Aow3 RTS

Introduction and some personal experiences

I played RTS very a little in my life. Maybe some 100 total hours when I was under 20 (pre 2006). Unless one considers hearts of iron and variations RTS, that I liked a lot.

My main problem, that is actually silly, is that I couldn't stand the fact that a tank could be taken down by one infantry unit without bazooka. I did miss the abstraction that the tank was just a heavier unit. Later games with counters (see hearts of iron) improved my appreciation for abstraction.

Fast forward after tens of years, I don't play much. Therefore I bet on android as a viable economical gaming platform. Android has plenty of games and more are coming. Devices are really little monsters. If a pentium 3 could run hearts of iron 2 or 3, so can a android device since 2013. The only problem is: the hardware is there but the games need time to be developed.

Now I liked FPS on pc, see ql:root or also return to castle wolfenstein, call of duty, warsow, teeworlds and so on. Then I tuned down my ambitions but I wanted to play with a controller, so I discovered War thunder and Battlefield heroes that could be enjoyed with a gamepad and xpadder. See gamepad-mouseaim.

The nvidia shield k1 as android gaming platform was fitting as it has a controller. See also https://old.reddit.com/r/pireThoughts/wiki/pier4r_oc/2016/09/android_console_nvidia_shield and https://old.reddit.com/r/pireThoughts/wiki/pier4r_oc/2016/09/android_console_shooters . Not all FPS supported well the controller but some are ok and potentially really long lasting. See dz:root.

Anyway while FPS are really intensive, I slowly moved away from them. I'd like to have games of thinking like the hearts of iron, as I had in the past before discovering Return to castle wolfenstein and all related FPS.

There is the "tower defense" genre that is not bad on android. See TD bloons. Then there are other tinkering games like bad piggies. Then there are several attempts to have games a la hearts of iron back, like strategy and tactics or games from the publisher Herocraft (at least that was the name in 2014-2016). Anyway thanks to this search the google play store suggested me gladiabots. That's a huge game. Now gladiabots is great but due to its greatness is also a time sink and the older I get the less time I have.

Therefore I moved again trying to find a game where I can think "solving many small problems" but it is not a time sink nor too intensive like an FPS, aside from short burst maybe. I rediscovered chess. I always liked chess but I played some 300-400 games in my life without caring about the real tournaments and chess theory, therefore I started to watch around. Playing a little but mostly consuming enjoyable content. Chess is a very simplified wargames but still plenty challenging and interesting.

Instead of playing chess online, that I did tens years ago but now I don't crave for it as I want to have my pace, I combined my appreciation for calculators and chess using Tichess 4.17e on the ti-89 titanium. I'll write about it, hopefully, but suffices to say that is plenty enjoyable for my level.

Nonetheless it still lacked something. I was searching for a wargame - as chess, hearts of iron, close combat and others are in the family of wargames - that was offering me some thinking, some strategy, in 30-60 minutes of good games and that's it.

There I rediscovered the RTS. Actually it was from reading in a chess channel about alphastar. From there on youtube checking startcraft tournament matches (on youtube there is a lot of neat content if one has the keywords) and then saying "Ah yes RTS, I'd like to try them again". Then I searched on the playstore and I started to try all RTS with enough downloads. As enough downloads are hopefully a sign of a possible active online community. (another way is to check how many youtube videos there are about the game). And there I found art of war3 that was released in 2016, like gladiabots, and it is pretty alive today in 2019 with very neat game challenges.
(I wish I would know all this before, but that is what experience and reflection over it is for. Moreover before the possibilities were missing. That is, the infrastructure and games were not there)

The game is pretty simple in its mechanics, for example one manages enegy to get resources, resources to build military buildings, packing management to get most buildings there, then army units. Nonetheless the variety of challenges are plenty even with this simplified abstraction.

The multiplayer section is pretty alive and so far the matchmaking (after 20 games) is providing balanced and enjoyable challenges most of the time, even if one doesn't buy "improvements". I like proper pay to win as it takes away cheaters. Cheaters normally ruins the experience for the others, while pay to win offers soft legal cheats that avoid that people get more extreme cheats (dunno like: unlimited resources from the start) while also financing the game so the publisher keeps improving it.

The game is full of small challenges that either one picks from guides or one solve by himself (the more fun way, as copying solutions is not that fun). I want to share those observations if I can.

Oh and of course, why wargames? Because the examples of confrontations provides plenty of challenges that are difficult to translate in more neutral contexts (TD bloons does it a bit though). Also it seems that human are interested in war. I for example enjoy a lot reading about ww2 history, especially the organizational and logistical parts. Military production and technical challenges faced at that time. For example the 1500 kw aircraft engine problems in Germany (search on wikipedia for bomber B). Also designing a interesting wargame is not that easy, it has quite a lot of factors.

So playing an RTS combines those interests. In an RTS one has to have a good economy to overcome the opponent late in the game, still one has to care about principles that are valid also in chess, the game provides small but intensive logistical problems that can be improved with analysis later. Like with chess or gladiabots.

Last but not least, there are a lot of small but interesting math problems (even to figure out how much one won in a tournament) that keeps oneself active. And the upgrades, as they are timed, can be used as timer while one does something else! Pretty neat.

Additional resources

  • Many players, especially from russian and vietnam, upload replays on youtube that are interesting to see what plans one may employ or counter. A search string on youtube could be "art of war 3", then from the videos check the channels. There is also a youtube thema that automatically tries to collect videos (through which one can find channels). Then there are plenty of other channels that do not mention directly hte "art of war 3" string. One can find them logging in the official discord of aow3 and then asking for resources (or checking the video section)
  • The most active social network account outside Russia: search for "art of war 3" on facebook. https://www.facebook.com/aow3rts/ official social media channel as of 2020-08-30 (with lots of changelogs if one scrolls in it)
  • Then, as one gets deeper in the community, one can find resources that are also in other languages, like: http://aow3-dp.wikidot.com/ (a wiki in German)
  • A wiki with quite useful information, here: https://sites.google.com/view/artofwar3wiki/home

Explaining the game shortly 2019 09

Many players that enjoyed Gladiabots can enjoy RTS as well. This because in gladiabots one has to learn the general patterns of what happens in the arena (strategy) and also the single engagements (tactics) and then code this in a battle plan that will be executed during the battle without interactive changes.

The "coding" part is a very time consuming part, as it is coding or writing non trivial frameworks. Every time one has to fight with little bugs to bring the bots to behave as one wishes.

In a RTS there is no coding part. There is the game plan devised before the game through experience and reflection (as in other strategy games or problems) and then there is the execution (tactics). The execution has to be trained too but it is normally done in a real time way with the hands giving orders to the units and buildings.

Thus those Gladiabots players that do not have time to code their battle plan (me) can enjoy the RTS where one can play every now and then as the battle plan can "grow" in the player's mind between games. (In gladiabots one has to code it, otherwise the ideas do not get implemented out of thin air)

Thus I explained the game to one gladiabots player from Brazil (but living in Europe) that was dominating in Summer 2017 and Spring 2018.

Here is what I wrote.

Pier A, [19.09.19 09:07]
I think you can like it, if you ever like RTS (clash royale is a sort of RTS too indeed, but aow3 has more things to do in game)

Pier A, [19.09.19 12:20]
So, let's make a summary because I think you will like the game. Or at least, I did underestimate RTS for long time, but they are really appealing if one finds the right game and has little time.

Pier A, [19.09.19 12:22]
In practice in the RTS you have buildings and units with different strengths/weakness production times and output (for buildings).

Each unit and building has upgrades. Those costs virtual currency (from here on: credits). Credits can be either purchased, or collected slowly if one is logged in in the game (every day there are free rewards), collected a bit faster if one sees ADS (ads can be completely disabled though, it is great), or collected even faster if one plays and has a good win rate.

Pier A, [19.09.19 12:23]
The number of upgrades is massive (like 40-50 upgrades for unit or building) and they are limited by rank. Thus if a player as a couple of upgrades more than another, the difference is not that big.

Pier A, [19.09.19 12:24]
Ranks are practically a proxy for experience (although not completely, as one could intentionally sandbag. There are few players that do this). This because every multiplayer game costs credits, thus it discourages throwing away games.

Pier A, [19.09.19 12:25]
Normally each game costs a certain amount of credits, and in case of victory one gets 5 times the amount paid. Thus the idea is: you cannot play too bad, otherwise you lose credits.

Pier A, [19.09.19 12:26]
What determine ranks are stars. For each win one gets 5 stars (at the very high ranks, I guess it goes down to 3 stars, I am not sure). For each loss one gets -1 star.

games lasts maximum 1h.

Pier A, [19.09.19 12:27]
There are 4 main online modes of playing. 2 players against the AI (not easy at all, but quite repetitive). Tournament (1vs1), world war (1vs1 and 2vs2)

Pier A, [19.09.19 12:28]
The devs aren't fools thus they limited the AI, tournament and 2vs2 mode. Those can be played each hour more or less and not everytime. The only mode that can be played everytime is world war 1vs1

Pier A, [19.09.19 12:30]
Also the devs aren't fools so there is a matchmaking in place. For world war 1vs1 they say "we try to match players that have a similar rank and win rate". After 70+ games I can say that it is so. The higher the win rate the more you wait for the pairing in hope that some good opponent shows up. Of course it depends on the playerbase.

Anyway after a minute or so of waiting the matchmaking becomes less strict and enlarges the range. The most popolous ranks are up rank 15, thus until then one doesn't really wait a lot.

Pier A, [19.09.19 12:30]
Max 90 seconds.

Pier A, [19.09.19 12:31]
I am currently (2019-09-19) at 75% win rate and it is not rare to be paired with those that have 90%+ win rate. Thus it means that the matchmaker didn't find anything for them and relaxed the range to pick more players.

Pier A, [19.09.19 12:33]
Note about "I want to compete with the best". Those with 80%+ win rate for me are the best, and when I get paired with them I cannot say that they win because they paid. Rather they have also good plans.

Pier A, [19.09.19 12:33]
So far I never felt like "oh ok, the guy sucks but won because he paid". I always felt I had a chance or I messed up.

Pier A, [19.09.19 12:34]
For 2vs2 the matchmaking is different, as likely there are less pairs that are in the queue waiting. The queue opens 5 minutes before the full hour, and then the matchmaking tries to make the best pairings, but of course with less people playing and skewed pairs, the results are much worse than the 1vs1.

Pier A, [19.09.19 12:36]
For tournaments, the players are divided in leagues according to their rank. Rank 7-9 are in the lowest league. Rank 10-12 in league 4. Rank 13-15 in league 3. Rank 16-18 in league2 and rank 19 an above (around 5000 accounts only, inclusive the abandoned ones) league 1

Pier A, [19.09.19 12:38]
for tournaments one need to have at least a 40% win rate and have played enough world war games (the rule is: one needs to play at least so many world war games as tournament games). In this way the devs avoid too much sandbagging (to progress through ranks one needs at least a 21% win rate) and avoid that people play only tournament.

Also each player can play 30 times in a tournament in 2 weeks. Thus everyone has the same chances.

Wo wins the most (and win streaks helps) gets more points.

Pier A, [19.09.19 12:39]
Thus the difference is that in world war 1vs1 the matchmaking can find a balanced opponent. In tournaments the pairing is random with the players in the queue (since a game can be started every full hour). One can get players with more skill/experience or less skill/experience.

Pier A, [19.09.19 12:40]
With my little time I enjoy the world war 1vs1. There are so many maps (more than 30) that I didn't play all of them enough yet.

Pier A, [19.09.19 12:41]
I guess to play all the maps enough times (3/4 times per spawn point) one has to play around 400 games.

Pier A, [19.09.19 12:41]
Maps are not symmetric. The balance comes that one gets a random spawnpoint of those available in the map, the more one plays. Thus one learns to make the most of it every time.

Pier A, [19.09.19 12:46]
Then there are boosts. Boosts are "boosts" of performance for a limited amount of time during the game. They costs credit or they are collected via daily/playing rewards.

The blue boosts improve some attributes by 10%, the red ones by 20% the gold by 30%.

But the gold costs 35 times a blue one and the red costs 5 times a blue one. Thus red and gold are practically "sucker taxes" as they are not economically profitable. A gold boost has only 18% more effect than a blue one.

In tournaments 1vs1 and world war 2vs2 often players use the 30% boosts due to the prestige associated to victories in those formats (in 2vs2 you don't want to be less than your ally. In tournaments you have a leaderboard). Say between two players with similar skills and upgrades, the boosts may make a difference. Otherwise no matter how much boosts, in a poor play one is going to lose.

Pier A, [19.09.19 12:48]
In world war 1vs1 boosts are very rarely used, max blue. Only the suckers use them (but then they burn their credits quickly, and thus quit).

Blue boosts costs almost nothing. With one victory one can buy 20 blue boosts or more (each boosts is almost a minute, so one can cover 20 minutes).

Pier A, [19.09.19 12:50]
In a game the opening is crucial. As at the start of the game the map is sprinkled with resource containers (with resources one can build units/buildings).

Thus being slow means losing the initiative unless the other player sleeps.

then after the opening it is a matter of proper army composition and timing of attack/defence , decoy and baiting

Pier A, [19.09.19 12:50]
I would really suggest that you just try it. I think for problem solvers such games are perfect.

Pier A, [19.09.19 12:50]
of course unless one has time, then gladiabots is better.

Pier A, [22.09.19 21:07]
I don't think so. On a tablet with a mouse I can substitute my finger (that is fat), but the keyboard brings nothing.

On a pc you can have a couple of macro but for very minimal actions (as you don't have a program that can understand what is going on on the map).

In general who can use 2 fingers as in: one to direct the movement another to control mini map, production and boosts has the edge and some experienced players can play 2 fingers. I can't.

Theoretically with a finger I would be even more precise than my mouse (I don't have to pan the finger on the mousepad to reach a point on the screen), but I am so poor that a mouse is better

Pier A, [22.09.19 21:07]
So far I can tell that execution decided maybe 2 or 3 of my games out of circa 80

Pier A, [22.09.19 21:08]
it is mostly: place your buildings properly and timely, explore and pick the resources, have game plans for each map and major situations (against combination of enemy armies)

Pier A, [22.09.19 21:08]
anyway what I can suggest is do 20/25 missions and then go 1vs1

Pier A, [22.09.19 21:09]
at first many players are clueless, only few are hard but normally those already have experience on maps and co. Already knowing the maps it will take a bit.

Pier A, [22.09.19 21:11]
if you start as "I want to be the best from day 1" it would be an overshoot. After you get enough experience on units , hq 5, maps (some 200 games) then you can expect to do better.

Pier A, [22.09.19 21:13]
but always remember that there are people sandbagging or restarting thus they can have plenty of experience and that alone pays off. You should know from sc2

Pier A, [22.09.19 22:18]
no. Buy the starterpack (1 euro on sale). it is enough credit (unless you are unpatient and click to speed up the upgrades).

Then just do upgrades - the go nowhere and they are done while you play/work - and play without stress (aka take it slowly) and accumulate experience.

Pier A, [22.09.19 22:18]
I reached mission 20 and I played 1vs1 online (world war). Then after a while I needed anti air because people used airplanes and I had no chances and so on.

Pier A, [22.09.19 22:18]
make progress in campaing when needed.

Pier A, [22.09.19 22:19]
again at first you need to unlock units and accumulate experience so it is all good. The real thing starts around rank 13.

Pier A, [22.09.19 22:20]
then one has little excuses.

Oh, and don't use boosts to accumulate stars, stars are a proxy of "how much you play" but in every rank range there are enough tough opponents.

Pier A, [22.09.19 22:20]
for boosts to accumulate stars I mean: premium (where you can get 10 stars per win rather than 5)

Pier A, [22.09.19 22:21]
the more stars you collect without enough experience, the harder and frustrating it gets.

In game strategical or tactical observations (They may be valid for RTS in general)

Sub page here.

Math related problems

Sub page here.

Other observations

When doing stats using the in game stats leaderboards, identifying the faction quickly.

One can notice that the rank symbols are colored too. They have always a blue background for the blues and a non blue (normally red and green) for the reds.

Main rankings in the leaderboard, all players, even inactive ones.

I mean those where players have 20 plus ranks. I asked today 2019-03-11. They show all players, whether they are active or not. Therefore also the way to get stars should be consistent over time otherwise it is falsed.

2019-03-22 status of the majority of top players by country

Going through the global leaderboard as well as the top teams global leaderboard I noticed the following.
The russian speaking countries (Russia, Ukraine, Moldova, the countries in Asia that were part of the Soviet Union, The countries around the Caucasus, etc..), in which the game was available since spring 2016, have the most players over rank 19.

Then follows the south east asian countries. Vietnam, Indonesia, Malasia, Philippines, Laos, etc.. Those have the game since early 2017, although Vietnam likely got it in 2016 like with the Russian speaking countries. Those countries have quite some rank 19 and few players over this threshold.

Then comes the rest of the world, that had the game from spring 2018. Currently a lot of countries, in particular: Poland, Germany, Italy, France, Spain and latin countries, US/UK (not many players from there), India, Pakistan, Iran, South Korea, Japan, Turkey, China (although China may have got the game together with south east Asia) and others have the majority of top players around the rank 17-18. We start to see slowly more rank 19 but more than that is difficult as climbing ranks costs a lot of time, resources, commitment and sometimes is not wanted (as the higher the rank, the harder the players. I'll try to explain this later).

Update 2019.04.27.

On the 2019.04.30 the game will be 3 years old (in terms of: accessible to the public, although not everywhere from the start).
Moreover I noticed that one index that can be a proxy to tell for how long a player was active is the tournaments battle counts. Since each tournament lasts more or less 3 weeks and allows 30 games to be played, tournament games are relatively scarce. Approximating one tournament per month, a year has 360 tournament games so players with 900 or more tournament games played since the start of the public release.

And indeed there are few players like that from russian speaking countries (with 1000 or more tournament games). Then there is a block of players with around 700 to 800 tournaments, and those indeed are from Vietnam mostly. Then come players from indonesia, Thailand, Myanmar, Laos, China and the like with some 700 tournament games. Then the rest of the world with 300-500 tournament games.

Therefore this confirms that the game was released in 3 regions at different times. Likely 2016 in Russian speaking countries, late 2016 or early 2017 in south east Asia, then 2018 everywhere else.

And indeed in league 1 (rank 19 or higher), that few players reach because it requires a lot of commitment and activity, there are mostly: russians speaking players, vietnamese (that are quite strong), Indonesia (high ranked but not yet successful), few from Myanmar, laos or the like, few from China (I am not sure whether Chinese players actually live in China or they live elsewhere as the game can be blocked in China). From other countries there are very very few. Like some Germans, Italians, Indians, but the mass of other committed players is still around rank 17 or rank 18 (league 2).

Update 2019.09

One way to gauge the activity of players from all over the world is to select the rankings for world war for teams (the mode mostly played, in 1vs1 or 2vs2). In teams one can see which players, in terms of nationality, the team attract. For example the russian speaking teams have often random flags, as likely for a while they competed without competition from other nations. But one can spot them from the team description or from the most common flags of the team members.

Thus the majority of players in the top teams, teams that collected a lot of stars (thus played a lot of games), are from russian speaking countries. Then come the rest. It is healthy when a country has 3-4 top teams (in the top500) that are pretty active.

Let's say in terms of activity we define:

  • healthy: several teams on the top and active
  • almost healthy: (3-5 teams) on the top and active
  • barely healthy (less than 3 teams) on the top active
  • just a bit (1 team or maybe 2)
  • healthy
    • Russian speaking countries
    • South East Asia. Especially Vietnam, then Indonesia, Philippines, Malaysia, Myanmar, Laos, Thailand and the rest. Vietnam has several teams made by vietnamese players, the others have often mixed teams.
  • almost healthy
    • Spanish speaking countries (Spain, and then latin america) are also quite active although as conglomerate.
    • Germany
  • barely healthy
    • Brazil. Often not mixing with spanish speaking players.
    • Iran
    • Pakistan
    • India
    • China
    • Italy
    • English speaking countries have teams as well, but as conglomerate. There is no clear division and together they don't have much more presence as, say, Iran.
    • France speaking countries (not only france, but also african and other countries)
    • Poland
    • Arabic countries as conglomerate. Players from different arabic countries mix in one team.
    • Turkish speaking players (also from places outside Turkey) as conglomerate in teams.
    • Nigeria
  • just a bit
    • Korea
    • Japan
    • Taiwan
    • Israel
    • Rest of Europe

It is interesting to see how the rest of the world develops top teams (that are committed to the game) to balance the russian speaking countries

Update 2019.11

Ok I found a better metric to see the activity in game according to regions (based on teams). As exposed in another section of the info collected regarding aow3 (see the page about math problems or the like) the dominance points collected in total in the world war are an excellent metric.

For this metric I have a somewhat different view compared to what I observed (but not rigorously collected) in 2019.09. Data collected at the end of 2019.10 .

  • In the top200 there are:
  • Teams from russian speaking countries
  • 48 teams from South East Asia (Vietnam, Philippines, Indonesia, etc..)
  • 18 teams from Europe (Excluded Russian speaking countries).
  • 10 Teams that are English speaking (also international or Nigerian based)
  • 10 Teams related to arab countries.
  • 9 teams from Spanish or portoguese countries.
  • 6 Teams from around India (India, Pakistan, Bangladesh, etc..)
  • 5 From East Asia (Taiwan, Korea, China)
  • 2 French speaking based teams outside Europe.
  • 2 Israeli teams.

As I wrote, more info in another section in this wiki.

Update 2020-07-03

An even better metric was available in 2020, introduced in January. The "war points" that are quite a good proxy of the activity of clans, even when people lose but try to survive in game (so they don't lose too quickly), or they complete contracts and so on. In other words, people can be active already "watching ads". In this way one can detect activity that does not appear in the overall of the world war list as the world war list counts mostly points given by wins, while the war points can be collected en masse also by players losing a lot but also playing a lot. Thus it is really more a proxy of activity rather than "who wins more".

Data collected on 2020 04 19. Normally the time to accumulate warpoints is 2 weeks, I collected the data and still 6 days and 11 hours were left. Normally after so many days the data is quite stable.

  • In the top200 there are:
  • Teams from russian speaking countries (93 teams) top 5 positions (1,2,3,4,6)
  • 23 teams from vietnam (the top5 positions were: 5,13,17,19)
  • 18 teams from Indonesia (34, 45, 61, 64, 91)
  • 15 teams from the arabic countries (15, 25 TUR, 30, 52 TUR, 62)
  • 14 teams spanish or portugues speaking (24 BRA, 29, 53, 63, 74)
  • 12 teams from Europe (the top5 positions 1 ger, 2 ita, 3 ger, 4 pol, 5 ita)
  • 10 teams english speaking or international (49 NGA, 56, 68, 89 NGA, 101)
  • 8 teams from South east asia (22 PHI, 57 THAI, 75 THAI, 78 PHI, 81 PHI)
  • 3 teams french speaking (37, 82, 122)
  • 2 teams from india (124, 157)
  • 2 teams from east asia (42 CHN, 188 TWN)

Update 2020-09-27

Some months ago gear games (the developers of the game) licensed a game for the asiatic market (east asia mostly and china) that is very similar, thus it is likely to see less asian players on the aow3 main game. See https://www.youtube.com/watch?v=EG15AYiXU58 .

Dividing players by experience and strength or engaging players with more stats.

When games offer several ways to sort players on ranking (games played, win rate, achievements, tournaments, 2vs2 games and so on) , there is more chance to engage them on a long time as everyone can decide to be better at this or that stats.

In general having the main stats and ranking about "how much victories/experience you collected" is better than the Elo/win rate. Why? Experience and victories always increase, strength wise one eventually has a plateau against the active playerbase. If a player plateau by the main stats, surely he can find his personal challenge but it is harder and more unlikely to happen. Rather people would just quit because "more than this it doesn't get better ". Thus for engagement, having the players grouped by "experience" and within the groups by strength is quite reasonable to keep players playing. As every new players cannot just rank up to the top, but has to play also a bit.

Recruiting in Teams

It seems that the devs of the game are really good. They show active teams only to those that search for teams. Namely teams where the leaders (leader and vice leader) were online lately. Playing doesn't mean online for the game. Online means really: browser the game or engage in chat discussions.

Personal improvements. Things I need to test or train on that are weakpoints.

2019 H1

Changing composition of the army

As wrote in the observations and experience, I often have to much infantry against initial attacks. The infantry costs CP points for needed units for middle or late games (after 30 minutes). When the enemy produces certain units, having too much infantry is detrimental.

Finding a good army composition for the various stage of the game (depending on the opponent's skill)

This also is very important. The higher the rank the opponent is, the more likely certain army configurations (in terms of units) works better than others in certain moments of the game.
This because the opponent is not sitting idle, but likely is preparing his best attack, and that attack may come with different type of units in different moments of the game.

Trying to develop a guideline which army works in which situation is needed to avoid critical encounters. And to switch from one composition to the other one needs to be trained in changing composition of the army.

66 cp (no airplanes, no camouflage/shield) vs well defended base

One needs artillery against artillery.
Helicopters are an easy prey or airplanes. Therefore mobile AA is needed.
The artillery need to be defended by heavy armored units.
Heavy armored units need to be defended from heavy infantry.
The base need to be defended from airplane while missing AA defensive buildings.

Therefore one may try attacking (slowly and in a sustained way) with.
7 mobile AA, 2 in base 5 in attack.
2 artillery (or 3) to move every time.
6 or more heavy units to protect the aa and the aa may defend them from heavy infantry.
the rest in the base to form new armies.
Then moving every time close to the enemy while applying pressure.
In the meantime regrouping the army before the enemy base as soon as losses mounts.

Before the base is very well defended one can try counterattack as exposed above. Counter with the entire army if the enemy just attacked and it attack failed while your CPs are quite filled. His base should be almost empty and the lack of defensive buildings will help. Another thing to try before the base is too well defended is a mass helicopter attack if via ground it takes too much. One can stry at the side of the base not well defended trying to take out HQ or supply centers or powerplants.

It seems an approach very hard to coordinate but one may try it until it is clear that one cannot master it. Losses do not matter much on the long run.

Resistance units have a plus if moved (2019-04-01)

Especially cheap units (aside grenadiers), such as vehicles, can deal quite some damage if moved around (and not against) the enemy units. As the enemy units often cannot shoot on the move and can poorly aim moving units.

2019 H2

Getting a streak of only wins

With my rate of play , since I can barely play world war , reaching two or three times a current win rate of 100% (provided a certain difference in games in between otherwise one just keeps having one streak) would be a satisfying and hard to reach goal.
This objective has to be reached given that I do not use boosts - or if I really use them they need to be only blue (or the lowest level) - as I am too busy managing units or thinking about the next move; and given that the current win rate is based on the last 50 games of whatever type (although with my little time I play only world war, it is enough for me).

Thinking about it, even if I have one time 100% win rate, but I can hold it for 25 games (the half of the games used in the current win rate), then it is also as satisfying as reaching two or three times the 100% win rate.

For this keeping a journal, to record the level of the opponents after the battle and ones level, helps to see against which players one is weak.

What I noticed is that the current win rate, that for lower ranks matches practically the global win rate due to the low amount of game played (today, 2019-08-27, I played around 70 games in PvP, after 7 months in the game), is likely used by the matchmaking at least in 1vs1 and tells a lot. If one matches a player with a similar win rate, the game is going to be enjoyable and likely long.

Update 2019-12-21. The matchmaking is very likely using the actual win rate for opponents. At least world war 1vs1 is really well done. Also to ensure that people are really worth their 70% plus win rate, the matchmaking every now and then pairs players with "weaker" players (according to the win rate) and "harder" players. Otherwise if players would be paired according to their actual win rate only, everyone would have around 50%. Why? Because everyone would play only similarly strong players.

Also the higher one gets, the stronger the players are. Why? Well only players that are very dedicated to the game stay. Thus the higher one get the harder is to hold the previous (high) win rate. For a while I was fluctuating between 74 and 78 for the actual win rate, but now (130 games played 1vs1) I feel that is quite an objective to hold over 70%. Thus the fabled 100% is likely impossible to reach. I will see for some weeks and likely I will reassess my objective to make a challenging but realistic one.

Update 2020-02-14: first time actual win rate of 80%. It is really hard to reach for me (with blue boosts) after an experience of 182 games. 1vs1 is really hard (tournament or world war), 2vs2 is sometimes easy and sometimes hard. Since the devs introduced the 2vs2 (almost) random at the end of Dec 2019 I am playing mostly 2vs2. It is much more fun and still hard. So let's refactor the previous objective of 3 times reaching 100% win rate to 3 times reaching 80% win rate. If I already have 80% win rate (or higher) then I need a streak of 25 games with at least 80% actual win rate to consider the objective achieved. To consider if I reached again the 80% (or more), 10 games (20% of 50) need to pass from the moment I go under 80, otherwise I can easily fluctuate doing 80, 78, 80, 78, 80. Thus if I get 80 or more again in those 10 games, they won't count unless I am keeping 80 or more (that is a streak).

If I reach 80% under those conditions then I can either refactor it or see if I have enough. Surely 100% is unfeasible for me with blue boosts.

For example I am playing mostly the world war mode 1vs1 or 2vs2in confrontation. I can then extend to play also the world war with flags and I can start playing more tournaments.

Update 2020-08-06: With a lot of 2vs2 with good partner I was able to reach once 80, getting as high as 84. I fluctuated once, I got 80 then 78 then again 80. Then I stayed at 80 or more for 30 games (I was never expecting such a streak). Thus I could consider the objective previously stated as achieved. That is, either I reached 3 times 80% winrate, excluding fluctuations, or I got 80% and I kept it for 25 games at least. But I have to say that having a streak - especially with 2vs2 and good partners - may make it somewhat easy to keep over a certain winrate, although high. Anyway, since I reached the objective I can refactor it as stated.

Thus I started to play tournaments, also because I need the superweapon. After rank 13 the amount of people that defends really hard and then destroy you with the superweapon grows - once again the higher one get, the harder the opponents - and to unlock the superweapon one has to play tournament. Tournament is less forgiving than world war battles. World war battles aren't that easy, but every now and then the opponent is weaker than yourself (unless your winrate is between 45 and 55%), in tournaments that happens less.

Tournament pairings - although I am still analysing them - pick mostly players with similar or superior win rate, thus it is the hardest mode against other players so far. Moreover tournaments are different from the only mode I played, confrontation. Each tournament has its own game mode with its nuances and it takes time to master it - while other players may have more experience with it (I started tournaments quite late). And it shows! I played about a dozen tournament games, especially when I was away from the game for weeks or more (I reached my 80%+ winrate streak when I played often. Practices makes the master, thus inactivity hurts), and my winrate in tournaments is barely 50%.

It is true that tournaments are the most expensive gamemode, players tend to throw at it everything (especially costly boosts), but I decided to use only once the gold scouting boost and then, as in world war, continously blue. Well sometimes I really feel it is enough but it is ok so. Experience and I enjoy it nonetheless. Thus my actual win rate is going under 70%.

Therefore factoring tournament play, mostly tournament play (and occasional 2vs2), I again set the objective to 3 times reaching 80% (aside from fluctuations) or reaching 80% once and then keeping it for 25 games without interruption. This will be really hard if my main focus is on tournament (so far the win rate for tournament is like 50%, far away from 80%). I may easy it a bit, instead of asking myself 80% I can go down to 70% (considering that I use mostly blue boosts), we will see. Ah yes, let's also throw in a minimum amount of tournament wins, otherwise I could give up too early.

Therefore 3 times 80% (maybe 70% if it is really hard) or 80% kept for 25 games (70% accordingly), mostly focus on tournaments and at least 100 tournament wins.

Update 2020-08-24: In addition to what I wrote before, I had enough indirect confirmation that for who wants to play among the top (top tournament placements, independently from the league, top winrates from rank 12, top ranks from rank 18) one has to incresingly rely on p2w. In the game, as it is right so, the top performance include top performances between skill and financial performance in game. It seems that from rank 19 (or 20), as long as there will be only a hundred players active in a month with such ranks (and not thousand or more), it is incresingly hard to make progresses - that is, improving one's performances compared to the past - without constant financial investment.

Furthermore since the progress in ranks will take more and more stars (each new rank is 30% more of the priovious rank, thus if one is rank 18, that is 3000 stars, rank 21 would be 6900, or like half (and more) the time it took to reach rank 18. In other words, between rank 18 and rank 21 it takes around the same amount of time than starting again and reach again rank 18. The only difference is that from rank 18 to rank 21 the game is incresingly difficult and one, to improve, has to put the hand in his pocket more often than before. The game between rank 1 and 18 is as difficult as before.

Said that, since I like the game but I don't like it when the p2w part becomes increasingly noticeable; about that the devs did a very good job. When the p2w is quickly noticeable, it is discouraging, instead in aow3 is noticeable only if one wants to be at the top and that is ok so. People think that money alone buy great performances, nope, one has to be good as well. Money alone will help only that much. Anyway amonhg many people that have enough skill, money makes the difference.
Back to the point. I like the game, I can extrapolate from the info that from rank 19 (or league 1) the p2w part becomes increasingly noticeable, thus what one can do to not leave the game is: start with a new account. That is, the old account stays, unused, at rank 19, and it will be as a reference to beat (better stats) with the new account. As the effort to put in after rank 19 is similar to the one from rank 1 to rank 19. In this way one will see its progress in performances across diffferent accounts.

Update 2020-09-27. As I maybe wrote, the difficulty of modes (given the same winrate) in aow3 is the following:

  • tournament 1vs1. In tournaments, aside few people that lose intentionally, people give really their best and are not shy to use costly boosts. I use 1 gold boost (scout) and the rest is blue like in world war.
  • 1vs1 world war at 65%+ win rate . People do not want to lose but they aren't willing to push at max like in tournaments, well maybe aside from league 1 where at the top people give almost always their best.
  • 2vs2 . Those are a bit more volatile as it is difficult to get balanced matchmaking with 2vs2 pairs that can be unbalanced in themselves. Although in 2vs2 people do not want to be less than their ally, so they can push too.
  • AI. Unless one goes for a medal, AI can be quite relaxing.

Given that, and collecting some direct and indirect experiences, the best way to lower one's winrate while still doing one's best is the following:

  • get near to a good actual winrate. For example I reported below that my objective is 80% winrate, thus a good one for me is 70%
  • play tournament relying mostly on blue boost. Tournament is mostly merciless, and one will tend to be matched against people equally strong at least (see also: aow3:root:mathproblems:tournaments and the stats collected there about the matchmaking) . Being equally strong people, the winrate will tend to stabilize around 50%.
  • Once the actual winrate goes down, one can play world war. The matches will be somewhat easier due to the collapsed winrate although not representative one.
  • Once the actual winrate rise above a certain threshold, rinse and repeat.

With somewhat easier matches one can lose the practice against strong people but one may practice to get better in some other points harder to practice against equally good or better players.

Another thing I noticed is that there are different ways to approach the game. One is to tend to be always aggressive, one is to be defensive and then attack via air/superweapon and so on. Blue actually is more fitting for defensive players, red is more fitting for players that prefer attack and movements. I would be actually a good blue player but I will try to adapt. One thing I realized is that my approach often way, and I will try to keep it so: I play without forcing my hand, and maybe losing the game from overextension (trying to do too much compared to what I can). Rather I will try to put constant pressure on the enemy, without overdoing it, and then capitalizing on his mistakes (trying to do less mistakes myself).

Competitive games are mostly about "who does less mistakes" and in human pressure helps creating mistakes. That is: I pressure as much as I can and still being safe, and then I get more and more ahead the more the opponent does mistakes (unless I do more than him). Such approach would be better on blue but I guess it is feasible on red too and I didn't yet see any player - that shares libraries - doing this consistently on the red side. Maybe I didn't watch enough games of some players that share videos or maybe I didn't read their games properly.

2021-02-12: Barely time to play. it is great under such conditions if I keep over 65% wr. Because being rusty and slow in micro or slow recalling plans does not help.

2021-08-22: Barely time to play. I am happy if I will ever get 400 victories against other players. This because for one victory it takes at least 30 min of play on average, thus 400 victories are 200 hours (or maybe much more, considering overhead around the game) and that is plenty enough for the game. The game is good - although the higher one gets the more the p2w factor is felt - and the entertainment it gives is worth it (not necessarily with toxic elements of the community though), but 200 hours (+ a lot more overhead) are enough at first.

2022-03-05: what I wrote in late 2021 stays true, only I noticed that if one wants to slow down, beside solo speedruns, there is also AI that is less intensive (if one tries to win without setting a record time). Further with the war in Ukraine, as the publishers are russian, one shouldn't give them extra money (although it is unsure whether they are pro war, most likely they aren't as it is not good for business).

RTS observations

201905

I checked the popularity of rts . One way to gauge this is the amount of videos in the last year on youtube . As players normally tend to share (especially casual ones).

Aow3 is quite active with some 5-10 videos per day but the most seen video in the last year that is about the multiplayer battles has under 10k views.
I was not able to find any serious multiplayer battle for command and conquer (pc).

Command and conquer rivals has many multiplayer battles with over 20h views if not over 100k.
Age of empires impressed me. The game is still going strong (well, it is a nice game) with world cups and co.

Starcraft and war craft are obviously active . Empire earth is barely active .

I didn't check any other rts. But I would say that among all those I really like aow3. The context the units the buildings the fact that there aren't too many options (but still not a little ). It is really great.
And on a tablet one doesn't even need a pc (although I'm sure some player tries to emulate android on a pc). Plus I realized by some videos that who has very good thumb coordination has the edge as it is like having two mice pointers at once .

Then I also asked on reddit about active and competitive RTS in 2019. Here is the link.

I like when games with good mechanics survive long . As slowly players formalize game actions and that helps in other knowledge branches over time .
See chess for example. I feel that the early chess masters in the 1700s were dominating mostly based on intuition rather than theory . Or better they built their theory. As in aow3 now the best players likely discovered things but there is a lot more subtleties to discover and learn.

Update 2019.06.10. A follow up about RTS formalized principles here. Searching for "real time strategy guide" or "<insert RTS game name> guide" helps a bit.

The theory is missing.

In many RTS, at least in aow3, a lot of small but crucial observations are missing. I do not mean precise information as in "do the action X,Y,Z and then you have an advantage", rather ideas. "Do not build more than this, because it is detrimental", "avoid building there, because it is detrimental", "control chokepoints" and so on. Concepts that cover many critical situations that may lead to a consistent advantage. This is missing. In other games (chess) the theory, instead, is quite developed and preserved over time.

People try to compute precisely, say, the damage done from a unit to another in 1vs1 (where 1vs1 in battle are ultra rare), without going collecting and consolidating general observations about what to care in a battle, which mechanics may bring some advantage and so on (and not mechanics about things that are difficult to control in real time, like this unit against this other unit in a mass battle). Thus every player has to discover those points more or less alone, as fruitful discussions about those points are rare (most discussion are about useless 1vs1 between units).

Therefore the players that analyze their game and videos from others and realize those points are going to accumulate over time more and more advantage, as long as they try to make use of those observations. This means also that strong players need hundreds and thousands of games to realize marginal advantages all by themselves as few share and everyone is rediscovering similar things. It is fun on one side, on the other side we are far from very strong playing. The few that are barely seeing are dominating the blind ones.

RTS observations 201907 . In Starcraft and age of empires 2 is harder to defend

Analysing some other common rts like starcraft. I noticed three things : the economy can be harassed easily . Simply forcing the units to stop collecting resources with an early attack. That could be simulated in aow3 in a way: if your SC gets shot at it doesn't produce (or if it gets damaged and it has only 70% HP, it produces 70% of the supply output).

The defensive buildings are minimal and the production time is not that short . Thus if one loses the majority of his army it is difficult to defend .

Resources are not endless and are location based. They are going to be exhausted quite quickly thus a player cannot just sit and wait indefinitely.

Those three factors in my opinion contribute to the fact games that last over 20 minutes in starcraft are rare.

The same harassment can be done in age of empires 2, but there the defensive buildings helps a bit more thus games are a bit longer.

Update 2019-08-27. Since some weeks it is harder to defend also in aow3. A change in the balance (search balance 2.0 on the aow3 official facebook page) meant that now one cannot spam defensive buildings, so the army is needed to defend the base as well. If the few defensive buildings are destroyed, the base is practically undefended. Still the production of resources is not affected by any partial damage, only total destruction.

2020-09-27 aow3 is great because it is slow

In starcraft 1 and 2 , command and conquer and other competitive RTS games, the game is much more fast paced. Especially moving units across the map and producing them is much quicker than in aow3 where one has to keep in mind the unit production speed and the manouvering speed much more than in the mentioned RTS. Thus aow3 is more appealing for players that wants to play slowly.

One way to play other RTS is to use the slower mode, but competitively that is not possible, the majority of players play on faster mode and thus aow3 once again keeps being more attractive for the quieter players. Although at higher ranks the upgrades can speed up the game quite a bit.

Aow3 can be "slow" thanks to grouping players by experience.

There are several ways to group players for matchmaking purposes. One, for example, would be based on winrate alone (or alternative ways, like Elo and its variants). In that case the game would be much harder. Why? Because players that improve, quite quickly will meet the veterans and they will have an hard wall. Many people would be frustrated by this.

Instead grouping people by stars is like grouping them by experience (although there are people that "smurf" around, having multiple accounts, but they are a minority except in tournaments in the top100), and this means that players with similar experiences play against each other, instead of encountering soon veterans that smash them with no hope. It is a smart move by the dev, as it ensures that for a while people can enjoy the game at their pace, as there are many others at their level.

Only in the legendary league, since very few stick around in the game to play thousands games, there are relatively few active players and for matchmaking needs then the lower ranked in league 1 (ranks 19 to 21) are every now and then paired against the real veterans (rank 22 or more) and have an hard time to get through this time, to become veteran themselves.

Unless otherwise stated, the content of this page is licensed under GNU Free Documentation License.