Advanced Tactics 3: Prediction Modelling

Upon agreement from (now former) Order of The Emperor members, we have decided to release all of the guides, expansions & other helpful threads that we used inside the clan. Garde throughout the past year or so has gone threw the forums and brought public (mostly) information together to one thread that makes it easy for someone to learn what the thread is talking about.

These guides are endorsed officially by the Supporters Team (est. 8/8/2020) as they are made by the community as a whole and includes different points of view on the game- all while covering one basic topic to allow the player to easily read and understand what they came to the thread to prey their eyes on.

Do be mindful that these guides may still be updated over time and have sections put on them ahead of time- Also be mindful of other peoples' views.

Advanced Tactics 3: Prediction Modelling was made by Garde.

---

One of the key elements of winning in competitive AtWar and elsewhere is the ability to accurately predict where your opponent will move next, and how you can capitalize off of that information. In this guide I will break down my thought process for doing so and teach you how to predict where your opponent could go, where he should go, how you can rig the scene to funnel him somewhere specific, and how you can remain unpredictable yourself. The information is mostly esoteric in nature however, ultimately speaking you must have a grasp of how your opponent plays, what strategy they're playing, and what your holdings look like in their point of view. Without these things, prediction is futile. At the end of the thread, I'll go over a lot of different situations where you have to predict something in order to determine the outcome as well, just to give you some ideas and insight into the process.

One should note however, that ultimately you nor I are Nostradamus; Your objective in correctly predicting your opponent's moves should be to game him into making them more than anything else, otherwise you're doing nothing other than gambling.

I. Prerequisites to Prediction:

As I stated before, you need to be able to identify how your opponent plays- or at the very least have a guess after their expansion- as well as what strategy they're playing, first and foremost. If you've played against your opponent numerous times before or watched them play, you'll start to retain their playstyle and develop tactics to combat them naturally. I played against Kaska's Ukraine for a long time when learning how to play Turkey, so I've always prepared myself for T1 rushes there by Doublewalling and rushing only a few countries at a time, for example. But being flexible is key here as well, as not all players will function the same way in this regard. I like to memorize 2-3 different expansions at a time so I myself can remain more unpredictable to people who watch me play or play against me, and this will hold true for everyone else you play against, so stay on your guard.

In regards to determining your opponent's strategy, there are a few factors to think about: I. What pick is your opponent playing? What picks are the rest of his team, as well as your own? II. What strategy would you chose if these were the picks you were stacked against? III. After T1, where did he expand to? How did he get there? What units form a plurality in his stacks? These are the things to think about when figuring out someone's strategy, but be warned: Some players will purposely dupe you here by spamming units not commonly used in said strategy that still have their merit to throw you off. A classic of mine is using Marines and Militia early for a conservative expansion as DS Ukraine so my opponent prepares for a GW-style long game, only to weaken himself to stacks of Helicopters shortly after.

You yourself as well have to take a look at where you've expanded to, what walls you have, and what strategy your opponent thinks you're playing in order to really determine what actions your opponent will make going forward. I good analyst will always manage the variables he himself creates on the battlefield in order to determine the appropriate solutions his opponent will come up with. Occam's razor is extremely important here as well: The simplest solution is typically the correct one. If you've pressured Ukraine as Turkey and placed 2 Sea Transports in Varna and Instanbul respectively in preparation for a T3/4 rush on Kiev, what do you think he's going to Turnblock? If you've left your cities open for Raiding, how wil he get there? If you've walled Romania and Bulgaria but not Serbia, where will he rush?

II. Laws of Prediction:

As I stated previously, Occam's razor is the primary concept you need to grasp in order to accurately determine and even- later on- force him into certain moves. Besides this though, there are a few other aspects I adhere to when predicting someone else: Is he low on cash/reinforcements? Where does my opponent have to expand to in order to secure more funds/reinforcements? If he has a plurality of both, how can I best defend myself against his incoming armies and shift the tide back towards him?

Registering your opponent's condition during the game can either be very simple or very tricky depending on the player, pick, strategy, etc. Key factors I always look out for are the types of units they're spamming (seeing a lot of Tanks are a sure sign they've got buckets of gold laying around somewhere), the types of mid-game moves they're making, how many AT's they have, and (if you're playing in East) what cities they control or are vying for.

I asked before "Where does my opponent have to expand to in order to secure more funds/reinforcements?", I would like to also add the following: "Where can I force my opponent to go? The concepts behind Fabian Strategy play a very crucial role here: If you can disrupt your opponent's means of production (reinforcements/funding) as well as beat back his expansion and turnblock everything else, determining where he will proceed to go becomes piecemeal. Fabian tactics in AtWar involve Raiding, Walling, and Wallfucking. These actions funnel your opponent into precarious positions, increasing the odds they will attack/stack/rush/etc certain destinations over others. It's crucial you funnel your opponent in such a manner that benefits yourself unknowing to him; Most of the details about these sorts of Fabian maneuvers are better summarized as a concept called Pressure.

If your opponent has money and reinforcements, you've entered a defensive state and have to break this trend to shift the tide back in your favor. At some point I'll do an entire writeup about this topic as it interests the hell out of me, but to summarize it: You have three keys to winning, with one key differing depending on if you're West, East, or Third: I. Secure wealth; II. Secure Reinforcements; III (West). Win the war of annihilation either by outplaying or outstacking your Western adversary; III (East). win the war of attrition by outplaying or outstacking your Easter adversary, or conversely bide your time against him and focus West if need be; III (Third). Decide who your prey will be early on and seek to disrupt their ability to play by any means necessary, as well as beating back opposing forces in synchronicity with your allies. Past this and back to Prediction Modelling, Tideshifting plays a key role in determining objectively where your enemy should go to destroy you quickly; Those are your predictions, and you must commit to them and stack the hell out of them, or use some alternative method to deny your opponent any luxery.

I know what I said is confusing, perhaps I'll rewrite it in the future when I can more concisely explain these sorts of things. But if you're asking yourself right now "But how do I predict where he will go?", you've missed the entire point of my dialectic viewpoint on the matter: You're not actually predicting anything, you're reading your opponent and basing your actions off of objective truths you believe are sound in so far as how he will conduct his maneuverings and attacks towards you and/or others.

III. Maintaining Unpredictability:

Having a varied tactically skillset, knowing at least three different expansions per strategy for each pick, and feinting your opponent.

My understanding of this:

Watch their other games/duels
-where do they tend to expand to?
-how did they get there?

When Playing them
-how did they get there and WHAT did they get there? (what units they sent)
-what do the units indicate?
--transports means rush (not always but u can tell when)
--tank stacks = lots of money
--marines = money OR stealth friendly strat (mos, gw?)

After Playing them (or playing again)
-WHY did he go there?
--high income?
--high reinf?
--strategic positioning?

-How can I MAKE him go there? (well not really but you know)
--where would he WANT to go?
--how can I deny him certain movements?
--what does he want? what DOESN'T he want?

Practicing
-learn multiple effective expansions
-memorize that George bush chart abt taking neutral cities
-think about what HE would think
--how I make my methods of expansion obscure?
--how can I make my intentions obscure? (they can see what and how much is in my cities but still can't tell what I'm up to)
---where do I place or HOW do I place units to be obscure, yet in range?
--what do they think I'm up to? what do they remember me doing? what will they do with that information?


If I got anything wrong lmk, I j made this cuz I suck at reading so I j made notes. hope this helps

Kirjutas ss_uniform, 12.09.2020 at 20:50

My understanding of this:

Watch their other games/duels
-where do they tend to expand to?
-how did they get there?

When Playing them
-how did they get there and WHAT did they get there? (what units they sent)
-what do the units indicate?
--transports means rush (not always but u can tell when)
--tank stacks = lots of money
--marines = money OR stealth friendly strat (mos, gw?)

After Playing them (or playing again)
-WHY did he go there?
--high income?
--high reinf?
--strategic positioning?

-How can I MAKE him go there? (well not really but you know)
--where would he WANT to go?
--how can I deny him certain movements?
--what does he want? what DOESN'T he want?

Practicing
-learn multiple effective expansions
-memorize that George bush chart abt taking neutral cities
-think about what HE would think
--how I make my methods of expansion obscure?
--how can I make my intentions obscure? (they can see what and how much is in my cities but still can't tell what I'm up to)
---where do I place or HOW do I place units to be obscure, yet in range?
--what do they think I'm up to? what do they remember me doing? what will they do with that information?


If I got anything wrong lmk, I j made this cuz I suck at reading so I j made notes. hope this helps

You don't need to memorize any chart. Just send Militia+1 unless the city has >5 Militias inside or your opponent is going there.

So THAT'S what 'expansions' means. Was very confuzzled for a little bit.

Is there a good reason to use specific Expansion turn patterns, or is that just simplification?

Kirjutas TriusMalarky, 30.10.2020 at 08:45

So THAT'S what 'expansions' means. Was very confuzzled for a little bit.

Is there a good reason to use specific Expansion turn patterns, or is that just simplification?

No, this is all meaningless bs from noobs.

You can pretty much be a top player just by knowing one solid first turn expansion.
You should expand depending on how your opponent plays and test what works and what doesn't.
Memorizing stuff like this is useless. It's all about adapting to different situations.

Trying to predict your opponents moves is useless as well. Its only worth it when its painfully obvious.
Just play safe and don't take too many risks lol.

Most 1v1 games in the east are decided by how fast you expand to wealthy countries while defending your cap.
West games are decided by the same stuff, but there are more rushes (and its boring).