There’s a mistake in the characterisation, though. KotD 4 featured a lot of initially 1TC play by players during Castle Age. That’s absolutely not the same as all-in 1TC play. Players would try to add TCs later. Part of the old problem was that starting Castle Age 1TC set you so far behind in economy that, as you say, adding a second TC would not be strategically viable, so you were forced into a 1TC all-in. I’m going to say that for my money, this kind of automatic “am I first up? yes/no” decision wasn’t actually a sign of meaningful strategic diversity.
1 Like
Now we get a actually interesting in-depth conversation here.
I also agree that “all-in” and greedy plays don’t have to be that kind of “rigorous” divisor of strategy. Fact, it historically was and we now witness some met change in the current kotd there.
For me personally it looks like a deviation from the meta we saw already before - when an “agressive” player gets some kind of lead it was common to add TCs behind the agression (this was the “safest” way to snwówball the lead as an overextension in military is a potential risk, you want to diversify your lead to another dimension so it is harder for the opponent to come back.
IMO that’s exactly what we see in current kotd that because of the “agressive nature” of the map the one who thinks he has a lead tries to add TCs behind. And this btw forces the one with a disadvantage to do the same cause otherwise even if he pushes back he would be so far behind economically that he can’t have a comeback.
That’s how I have perceived it so far. But it’s clear that this is only my personal observation and i am probably biased by the metaplay I was used to see. It can totally be that we are in the process of developing a totally new mata with different strategems.
I have seen some interesting discussions about causality in this thread.
It is possible to “infer” causation with observational data. These criteria are commonly used in epidemiologic studies with non-experimental data.
Within those criteria, the most important one is temporality (the cause must be prior to the consequence). But discussion about plausible mechanisms may also be considered to infer causality.
For example, we can know for sure two things that are prior to winning: the civ selection and the player name/skill, because the match has not started by the moment a player pick their civ.
Thus, with more data available, we could run models that takes in account different things in order to predict the probability of victory:
- Player skill level
- Opponent skill level
- Player civ
- Opponent civ
- Feudal time
- Aperture used by each player
- Kill death ratio or score at 10 minutes
- Castle age time
This way, we could predict the probability of victory according to castle age uptime while controlling for the effects of prior confounders like player skill, civ selection and who was winning before castle age time.
There might still be some residual confusion explained by variables outside our model, but this model should be enough to provide enough confidence about the causation effects of castle age up time.
But we would need a better database for this.
5 Likes
Yeah as long as you don’t have experimental data causation can be infered. Absolutely agree.
But the Bradford Hill criteria just make it very unlikely that there is no other causation. It doesn’ proof causation. They just call it causation, but it actually isn’t. It’s just a very high likelyhood.
In general, medicine institutions use some vocabular very lazy. Like “clinical evidence” or whatever. And I can understand that cause would you believe a doctor who tells you: “Well, we tested this medicament and it is very likely to help against your desease…” .
So medicine institutions are actually somewhat forced by the society to “fib” a bit with their vocabulary.
One Idea, I think we will surely find a lot of different factors more. And then we would need to make a multi-factorial test. But as I said because of the high noise and double-dependencies it is actually very unlikely we would get any usable result from that.
Totally agree, much better and much, much bigger.
1 Like
That is the reason I use causal inference instead of proof of causality. We may asume that there is a causation, although it is never 100% sure.
You don’t need a regression, just reasoning. The main reason to be up to Castle Age faster in KotD4 is that you played better than your opponent in Feudal Age. This is because players are taking larger economic losses and making larger military investments than in the old walled-closed-Arabia meta, so your timing is more affected by gameplay and less by how well you can wall your map for a Drush-FC, broadly speaking. Furthermore, this superiority means you’re very likely to be a better player throughout the game.
1 Like
I agree wuíth that. It’s what I observed aswell.
And that’s actually problematic. Going up to the next age should be a strategical decision not a matter of snwoballing a lead.
At the same time I think the powerspike itself is als too strong. Castle age just gives too much value for a single aging up.
If you compare it to Imp that often is made just because of the access to trebs it just enables so many more tools.
I actually would even prefer if it was the other way around: The player behind should try to get up earlier for a potential comeback play. This way games would have much more variety as this scenario would lead to more diverse situational gameplay, depending on how this greed wents for the player “behind”.
What would you change to reduce the Castle Age powerspike? Move some Castle units/techs to Feudal? Or push them out to Imperial?
Before the last changes to arabia I proposed to just increase the research time by 20 secs. This would have made feudal agression more revarding as the feudal player would have more time to get value from his units.
But with the current design of Arabia this wouldn’t work anymore, the map is already too agressive and this change would only benefit the player who has a lead at this stage.
Other ideas were to give more “sieging” options to feudal, like maa dealing more bonus damage to buildings. I don’t think adding siege to feudal directly would work as it is so high of an investment at this stage of the game. (You also need a lot of army to protect your siege)
But I also want to again speak about KOTD. I just watched DauT vs Villese 1st game. And it was again way to easy to predict the outcome. DauT was way too greedy in adding his additional TCs. This was the perfect invitation for Villese to pressure him and add his own TCs behind his agression.
I think it’s really bad for the game if you can tell “ok that game is over, bad strategic decision” like on minute 20. In this case the game was gladfully over just 10 minutes later but I already saw a lot of comparable games where it then took like 20 minutes or more until the game was done.
What I don’t understand is why the pros make these clear tactical mistakes you can already see not working, but it’s really annoying to watch games you know are decided drag out like that.
I want to see variety - and I also want to play various games, that’s why I prefered playing on Arabia cause the games were the most diverse. It’s really bad the map has now be pushed to be so agressive. It makes gameplay way too repetitive. I am bored by kotd already.
Meanwhile we continue to see players sit back and play defense and win.
Yes let’s make men at arms opener even more powerful.
You read the game wrong. We see a lot of stalemate situation where nobody wants to overextend and possibly lose. And basically everytime somebody tries to be greedy he is punished heavily for it.
The current situation forces both players to mirror the army power of the opponent at least otherwise it becomes a snowball game.
And that’s exactly what I warned for. If there is a clear strategic preference both players have similar win conditions which leads to that kind of repetitive gameplay. And basically no comeback potential. How many real “comebacks” we have seen in this tourney?
The only player who actually mixed it up a bit was Viper, who “invites” the opponent to overextend by not walling. It’s a very ballsi play only viper can really go for because of his insane fancy damage control skills. The opponent at some point needs to go in aswell otherwise Viper could slowly snowball the eco lead he got by not having walled.
But that’s a play only Viper can pull of (Maybe Hera could do it aswell if he can play the map the same way viper does).
It became only powerful because of the recent changes, if we would revert arabia to the state it was at we proposed the change to maa it would have been perfectly fine.
With the current situation of Arabia it would actually produce more inreresting games if walls were completely disabled. Players have the same win conditions already, there is no benefit in anything that buys any time cause there are no long-term investments for greedyness to pay of anymore.
If this is the case walls are only unnecessarily dragging games to take longer than they need to be.
If your complaint is that DauT lacks your macro judgement, there’s nothing any map or patch can do for you to make the game enjoyable.
I liked the balance we had like 2 patches ago. Arabia favored agressive play, but it allowed for all strategies to be viable to some extend.
We already see it in kotd: Hera is out cause he was too greedy with his eco. Is that what we want to see? Whoever is greedy is out?
Or do we want the best players to win?
Well you can say Hera has to blame himself that he didn’t adapbted like Viper after he was almost out in one set already. But be honest, would you be able to adjust your whole game approach that fast like Viper did? I think most players can’t adapt like this.
I don’t want to reduce Vinch performance. He played a very strong series. But Hera just made too greedy strategical decisions and was punished for it. Heavily.
That’s a common motive I observed throughout the tourney. In almost all seriese who is too greedy gets punished, basically regardless of whom he is playing against. And that’s bad. The game thrives by greed vs agression play. It’s the most basic yet important strategic decision and if you can’t get away with greed anymore we will just see snowball games all the time cause both players have the same win conditions. It’s that easy.
I wonder why even Hera hasn’t learned that yet that in this arabia doesn’t you can’t be greedy with eco anymore, but I think to be a pro you can’t be that analytic also, you need to make fast decisions and hera didn’t had the experience with the new conditions yet.
Hopefully we will see devs reacting to this new meta development soon, otherwise I see no future for arabia anymore. The most versatile map everybody could agree on playing has become strategically restrictive like all the other maps, too.
Sad.
We’re watching different games. Hera was significantly earlier to Castle Age, now you are complaining that the famous “Castle Age power spike” wasn’t enough for him to win and it needed to be bigger so he could play a “greedy” game. I think the concepts are extremely confused at this point. Pros are good but they make game-by-game mistakes like the rest of us.
3 Likes
As I already told, military advantage is even more important at the current design of the map. Hera never could threaten VInch as he was too greedy with his eco. We seen a lot of this games already in the tourney aswell when the one with the faster uptime didn’t opted to use the military powerspike but was too greedy with is eco (and besides that happening frequently faster castle age still has insane influence in winrate). I agree that this makes it convoluted, but even with that insane amount of “too greedy” faster castles the castle age powerspike still has an insane influence on winrate.
If you would remove all the games where the player with the faster castle age opts for adding eco instead of using the tech advantage on his units we most likely would see a 75% winrate or even higher.
Both factors are too strong atm, military advantage to idle the opponent eco and the castle age powerspike. They both need to be reduced heavily for diverse gameplay.
(before the last patches only the castel age powerspike was too big, but with the recent changes they restricted arabia strategy even more (unnecesarily))
And as I said in the beginning it won’t even lead to the “intended” changes. We see the same amount of walling every time. Instead of removing walls the only thing we achieved is reducing gameplay diversity which inevitably leads to repetitive meta gameplay and snowbally matches.
As I also prognosed even before the recent changes. Nerfing walls and making maps less “wallable” actually don’t lead to less walling. It only leads to heavier snowballs and less comeback potential.
Hera losing because of super greed is funny poetic justice. He’s a big opponent against walling into economy strategies. And when he finally gets the walling nerfs he’s been waiting for, he tries for a non-standard greedy play anyway, then gets clapped because of the same nerfs he wanted 
. Just goes to show listening to pros for game design advice isn’t always the best idea.
3 Likes
Well not every set was epic but at least half of them created such great games and close sets that imo was better than anything we saw in the last year.
The openess of the map creates a lot of messy situations which leads to a lot off meta play. In contrast a more closed arabia may lead to more 3tc vs 1tc but usually these games are always the same. Aggression either works rather early or gg. This tourney has already produced so many back and forth games where both players respectively were in the offensive and defensive position. That is less one dimensional than the same old boom vs boom or boom vs one tc meta.
Oh and by the way kotd ara also has quite some generations that are super closed like mr yo went for naked fc with khmer (btw such a great set, if you are bored by that what more to discuss? ). Map gens are pretty consistent within one single game but not across different games. That’s how it should be.
2 Likes
I’m pretty sure that with how Hera played no amount of wall un-nerfing would have saved him.
So apparently in this tournament some civs were used not even once, some were not even drafted once.
Burmese, Portuguese, Spanish and Turks were not even drafted.
Slavs, Teutons and Bohemians were drafted at least once but never used.
Italians, Koreans, Malay, Persians, Poles and Slavs were used just ONCE and all but Koreans and Malay lost.
Does it seem that civs are fairly balanced for high level play on Arabia, which is the most played map ever, almost TEN times more played than the second most played which is Arena (which stands at less than 7% play rate)?
To me clearly doesn’t seem so. Some civs are in DESPERATE need of some buffs to even compete.
Meanwhile we have the top banned civs that are Aztecs, Mayans and Chinese, what a surprise!
Trust the pros’ experience they said, alright, here is the result.
The problem with buffing those civs is you’d absolutely break them on other maps. Like arena
2 Likes