Skip to Content
 

I googled: "what to do against the fastest unit with the longest attack range"

5 replies [Last post]
Fri, 03/27/2026 - 19:36
X3M
X3M's picture
Offline
Joined: 10/28/2013

Why did I google that?
It turns out that these units in their most advanced form... Kinda get into a loop for being more expensive, when it comes to super micro.
That is, if I add them into my tool.

At first, I thought: "My tool doesn't work as intended?"
But then I figured. I am doing something else wrong.
And are there examples out there?
Yes! There is this marines ball in Starcraft 2 that is rather similar to how to micro in a similar fashion.

Can it be countered? So...I googled. And compared the results to my board game.

google wrote:
~AI Overview~

Countering a fast unit with a long attack range (often referred to as a "kiter" or "hit-and-run" unit) requires prioritizing mobility, disruption, and terrain over brute force. Because these units are designed to trade efficiently by avoiding damage, you must force them into unfavorable engagements, trap them, or outrange them with specialized artillery.


Ok sure.... that can indeed be the case for my boardgame as well. But why don't I see the results in my balancer tool?

google wrote:
Here is a breakdown of strategies to combat them:

1. Hard Counters and Tactical Choices
Fast Cavalry/Units: Use units that are equally fast or faster to chase them down and force them into melee, where they are often weak.
Long-Range Artillery: If they have the longest range, you need something that can fire from further away. Use siege weapons (like Trebuchets, Falconets, or Long-range SAMs) to force them to move or die.
Anti-Air/Specialized Units: If the unit is flying or specialized, use dedicated counter-units (e.g., dedicated anti-air vehicles) that have higher damage output in specific scenarios.
Surround/Split Tactics: Do not charge in a single line. Split your forces into smaller groups to surround them, limiting their ability to retreat in a straight line.


Alright, lets break this down indeed.

Fast Cavalry/Units:
Long-Range Artillery
These, are not THE options. Sure they are... But only if I removed the most advanced version of our little kiter. But that would mean I banish a design. And I rather include all possible designs. "Why?", you ask... well, for making sure that if we have a "design feast" game. We cannot make an unit that is overpowered.
Anti-Air/Specialized Units:
Ok, this is just a bullshit comment of the google ai.
Surround/Split Tactics:
And here, the ai actually found a good point.
I should include surrounding, aka trapping, in the calculator. This will make the units less powerfull. In fact, the most advanced kiter has a score of 100% against all other kiters and most lesser forms in designs.

--> I need to include trapping in the calculator. But how?

There is also this side note of how the terrain will assist in this. And of course, if it is the terrain that is of influence. Then we need to work with extremes. NO escape!!
Ok, I might have answered my question already.

google wrote:

2. Disruption and Control

Use Terrain: Force them to fight in tight areas, such as forests or narrow mountain passes, where their speed advantage is negated.
Control/Slow Effects: Use abilities that slow movement speed or stun the unit, preventing them from moving far enough to escape or re-position.
Targeted Abilities: In MOBAs or RPGs, use gap-closing abilities (dashes) combined with burst damage to catch and eliminate them before they can react.


Use Terrain:
There it is! :)
Control/Slow Effects:
Ok, this is unfortunately bullshit. "Why?", you ask again...
Applying this effect can only be done by a "weapon" design that functions in the same way as the Fast Cavalry or Long-Range Artillery
And the same applies for....
Targeted Abilities:
Yamato gun...anyone?
I often see super fast units flying away. And the Yamato gun just ignores the distance.
This gives me one idea though. What about a missile that keeps following, and the only other effect is that the damage goes down on the moment of hitting. Then...nah, it is the Fast Cavalry in sheeps clothing.

google wrote:
3. Economic and Strategic Countering

Volume over Quality: If the unit is expensive, overwhelm them with cheaper, faster, or more efficient units. Even if your individual units are weaker, high numbers can surround them.
Force Bad Trades: If they are "poking" you, do not take the fight on their terms. Keep your troops defensive, such as setting them to hold position in good terrain (woods, hills), forcing the opponent to come closer if they want to cause damage.
Remove Support: Often, these units rely on defenders. Attack the "home" base or infrastructure that supports them to force them back, rather than chasing them across the map.


Volume over Quality:
Wut?!
Isn't this another way to say, "trap the bastards!"?
It might be yet another case to have fodder that is of the Fast Cavalry type. Then again, if that is the case. Then trapping is really only possible in a relative "slow" way in combination with terrain. My game needs terrain effects. There is no other way.
Force Bad Trades:
Pokey Pokey!! Yes, thats how I consider my Bram poker's Stalkula units to behave. Again, terrain is mentioned. This time, not only as a way to block it. But wouldn't you believe it? High ground is also mentioned here. In other words, if the situation arises, the "cheaper" unit will be more effective. As long as it is on high ground.

I am actually hoping for some other options though...
...
...
Remove Support:
I almost didn't even read this properly.
But when I did. Eureka!!!
Of course, simply damage something else. And that ISSS your victory condition.
There are even hit'n'run designs in RTS where the units are melee and have specific targets. While ignoring or even unable to target other units/buildings.
C&C; Engineer, Flamethrower/Flametank
WC3; Wolfrider
KKnD; Saboteur, Flame-ATV
Starcraft/Broodwar; Zealots, Zerglings (they ignore air units and destroy a building asap) etc.

--> I should calculate the inflicted damage, while dying. And compare that to the hit'n'run units.

But...

Shouldn't I calculate that for ALL units? A, "what if... I cannot attack? What I do then?"-Type of situation. Meaning, I should calculate the damage they can inflict while dying.
I already did something similar for my towers.
Situation A: towers are within each other attack range.
Situation B: 1 tower is placed later and outranges said target. But that target, can it inflict damage on something else? And if so, how much?

X3M wrote:
This is a strategy or tactic. That I MUST include in the calculator.

Only issue I have with this is that a "what if" situation arrises where we don't have a tertiary group.
How much should both weight? 1:1? 1:2? 2:1? I don't know.

google wrote:
Summary of Tactics:
The ultimate goal is to force them into a sustained fight (a "trade"), where they are forced to take damage in return for the damage they deal.
Or they are just ignored.

Well, I now know what to do.
I hope you enjoyed this read.

Top
Mon, 03/30/2026 - 09:20
#1
larienna
larienna's picture
Offline
Joined: 07/28/2008
Finally, somebody else using

Finally, somebody else using AI. Not sure if google is good for that kind of talk. So far, my AI ranking is:

1.Claude
2.ChatGPT
3.GoogleAI ( the one in the browser, not sure if its gemini)

Claude is good if you need accurate real life information. Getting ideas to fix games does not require to be up to date in the news for example. Knowing where to invest in the stock market does require accurate information. Claude is better for that.

While, I use chat GPT for game design. It helps finding issues, determining priorities, exploring possibilities, calculating probabilities, etc. I have more chances to finish a game now that I have AI. It helps me organize priorities too.

For video games, it also helps for programming, debugging, researching, and comparing different implementation path. We concluded yesterday that shipping my game without music for now is acceptable.

Ironically, for my upcoming stock ticker video game, I talked with AI how to design a simple AI based on simple probabilities and distribution of weights over the price of the stock. I was a good 2-3 hour talk, but now I have a good implementation idea, that have good probability of working, so I am not losing time working on something that will not work.

In overall, its about optimizing the time spent.

Top
Mon, 03/30/2026 - 11:14
#2
X3M
X3M's picture
Offline
Joined: 10/28/2013
AI did help, yes!

larienna wrote:
Finally, somebody else using AI. Not sure if google is good for that kind of talk. So far, my AI ranking is:

1.Claude
2.ChatGPT
3.GoogleAI ( the one in the browser, not sure if its gemini)


Google on 3rd place? Interesting. ChatGPT on 2nd, also interesting. But I have never heared of Claude before.

larienna wrote:

Claude is good if you need accurate real life information. Getting ideas to fix games does not require to be up to date in the news for example. Knowing where to invest in the stock market does require accurate information. Claude is better for that.
Since I like being in the stock market as one of my hobbies. I will give it a try.

larienna wrote:

While, I use chat GPT for game design. It helps finding issues, determining priorities, exploring possibilities, calculating probabilities, etc. I have more chances to finish a game now that I have AI. It helps me organize priorities too.

For video games, it also helps for programming, debugging, researching, and comparing different implementation path. We concluded yesterday that shipping my game without music for now is acceptable.

I know someone who programs a lot. But always double checks what ChatGTP has written. As for making beer, we also double check that. It seems that the chemistry often has mistakes. Oh, that reminds me, we need to measure pH of the next brew.

larienna wrote:

Ironically, for my upcoming stock ticker video game, I talked with AI how to design a simple AI based on simple probabilities and distribution of weights over the price of the stock. I was a good 2-3 hour talk, but now I have a good implementation idea, that have good probability of working, so I am not losing time working on something that will not work.

In overall, its about optimizing the time spent.


Yeah, I agree. I never did this before. But thought, let's give it a try. I did save me a lot of brainstorming time.
But now for the next step, I need to think of what to do next. Which is entirely new in the world for all I know. Let alone, how to ask AI about the next step.

Top
Mon, 03/30/2026 - 13:11
#3
X3M
X3M's picture
Offline
Joined: 10/28/2013
Trapping and Ingoring opponents, and their weight?

Trapping opponents
It is a simple concept. Instead of allowing a maximum micromanagment. The opponent is simply unable to perform this micromanagment. And thus, it got trapped.

But how can it get trapped in my game?
1. The terrain stops it from moving. In the worst case, they are already in a corner. Unable to flee. It turns into a tower defence mode.
This depends on the players decisions and the terrain layout.

The worst case scenario plays out. The better will play tower defence as if it got trapped.

2. Action Points are a limitation. A hit'n'run unit might make use of the hit'n'run tactic. Which costs all AP. The other unit will take slower. And all it needs to do is move, move then attack.
The difference between the 2 is that the H'n'R unit moves forward, attacks, then moves away. And the slower unit will move forward twice. If the H'n'R unit has R+S being more than the other units 2*S+R. Then it can still always outrun.

This should be included in the original calculation. A simple test. If the lesser one succeeds in firing. The better one will have less of a victory score.

Each situation needs to get an equal weight for the endscore. If they are an option. If the players apply a strategy, it depends on the map. Thus a spectrum. And I can only pick the absolute average. Thus 1:1 or 1:1:1.

Ignoring opponents
This one is much more difficult?
How much should it weight?

The initial concept is easy though.

You ignore you are being "removed" from the game. But by doing so, you are set on another goal. Removing something else from the game as well. A tertiary target if you will.

But how to weight this in victory?

Not only that, but ALL situations actually can add this concept of ignoring the opponent. And the opponent can do the same as well?

Now, here we have an issue.

Example:
An unit with 4 movement points and 4 attack range fights, idk, a freshly placed tower with a melee attack. This tower was placed in such a way that it can target a tertiary unit on the map. Clearly, it can never win from that HnR unit. The score in the original calculator is 1 for the HnR unit and no more than 0 for the melee-tower.

The melee-tower is more durable and can deal more damage though. Compared to the HnR unit. So, it can deal a lot more damage in total. If the durability is x2 and the damage is x2. It can deal a total of 4x more damage than the HnR unit.
How does this compare?

Do we pick this 4 to 1 as a secondairy victory score? Thus 0.8 and 0.2? That would make sense.
Then we pick the average with the original 0 and 1. We get 0.4 and 0.6.
Oh, actually, it isn't that difficult then?

In a sense, the cheaper an unit is, the better it scores in this regard.
A more expensive design that is simply in the same boat as the 4/4 unit. But can never hope to fight back. Would be a 3/3 design. Lets say the durability and damage here is x1.2 and x1.2 in comparison. Thus attacking a tertiary target, it yields 0.59 to 0.41.
And the average this time is 0.295 to 0.705.

What can we conclude from this? The 4/4 vs 3/3 is no longer a solid victory. And it isn't much better than vs 0/0 either. There is no loop cascade effect in the calculator this way either.

The main question here is...
How much does this situation have to weight? It is a yes/no situation?

Is there a tertiary target avaiable? Yes/No --> This victory condition weights only 50%.

Thus we got a 1:2 distribution?

That makes 0.8/3 to 2.2/3 or 0.27 to 0.73.
Instead of that 0.4 to 0.6.

But, we can replace the question too.
Will we attack and resolute the conflict?
Yes/No...depends on if there is a tertiary target avaiable.
If the latter is not avaiable. The 0/0 design is not even going to be placed.

As you can see, determining how much weight a situation might need depends on the practical choices of players.

Originally, the victory was a solid infinite for the 4/4 unit.
But with the practical choice in place. It changes to a 1.5x choice.
And if the practical choice is a gamble. It changes to a 2.7x choice.

***

All I need to do is add a few extra victory point calculators. And pick the right classes where they belong to.

All thanks to asking the right question to the AI. And having this train of thought going on the right tracks.

Top
Tue, 03/31/2026 - 17:13
#4
X3M
X3M's picture
Offline
Joined: 10/28/2013
I asked google: how to trap a fast unit in rts

google wrote:
Trapping a fast unit in an RTS game involves restricting its movement options using superior positioning, surrounding tactics, and environmental awareness to negate its speed advantage. Key strategies include using worker surrounds, utilizing choke points, and employing instant-lock or slows.

Ok....but what if you are extremely limited to....1 unit?

google wrote:

Effective Methods to Trap Fast Units

Surround Tactics (The "Surround"): In games like Warcraft III or StarCraft, command units to move behind the enemy unit, then switch to attack/hold position to create a circle that prevents movement.

Clearly the AI assumes we have a lot more units to our disposal. But let's say, we got something like a Reaper rush like in Starcraft 2. The only difference, it cannot jump? Well, and even if it can jump. How to counter it?
Most options will include something with a longer attack range. That is not our goal. We want to trap it.

google wrote:

Unit Blocking (Collision Mechanics): Use melee or sturdy units to box the faster unit in, exploiting the game's collision physics to physically prevent it from passing through.
This, is a thing in my board game as well. However, we need a production to be greater than the destruction that said units will apply. This can be easy in a sense due to the fact that the faster an unit is, the more expensive it will be. Not only that, but their attack range is also increased. Since we are actually trying to trap H'n'R units.

google wrote:

Utilize Terrain and Choke Points: Force the fast unit into narrow areas, such as between buildings or map obstacles, where its mobility is restricted and it cannot easily bypass your forces.

This only works on maps with a lot of choke points indeed. Preferably one where you approach the H'n'R unit from as many sides as nessesary. Let's assume the map is limited in size. Thus the edge counts as impassable terrain. That means we can produce as many units as the enemy makes H'n'R units. And all we need to do is close in the enemy. The threshold on this being succesfull is linked to the size of the map. Not just the terrain layout.

google wrote:
"Hold Position" Trap: Move a fast unit into a cul-de-sac or behind your own lines, then use the "Hold Position" command to trap it between your units and the environment.
Not going to happen in a chess paced game.
google wrote:
Use Crowd Control (CC) Abilities: Utilize units or abilities that slow, snare, stun, or freeze the target to negate their high speed.

Again. Not going to happen in a chess paced game.

google wrote:

Micro-Management Tips for Trapping

Map Control & Scouting: Know where the fast units are likely to raid so you can set up traps in advance, rather than reacting too late.

There is no fog of war my AI-friend...

google wrote:
Concave Formation: Rather than a direct line, spread your units in a U-shape around the target. This ensures maximum damage and prevents the target from easily running away.
I guess this should be in combination with the "Utulize terrain and choke points" section.
google wrote:
Stutter Step (Kiting): Move your melee units between attacks to close the distance and maintain the surround.
While this is a thing that is possible with the right units. We are facing an unit that is faster than us. I don't get how the AI got to this option.

google wrote:
Unit-Specific Strategies

Workers: Often the best for surrounding a hero or high-value unit early in the game because they are cheap, expendable, and available early.

I am starting to feel that I should have some sort of tech-tree in place. However, that is not the goal of my search for trapping H'n'R units.

google wrote:
Long-Range Units: Use long-range units to pick off fast units from a distance while keeping them cornered.
I knew this would come along. Perhaps I should rephrase my question.

Top
Tue, 03/31/2026 - 17:49
#5
X3M
X3M's picture
Offline
Joined: 10/28/2013
I asked google: how to trap a hit and run fastest unit in rts

NGL, clumsy question. But it suggested it being asked like this.

google wrote:
AI Overview
Trapping the fastest hit-and-run units in an RTS requires a combination of terrain exploitation, superior mobility (surrounding), and the use of crowd control (CC) abilities to stop their movement. The fastest way to trap them is to prevent them from moving rather than trying to outrun them.
Fastest Ways to Trap Hit-and-Run Units:

Ok, kinda similar. But let's see what other things the AI came up with.

google wrote:
Surround with Faster or Similar Speed Units: The most effective "trap" is to surround the enemy with your own units, often referred to as a "surround" or "concave". This blocks their pathing, preventing them from kiting backward.

Thought I asked, how to trap the fastest unit...?
google wrote:
Use Crowd Control (CC) Abilities: Use spells that stun, slow, or immobilize the unit. Examples include EMPs, stuns, or "widow" type units that can freeze enemies in place, allowing the rest of your army to destroy them.
This is an attack range solution.
google wrote:
Utilize Terrain and Structures: Force the units into narrow corridors, bottlenecks, or against impassable terrain, which prevents them from maneuvering. You can use your own buildings to wall off sections of the map to trap them in a corner.
Hmmm....trapping them by producing impassable terrain yourself. That sounds reasonable. Didn't think of that. It is more or less a mirror of actually ignoring the buggers when attacking the enemy base.

Still, production has an "attack range" as well.

google wrote:
"A-Move" and Stutter-Stepping: If the unit is fast, instead of right-clicking (which triggers the pursuit), use "Attack Move" (A-move) towards the direction they are retreating to. This forces your units to attack any target in range while moving, reducing the effectiveness of their hit-and-run.
Lol, this only works for certain RTS games. Not a chess paced and chess like board game.
google wrote:
High-Damage Burst Units: Use units that have high damage output in short bursts. If a unit cannot take many hits, a fast, decisive strike from high-DPS units can destroy it before it can flee.

This....is true. It does matter for equal ranged units. The problem here though, is that the attack range may NOT, be the solution.

google wrote:
Key Tactics for Specific Scenarios:

Against Kiting: Use "Hold Position" to keep your melee units from chasing and instead wait for the enemy to come into range, which can disrupt their kiting pattern.

Ok, this AI needs to train first...
google wrote:
Flanking: If the enemy is using hit-and-run, flanking them from two sides at once can trap them between your forces.
Well....this sounds again, similar.

***

What other question could I ask the google AI? Or, how should I rephrase my question for a better answer?
Or could it be that trapping is nearly impossible, with all the practical choices a player can make?

***

So far, asking this question to the AI. Has resulted for me into believing that:
1. Trapping is a yes/no situation for certain now, based on terrain.
Some maps are perfect for it. Other maps are useless for it.
Thus trapping victory weights count indeed only 50%.
2. I should consider my 1 on 1 unit design. Also as a fodder vs single unit design. As if the whole fodder group exists in total for the entire combat resoltuion calculation.

Either way, the victory weights are as following:

100% Normal combat including perfect micro.
Best H'n'R unit shows the most [1 to 0] victory weights.
Action Points limitation must be included here. (Need to add this one)

50% Trapping.
Best H'n'R unit will show a reduced victory weight, from [0.5 to 0.5] to ["0" to 1]. But this only counts for 50% due to situation events.

100% Ignoring opponents.
All designs get this score added. They produce damage during their lifetime. The lifetime-damage is compared to each other. The choice is simple: they ignore the fact that they are being destroyed. So, there is no infinite lifetime-damage. We are only considering all-round units for this calculation. So, in a sense, H'n'R units are bad defenders.
Best H'n'R unit, with infinite speed and attack range, shows a [0 to 1] victory weight here.

***

While some H'n'R units showed a lot of [1 to 0] victory weights. It is these same units that will receive a [0 to 1] from both trapping as from being ignored.
(1x [1 to 0] + 0.5 x [0 to 1] + 1 x [0 to 1]) / 2.5 = [0.4 to 0.6]

And this actually means, they are worse now, against the worst designs. They are "over powered" in such a way, that it went over their head.

I wonder what results I yield, next time, when adding these extra calculations. But I need a clear head, and lots of time. Luckily, I only need to add stuff to one calculator. Then copy and drag to the multi-calculator. The third step has already been automated.

Another funfact: As soon as the multi-calculator is up and running. I could automate the matchups. Then change the primairy contestants. And see how it all evolves.
I can have 25 different contestants in the current tool that I am working on. And the effects of skipping an attack range or movement speed, will be made clearly visible to me.

Top
Syndicate content


forum | by Dr. Radut