw is building damage. Arrows and slingshot cannot damage buildings. Pila can. I've destroyed more than one ram by stupid infantry volleying their pila at the enemy through the ram.
4. Skirmishers and missile troops are robbed of their greatest attribute in pitched battles. They would make the enemy line crouch under their shields and raise their shields in order to guard themselves, tiring and losing confidence in the process, even though they may not lose a great number of troops. I believe until Rome II this wasn't really well reflected. Ergo, the way we maintain skirmisher usefulness would be to make them more lethal in some way that does not make them overly powerful.

Personally I believe in a holistic approach. Just by tweaking the EDU we can't fix everything. We can still do a lot, mind you, but we should look at the other options as well. For example, I believe we may be able to add a hidden trait that is activated if the army has say three or more skirmisher units, a trait that adds logistics and scouting ability, i.e. starve slower, see farther. This would make them useful even if they suck on the battlefield.

Second, we should analyse the function that each class is supposed to fulfil.
Foot skirmishers are supposed to be a nuisance, lethal but fragile, lithe and agile, good for molesting the enemy line and chasing down routers, but not good for more.
Skirmisher cavalry should be an equivalent with an added bonus - since it has its own transport, it should be able to run down foot skirmishers and decimate them, as well as lock enemy cavalry in a skirmish duel and fixate their attention on them instead on the backs of friendly infantry. Skirmisher cavalry should be very good in chasing routing enemy cavalry, especially heavy cavalry that tires in a short period of time. The downside would be that skirmisher cavalry, if professional and not levied, needs more maintenance than equivalent medium and maybe even heavy infantry unit.

I won't be able to test the changes, but I believe just adding some distance to javelins thrown by skirmisher cavalry, and making it more accurate, may solve a lot of problems. As always, the minimum skirmish distance shouldn't be too low, or skirmisher cavalry may end either charged by enemy cavalry, or even javelin-stormed by opposing heavy infantry.

[QuintusSertorius]

I was actually more after your brain and understanding of the mechanics, than for testing.

1. As it stands in my latest update, cavalry javelins (0.045) are more accurate than arrows or infantry javelins (0.08 for both). That's without tweaking the descr_projectile.txt for myself, which I'm trying to avoid (people will switch out an EDU, but change other files as well and they may lose confidence). Can I use prec_javelin80 for cavalry so they'd be the same weapon as the infantry javelineers? Or will that do funny things? If all else fails, cavalry units are much smaller, so while they'll pack more punch for their size, it should be about the same as an equivalent infantry unit (and much more expensive, with a shorter range).
2. I'll target speeds as my next round of edits, cavalry at the very least, perhaps light infantry too. Do you have a feeling at all for the numbers here? Currently they range from 0.74 to 1.13, from Parthian Cataphract to Numidian Skirmisher Cavalry.
3. Thanks for that, makes sense now.
4. Indeed, this is the best we can manage without more sophisticated effects built into the engine.

I agree with you on a more holistic approach, but I'm assuming the team will manage that when they get round to tinkering some more. All I'm aiming for here is a workable EDU with units that function the way they're supposed to in the meantime that other people can share. I'd be curious to hear if you get anywhere with hidden traits for logistics. I've heard some other mods use them, so I think there's scope.

[QuintusSertorius]

I think I've settled on what I'm going to do with cavalry speeds. I'm going to multiply them all by 1.6 - ie increase them by 60% - and see how that affects things.

[QuintusSertorius]

Right, I've now updated my EDU tweaks, heavy cavalry are 30% faster (and so generally faster than most infantry, though not the lightest and most fleet footed in the case of the very heaviest cataphracts), light/missile/skirmish cavalry are 60% faster. I'd be keen to hear from people how it impacts battle dynamics.

[QuintusSertorius]

If anyone is still following this, I've finished my speed-tweaks and I think they're about right. Heavy cavalry are 25% faster, light cavalry are 35% faster.

[malibu.stacey]

Just tested your new update in a couple of custom battles, cav behaves properly now, in my opinion. Good changes and + rep!

[QuintusSertorius]

I also note charges seem to be a little more effective, that might be a speed-momentum-related thing.

[alex86]

Awesome work on this!

I haven't played EB2 in a little while, but it sounds like when I do I'll have to try out your updates.

[Mouzafphaerre]

​
Thanks Quinte! Followed your thread and now downloading your EDU
​

[QuintusSertorius]

Awesome work on this!

I haven't played EB2 in a little while, but it sounds like when I do I'll have to try out your updates.

I'd be keen to hear what you think.

​
Thanks Quinte! Followed your thread and now downloading your EDU
​

Great! I'm thinking of putting it forward as an unoffical hotfix for the myriad problems a lot of units have in battle, at least until the team get round to addressing them in the next release.

[Sarkiss]

I'd be keen to hear what you think.

Great! I'm thinking of putting it forward as an unoffical hotfix for the myriad problems a lot of units have in battle, at least until the team get round to addressing them in the next release.

go for it. im looking forward to giving it a go once my PC is sorted. one nubbish question though. from your posts i got an impression that theres only light - heavy cavalry divide. isnt there medium class that could use its own % value of speed increase?

keep it up cheers

[QuintusSertorius]

go for it. im looking forward to giving it a go once my PC is sorted. one nubbish question though. from your posts i got an impression that theres only light - heavy cavalry divide. isnt there medium class that could use its own % value of speed increase?

keep it up cheers

There isn't actually a medium classification in the EDU: there's Light, Heavy, Missile and Skirmish. Missile and Skirmish are actually all over the place; there's some heavily-armoured horse archers classified as Missile, and heavy chariots who are Skirmish, for example, and anything we might class as "medium" is really light with a bit of armour. Using a rough light/heavy split divides the total cavalry roster almost in half, the heavy end bulked up by the volume of FM bodyguard cavalry who are heavy.

Furthermore, there's quite a lot of variability in speeds anyway, so Hippeis (who are archetypical medium cavalry) are classified as "light" but they're nowhere near as fast as Hippakontistai, for example. So within each of those two categories, there's enough of a spread that we don't really need to distinguish a medium class. Besides which, that would complicate things a lot, since I'd have to come up with criteria of my own to determine what's light and what's medium.

[Sarkiss]

i see, thanks for elaborating on this. the team should really get you on board so you can assist Ibrahim in getting this area improved.
many thanks for the fix.

[Col.KanKrusha]

Hi

Trying this out. I didn't think the speed change was neccessary as I was having no trouble chasing down routers and i liked the limited cavalry. But having tried it i do notice that skirmishing cavalry are getting caught in melee much less often and seems like a good change.

i am on your update 4 which I am finding a little too fast so your changes for update 5 are probably right on the mark.