Hi everyone,
a few weeks ago I’ve published my AoE2DE Combat Simulator here and on reddit: Post
Since then I have…
…reworked the simulation process (especially cases with trample/cleave/AOE damage, and very uneven army sizes)
…brushed up the UI.
…added cost efficiency matrices for different scenarios (different Hit&Run modes and different stages of the game (e.g. during the end of 1v1 matches, gold is often worth more than five times as much as food or wood <=> common market exchange rate is 100F=100W=17G)).
Below you can see one exemplary cost efficiency matrix for the specified setting (no Hit&Run and resources of equal worth).
I’ve automatically generated this matrix (and all other matrices) using my Combat Simulator.
The matrix basically gives information about the cost efficiency (or inefficiency) of each unit type vs. each other unit type.
For matrices with different settings (especially Hit&Run behaviour for ranged units) and further information just visit Combat Simulator .
Or have a look at https://www.youtube.com/watch?v=QZVT7kQYvoM&t if you want to know further details about the simulator and the simulation process.
Feel free to send feedback, suggestions for improvement or report bugs.
Important: The simulator hasn’t been optimized for mobile devices yet!
It reports for 100 villagers vs 1 paladin fight, that 88.9 villagers survive, which is almost the same as my scenario editor result of 90
Definite improvement over previous version result of 98.
Awesome job! The simulator is way cool and is a way more efficient way than testing different unit combos 20 different times in the scenario editor! I don’t know how realistic or difficult these suggestions are, but here they are: Search bar feature (to make it easier to find units), adding civ bonuses (i.e. more HP archers for Vietnamese), adding ships vs. ships, and less upgraded versions of units (i.e. cavaliers instead of paladins). Overall way cool website and I’ll use it all the time for unit testing!
Awesome work! Looking through the matrix now. Will have to try the rest later
Edit: This is some great work but im sorry to say this first table is almost meaningless. Resources don’t have equal value, and some degree of hit and run should always factor, for example there’s not a remote chance that TK are that effective. How unskilled must someone be to not even remotely kite TK’s?
This is an impressive piece of work that must have taken considerable time to develop.
I’d find it useful to have text next to the pictures of units to make it easier to find something. Also, it looks like it only has the highest version of each unit? Do they all correspond to Post-Imperial in the Scenario Editor? It would be useful to be able to specify any unit from any age and also individual upgrades to see the effect of e.g. Chemistry or Ballistics, but I appreciate this is not easy to do in the Scenario Editor either.
If you’re interested in an idea for something else to do, I’ve been contemplating something along these lines to simulate build orders. So you’d assume a continuously utilised TC, and it would allow you to choose which resource each villager goes to collect, and that would produce a timeline of resource availability, and then you’d be able to choose things to spend it on at different points in the timeline. This would let you optimise the number of villagers on each resource to maintain a given number of military production buildings constantly working, for example. It would also need to allow specifying movement of villagers from a resource to a different one at specified points in the timeline. It would need to model the reduction in rate of collection of wood as the trees become farther from the lumber camp, and the effect of Wheelbarrow and Hand Cart on this. All quite a bit of work, for sure, but I watched your YouTube video on how the combat simulator works, and I think what you’ve already done is around the same level of sophistication that would be needed to make a reasonably accurate build order simulator.
been running the sims, they seem good, interested to see if someone confirms them but they seem accurate, but at the same time seen some remarkable results… berserkers are simply insane according to the sims… they trade cost effectively against almost everything yet theres a thread going asking for a buff… like how far off people can be…
edit: i wish your method of target selection was used in the game… it really seems like the game one is worse at times…
@CausingDart777
Absolutely agreed, but implementing that is very time consuming. I have to see if that’s possible during the next weeks.
@phoenix1089
Obviously you’re right that TKs are insanely strong without Hit&Run.
However, I’ve created different matrices for different Hit&Run settings and imho it’s absolutely reasonable to include a matrix without any sort of Hit&Run as well. No player can be at all places at every single point of time. There will always be situations when TKs will approach ranged units. Therefore matrices without any sort of Hit&Run are as reasonable as matrices with medium or perfect Hit&Run. They are just “theoretical guidelines” that can help you when the respective situation arises.
I actually don’t agree with your general statement that resources don’t have equal value.
In some situations they definitely might have equal/similar value, in some situations wood is the most valueable resource (especially late game on island maps), in some situations gold is the most valueable resource and in some situations stone is the most valueable resource. You just can’t make general statements about the value of resources.
Therefore I have different matrices with different resource value settings.
Elite Berserks are indeed a very strong unit imho. They are cost efficient vs. the vast majority of melee units and even some ranged ones and they are just insane in trash wars if you can afford them.
@breeminator
Thank you! It indeed took quite some time to implement the current version
Yes, the previous version had text (unit names) instead of their icons. Maybe I’ll add text again and keep the icons too.
All units are Post-Imp teched.
Regarding individual techs/bonusses: As I’ve mentioned before I think that’s a very nice idea, but it will take a lot of time to implement that.
Your suggestion for build order simulation is very interesting but I just fear I won’t find enough time to do that :-/
with no micro and evenly splitting the halbs between the elephants? not unlikely.
it would take 5 to 1 halb to war elephant odds to kill a war elephant without losing a single halb (and that assumes just a 1 on 5 fight, with no other splash from other nearby elephants)., so when you have only slightly over 1 to 1, its going to be in the elephants favor, especially when you factor in the trample.
I suggest everyone who finds some of combat simulator results incorrect, do corresponding testing in scenario editor. If scenario editor results are significantly different, then come post about it in here. Else it is not even known, what result should have come out of combat simulator.
@Parthnan
I’ve just tried out 40 Elite War Elephants (Persians) vs 50 Halberdiers (Spanish) three times in AoE2 DE:
First 38, then 39, then again 39 elephants survived.
To be honest, I also expected more elephants to die, but that’s the result, so actually also in this scenario my simulator is pretty accurate.
If there’s a problem at all, then it’s not about overestimating cavalry, but overestimating trample damage. But apparently my simulator is quite good calculating trample damage as well now.
@Yorok0
I totally agree with your suggestion.
It would be really nice if people could check there assumptions first (at least once).
Checking all the assumptions and proving them with ingame screenshots is also quite time-consuming^^
I think I’ve already checked dozens of scenarios people doubting the correctness of the simulator have come up with, and (if I remember correctly) in all but two (one of which were you, @Yorok0 ) my simulator was right or very close to the results observed in AoE2.
So, yes, feedback is very appreciated to improve the simulator’s quality, but plz check a new assumption first
Just as a side note: The two mentioned major issues in the previous simulator version were overestimating trample damage and not having a maximum number of attackers for each unit (which led to incorrect results in fights with very uneven army sizes). But that should be fixed by now.
Very nice work! I’ll be using this in future rather than setting up scenarios, except in cases of high doubt…
I agree about civ bonuses being a priority for the next update. One of the main things I like to test is “How can civ X beat civ Y?” For example, today I was looking at Goth late-game options vs. Teutons (lol), and it would be amazing if I could use this to test whether the cheaper units < or = or > the extra melee armour.
) my simulator was right or very close to the results observed in AoE2.