ive watched a few games from both hera and viper in the past on stream stating they get the arch upgrades as cav civs all the time in very late game, they mostly do this in post imp games where their opponent is hussar raiding, most of these games ive seen in tourneys dont end up in this situation to begin with most tourney games are won with an all in attack approach either before booming or after booming because obviously why would you need the arch upgrades as a cav civ when you have massed knights about to knock on his door. watch heras stream and you see often 1.5 - 2hour long trash war games, ive noticed youve cherry picked all your games. A 5 game set is played in 1 hour 40 ish mins and never reaches the point that pros even get these upgrades and in the 7 set of hera vs viper in malay vs turks hera does get the arch upgrades in the late game once he is POP CAPPED and CANT SPEND HIS RESOURCES ANY BETTER and finally the way that map is layed out and given how hard hera pushes in this game he doesnt need arch upgrades because hed weaken his push and he already has one side locked down with push other side has buildings in front which will warn him of attack he does not have any need for arch attack upgrades yet and if attack switched other side he could easily get them before attack hits his main base
game 6 of same set celts vs japs theres obvious reasons for them not getting arch upgrades like: when facing masses of infantry that destroy buildings fast do you prioritise army to kill other players army fast or at least match it or do you spend your limited non post imp resources on attack upgrades for buildings that will go down fast before they pay for themselves? come on this is obvious strategy of eco expenditure you are cherry picking examples and they are poor examples to begin with, casusincorrabil stated he gets these techs late game like all the other pros do but none of your games showcase this because they end too early before pop cap and with limited resources at time of finish and even in one of your referenced games hera does indeed get the arch attack upgrade in the late game lol, at least give good examples of games if your going to make a point and not contradictory/ irrelevant examples
Viper didnāt get archer techs for his fortifications. Do you think he misplayed? Is this your argumentation? Or are you just saying stuff. You havenāt contradicted my pointā¦ youāve just kinda said stuff about the game. Gratz.
So youāve got a reason to think thereās a specific argument as to why they didnāt get archer upgrades in one particular game. What about the others I showed in which nobody got archer upgrades, or one player decided against them? Or, did you literally print twenty-three lines of text to say āwell Iāve seen some stuff and while I provide no evidence itās totes real dudeā and then complain about one game?
If heās getting them that late they arenāt reducing the cost of his Archers. Which is his claim. His claim is that youād already have them. If thatās your argument, you agree with me.
Honestly, I donāt know whether Iād have Casus acting like I donāt exist because he thinks people who disagree with him are trolls, or having you rant on about two examples that prove my point, act like they donāt, and then ignore the other five, to complain about a unit that was buffed a week ago being too weak. Theyāre both pretty quality feedback.
How about you find games where the Archer upgrades are researched specifically for fortifications rather than being mad at me for being able to find examples of games that support my position, howabouts? Since it happens so often you should be able to find plenty of them. Iāve found at least five from the last two months in which it didnāt happen, and I found one game that was completely over in which someone got fletching for that purpose about ten seconds before his opponent CCād. Go ahead and find four more. Weāre waiting. Prove your point.
No, Iām giving games where the archer upgrades arenāt upgraded for the sake of defensive fortifications, against raids, etc. Thatās the claim @casusincorrabil makes as justification for claiming that the Hand Cannons are not as cheap as I make them out to be, and is the justification for them needing a buff.
Itās his claim you will get the archer attack upgrades for your defenses, and as such already have them for Arbalest, making HC less inexpensive by comparison. This is the rebuttal to that. In every one of those games I mentioned, at least one player (most of them both) never got the archer attack upgrades for their defense and a majority of them went to Imperial.
A.K.A. Heās wrong. Which means the foundation for his argument doesnāt exist. Which means his example simulation of:
is absolute nonsense, and can be safely thrown out. Since heās sticking to it in face of all available evidence, his point is bunk. His numbers make no sense so long as you assume the archer techs wonāt be taken before an HC transition is possible, and as Iāve shown, that is *overwhelmingly the case in games without an archer/skirm opening from a player.
If you want to show even the barest attempt at an honest evaluation and discussion of the unit, youād show both examples of the HC with two simulations. One where you have ascertained ranged attack as you claim would be normal (maybe itās a hybrid map and you went war galley play, maybe you went skirm and had to tech-switch after?) and one where you donāt get the ranged attack, and have to evaluate the two options from a bare start.
That way, weād get to see how it performs in terms of cost in the best case and the worst case, and really see where on that line we want it to be. As you are doing now, you are (knowingly) providing us with the worst case and using that as a justification that it needs a buff. If you wanted to have absolutely any semblance of a fair and meaningful discussion, you wouldnāt slant the data in this way, nor would you castigate those who are trying to point this dishonesty out as ātrollsā and ignore them for the sake of being able to ignore the repercussions of tainting a discussion with biased samples.
Do the test again with the archer upgrades completely clean, and letās see how they do. Whatās the worst it could spit out? The only thing we stand to gain is a more rounded discussion. If you want that, youāll agree.
NO i settled in the middle ground. And I told before hand to do so. You just repeatedly claim I would chose the worst case or whatever, but I definetely didnāt. And I said it, several times that I tried this to account for yourā¦ whatever. I was trying to get this ādiscussionā to a level of equality by just accepting that there might be different perceptions of the game besides I donāt understand them Instead of accepting my hand given to you to settle, you just took it and ranted, repeatedly. And I really, really dislike that behaviour in a discussion. It just showing that you donāt agree on equal terms, you just want to āwinā the āargumentā whilst we are still just trying to figure out why HC have that bad state and what might be the reasons behind it.
I can make worst and best case including ballistics and chemistry if you like.
I donāt agree what youāve decided is the middleground and I think itās far from it.
Iād be more than happy to see you add Ballistics and make a true worst-case scenario, so long as you also include a true best-case scenario, so we see the absolute worst, and absolute best case for the unit. I believe that range would be invaluable to see where it actually sits in itās role. Itās certainly much better than the alternative.
Ok if you have to make the complete tech switch, didnāt made any upgrade before it would be 30 arbs of which 23 would survive. (7 lost)
If you already had chemistry + ballistics and the attack upgrades it would be 38 arbs of which 35 would survive. (3 lost)
Looks still better than HC in my opinion, even the worst case.
Edit: To Berbers x-bows: It would be 32 berbers x-bows of which 20 would survive (12 lost) - still better than the HC switch.
Because I took your āworst caseā cenario.
It is not a āone timeā interaction, the whole scenario is set up to account for the tech-in cost in the process of the game. Thatās the principle. As the game doesnāt is about one single big engagement but more about several, the tech-in is distributed over several engagements. Thatās the Idea behinf the calculation.
If you spread out the cost along multiple engagements, you then must also concede that for earlier portions, while the tech-in is being made, your units are worse and as such, will take worse average trades. How are you factoring that in?
Yes, thatās how I do it actually, but in a floating manner.
Earlier engagements are weightened much more than later ones, that is the principle of the calc.
But why do I have to explain it here again, itās very well described in the explanation how I calc itā¦
The calc is set up basically to calc when itās worth to get a specific tech, thatās the basic idea behind the whole calculation.
BTW in the case the āworst caseā upgrades make actually 38 % of the effective Arbs cost, just to say. Itās quite massive even for a hard switch. As we all know that having 38 % less of a unit can completely switch the outcome of a battle from big win to big loss.
Edit just an example: 20 franks paladins vs 100 britons arbs give an average of 58 arbs surviving, but switching this to 20 palas vs 62 arbs makes this average 13 pala survive.
So, 38 % less bec of tech-in is massive in this engagements.
The last two blacksmith techs alone take more time than Chemistry, which, aside from Armor, is the only tech the HC needs. Iām not talking mass, or anything. Iām strictly speaking that for the next two minutes, the Archers are going to be lacking big upgrades. How is that accounted for, specifically?
I get this. But youāve not pushed the units back by 950 resources, which is the excess. Youāve pushed them back by 280, and of a more scarce resource in food. So I donāt understand the methodology at all.
WTF?
I explained it -.- I made it exactly how you wanted to, i made the worst case and made a complete switch -.-.
Man, itās becoming really embarrising now -.-. I do what you want and you just pose like I did something different -.-
Iā, out of that, thatās no discussion, thatās full troll mode.
I will only discuss with people which are willing to in this thread, sorry.
Iām not trolling. You just arenāt explaining it very well or Iām missing it. I could see both. Because youāre doing a long-term evaluation of the cost after the switch and that uses different metrics, but Iām not sure that translates well to this circumstance.
Wait, did you reduce the number of Arbs by an equal amount to the percentage increase of upgrade costs?
i think we speak about the same principle, but from the different angle. Whilst my proposition is to assume basically a decay of worth by time, you come from a more economic perspective. And I thiink itās my fault cause I told you itās basically the same thing, cause in the aftermath it is - but specified.
The way I do it allows me to insert all kind of investment, current value etc. at various points, but I have to chose them wisely - it doesnāt restrict me to a pure economical POV, if you understand what i mean.
In the aftermath it may be equal, but it is also more complicated to insert more āsoftā elements like any kind of delay eg through tech-in timings. As well as I admit, that I havenāt included adding up of timings cause of dependencies.
But I think generally our methods would come to the same results if we would put in the same values.
Not directly, but I can insert a delay for a unfinished tech switch.
Because I canāt predict when an engagement will happen, I need to decrease units worth at same speed, but ofc for both sides. Itās an equal decrease of value, so it doesnāt really matter if the engagement is delayed as both sides will lose the same relative value during that delay.
Iām a little skeptical how you calculating food cost or discounting. Take the food cost example first.
For example if you are amortizing the cost of the farm using exponential discounting you will have for the cost:
Cost function(t) = (time_cost_of_construction/wood)*e^(-rt)/normalizing_constant +1 where the normalizing constant is the integral from 0 to farm_life of e^(-rt). This is what I was explaining early about amortizing units based on their occurrence. Obviously this converges to the function f(x)=1, because each second of food gathering takes 1 second of work time. The earlier seconds cost more than 1 villager second per second spent farming because they amortize the cost of the farm.
and for the benefit: Benefit function = 0 for 15 seconds, 0.4 after since handcart farmers farm at about 24f/minute = 0.4f/sec
This gives you the marginal cost of food for any farm at any point in time. Since we donāt know how long farms have been around we treat it like a random variable. Since no new farms need to be planted we treat it like a uniform random variable. This is what you should be doing, but I have no idea what you are doing in place of this. Take the expectation over this function of this distribution using benefit/cost as the function of time (this is valid by the law of the unconscious statistician).
See Graphing Calculator for the work on an average food cost of 0.385 food per villager second. You can change the parameters however you want but a value below around 0.37 is not realistic for this example.
For discounting the hand cannons come out first (see example above about the time constraints) which means they should have the highest value. Their upgrade costs and stuff arenāt āwaiting aroundā to be used. The arbalests need to have their cost inflated, making the problem Iām having reconciling their cost even worse.
As far as I can tell 34 and 34 is much closer to reality.
Well, as I canāt distribute the cost unequally over the time of the farm, because, yeah this would make calc just impossibleā¦ I dsitribute it equally. Luckyly the whole approach allows this - not only allows it, itās the base concept of the approach.
So instead of paying of the debt of the invested 60 food in the first minutes of the farm, I distribute it equally. This also makes more sense as Farms expire at different times, so an unequal distribution makes no real sense.
So for every piece of food I get from a farm I assume a certain amount of wood needed to be gathered (at that moment) to supply for that farm instead of a one-time investment. I chose it a way that the sum of these supply wood, corrected by the time correction factor (e^(-rt)) equals the upfront investment.
There is only 2 small problems with this method which can occur occasionally:
A) Wood is usually floating anyways, as it is the easiest ressource to convert or switch out. So investing into more farms doesnāt hurt the eco as much as going of from them, it often benefits the eco because of less floating wood. And ofc in the lategame you also need less wood to sustain your economy anyways, so investing it into farms is quite reasonable to get a āmore balancedā eco.
B) There are some breakpoints in the game where basically nothing happens because players try to get some powerspikes. How they work is that they basically reset all investments done before, meaning all tech-ins before these powerspikes will be paid of the moment you hit that powerspike.
EG The Imp powerspike will basically reset all farms placed in castle age meaning all food collected from them is basically debt-free in imp.
But I donāt know how I could even barely adress that in my calculation, as I donāt know how many farms would be placed before hand.
I think the key thing is the r value. I chose to have a HWP of 600 secs (HWP*ln(2) = 1/r), itās more realistic for the lategame as it is slower snowballing than the midgame which is dominated by booming (yes booming needs to have the best HWP in a strategy game). And the best (lowest) HWP I found for full booming is 240 s (Indian full boom). WIth that in mind the r value I use is .0025 / s. Thatās kind of a difference maker.
Ofc you can chose to use a higher r value, but everything what exceeds a full boom canāt be realistic, as it would mean the game would be disbalanced towards more military and we all know thatās not the case.
I know there are a lot of people here looking at the numbers of resources and units and where they find equivalence, as well as time and investment to tech into HCs vs other units. But to me, all of that is kind of missing the point. The rarity of seeing HC compared to other imp units is in and of itself all the evidence we need to know that they need a buff. Because clearly, players are not finding them to be useful, even in the scenarios that they are designed to fit.
You can use simulations or run the numbers to try and figure out exactly what buff they need, but that they need a buff is clear.
Personally, I would like to see them changed rather significantly. I donāt believe the often-stated buffs like increasing accuracy or decreasing cost are right. What Iād like to see is them ignore armour. This would make sense in terms of flavour (gunpowder weapons being very effective against armour that hasnāt been specifically designed to counter them) and would also make them much more effective against a wide range of units.
Decreasing training time would also help, since one of the primary advantages of gunpowder weapons was that they are easier to pick up and use than bows.
Yes, and thatās also one of the reasons why I donāt wnat them armor piercing, rather actually have less raw power. As high quality armor was indeed quite effective against gunpowder weapons in the early days.
Gunpowder actually partwise lead to better armor devolpment (together with better melee weaponry ofc).
