QOL - Military Strength for Groups of Entities
Moderator: MOD_DW2
-
StormingKiwi
- Posts: 336
- Joined: Thu Feb 11, 2021 6:35 am
QOL - Military Strength for Groups of Entities
Something I'd like to see in DW2 is a number for how many entities the military strength is divided by.
E.g. is it a single ship with a strength of 100, or a group of 8 escorts?
E.g. is it a single ship with a strength of 100, or a group of 8 escorts?
-
StormingKiwi
- Posts: 336
- Joined: Thu Feb 11, 2021 6:35 am
RE: QOL - Military Strength for Groups of Entities
This matters because, in theory, both groups of ships have the same military strength.
In practice, quantity is better than quality, so more ships will overcome a single ship even if that single ship has greater military strength than the fleet combined, to a point. This makes the military strength convey misleading information to the player at best.
An example of this can be seen in this video of TortugaPower's, where a fleet of strength ~580 with 27 ships easily defeats a single base with strength 850.
In practice, quantity is better than quality, so more ships will overcome a single ship even if that single ship has greater military strength than the fleet combined, to a point. This makes the military strength convey misleading information to the player at best.
An example of this can be seen in this video of TortugaPower's, where a fleet of strength ~580 with 27 ships easily defeats a single base with strength 850.
Last edited by StormingKiwi on Tue Mar 01, 2022 4:08 am, edited 1 time in total.
RE: QOL - Military Strength for Groups of Entities
This is absolutely not true. I think you are forgetting how complex ship design is in Distant Worlds, and how so many different factors about each particular design (weapons, shields, sensors, range, speed, etc.) can contribute to the outcome of an engagement. Hell, in DW2, even galactic terrain can alter the course of engagements (most notably nebula).ORIGINAL: StormingKiwi
In practice, quantity is better than quality, so more ships will overcome a single ship even if that single ship has greater military strength than the fleet combined, to a point.
To elaborate, I am fascinated by ship design in Distant Worlds, and spent a great deal of time creating various warships catered towards specific use-cases in DW:U. There are absolutely instances where X ship can beat Y, but not Z. Yet Z cannot beat Y, even though it can destroy X, and X is capable of beating Y.
Point being, ships are very complex, and there are too many factors contributing to the outcome of engagements for you to generalize combat as "quantity is better than quality," where "more ships will overcome a single ship even if that single ship has greater military strength than the fleet combined."
That is not at all how it works.
-
StormingKiwi
- Posts: 336
- Joined: Thu Feb 11, 2021 6:35 am
RE: QOL - Military Strength for Groups of Entities
I too am absolutely fascinated by ship design and fleet engagements, nor am I forgetting how complex ship design is in Distant Worlds.ORIGINAL: Galaxy227
This is absolutely not true. I think you are forgetting how complex ship design is in Distant Worlds, and how so many different factors about each particular design (weapons, shields, sensors, range, speed, etc.) can contribute to the outcome of an engagement. Hell, in DW2, even galactic terrain can alter the course of engagements (most notably nebula).
To elaborate, I am fascinated by ship design in Distant Worlds, and spent a great deal of time creating various warships catered towards specific use-cases in DW:U. There are absolutely instances where X ship can beat Y, but not Z. Yet Z cannot beat Y, even though it can destroy X, and X is capable of beating Y.
Point being, ships are very complex, and there are too many factors contributing to the outcome of engagements for you to generalize combat as "quantity is better than quality," where "more ships will overcome a single ship even if that single ship has greater military strength than the fleet combined."
That is not at all how it works.
However, I have a video of actual gameplay providing a concrete example to support what I have written, and the principles are well supported by a century of OR.
Unless you're going to provide evidence supporting your point of view, that is a good generalisation of how this works.
Further, this is a tangent to the actual point I am making, which is that the estimate of military strength conveys misleading information to the player.
RE: QOL - Military Strength for Groups of Entities
ORIGINAL: StormingKiwi
However, I have a video of actual gameplay providing a concrete example to support what I have written
My initial argument was against the statement of yours which describes "...quantity [being] better than quality."
I will not agree quantity is objectively superior to quality. There is a time and a place for both. Any player of Distant Worlds, or observer of warfare in general, should understand this concept. If you need "evidence," I can fire up DW:U and create a single ship capable of destroying a hundred others at once. Your statement is not a rule Distant Worlds strictly follows, and a single five-minute engagement in a preview build will not make it so.
Regardless, you are correct in saying "this is a tangent to the actual point I am making." In an effort to remain on-topic, I spent some time thinking about the current state of strength in DW2. Closer inspection of Tortuga's battle has led me to believe the value of strength in DW2 is so inaccurate, it's practically useless. Furthermore, your original point is equally as useless of a solution, as defining the quantity of entities within a given value of strength will not fix the inherent inaccuracies nested within said value. Simply put, telling the player 800 strength is one ship or seven ships doesn't change the fact 800 was an inaccurate measure of strength to begin with.
You tell me you want evidence.
ORIGINAL: StormingKiwi
Unless you're going to provide evidence supporting your point of view, [my point] is a good generalization...
To elaborate, let us simply look into the evidence you provided.
The engagement linked in Tortuga's video suggests strength in DW2 is a poor indicator of an entity's prowess in combat.
Allow me to demonstrate:
Tortuga had two fleets, the 1st & 2nd Defense Force, of twelve and fifteen ships respectively. Combined, these two fleets totaled to eight frigates and nineteen escorts. Both Tortuga's frigates and escorts have 228 shields and 0 armor, but differ in their damage per second. Each frigate deals 2.9 damage per second, whereas each escort deals 1.8 damage per second.
Let us find the combined hitpoints and damage of Tortuga's two fleets (which DW2 projects at a strength of 568).
228 * 27
= 6,156 hitpoints
2.9 * 8
= 23.2
1.8 * 19
= 34.2
23.2 + 34.2
= 57.4 damage per second
Let us list the hitpoints and damage of the Pirate Base (which DW2 projects at a strength of 850).
= 1,384 hitpoints
= 9.8 damage per second
With the numbers above, it should be clear as day Tortuga's two fleets are significantly stronger than the Pirate Base. So why does DW2 suggest the Pirate base has more strength? Because the value of strength is broken.
I suggest strength should be a combination of only hitpoints and damage. You might ask, why only hitpoints and damage if I previously argued "ships are very complex?" The answer lies in what followed the aforementioned quote: "there are too many factors contributing to the outcome of engagements for you to generalize combat" into a single value of strength. The rest of combat is far too unpredictable; for instance, the range of weapons or speed of ships will always be too situational to reliably define. If we want a single, simple value like strength to quantify an entity's combat prowess, we must use concrete values like hitpoints and damage. This way, the player will always get a reliable, ball-park estimate of the strength of any given ship, fleet, or base.
My suggestion fixes strength, creating a predictable value to estimate the power of entities. This value paints a more realistic picture of an entity's prowess in combat than whatever DW2's strength is currently providing.
TL;DR: The problem with strength has nothing to do with the number of entities, but rather how strength is calculated. It is currently highly inaccurate, and needs to be more predictable.
RE: QOL - Military Strength for Groups of Entities
As someone who never played DW1 beyond a few hours the above discussion is really useful.
Thank you .
Thank you .
"I do not agree with what you say, but I shall defend to the death your right to say it"
-
StormingKiwi
- Posts: 336
- Joined: Thu Feb 11, 2021 6:35 am
RE: QOL - Military Strength for Groups of Entities
I agree based on your presentation that the military strength per ship number is inaccurate as DW2 currently calculates it, and I too would like to see it more accurate. However, accuracy on an individual (per ship) basis is meaningless if it is still predicting outcomes incorrectly on a group (per fleet) basis.ORIGINAL: Galaxy227
TL;DR: The problem with strength has nothing to do with the number of entities, but rather how strength is calculated. It is currently highly inaccurate, and needs to be more predictable.
Now, I wrote my post with the view that Distant Worlds 2 is a game where combat primarily occurs between groups of ships/bases rather than as duels between only one of each.
This brings us to the laws of concentration. Focus fire exists in the complexity of the simulation.
Lanchester's Square Law predicts that where combat is an all-against-all brawl, as is the case in Distant Worlds, the group's military strength is proportional to the square of the number in the group and only linearly to the strength of each individual member.
Even if the military strength value was an accurate number per ship, it would fall apart when groups are concerned because of the ability that ships have to 'gang up' and focus fire. Hence why the reporting needs to reveal the overall strength of the fleet and how many ships are in the fleet.
-
Jorgen_CAB
- Posts: 895
- Joined: Wed Mar 17, 2010 7:53 pm
RE: QOL - Military Strength for Groups of Entities
It is practically impossoble to decide the actual strength of ships... I'm pretty sure that strength is more or less a ships DPS value.
It does not take into account range, fall of damage, ship speed and manoeuvre rating, shields, armour, hull etc...
A missile ship will be considerably weaker than a Maxos Blaster ship for example, in practice it can be allot stronger depending on other factors.
All of these factors can greatly contribute to the ships actual strength in practical terms.
In my opinion this is not an issue in and of itself.
In my opinion the original idea with the OP of listing both strength and number would give a slightly better idea of a fleets actual strength. I don't think the strength number is broken, if just reflect your fleets overall offensive firepower, but this is not everything that is important, the rest you will have to experience.
It worked the same in DW1 so there is no big surprise there. I don't like the strength numbers at all, we should not even know that. The game should only tell us what we actually know. Such as there is a base and its size, how man escort, frigates etc a fleet has. We should not have the strength numbers at all, at least not unless we know what weapons the ships are carrying. The numbers are highly misleading anyway.
It does not take into account range, fall of damage, ship speed and manoeuvre rating, shields, armour, hull etc...
A missile ship will be considerably weaker than a Maxos Blaster ship for example, in practice it can be allot stronger depending on other factors.
All of these factors can greatly contribute to the ships actual strength in practical terms.
In my opinion this is not an issue in and of itself.
In my opinion the original idea with the OP of listing both strength and number would give a slightly better idea of a fleets actual strength. I don't think the strength number is broken, if just reflect your fleets overall offensive firepower, but this is not everything that is important, the rest you will have to experience.
It worked the same in DW1 so there is no big surprise there. I don't like the strength numbers at all, we should not even know that. The game should only tell us what we actually know. Such as there is a base and its size, how man escort, frigates etc a fleet has. We should not have the strength numbers at all, at least not unless we know what weapons the ships are carrying. The numbers are highly misleading anyway.
RE: QOL - Military Strength for Groups of Entities
There appears to be a number of somewhat related issues to converse about in this topic.
The first issue is essentially what "strength" means. It is not at present an accurate number, nor do I think it is meant to be. It is a very rough indicator that gives a very ballpark idea of whether you're dealing with a housecat or a sabretooth tiger. That is really all it does. As I recall from DWU, it was really just the basic weapon damage from each weapon in the group. It tells the player something but that something is by no means an actual metric for performance. It's just a very rough indicator.
The second issue is that adding the number of entities in the group really doesn't improve the level of information. If anything it makes it worse, as now a ramshackle back-of-the-envelope number suddenly looks like something more meaningful than it really is. Consider the situation with two dreadnaughts and 20 escorts. The 20 escorts have a combined strength of 1200. Add the last two ships and the strength jumps to 3200. What do we learn through a simple average? Not a thing, really.
The third issue is how one could even begin to offer a single number as a remotely reasonable indicator of ship strength without having it turn into meaningless nonsense. What determines the strength of a ship in a particular engagement? Alpha strike vs effective dps? Engagement range and damage profile of the weapon vs evasion ability of the target? Weapon type vs target's defenses? Accuracy? Rate of fire vs number of targets? Damage type vs target's damage mitigation and recovery? And now that the game is in 3d, there's also engagement angles to consider. How many weapons can even be brought to bear? And an often overlooked detail in DW ship design, does the ship actually have reactor power to keep firing all weapons while at sprint speed? Does it have fuel to keep the reactor power up?
I humbly submit that there is no single number that can convey all this information and that even trying to construct one would be a serious mistake.
The fourt issue is the many vs one discussion. I have zero clue what the OR conclusions past and present would suggest. Some references on that might be helpful if OR is to be part of the conversation. But I did play DWU quite a lot and in that particular model of reality, the answer was not a simple binary. In the most general case, if the many ships lack sustained dps to overcome the self-repair and shield recharge of the big single target then the single target wins. If the single target has a low rate of fire and massively overkills each individual member of the swarm then a lot of the strength is wasted and then the swarm has an advantage. If the swarm have weapons that bypass shields then the swarm might be able to overwhelm the single target and deal crippling damage before it can handle the swarm.
The first issue is essentially what "strength" means. It is not at present an accurate number, nor do I think it is meant to be. It is a very rough indicator that gives a very ballpark idea of whether you're dealing with a housecat or a sabretooth tiger. That is really all it does. As I recall from DWU, it was really just the basic weapon damage from each weapon in the group. It tells the player something but that something is by no means an actual metric for performance. It's just a very rough indicator.
The second issue is that adding the number of entities in the group really doesn't improve the level of information. If anything it makes it worse, as now a ramshackle back-of-the-envelope number suddenly looks like something more meaningful than it really is. Consider the situation with two dreadnaughts and 20 escorts. The 20 escorts have a combined strength of 1200. Add the last two ships and the strength jumps to 3200. What do we learn through a simple average? Not a thing, really.
The third issue is how one could even begin to offer a single number as a remotely reasonable indicator of ship strength without having it turn into meaningless nonsense. What determines the strength of a ship in a particular engagement? Alpha strike vs effective dps? Engagement range and damage profile of the weapon vs evasion ability of the target? Weapon type vs target's defenses? Accuracy? Rate of fire vs number of targets? Damage type vs target's damage mitigation and recovery? And now that the game is in 3d, there's also engagement angles to consider. How many weapons can even be brought to bear? And an often overlooked detail in DW ship design, does the ship actually have reactor power to keep firing all weapons while at sprint speed? Does it have fuel to keep the reactor power up?
I humbly submit that there is no single number that can convey all this information and that even trying to construct one would be a serious mistake.
The fourt issue is the many vs one discussion. I have zero clue what the OR conclusions past and present would suggest. Some references on that might be helpful if OR is to be part of the conversation. But I did play DWU quite a lot and in that particular model of reality, the answer was not a simple binary. In the most general case, if the many ships lack sustained dps to overcome the self-repair and shield recharge of the big single target then the single target wins. If the single target has a low rate of fire and massively overkills each individual member of the swarm then a lot of the strength is wasted and then the swarm has an advantage. If the swarm have weapons that bypass shields then the swarm might be able to overwhelm the single target and deal crippling damage before it can handle the swarm.
-
StormingKiwi
- Posts: 336
- Joined: Thu Feb 11, 2021 6:35 am
Re: QOL - Military Strength for Groups of Entities
Thank you for the support, Jorgen.
In response to Spidey's comments:
2) With a strength number, reporting the number of entities in the group, e.g. [Strength: 2000, (5)] gives the player valuable information as to whether or not they want to fight or flee from the reported target, while also being information that makes sense to report in-universe. I'm not thinking of a simple average, nor have I requested one.
Without quantity numbers, the quality number is meaningless. As soon as you have more than two entities engaged in combat, the strength estimate becomes useless because of focus fire effects and so on.
3) I'm not at all convinced that a number that presents military strength is unreasonable, as the computer's AI and the player both need some objective way of determining in which engagements to fight and in which to flee from at the highest level. From this perspective, the number is entirely fit for purpose. However, I do agree that a single number is inadequate as the number of entities involved in combat increases, which is why I have suggested my solution.
KISS principle, of course, applies, which is why I have suggested presenting a pair of numbers (strength and number of entities) rather than a more complicated solution.
4) The math concludes that an n-fold increase in quantity requires an n-squared increase in quality.
A link to the Lanchester's law discussion, as it relates to video games, can be followed through here: The ABC article by the mathematician is available through search engines, and numerous other recent articles can be found in the scientific literature discussing this topic, for those who have access.
The equations can be found on Wikipedia; however, they are also provided below:
The Lanchester system of equations is:
Where q_a, q_b is the quality of a unit of army a, army b;
while A, B is the number of units in army a, b.
I have found the most informative way to think about the equations has been as a ratio:
In response to Spidey's comments:
2) With a strength number, reporting the number of entities in the group, e.g. [Strength: 2000, (5)] gives the player valuable information as to whether or not they want to fight or flee from the reported target, while also being information that makes sense to report in-universe. I'm not thinking of a simple average, nor have I requested one.
Without quantity numbers, the quality number is meaningless. As soon as you have more than two entities engaged in combat, the strength estimate becomes useless because of focus fire effects and so on.
3) I'm not at all convinced that a number that presents military strength is unreasonable, as the computer's AI and the player both need some objective way of determining in which engagements to fight and in which to flee from at the highest level. From this perspective, the number is entirely fit for purpose. However, I do agree that a single number is inadequate as the number of entities involved in combat increases, which is why I have suggested my solution.
KISS principle, of course, applies, which is why I have suggested presenting a pair of numbers (strength and number of entities) rather than a more complicated solution.
4) The math concludes that an n-fold increase in quantity requires an n-squared increase in quality.
A link to the Lanchester's law discussion, as it relates to video games, can be followed through here: The ABC article by the mathematician is available through search engines, and numerous other recent articles can be found in the scientific literature discussing this topic, for those who have access.
The equations can be found on Wikipedia; however, they are also provided below:
The Lanchester system of equations is:
Code: Select all
a) dA/dt = -q_a x B; b) dB/dt = -q_b x A while A, B is the number of units in army a, b.
I have found the most informative way to think about the equations has been as a ratio:
Code: Select all
q_a/q_b = [ B(0)^n - B(t)^n ] / [ A(0)^n - A(t)^n ]Re: QOL - Military Strength for Groups of Entities
You're not really responding to the issue of simple averages being largely meaningless without additional context. 5 ships and 2200 strenght means what exactly? One dangerous boat at 2000 and 4 worthless boats at 50 each? Or five significant ships at 420 each? The more ships we add to this mix, the more useless any single number becomes and the more necessary it will be to pause the game and investigate the specific details of the force.
Additionally, the strength number is largely an aggressive simplification. It is the sum of all weapon strength without any consideration of how said weaponry can actually be employed and without any consideration of the durability of the frames that carry said weapons. Put 20 million riflemen in rowboats, if you will, and you have quite a force, but what can they do versus a single destroyer?
And regarding Lanchester, it is not my understanding that his math is a universal truth that is valid for all forms of combat without any qualification. He was considering first classic melee infantry vs classic melee infantry in a frontal engagement and then later riflemen vs riflemen in a similar scenario. His math assumes that all the combatants can actually fire on each other. If that is not the case then I'm not sure his math works.
Additionally, the strength number is largely an aggressive simplification. It is the sum of all weapon strength without any consideration of how said weaponry can actually be employed and without any consideration of the durability of the frames that carry said weapons. Put 20 million riflemen in rowboats, if you will, and you have quite a force, but what can they do versus a single destroyer?
And regarding Lanchester, it is not my understanding that his math is a universal truth that is valid for all forms of combat without any qualification. He was considering first classic melee infantry vs classic melee infantry in a frontal engagement and then later riflemen vs riflemen in a similar scenario. His math assumes that all the combatants can actually fire on each other. If that is not the case then I'm not sure his math works.
Re: QOL - Military Strength for Groups of Entities
Regarding simple "Averages are pointless"Spidey wrote: Thu Feb 24, 2022 8:14 am You're not really responding to the issue of simple averages being largely meaningless without additional context. 5 ships and 2200 strenght means what exactly? One dangerous boat at 2000 and 4 worthless boats at 50 each? Or five significant ships at 420 each? The more ships we add to this mix, the more useless any single number becomes and the more necessary it will be to pause the game and investigate the specific details of the force.
Additionally, the strength number is largely an aggressive simplification. It is the sum of all weapon strength without any consideration of how said weaponry can actually be employed and without any consideration of the durability of the frames that carry said weapons. Put 20 million riflemen in rowboats, if you will, and you have quite a force, but what can they do versus a single destroyer?
And regarding Lanchester, it is not my understanding that his math is a universal truth that is valid for all forms of combat without any qualification. He was considering first classic melee infantry vs classic melee infantry in a frontal engagement and then later riflemen vs riflemen in a similar scenario. His math assumes that all the combatants can actually fire on each other. If that is not the case then I'm not sure his math works.
This is a contrived scenario you are cooking up.
How would you ever even get close to that? A repaired battleship supported by Escorts?
How often will that happen in the average game?
Have you conidered that repair times scale with tech difference of a wreck?
Are you running under the asumption the Planet killers anti-ship power was not nerfed from the streams?
Regarding "Additionally, the strength number is largely an aggressive simplification"
Unless you got a better solution to evlaute ship strenght that can actually be done by a AI - and I mean a actuall working example - that is all just wishfull thinking. Not a basis for a meaningfull discussion.
Until then we have one figure for strenght.
This figure should represent damage output, durability, range and mobility as closely as possible.
It does not need to be perfect. Just better then trying to figure out "is [undefined] bigger or smaler then [undefined]?"
Regarading Lancasters law
Unless I am missing something, those ideal scenarios are exactly how combat in DW2 works.
Or do you have any knowledge about trenches being used in space?
Flanking maneuvers that actually impact combat performance of the flanker and flankee?
Re: QOL - Military Strength for Groups of Entities
The problem with averages is by no means a contrived scenario. You have no idea what composition a force has and what you're advocating is seemingly that we can blindly assume all ships in a force are equal. That leads to bad conclusions. We don't know what we don't know, and just assuming a convenient answer doesn't change that.
I'm using an aggressive example to showcase why simple averages are fundamentally a bit rubbish but compare with a regular carrier battle group, if you wish. Are all those ships equal? If you give a strength score to each ship, will the average value really tell us something useful? I suspect not.
As for the usefulness of the strength number itself, I obviously don't yet know how it is calculated in DW2, but back in DWU I'm fairly sure it was just the sum of the nominal damage values of each weapon on a ship, with zero regards for rate of fire, accuracy, speed, maneuverability, et cetera. Didn't even consider if the ship had enough reactor surpluss energy to actually fire the gun. That's strength in DWU.
You might think it is possible to construct a better strength score by adding more factors into how it is calculated. I don't think that is the case. We're basically pushing into KPI territory here, and the problem is that the more factors you put into an indicator, the more context is lost and the less understandable the number becomes. Yes, it is easy to see the number at a glance, but can that number be trusted?
If you're into soccer then check out Whoscored.com's player ratings for a brilliant example of a meticulously constructed indicator of performance that basically doesn't tell anyone anything worth knowing. It's the same thing if you try to make a great strength value from damage, damage type, range, accuracy, engagement angles, acceleration, maneuverability, shields, armor, self-repair, speed... It's too much context to force into a single number. You get much more complexity but at the end of the day the number is still not reliable.
Thus strength is what it is, a rough but hugely flawed guideline, and as my old stat teacher used to say, garbage in, garbage out. What is the average of a garbage number? Still garbage. The AI uses "strength" for decision purposes and consequently the AI will predictably make suboptimal decisions. We can do better, if we take the time to do so, but it is hard for us to be more than one place, it may even be hard for us to even be that one place, and it is trivial for the AI to be everywhere, so basically swings and roundabouts.
Lastly, as I understand Lanchester's square law, the assumption was two forces of WW1 riflemen where each rifleman can shoot at any enemy rifleman. Are all ships in fleet 1 shooting all guns at optimal range at the entirety of all the enemy ships in fleet 2? That seems very unlikely. And so we're already getting beyond the assumptions of the square law.
This is before we consider that Lanchester's riflemen didn't have different types of bullets with different accuracy and damage profiles, different types of self-repair, different types of damage mitigation, and his riflemen were not teleporting in and out of combat nor did they need to run up under fire to get into engagement range, nor were the opfor running away to make engagement harder.
Basically, Lanchester did the math on more shooters versus better shooters over time in open battle, all things equal, and in that context a given increase in number of shooters provides a bigger advantage than a similar increase of shooter quality. If we're talking about open battle where both forces just trade blows and suffer attrition until the battle is won or lost, that is. How often is that really the case, though?
Or to put it in DWU terms, I don't need 20 ships to remove 10 kaltors from a particular spot. I really only need 1 ship with missiles and better mobility than the kaltors.
I'm using an aggressive example to showcase why simple averages are fundamentally a bit rubbish but compare with a regular carrier battle group, if you wish. Are all those ships equal? If you give a strength score to each ship, will the average value really tell us something useful? I suspect not.
As for the usefulness of the strength number itself, I obviously don't yet know how it is calculated in DW2, but back in DWU I'm fairly sure it was just the sum of the nominal damage values of each weapon on a ship, with zero regards for rate of fire, accuracy, speed, maneuverability, et cetera. Didn't even consider if the ship had enough reactor surpluss energy to actually fire the gun. That's strength in DWU.
You might think it is possible to construct a better strength score by adding more factors into how it is calculated. I don't think that is the case. We're basically pushing into KPI territory here, and the problem is that the more factors you put into an indicator, the more context is lost and the less understandable the number becomes. Yes, it is easy to see the number at a glance, but can that number be trusted?
If you're into soccer then check out Whoscored.com's player ratings for a brilliant example of a meticulously constructed indicator of performance that basically doesn't tell anyone anything worth knowing. It's the same thing if you try to make a great strength value from damage, damage type, range, accuracy, engagement angles, acceleration, maneuverability, shields, armor, self-repair, speed... It's too much context to force into a single number. You get much more complexity but at the end of the day the number is still not reliable.
Thus strength is what it is, a rough but hugely flawed guideline, and as my old stat teacher used to say, garbage in, garbage out. What is the average of a garbage number? Still garbage. The AI uses "strength" for decision purposes and consequently the AI will predictably make suboptimal decisions. We can do better, if we take the time to do so, but it is hard for us to be more than one place, it may even be hard for us to even be that one place, and it is trivial for the AI to be everywhere, so basically swings and roundabouts.
Lastly, as I understand Lanchester's square law, the assumption was two forces of WW1 riflemen where each rifleman can shoot at any enemy rifleman. Are all ships in fleet 1 shooting all guns at optimal range at the entirety of all the enemy ships in fleet 2? That seems very unlikely. And so we're already getting beyond the assumptions of the square law.
This is before we consider that Lanchester's riflemen didn't have different types of bullets with different accuracy and damage profiles, different types of self-repair, different types of damage mitigation, and his riflemen were not teleporting in and out of combat nor did they need to run up under fire to get into engagement range, nor were the opfor running away to make engagement harder.
Basically, Lanchester did the math on more shooters versus better shooters over time in open battle, all things equal, and in that context a given increase in number of shooters provides a bigger advantage than a similar increase of shooter quality. If we're talking about open battle where both forces just trade blows and suffer attrition until the battle is won or lost, that is. How often is that really the case, though?
Or to put it in DWU terms, I don't need 20 ships to remove 10 kaltors from a particular spot. I really only need 1 ship with missiles and better mobility than the kaltors.
-
StormingKiwi
- Posts: 336
- Joined: Thu Feb 11, 2021 6:35 am
Re: QOL - Military Strength for Groups of Entities
Re: averagesSpidey wrote: Thu Feb 24, 2022 8:14 am You're not really responding to the issue of simple averages being largely meaningless without additional context. 5 ships and 2200 strenght means what exactly? One dangerous boat at 2000 and 4 worthless boats at 50 each? Or five significant ships at 420 each? The more ships we add to this mix, the more useless any single number becomes and the more necessary it will be to pause the game and investigate the specific details of the force.
Additionally, the strength number is largely an aggressive simplification. It is the sum of all weapon strength without any consideration of how said weaponry can actually be employed and without any consideration of the durability of the frames that carry said weapons. Put 20 million riflemen in rowboats, if you will, and you have quite a force, but what can they do versus a single destroyer?
And regarding Lanchester, it is not my understanding that his math is a universal truth that is valid for all forms of combat without any qualification. He was considering first classic melee infantry vs classic melee infantry in a frontal engagement and then later riflemen vs riflemen in a similar scenario. His math assumes that all the combatants can actually fire on each other. If that is not the case then I'm not sure his math works.
I am not responding to that issue because it is utterly irrelevant to the discussion. I am not saying report the average military strength and the number of ships. I am saying: report the total military strength and the number of ships.
You're right; reporting a single number is useless, which is why I'm advocating to report two numbers - quality of ships and quantity of ships within that fleet. At the moment, only one number is reported.
Re: strength number,
Yes, all models are wrong, but some models are useful.
Re: your comment on Lanchester's Model,
If you think the assumptions made to apply this attrition model are invalid, then show how that is so. As has already been alluded to, ranges in DW2 are long and cover is scarce.
It would take a bit of work to determine the value of the attrition model exponent (1, 2, some intermediate value, more than 2, etc.) with any certainty. I think it is reasonable to estimate it to be 2, and unreasonable to think it's 1.
Edit: to avoid a double-post.
sorry Spidey, I didn't see your post before I replied.
Just a quick heads up that Zgrssd and I are two separate users.Spidey wrote: Fri Feb 25, 2022 2:44 am Lastly, as I understand Lanchester's square law, the assumption was two forces of WW1 riflemen where each rifleman can shoot at any enemy rifleman. Are all ships in fleet 1 shooting all guns at optimal range at the entirety of all the enemy ships in fleet 2? That seems very unlikely. And so we're already getting beyond the assumptions of the square law.
This is before we consider that Lanchester's riflemen didn't have different types of bullets with different accuracy and damage profiles, different types of self-repair, different types of damage mitigation, and his riflemen were not teleporting in and out of combat nor did they need to run up under fire to get into engagement range, nor were the opfor running away to make engagement harder.
Basically, Lanchester did the math on more shooters versus better shooters over time in open battle, all things equal, and in that context a given increase in number of shooters provides a bigger advantage than a similar increase of shooter quality. If we're talking about open battle where both forces just trade blows and suffer attrition until the battle is won or lost, that is. How often is that really the case, though?
Or to put it in DWU terms, I don't need 20 ships to remove 10 kaltors from a particular spot. I really only need 1 ship with missiles and better mobility than the kaltors.
Respectably, your understanding is incorrect - the book is titled "Aircraft in Warfare", and discusses mounted machine guns vs rifles, aircraft, naval ships, artillery. It discusses the effect of heterogeneous forces and differing attrition rate constants.
Most combat in DW is an open battle, and of course, disengagement is a form of attrition. Self-repair is also a form of attrition - it has the effect that it reduces the rate at which attrition occurs.
Re: QOL - Military Strength for Groups of Entities
Then show us a stream where such a combination happened.Spidey wrote: Fri Feb 25, 2022 2:44 am The problem with averages is by no means a contrived scenario. You have no idea what composition a force has and what you're advocating is seemingly that we can blindly assume all ships in a force are equal. That leads to bad conclusions. We don't know what we don't know, and just assuming a convenient answer doesn't change that.
Show us a Fleet where 90% of the total Fleet power was in one ship.
If it is in any way, shape or form common enough to consider there should be at least one case in the Streams.
And since you argue they exist, you definitely have to know of them. Meaning you could easily find them to show us.
-
Jorgen_CAB
- Posts: 895
- Joined: Wed Mar 17, 2010 7:53 pm
Re: QOL - Military Strength for Groups of Entities
There are a few streams where they repaired some high tech ships and sent them off to fight on their own. Only to experience they would loose against far weaker ships that appeared in numbers. Such as a small fleet of very weak scouts with less than half the fire-power of the stronger ships.zgrssd wrote: Fri Feb 25, 2022 7:23 amThen show us a stream where such a combination happened.Spidey wrote: Fri Feb 25, 2022 2:44 am The problem with averages is by no means a contrived scenario. You have no idea what composition a force has and what you're advocating is seemingly that we can blindly assume all ships in a force are equal. That leads to bad conclusions. We don't know what we don't know, and just assuming a convenient answer doesn't change that.
Show us a Fleet where 90% of the total Fleet power was in one ship.
If it is in any way, shape or form common enough to consider there should be at least one case in the Streams.
And since you argue they exist, you definitely have to know of them. Meaning you could easily find them to show us.
The stronger ship generally managed to flee in most of the situations I saw though and often managed to destroy one or two weaker ship.
I could find myself finding a high tech ships and then adding some frigate escorts to it as more meat means allot in total.
Re: QOL - Military Strength for Groups of Entities
The accurancy or inaccuracy of the strenght number is not part of this discussion.Jorgen_CAB wrote: Fri Feb 25, 2022 11:39 pmThere are a few streams where they repaired some high tech ships and sent them off to fight on their own. Only to experience they would loose against far weaker ships that appeared in numbers. Such as a small fleet of very weak scouts with less than half the fire-power of the stronger ships.zgrssd wrote: Fri Feb 25, 2022 7:23 amThen show us a stream where such a combination happened.Spidey wrote: Fri Feb 25, 2022 2:44 am The problem with averages is by no means a contrived scenario. You have no idea what composition a force has and what you're advocating is seemingly that we can blindly assume all ships in a force are equal. That leads to bad conclusions. We don't know what we don't know, and just assuming a convenient answer doesn't change that.
Show us a Fleet where 90% of the total Fleet power was in one ship.
If it is in any way, shape or form common enough to consider there should be at least one case in the Streams.
And since you argue they exist, you definitely have to know of them. Meaning you could easily find them to show us.
The stronger ship generally managed to flee in most of the situations I saw though and often managed to destroy one or two weaker ship.
I could find myself finding a high tech ships and then adding some frigate escorts to it as more meat means allot in total.
Nor was it any part of the discussion I am having with @Spidey.
-
Jorgen_CAB
- Posts: 895
- Joined: Wed Mar 17, 2010 7:53 pm
Re: QOL - Military Strength for Groups of Entities
But you were asking if there were streams in which there was one ship that had the majority of the combat power and there are such streams. I don't have the energy to go back and point any particular episode out but they are out there.zgrssd wrote: Sat Feb 26, 2022 10:56 amThe accurancy or inaccuracy of the strenght number is not part of this discussion.Jorgen_CAB wrote: Fri Feb 25, 2022 11:39 pmThere are a few streams where they repaired some high tech ships and sent them off to fight on their own. Only to experience they would loose against far weaker ships that appeared in numbers. Such as a small fleet of very weak scouts with less than half the fire-power of the stronger ships.zgrssd wrote: Fri Feb 25, 2022 7:23 am
Then show us a stream where such a combination happened.
Show us a Fleet where 90% of the total Fleet power was in one ship.
If it is in any way, shape or form common enough to consider there should be at least one case in the Streams.
And since you argue they exist, you definitely have to know of them. Meaning you could easily find them to show us.
The stronger ship generally managed to flee in most of the situations I saw though and often managed to destroy one or two weaker ship.
I could find myself finding a high tech ships and then adding some frigate escorts to it as more meat means allot in total.
Nor was it any part of the discussion I am having with @Spidey.
I also said I could see myself having such a fleet in the game as well.
Re: QOL - Military Strength for Groups of Entities
That question does not deal with the accuracy of the combat strenght figures. So why do you keep talking about it?Jorgen_CAB wrote: Mon Feb 28, 2022 11:03 amBut you were asking if there were streams in which there was one ship that had the majority of the combat power and there are such streams. I don't have the energy to go back and point any particular episode out but they are out there.zgrssd wrote: Sat Feb 26, 2022 10:56 amThe accurancy or inaccuracy of the strenght number is not part of this discussion.Jorgen_CAB wrote: Fri Feb 25, 2022 11:39 pm
There are a few streams where they repaired some high tech ships and sent them off to fight on their own. Only to experience they would loose against far weaker ships that appeared in numbers. Such as a small fleet of very weak scouts with less than half the fire-power of the stronger ships.
The stronger ship generally managed to flee in most of the situations I saw though and often managed to destroy one or two weaker ship.
I could find myself finding a high tech ships and then adding some frigate escorts to it as more meat means allot in total.
Nor was it any part of the discussion I am having with @Spidey.
I also said I could see myself having such a fleet in the game as well.
-
StormingKiwi
- Posts: 336
- Joined: Thu Feb 11, 2021 6:35 am
Re: QOL - Military Strength for Groups of Entities
While I can see how this part of the OP quoted below is slightly confusing, I am advocating for a report of military strength to be total (number), not average (number)
Any suggestions to phrase this better? The sticking point seems to be "how many entities the military strength is divided by", which should probably read something like "how many entities the military strength is spread between"StormingKiwi wrote: Sat Feb 19, 2022 9:25 pm Something I'd like to see in DW2 is a number for how many entities the military strength is divided by.
E.g. is it a single ship with a strength of 100, or a group of 8 escorts?

