M'Hunters WarLight Strategy Guide

About This Guide

The goal of this guide is to formalize sound competitive WarLight / Warzone play. It covers 1v1 and team games but not FFA. Initially, this document was just a collection of forum posts by me in the M'Hunters clan internal forum, and this document is mainly a result of copying and pasting them to make them more accessible.

I believe that the basic building blocks of sound WarLight play can get formalized in a way that you can just turn your brain off, play by those simple rules, and get quite some results. I had to put some thought into how to formalize good WarLight play as I wrote bots for the game. In some way, those bots are just a collection of best practices stitched together, and the sum of those best practices leads them to play WarLight in a relatively sound way.

I have also made a YouTube video guiding you through the main topics of the guide: Strategy Guide Video

Efficient Expansion

Prerequisite: 0% luck straight round game, standard settings

Goal: Find out about good and bad amounts of armies to attack the opponent.

The formula for your kills is: "Attacking armies" * 0.6. The result is then rounded to the closest natural number with .5 rounding up.

Let's assume you attack with X attacking armies:

So in conclusion:

If you attack the opponent, for example with 4, you won't kill more armies than when attacking with 3. You might find a better use for that extra army. The problem is also that if the opponent deploys heavily there, you are putting 4 armies at risk instead of only 3.

If the opponent has more than 1 army on a territory, you can attack him with 1. That way you both lose exactly 1 army.

Expansion Move Order

Prerequisite: 0% luck straight round game, standard settings

Goal: Find out in what move order to perform your expansion steps.

Often when you take a territory, you assume that the opponent might also try to take the same territory in the same turn. If both you and your opponent use the same amount of armies to take the territory, it depends on the move order who gets it.

Case 1: Both you and your opponent attack the territory with a 3v2 attack.

Case 2: Both you and your opponent attack the territory with a 4v2 attack.

In Case 1, the player who moves first takes the territory, loses 1 army and has 2 armies remaining there. Then the player who moves second attacks those two armies with a 3v2 attack and takes the territory.

In Case 2, the player who moves first takes the territory, loses 1 army, and has 3 armies remaining there. Then the player who moves second attacks those 3 armies but only manages to kill 2 armies, so he won't take the territory.

So to sum things up:

Attacks With 1

Prerequisite: 0% luck straight round game, standard settings

Goal: Find out about the advantages and disadvantages of attacking a big opponent stack with 1 army.

If you read the previous rules, then attacking a big stack with 1 makes both you and your opponent lose exactly 1 army. So what are the advantages/disadvantages?

Disadvantage 1: The opponent stack might move away before your attack with 1. In this case, you will lose your attacking army, and since the opponent only has 1 army there, they will lose nothing.

Disadvantage 2: It might get more likely for the opponent to break your territory. Let's assume the opponent has 5 armies there initially (so they can attack you with 4). You have 3 armies there. That way, they can't kill your 3 armies. But if you first attack him with 1, then you only have 2 remaining armies while they have 4 left (so can attack you with 3). 3 attacking armies are enough to kill 2.

Advantage 1: You get an extra move to delay a crucial attack of yours.

Advantage 2: You are disturbing a precise opponent calculation.

Advantage 3: You might kill more armies if the opponent attacks you. Let's assume the opponent has a huge stack bordering a stack of 2 armies of yours. If you don't attack with 1, the opponent will only lose 1 army while you lose 2. However, if you first attack him with 1, then both you and your opponent will lose 2 armies.

Avoiding Random Picks

Goal: Find out how many picks you need to avoid random spots.

If both you and your opponent (or a teammate) go for the same pick, obviously only one of you can get it. Because you might lose your main picks, you need backup picks to avoid random spots. This raises a couple of questions. One of them is how many picks you need in total.

The easiest formula is:

"Amount of needed picks" = "Amount of starting spots per player" * "Amount of players in game"

This formula is true for all game types (1v1, team games, FFA).

However, in team games, you might decide to pick other spots than your teammates. In that case, you need fewer backup picks, and the formula is then:

"Amount of needed picks" = "Amount of starting spots per player" * "Amount of players in game" - "Amount of your teammates" * "Amount of starting spots per player"

This formula is true since your teammates won't steal your picks in this scenario.

Examples:

Possible Pick Combinations

Prerequisite: 1v1 game

Goal: Find out about the possible combinations of your picks with which you might end up.

The pick order works just as the move order. There is an option called "no luck cycle" where the player who commits faster gets their first pick. However, usually, it's random. Let's assume Player 1 goes first. Also, assume that there are 3 starting spots for each player. The picks are then allocated in what I call "rounds" while each round the players alternate.

After round 3, both players have their 3 starting spots. As I point out in Lesson: Avoiding random picks, both players need 6 picks when there are 3 starting spots. However, you can't end up with every combination of your picks. If you ended up with your picks 4, 5, and 6, this would mean that your opponent stole your picks 1, 2, and 3. Let us analyze this:

Round 1: The opponent goes first and steals your pick 1. Then it's your turn, and you get your next highest-ranked pick, which is 2. This means that the opponent can't steal both your 1 and 2.

Possible combinations for 1v1 and 3 spots (Autogame / Ladder settings):

1st / 2nd hereby means that you moved first or second during the picking stage.

There is some interesting information about these combinations:

For the possible pick combination in case of an arbitrary amount of starting spots, see Possible Pick Combinations

Board Dependent Moves

Goal: Finding out how to play "Transfer only," "Attack only," "Attack by percentage," and "Marking armies as used"

All those mentioned moves have one thing in common: you use them if you don't know how the board will look when the move gets executed (or not executed).

If you use Attack-or-Transfer only, then your current play is that you look at the board and play accordingly. This is also how the AI plays: it always attacks when "AI armies after deployment" >= "Your stack at beginning of turn" * 2. However, you shouldn’t follow the AI here. Instead, you should have a model in your mind about how the board will look like when your moves get executed. If this is your approach, then you will find many circumstances where you don't want to go with the default Attack-or-Transfer.

Transfer Only

You mainly want to use Transfer only when you are afraid of running into a bigger opponent stack. Let's say 11 territories are all bordering each other. The opponent holds 1 territory with 100 armies, and you hold 10 with 10 armies each. This is a bad situation for you, so you want to join your stacks in region A. However, if the opponent attacks A first turn full force, then you will crush 9 times into their stack in A, giving you 9 bad fights.

Attack Only

Attack only mainly comes into play when you are playing defense. Let's say the opponent has a stack that is bordering your territories A and B. A is part of your bonus, while B isn't, but B has multiple borders to your bonus. You don't know whether they will hit your bonus directly or flank it via B. In that case, you can deploy in A and use a late order Attack only from A to B in case your opponent went for the flanking option.

Attack By Percentage

Attack by percentage comes into play when you don't know how many armies you will have when your move gets executed. This is only the case when your opponent attacks your territory. In all other cases, you can achieve the same with a regular attack. Attacking by percentage is useful when you want to hold some armies back, for example, to blockade the territory.

Let's say you have 11 armies in region A. You want to attack region B with 5 armies and use the remaining 6 armies to blockade A at the end of the round. Before you move from A to B, the opponent attacks A and kills a total of 4 armies.

Case 1 - You don't use attack by percentage:

After the opponent's attack, you have 7 armies left. You move 5 and end up blockading A with 2 armies.

Case 2 - You use attack by percentage:

After the opponent's attack, you have 7 armies left. You move 50% (= 3 armies) and blockade with 4 armies.

Conclusion: You want to use attack by percentage if you care more about how many armies remain than with how many armies you attack. If the opponent attacks you, then attacking by percentage has the effect that more armies remain while your own attack gets smaller.

Attack By Percentage For conditional Delays

You want to use attack by percentage for conditional delays if you find yourself in the following situation:

You are planning on hitting a stack of your opponent with your own stack. If the opponent hits your stack first, you want to hit him later with your whole remaining stack. However, if the opponent does not hit your stack first, you want to cram in some additional delays by moving armies of 1 away from your stack.

The solution here is to move those armies of 1 via attack by percentage, setting the percentage so that the move barely rounds to ≥ 0.5. Only moves that move more than ≥ 0.5 armies get executed, while the other moves round to 0. An attack by percentage that rounds to 0 neither gets executed nor marks any armies as used, so you can use all your remaining armies for your big attack.

As an example, let's look at turn 9 of the following game: https://www.warzone.com/MultiPlayer?GameID=13572606

The dark blue player here almost managed to wipe teal from the east and thus most likely win the game. However, he failed with teal having only 1 army left. Not knowing how many delays were exactly needed, the dark blue player added many unnecessary delays far outdelaying teal. If he had used attack by percentage for those delay moves, he would have been able to hit teal stronger.

Mark Armies As Used

Let's look at the same example as before and see if we can find a better way to blockade with exactly 6 armies (prio A) while attacking region B with 5 armies (prio B).

Step 1: Mark 5 armies in territory A as used.

Step 2: Attack territory B with 5 armies. If the opponent has attacked us before, we won't attack him with armies that have already been marked as used.

Those steps explain the "what," but not the "how" to mark armies as used. The way to do this is by using them for a move that can't get executed. There are two ways to do this:

Despite such moves not getting executed, the WarLight game mechanics forbid using the same armies multiple times, so those moves mark the armies as used.

Income / Costs Ratio

Goal: Finding out about efficient bonuses.

When checking the WarLight forums, you will find the terms "Income / neutrals ratio" or "Income / territory ratio" to describe the efficiency of a bonus. Both ratios help to give a quick overview, but I'd rather think in terms of cost for the following reasons:

Income / territory ratio:

Income / neutrals ratio:

I recommend seeing the cost of taking a bonus as "Amount of territories" + "Neutral kills" in case of 1 army must stand guard and as Cost = Neutral kills in case 0 armies must stand guard.

Deployment Order

Goal: Finding out in which order to deploy your armies.

When playing, you want to have as much information as possible about your opponent while giving him as little information as possible. An often neglected aspect of information you are exposing is the order in which you deploy your armies.

This is how the game mechanics work: The way the deployment orders get executed works the same as the move orders get executed, and the player who gets first move order also gets first deployment order.

There are two possible ways to deploy your armies:

The disadvantage of Option 2 is:

By default, Option 1 is the way to go, so you want to first deploy next to your opponent and afterward in the background. Also, you want to watch closely what your opponent reveals with their deployment order.

Here is an example about intelligence your opponent might give you when deploying as in Option 2:

You are playing a Strategic 1v1 game and are asking yourself if your opponent took South America or Southeast Asia. They did take South America, deployed 8 at your border, 1 in the background, and moved first:

  1. Opponent deploys 1 in the background.
  2. You deploy something to a spot.
  3. You deploy something to another spot.
  4. Opponent deploys 8 at your border.

→ You know that the opponent has more than 8 income, so they probably took South America.

Aggressive Late Order Play

Goal: Play more aggressively when you know you get last order.

Let's assume the game has opened up and both you and your opponent are bordering multiple important spots of each other. You don't know where your opponent will focus on. In such a situation, you should play with very aggressive last order attacks.

Let's assume you are bordering the opponent in your territory A and the opponent's territory B (amongst probably many other spots). You want to break B but are afraid to run into a bigger stack than yours, and the opponent might also attack you from B. Let's assume you have 24 armies (attack with 23) and the opponent has 28 armies (attack with 27).

Case 1: First your attack happens, and then the opponent attacks.

Initial:

Situation after your attack:

Situation after opponent attack:

Case 2: First the opponent attacks, and then your attack happens.

Initial:

Situation after opponent attack:

Situation after your attack:

So the opponent only kills 6 armies more and won't take the territory.

As you see, running into a bigger opponent stack is still a bad thing, even if you get last order. However, it's not as bad as when you don't get last order.

Do Not Expand By Default

Goal: When you are having a couple of leftovers, you might want to use them for an "unplanned" expansion step that you might not have done without those armies. However, this is very often not a good decision.

Let's say you have 3 leftovers in a territory and can use them to just randomly attack a neutral territory in a bonus that you might get in a couple of turns. This is usually a bad decision, and you have to ask yourself a couple of questions before doing so:

Fighting > Expanding

Prerequisite: 2 neutral game

Goal: Often you have the option to either gain your own bonus via expansion or to break an opponent's bonus. The latter is better.

Let's assume you have the option to either take a +3 bonus or to break a +3 bonus of your opponent.

There are some benefits of breaking your opponent instead of expanding:

Offense > Defense

Goal: You want to be on the offense, not on the defense.

With offense/defense, I mean whether you are threatening your opponent (offense) or the opponent is threatening you (defense). There are 3 possible combinations:

One of the most important parts of my gameplay is that I always try to be on the offense instead of on the defense. Of course, there is also the aspect of out-predicting your opponent. However, this only works when there are flaws in your opponent's play on which you can build your own strategy. It's best to never get into a situation where you need to out-predict your opponent.

There are many problems with being on the defense (and vice versa, benefits of being on the offense):

There are a couple of rules that you can follow to ensure you are on the offense and not on the defense:

Play For The Team

Prerequisite: Team game

Goal: You win as a team and lose as a team. There is probably no correlation between playing strength and your income position at the end of the game compared to your teammates.

I guess the biggest reason players go for a cluster-pick on the Rise of Rome map isn't to get a mega-bonus faster but to avoid running into multiple opponents from the start. If you are really going for a team victory, this is a terrible reason to cluster-pick.

The formula for when it is a good thing to run into multiple opponents is:

"Damage that you are doing" > "Damage that the opponents are doing to you"

"Damage" here is an abstract concept that involves all kinds of factors like:

If you are just looking at the first two factors (Broken bonuses and Killed armies), running into two opponents is usually a bad thing. However, there are such huge other benefits to running into multiple opponents that (if you pick and play correctly) it is usually a very good thing. You should be capable of being on the offense for a couple of turns, or at least not just defending. The worst mistake in such a situation is to expand.

Look at the following game with me playing with random teammates: https://www.warlight.net/MultiPlayer?GameID=8113103

Since I saw the rather nonsensical Partha cluster-pick and I know that "Partha only" loses to "Partha + Asia only," I decided to pick Alexandria and hit Asia as soon as possible. I also ran into an opponent in Macedonia. At first glance, I didn't do that much damage to the player in Asia, and me fighting him also would have led to a long-term defeat against the player in Macedonia. However, me showing up in Asia was enough to give the Partha player (who would have lost otherwise) a very easy win against the Asia player. A little help from me shifted the game towards a team victory.

Analyze

Goal: You want to analyze the game while it is going on and also after it has already finished.

Here is some advice:

During the game, you have to be fully aware of what is going on:

Use your calculator to keep calculating stuff. For example, you have about a 50% chance for three 80% attacks to all succeed. Also, if you know your opponent's income, you can just deploy the minimum amount of armies to avoid getting broken.

Also, after the game you have to analyze:

In the following video I show how to remove the fog in 1v1 games: WarLight - how to remove the fog.

First Turn Bonus Play

Goal: You took a first turn bonus. Therefore, you had to use 2-3 picks (Autogame settings). What now?

First of all, putting all picks into one area is problematic, as I point out in Lesson: Offense > Defense.

You going for the first turn bonus (by probably putting all 3 spots into one area) has some advantages:

The disadvantages are, however:

Those advantages and disadvantages should make it quite obvious that there is only one sound way to play a first turn bonus:

  1. Step 1: Take the first turn bonus.
  2. Step 2: Kill your opponent with your initial momentum.

So, if you manage to play "Ultra offense" with your first turn bonus instead of "defense," you don't violate my advice "Offense > Defense."

Working Armies

Goal: Armies are like money. You want to work with them instead of putting them under your pillow.

The term "working armies" needs a definition. I define working armies as armies that you are putting to use this turn, more precisely:

However:

As a rule, you want each and every army to work. Only in well-defined exceptions might you consider playing with non-working armies.

Possible exceptions:

The part I'm mostly interested in is "Possible exception 2" since I often see gameplay that looks like this exception but is, in fact, rather nonsensical. This can have various reasons:

Relative Safety

Goal: Finding out about safety and how to make your income safe.

Mostly there isn't any "real" safety for you. There is fog, and theoretically, the opponent can be almost everywhere. However, practically, there is some relative safety, and during the picking stage, you can build your own safety zones.

With relative safety, you assume that your opponent didn't go for a badass counter-pick but picked normally.

Look at the board of the following game between Renegade and UniVersal: https://www.warlight.net/MultiPlayer?GameID=8065702

If you got a spot in Indonesia after picks, then there was no way for your opponent to be close. Indonesia was safe on its own.

Scandinavia isn't safe on its own. If the opponent sits in West Russia, they can attack you. However, you can make Scandinavia safe by also picking West Russia. If you got West Russia, then you know you can safely expand in Scandinavia, and if you haven't, then you know that you aren't safe at all.

Relative safety is especially important during the following stages:

The "Norman Way" Of Thinking About Picks

Goal: This advice aims at giving you an easily comprehensible explanation about how the creator of this guide evaluates a picking outcome.

I introduced the following 4 terms to explain to my clanmates what I think about their picking outcomes:

Advantageous neighbor: Each bonus A has some neighboring bonuses X, Y, Z, ... X is advantageous over A if, when one player has A and their opponent has X, then the player having X has some sort of advantage. This means mostly one of the following:

Advantageous distant neighbors: You want to extend the advantageous neighbor definition to bonuses that have bonus coverage over your bonus (and vice versa, of course).

Regarding bonus coverage and map coverage, I found the 2 steps to reach the opponent bonus as quite reasonable for defining bonus coverage when it comes to 1v1 ladder settings.

Generally speaking, you want to spread out your picks in order to get full map coverage, so there is no opponent bonus out of reach.

The following graphic helps you to understand bonus coverage:

Bonus Coverage

The ‘x’ markers represent your 2 pick outcomes, and the surrounding circles show which bonuses are covered by those picks. Since the wasteland in Scandinavian Peninsula blocks your path towards West Russia, the pick in Greenland does not cover West Russia. Also, I accepted here the West African pick covering Central America; however, then again, you have to pass a +4 on the way and a wastelanded South America, so it’s pretty much a “meh” coverage.

As you can see from the circles, we have bonus coverage over all reasonable opponent bonuses, so we have full map coverage.

The following graphic shows you the same picking outcomes from the hope play perspective:

Hope Play

Why Attackers Should Flank And Defenders Should Defend Their Bonus

Goal: I cover a very difficult topic here; however, even if you don’t understand everything, it should enhance your understanding of optimal WarLight play.

I asked my clanmates in our clan forum for a solution to the following riddle:

The Riddle

Riddle

The players are named “Red” and “Green” and own the territories according to the colors of the graphic.

If Red manages to break the bonus C within 2 turns, he wins. Else Green wins.

For simplicity, we are assuming no split mode. Also, Green has enough income to defend 1 territory and not 2. No fancy moves allowed, just 1 badaboom hit, either into a wall or into an unprotected territory.

We are only interested in turn 1:

Both Red and Green are WarLight bots. Their default behavior is to play in an immaculate way. However, if they realize a flaw in their opponent's algorithm, they exploit it. Both bots are capable of reading the source code of the other bot.

I have 2 questions:

  1. What is Green's immaculate decision of which Red can’t take advantage?
  2. What is Red's immaculate decision of which Green can’t take advantage?

As for the answer, I expect something like the following:

Green has to defend B1 10% of the time and C1 90% of the time.

Red has to attack B1 100% of the time and C1 0% of the time.

The percentages above are just random numbers adding up to 100%. Your task is to find the correct numbers.

The Solution

Breaking C1 has double the value of breaking B1 since when you break B1, you still have a 50% chance of messing it up. That’s why we have to solve 2 * B1 = C1, which boils down to a 33% to 66% ratio.

Green has to defend B1 33% of the time and C1 66% of the time.

Red has to attack B1 66% of the time and C1 33% of the time.

If both players play optimally, the attacker has a 33% chance of breaking the bonus.

Now let’s make some examples where one of the bots plays imperfectly and the other bot abuses him:

General Takeaways

As for myself, I find the results extremely fascinating. The actual numbers of this simplified example don’t apply to WarLight, but we can extract some general rules which do apply. With great caution according to your particular game, you can (to some extent) extract the following rules:

And more generally: The defender usually wins in a symmetric situation, and the attacker in an asymmetric situation. By this I mean that the defender wins if both players choose the same territory to focus on, while the attacker wins if both players choose different territories.

If you find yourself winning in a symmetric situation, you want to prefer the option which gives you more value, and if you find yourself winning in an asymmetric situation, you want to prefer the option which gives you less value.

Dirty Gifting

Prerequisite: Team game, gift card enabled

Goal: Finding out about gifting stuff to your opponent

By default, "Attack teammates" is set to false. Therefore, when an opponent attacks a territory that you are gifting the same turn to one of their teammates, instead of an attack, a transfer order will happen and the other opponent will get those armies.

There are some common examples when this is beneficial:

Appending: Game Mechanics

This section explains some of the different game mechanics in a concise manner. The goal isn’t to give a complete overview but mainly to explain some mechanics which even experienced players often don’t know exactly about.

This guide uses the following conventions:

Move Orders

There are 3 types of move order settings:

With all of these settings, the move order cycles each round.

Example no luck cycle, team game with the following picking speeds:

During picks the cycle is then:

First turn then the cycle swaps, so player B2 moves first and player B1 last. Since the player moving last in round 1 always moves first in round 2, they get 2 successive moves, which can be beneficial.

Move Orders With Order Priority And Order Delay

Both the order priority and order delay cards affect the round in which moves get executed.

Playing order priority cards introduces the new rounds O1, O2,...On, depending on the number of order priority cards played. O1 is executed before the regular round 1, O2 before O1, and so on. Each played order priority card in a certain turn pushes the first order one step higher into the O rounds. Since the first order no longer exists in the regular rounds, all other orders move 1 round up. However, when playing multiple order priority cards, they don’t move up any further. The cycle in the O rounds is the same as the cycle in the regular first round.

Let’s say, for example, that player A is moving before player B in the cycle. If both player A and player B play an order priority card, then the first order of player A gets executed before the first order of player B. However, if player B truly wants the first order, they can play 2 order priority cards.

Playing order delay cards introduces the new… let’s call them “sequence of rounds.” Let’s call the standard round sequence without any order delay cards involved the “standard sequence,” and each order delay card then adds the “OD sequence n” for the nth played order delay card. The semantics are now that first all moves in the standard sequence get executed, then comes the OD sequence 1, then the OD sequence 2, and so on.

Let’s say that both player A and player B have a critical attack which they want to get performed after all opponent moves.

Armies Per Turn

The extra armies per territory setting adjusts your base income depending on the number of territories held. It’s independent of income from other sources like bonuses. The formula is as follows:

‘Adjusted base income’ = Max(‘base income’,’extra income from territories’)

The extra income from territories is rounded down. For example, if you own 11 territories and the settings are extra armies for every 2 territories, then every even number of territories held gives you extra income, and your 11th territory doesn’t give you any income. As the formula states, you get either your base income or the extra income from territories, depending on whichever is higher.

The army cap setting reduces your total income if your stacks get too big compared to your initial income. Let’s consider the following example:

With this income, you have an army cap of 2*10 = 20, which means that the maximum amount of income you can receive is for you to have 20 armies in total on the board. Since you already have 15 armies on board, your income is reduced to 5. Reinforcement cards are not affected by the army cap, and the minimum amount of income you can receive is 0.

Luck Settings

WarLight offers two different dimensions when it comes to luck:

Luck now works in the way that first the luck modifier gets applied to the number of kills, and afterwards the rounding luck gets applied to the result.

The straight round rounding mode causes the number of kills to get rounded to the nearest number, with .5 getting rounded up. For example, if you expand 3v2 with 0% luck, you kill a total of 3*0.6 = 1.8 armies, which then gets rounded to 2.

The weighted random rounding mode interprets the decimals as a chance in percentage to get rounded either up or down. For example, with the mentioned 1.8 kills, the .8 is interpreted as an 80% chance to get rounded to 2 and thus a 20% chance to get rounded to 1.

The luck modifier is more complicated than the rounding luck. The formula for the kills is:

Kills = (SR * L) + (EXP * (100%-L))

Definitions:

The following examples should help to illustrate the formula: