For fun and curiosity, I tested nearly all cavalry units against each other in equal-resource battles in the scenario editor. thought i’d share the results. these tests were done on the august patch. I used 2000 resources as a common budget so there are slight rounding errors on the amount of units which I’ll point out.
I didn’t notice any RPS dynamics (tho I didn’t exhaustively test every possible matchup to find them) so I’m assuming there’s a strict ranking.
Castle Age:
Leitis (17 units, 2040 resources) (0 relics!)
Boyar (15 units, 1950 resources) (really super close against Monaspa, they actually went 3-3)
Monaspa (19 units, 1995 resources) (no regen because I don’t have georgians, so i can place monaspa in the editor but can’t make them georgian)
Magyar Huszar (25 units, 2000 resources) (I think it’s insane that a light cav unit that can be turned into a trash unit does this well against heavy cav)
Konnik (15 units, 1950 resources)
Keshik (20 units, 2000 resources) (but this is discounting the gold gen; for example it made ~200 gold in the fight against Coustillier)
Coustillier (18 units, 1980 resources) (note this is a straight fight, no micro, underselling the Coustillier charge)
Tiger Cavalry (14 units, 1960 resources)
Knight (15 units, 2025 resources) (kinda crazy just how badly this loses to every cavalry UU)
Hei Guang Cavalry (15 units, 1950 resources) (they have exactly equal hits to kill and reload time vs knights, so why do they consistenly lose? i think because their attack delay is longer)
Tarkan (17 units, 2040 reseources) (expected that this loses vs all heavy cav, but put it on just for completeness)
Imperial Age:
Leitis (again, 0 relics!)
Boyar
Keshik (really moved up the ranks, they gain more of a boost from elite than some other units i guess)
Magyar Huszar
Konnik
Coustillier
Paladin
Tarkan (much bigger hp boost from elite than most other units)
Cavalier
I could not test elite Monaspa, Tiger, or Hei Guang since I don’t own the DLCs, so I can place those units in the editor, but since they’re not the actual civs, they downgrade to non-elite on starting test. Likewise I didn’t put Iron Pagoda on here at all because only one civ gets it and it gets +20% attack speed from that, which I couldn’t apply.
Castle UUs should be better then generic units especially in Castle Age since they require a Castle which is pretty expensive.
You can instantly train Knights when reaching the Imperial Age.
Also almost all Elite Unique Cavalry Units have less pierce armour or HP then the Paladin (with the exception of the Tarkan)
You forgot the Boyar btw. they should win most battles in melee despite the one against the Leitis of course.
But yeah I’m also kinda surprised that the Magyar Huzars got buffed in the last patch.
A Castle Age UU must either fill a gap in the civ’s weaknesses or bring an exclusive combat advantage over their generic counterparts. Otherwise, nobody would bother training them. Even the unupgraded ones should offer at least one strong combat advantage over their Imperial Age generic counterparts, even if they’re weaker overall.
Luckily, almost all unique units fit that description, except for one.
Longbow is that unit right ? Why am I making 5 range longbows from an expensive castles when faster working archery ranges can pump out crossbows that do slightly less per shot but still get the civ bonus range and more importantly have higher base accuracy
Franks and Teutons paladins do beat coustillier with similar res (teutons might be even higher on the list).
I usually think new uu are too strong and overshadow old ones, but I think paladin still hold its place. Konniks are good and leitis are absurds in melee, but they are quite poor against archers.
Though most uu have a lower gold cost and are a lot more sustainable.
I’m surprised keshiks are that good in melee, considering they also have good pa and are only 30 gold each. That would require some tweaks, if some other units were not far more problematic.
Is 40g each, the fact that Keshik cost just 100 resources and have good durability+gold generation is what make them cost effective vs many melee units.
None is exactly that weak vs archers, Konniks so much better due to the dismounted form, and Leiciai with 2 relics can kill arbalests in 2 hits, so being overall strong when archer masses aren’t high.
You forgot the Boyar btw. they should win most battles in melee despite the one against the Leitis of course.
Oh yeah I did forget the Boyar. I should add that one
None is exactly that weak vs archers, Konniks so much better due to the dismounted form, and Leiciai with 2 relics can kill arbalests in 2 hits, so being overall strong when archer masses aren’t high.
I’d just like to add that Leitis also moves faster than Knight which is a nice advantage vs archers
okay so i tried the Boyar. i thought it’d beat Monaspa in castle age, but it actually was super close, i ran the test 6 times and they went 3-3! but since the Boyars actually had slightly less resourcs than the Monaspa (almost half a unit), I’ll say the Boyars win.
They were able to beat Magyar Huszar soundly both in castle and imp, as expected.
It’s a genius historical touch Lithuanians having a hard-counter unique unit against their western (Teutons) and eastern (Slavs) neighbors’ unique units. Playing Lithuanians against Teutons is extremely frustrating, though.
The dismounted form barely does anything against range units, so konniks are only slighty better than cavaliers, let alone bulgarian cavaliers. Leitis may need 2 shoots instead of 3, but they also tank 30 arrows instead of 60, lithuanian paladin are still a lot more well rounded in that role.
I’m still surprised they beat strong melee units like konniks. They are high hp, high pa, low attack, so they are already very cost efficient against range units. Kinda like magyar hussars, it’s weird they can also brute force other heavy cavalry.
That’s why I said when the archer mass isn’t big enough, ofc if you throw 25 Leiciai vs something like 40+ Arbalests or Heavy Cavalry Archers relic boosted or not, then you 80% good chances will have horrible time.