I made a reddit-post to take a look on AoE's state of the art, analysed all openings and follow-ups of the recent kotd stage game-by-game. Would like to hear your thoughts, especially from the top players.
I found a quite common narrative in the community, that the meta of AoE would be boring, that (pro-)games were all the time the same, often times referring to Archer-plays. More generally, many claim that the meta of AoE is figured out or claim to know what the meta is. I was under the impression though that the meta of AoE is not really figured out and that it varies a bit with every civ-matchup and dependend on the map, which contributes in my eyes a lot to the beauty of the game. I think that you have to make "new" decisions every game and you end up very rarely in situations where you just mindlessly execute a known pattern.
To get a clearer picture of what's the meta and how stalemate and figured out it actually is, I took an in-depth look on the KotD4 quarter-finals from last weekend. I noted every opening, every Feudal-Age-follow-up and then also the Castle Age approaches of the players.
Results
All games of the King of the Desert 4 quarterfinals.
Explanation of the used colours.
Findings
Conclusions
- In general, there are many different options. In 21 games, we had basically 8 different openings while none of these appeared more than 10 times (= 24%). Also, that covers almost everything that theoretically can happen. The things we didn't see are a straight-up Stable-Range-opening (not a strat bc unaffordable), a TC-Douche, a no-army Tower-Rush (but we did see towers) and Eagle-openings. Mass-militia or continuing M@A-production are not a thing either.
- The most popular opening was actually something that is kinda new and wasn't meta at all before the tournament: 2-Militia-Drush was pretty dominant and got a lots of wins (lost only in the very uneven matchup between Lierrey and Sitaux + when the player was behind in Dark Age due to losing starting ressources). That is a big indication for a lively meta, which is not figured out yet.
- There is the obvious and expected tendency that Barracks dominate the openings and Ranges dominate the rest of Feudal Age. Still, they are far from being without alternative: Barracks-openings happened 21 times, so it was exactly 50/50 (with 2 out of 21 being a Spear-opening to counter Scouts). Ranges in Feudal Age occured 33 times, which is a solid 79% (with 10% being Meso-civs which obviously have no Stable-option). Stables occured 14 times (37% when available).
- There were lots of variations though how the ranged units were used:
- 9 times (21%) players opened straight with ranged units (6 times with Archers, 1 Skirms, 2 mixed, 1 time double range)
- 5 times a second range was added (12%), so that 28 times it was a one-range-play (67%)
- 12 times Archers and Skirms were mixed
- 6 times it was purely/mostly Skirms
- 15 times it was purely/mostly Archers.
- The clearest pattern was that there were hardly any situations of players investing really heavily into Feudal Age with something like triple range or Archers-Scouts or forward towers and such. There were two kinda wild games between Vinch and Yo and then Sitaux tried to make something happen with Feudal pressure against a superior player, but even he didn't commit to 2 ranges even for a single time.
- The best players seem to approach Feudal Age more defensively than others. Viper and Lierrey opened only three times with Men ar Arms or Militia which is 33% of the time. Jordan did it twice in seven games = 29% of the time. The rest of the players did it 14 times on 26 occasions = 54% of the time, with Sitaux losing his games while doing it 100% of the time. Lierrey, the master of the Archers, only built Archers in one of his four Feudal Ages (other players: 68% of the time).
- Except that, there seems to be no clear pattern about what opening is more and which is less succesful. All options seem close to 50/50 when it comes to winning the game in the end. (Data is bad here though with the majority of Scout-openings being mirrored.)
- Unusual strategies were remarkably succesful, another indication for a meta which is not stalemate:
- Fast Castle is still a viable option; it was played three times, every time with a different opening before (2-Militia-Drush by Tatoh, 3-Militia-Drush by Lierrey, walls with no army by Yo). While played only 7% of the time, it won all three games.
- Early Castle drop into unique unit was also just played once (by Vinchester with Camel Archers) - but won the game.
- Scout opening wasn't particular popular and was played 9 out of 11 times only when the opponent did play Scouts too or was expected to do so. Still, it lead to the 2 fastest wins of all the games (Viper Feudal Age win vs Villese; Lierrey gg-ing Sitaux on reaching Castle Age). Maybe the civ-matchups were a bit unlucky to create useful data here.
- In Castle Age, the variety of strategies is actually kinda mindblowing on first sight. There is rarely any approach which occured more than a single time. That only holds when looking at details like adding TCs, Siege, Ballistics, etc. When just looking on the main-unit, you mostly had Xbows and Knights/Eagles, sometimes Skirms. Not many UUs, Longswords were only used once as an Eagle-counter. Cav Archers were probably unlucky due to matchups.
- Kinda disappointing is also the impression that there still wasn't too much decisional freedom for the players, because the approach often seemed dictated by whatever happened in the Feudal Age. Often, both players went into quite similar approaches and the player with an advantage just had the better eco behind it, being able to invest into TCs earlier.
- Many players tried to go for the Xbow-powerspike, sometimes even when they did not really play heavily into Archers in Feudal Age.
Some reflection about the logic behind the observations and some thoughts about what could be changed in the game to create even more variety:
My summary here would be that the AoE meta is very complex and not really figured out and is in a spot where it permanently forces the players to gather informations and adjust their plans and decisions to that. Correctly anticipating the value of military investments and the chances of proper damage control seem to be a key aspect in that decision making process. Still, there are a few things which are stronger than others and have a high priority: Ranges in Feudal Age, Castle Age research and early Xbows, adding Town Centers (as long as you keep a competitive army).
- One thing I was wondering myself is, how good Scouts actually do when you consider that they tend to lose against a Men-at-arms-opening and also struggle against a Spear-Archer opening. On most occasions, they were played as Scouts-v-Scouts which creates the impression that they are a kind of silent agreement between the players to focus on economy and do without M@A; this seems genuinely bad, because both players seem to have an option to get an advantage here and won't use it for comfort. On the other hand, shouldn't a Scout opening do pretty well against the 2-Militia-Drush? I figure this option might actually lead to a even more complex meta with the classic Archer -> M@A -> Scout (-> Archer) triangle being basically mirrored with 2-Drush turning the order around (winning against straight Archers and M@A, but potentially losing against Scouts).
- I believe that the success of the 2-Militia-Drush is also a bit due to the players being not so much used to play against it. Multiple times, they were caught off guard by not small-walling their woodlines for example which seems much rarer against Scouts and M@A. That points toward a general thing about meta(-development): New strategies have the edge of opponents being badly prepared for it and unexperienced with it. It can create mistakes even when it's a little worse in theory. (Something we see more often in more diverse tournaments like Hidden Cup.)
- Reaching Castle Age seems just to be a little bit too valuable overall. I think the tales of walling being meta or Archers being meta or anything can be very much just substituted by: Reaching Castle Age earlier is meta. The decision making logic in Feudal Age seems very much: How can I get up earlier without taking damage. And not much of "how can I do damage". With your Castle Age timing advantage, you usually do more damage than any Feudal army can do and then this is the exact moment where you would invest into TCs, so that the advantage amplifies immediately. Market usage and selling stone to click up becoming more popular clearly underlines this aspect. Few ideas how to make this factor a little bit more balanced and less defining:
- Make the Castle Age research more expensive or a bit slower.
- Make Xbow/Bodkin Arrow more expensive or slower researched. Maybe do separate upgrades for Archer range and attack.
- Create an option to upgrade a specific tower in Feudal Age (as in AoE4), this would buff tower defense while tower rush still has the risk of the towers being beaten down.
- Give the option (for some civs?) to build Siege Workshops (and maybe Monasteries) in Feudal Age, just without the option to build anything in it. Maybe a tech in the Blacksmith that allows for that?
- Barracks unit fall off heavily after their initial power spike, Feudal Eagles are not a thing which makes the meso civs more boring than they needed to be. Also adding Scouts to your Archer army or making upgrades for Scouts is rare.
- All of that can be adjusted, without any risk really, by making the Feudal Age melee upgrades cheaper. The infantry armor should be definitely cheaper in my opinion, for Forging and the cav armor there can be made a case. If you add the saved ressources as an additional cost for the castle age upgrades, you would be super safe to not have any unwanted balance-snowballs later on.
- Another approach might be to just make Fletching and/or the Archer armor more costly. Fletching is an insanely good upgrade for its price. (I still think it would be an absolutely great idea to have range and attack as separate upgrades. It adds to the decision making.)
PS: The Malians vs Byzantines game between Jordan and Tatoh demonstrated very well that flexible, decision making based Civs are just a better idea than one-strength-powerhouses.