The 21 tiles rule official topic

[Iluvalar]

Hello guys, this topic is to explain to everyone what is the 21 tile per power fraction magic rule that I found in the game.

In a shell :
The frontier for a given defending position should be 21 tiles wide time the proportion of power the defender have against his attacker.

For a quick exemple a player with 5 oil fighting a player with 10 oil should have frontier of about 10 tiles wide to be able to defend itself (at least for a while if it's a duel... He gonna be doomed because of the research).

Where that come from ?
The mg towers are around 255% stronger (price included) than the viper mg h-track tanks at the very beginning of the game. If we consider that a mg tower can cover about 8 tile wide of passage, we end with the magic number 20.4 tiles. The defending vs attacking units keep that 255% ratio during a long while in the game.

I expect that this is due to the trial and error tuning of the defense buildings that occurred in the map 4c-rush that follow that 21 tile rule in not less than 3 different positions. If you pay attention carefully, you'll discover that most bundled maps have at least one frontier that follow the hidden rule. Would you be surprised to discover that even NoQ's vision map that was recently added have 21 tile per oil ratio frontiers on 3 different spot ?

Why is that important and what happen if I don't follow it ?
That's what create the balance between the structures and the other units. At 21 tile it guaranty that both strategies are viable. If the choke points are wider, the building of structures will become difficult and rare. It gonna cause a domino effect that will turn the map into a mono strategy of spamming. It will end in a battle of flamers everywhere since it's the weapon that benefit the more from the absence of structures. Because of wide open space with no "structure" players might not meet each other in the middle and cross each other and fall into an empty base.

On the other end, if the choke points are too small, it will cause all players to turtle in a boring stalled game. If a player can reach more oil, he will have to research more during half an hour were no action happen on both side. Than, he will give a final assault and the defending player might not even see the units that crush him.

conclusion
In other words, for the map to fit with the weapons and structure stats of the game and create a relevant game experience, there is a need to have some 21 tile per power ratio choke points more or less in the map.

Of course, a map maker can cheat. Something in between 17-25 tiles would fit nice into the weapon modifier margin (30%).

And sure, it also work for players that could determine if the point is suitable for defense or not using the same rule.

Have fun.

[Andrie]

Very Interesting!

[Iluvalar]

The "m", "H" and "T" perks exceptions
When a choke point is pretty small 3-5 tile (depends on the oil quantity) the units fighting inside it will have smaller number than usual. If on top of that, the path quickly open like a "T" shape it gives the defender player the opportunity to use the group advantage even if technically he have the same amount of unit on the map. It make him roughly 2x as strong. 1.5x if only one side of the T is buildable.

However, for this to be true, the attacker must not be able to reach the side of the T and shoot only one side at a time. You could fill the side with inaccessible ground like mountains. But you will find out that it is pretty hard and space expensive to make the mountain create a T shape in only one direction. You will end up most likely with a H perk.

In a H perk, both side players will have more facility to defend than attack. There is a lot of H perks in 4c-rush for example. Any narrow bridge would do the job as well. Sadly, despite the high potential of this, you will find that in most of your designs, you don't want at all both side to be defendable. Unless the H perk is in the middle of the map, you will most likely not want the dominating player with more oil to have that capacity to defend easily a position.

The other, and more useful possibility is the m perk which is giving the defender access to both side. Not only that allow you more compact design (by using that extra space) but it will also create a uni-directional T perk. It boil down to a ramp digging inside a cliff. If it's narrow enough, it's a m perk.

While players that are not used as making maps might not grasp the potential of it, it is an incredibly powerful tool to distribute the oil in another way without breaking the sacred 21 tiles rule. Because the chokepoint tiles in those perks only count as half. The m perk is the LED of the map flow ^^.

[zydonk]

Solid work, Il. I'm assuming both the H are T are to be viewed side on - on a north-south axis, that is. But isn't there an implied symmetry to the T layout, so that the attacker might also have an inverted T layout in his approach. The Pumpkin map "Mountain", for instance, is a complete bitch in this regard.The m layout may also be problematical, esp in your example. Trying to take advantage of the overlooking wings exposes the defender to the enemy units not yet in the defile. They will be massed with nothing else to do but pop the ducks up on the ridges. Mortars would work best here.

As for the 21 rule, I'm sure it's true and very useful, but getting numerical is a slippery slope - once everything gets numbered then its possible to get lost in detail. So we get an infinity of rebalancing. But it certainly makes me aware of the significance of the width of approaches.

[Iluvalar]

That's why I say it's rarely a "T" it's more likely 2 opposite "T" that I then call an "H".

When you defend on a m perk, you don't actually use the wings. You would only if the enemy was trying to turtle you in. Then you would be able to move your units on one side first, shoot a little part of the defense and then go around the other side.

Of course, I'm not implying that each maps should absolutely follow exactly 21 tiles per power ratio everywhere. Creativity is a key here. But it sure can be a nice guideline. If you are going too far astray, you can be sure your map will have annoying problems. Also It conveniently work the 2 ways around. So if you have designed an entire map and you just wonder how much oil you should put in the base or at a particular position you can still use that rule as a guideline.

[zydonk]

Ah, that explains the H layout. I didn't comment on that because I wasn't sure of the actual arrangement. And I see more clearly now that the M plan is a modification of the T. In fact, there could be any number of layouts with varying angles for the wings in between. Let's call this the Birdy scheme (flapping wings?)!

The 21 rule. Are you thinking mp here, with a relatively limited variation in oil distribution between players? I usually lay out choke points with an eye on the relative importance of the choke itself, because in sk the distribution of oil can change significantly among players during the course of a game. I conceive of defense as layered, rather than a simple line across the choke. Fixed weaponry will always fire in salvo at the nearest enemy, so staggered lines of firepower are necessary. My model for this is the Eighteenth century musket lines, which can advance or retreat in fire, load, aim, fire rows. (Took a lot of skill and courage then.) Properly organised, even lancer bunkers can hold off pretty sizable heavy track assaults.

[Iluvalar]

In fact, there could be any number of layouts with varying angles for the wings in between. Let's call this the Birdy scheme (flapping wings?)!

No really, the defender must have access to the bottom side of the T . Otherwise it will most likely be an H (bi-directionnal) if the defender can't access it or not a perk at all if both players can move there.

The rule would work if the AI is a real challenge in SP. But I know many players in SP tend to play with their food and take lot of time to win when they start to dominate the AI. I'm not really sure what is the funniest map layout for them.

[raycast]

The mg towers are around 255% stronger (price included) than the viper mg h-track tanks at the very beginning of the game. If we consider that a mg tower can cover about 8 tile wide of passage, we end with the magic number 20.4 tiles. The defending vs attacking units keep that 255% ratio during a long while in the game.

Does this still hold with the balancing changes we saw end of 2.x (when armor was broken) and with 3.x?

Plus, I cannot follow your reasoning. When one tower is 255% than a viper, that mostly says that you need 3x as many vipers to destroy a turret. But it doesn't really relate the passage widths for me.

[raycast]

Another followup:

Maybe we can turn some of these thoughts into a proof-of-concept type of map.

Player 1 has the oil advantage, at least in the beginning. His base is equipped with quite an amount of oil, but in a very open area that is hard to defend. It also is not yet highly developed.

Player 2 has little oil in the beginning (maybe there is some that he can explore during the game, but that requires some distance to travel to get to, plus it will be harder to defend - say, maybe it needs air or water transport - we BTW need hints for the AIs to make use of such oil!). However, he has a very good defensive position, maybe with some initial MG bunkers and a sensor already set up, and some good air defenses at his base.

Due to the defensive position of #2, an early rush will not work for player #1; due to the higher oil of the other, it also will not work for player #1. Time will first give #1 the oil and research advantage, but #2 can use this time to develop his defenses and reach out to the extra oil. So if player #1 waits too long without making progress, he loses the early advantage and will have to defend in his open position.

Here is a rough sketch of such a map. o are oil locations, # are cliffs or walls, _ is water.

Early on, the player in the top left has a 6 : 4 oil advantage (depending on how the other player needs to reach the 2 out of base oils, even 6:2 for some very early time. There is some race to get the two air-only oils - one strategy for player #1 might be to get air transport quick, then fly along the edges of the screen to the highground and establish a second base there. Maybe we can even have some scavenger air defenses there, or give the second player a chance to set up air defenses in the oil corners to prevent this strategy. On the long run, it will be next to impossible for him to reach that spot.

The defensive player will need to secure the back water entrances at some point in addition to upgrading his M choke points. Plus, at some point, his main base may come within reach of artillery.

Code: Select all

ooo     _
oo      #########
o       #  #  #o#
        ## ## # #
                #
        ## ## # #
        #  #  # #
_###  ######### #
 # #  # #  o    #
 #      #     ###
 ###  ###o    ___
 # #  # #     ####
 #      #     #
 ###  ### #_###  o
 #o       #_#
 ##########_#
            # o

[Iluvalar]

Plus, I cannot follow your reasoning. When one tower is 255% than a viper, that mostly says that you need 3x as many vipers to destroy a turret. But it doesn't really relate the passage widths for me.

If you try to cover 20.4 tiles with mg towers, you will soon discover that you have weakpoints (starting on a corner) where the opponent can attack roughtly 1/3 of your defense and get trought the frontier by avoiding the range of roughlty 2/3 of your defence.

[raycast]

But at, say, ~16 tiles width covered by two turrets, he can attack 1/2 and avoid range of the other half?
I don't get the math quite yet. Do you mean that at 20.4 tiles, there is the break-even point between taking down the whole defense and trying to punch a hole? When it's wider it pays off to punch a hole (i.e. rushing), when it's smaller it pays off to take the defenses down completely (so turtling is better?)

But doesn't this change completely when weapons get a different range?

[Iluvalar]

I mean that at 20.4, 1 tower out of 2.55 will be used. Each tower can fight 255% of their worth in tanks. So the power of the defender and the attacker will be even.

[raycast]

What is the setting? How are towers "used"? Are you assuming an uncoordinated attack by a single viper? Then it will either lose or avoid the tower, I guess. What if the defender builds 3 turrets and the attacker attacks from one side with 3 vipers?

Isn't the math more on the lines: if you build a tower every 2.55 tiles, this row of towers will draw against a row of vipers - because there are 2.55 as many vipers, but the turrets are 2.55 times as strong? I don't see how the width gets in.

[NoQ]

Rather if you spend 2.55 times more $$$ on vipers, then it will be a draw. This allows a defender to split his tower equally between 2.55 standard chokepoints ("s.c.") and feel good. Now we need to define what an "s.c." actually is, that is, when a passage counts as one s.c. and when it is wide enough to count as two s.c.'s, and when it's roughly 1.55 s.c.'s, etc. That's where the tile count and choke structure becomes relevant.