This game vs Armored Warfare

[Sabrewolf]

I haven't made a purchase of the game myself either, but when I posted on AW's Discord the penetration values of the various weapons for this game, a so-called consultant actually had the audacity to laugh and make fun of this program. He actually had the arrogance to think that AW was a better more accurate platform.

I was also a Beta Tester for the AW game and have been playing it for 8 years now, it is amazing how much the game has fallen. Version0.18 was the last best version they had before the Russian devs moved in and screwed it all over. These people actually believe that the armor of the M1A2C is so weak it can be drilled with a simple 30mm gun as if it has zero armor at all, which we all know is so far from the truth.

I'm just looking for a more accurate tank simulator to play with because I am getting tired of the arcade style crap of AW and WOT of which I have been playing off and on for 11+ years

[Ssnake]

Generally, we don't comment on other products (mostly because I don't have the time to test them extensively, partly because we know very well what we want to do, so we don't have to imitate others, partly because it's just bad style). We let the documentation of our work and the simulation results speak for itself.

 

A lot of people choose to believe whatever causes the least cognitive dissonance to them. We try to eliminate biases from our estimations as much as we're aware of them. But of course, estimates remain estimates, and nobody should use our figures as an authoritative source.

Knowledge is classified. Those who know can't talk, those who talk might occasionally know a thing or two, but with overwhelming likelihood they don't possess the full picture. And then there's that class of particular idiots that publish classified material just to assure their Alpha Nerd status. I never envied that position.

[Volcano]

So the person thought that AW's armor or ammo data was more realistic than SB. That's a good one! 🤣

 

Jokes aside, what it is really about is how confident does a developer feel with their numbers? We feel very confident in ours, and in the few areas we don't we revise periodically. If the AW developer feels confident in their values, then good for them.

[Arch]

SB is a simulation software that just happens to have a consumer version. A comparison to AW is a bit weird seeing as they're not in the same genre.

 

I guess with the benefit of a doubt, the AW person is not aware that their numbers that they likely picked off Google are not necessarily accurate and would need to be correlated to more data and thus don't match up with more simulator oriented products.

 

The last time I revealed some Forbidden Knowledge, almost everyone using the same product had to revise everything that had to do with what I talked about, but I can understand that military-related professional projects might not have that kind of communication SOP. They can just figure out the right numbers themselves, anyway.

[Ssnake]

1 hour ago, Arch said:

SB is a simulation software that just happens to have a consumer version. A comparison to AW is a bit weird seeing as they're not in the same genre.

That distinction line appears somewhat blurry to me. Since I don't know anything about that AW title, I can't really comment on it. But I assume that both their developers and we are working from similar, unclassified sources. I always felt compelled to demonstrate to the public our methodology (at least up to the limit where it might start hurting our business). That way, people can make up their minds whether they find our assumptions reasonable or not - unlike completely closed source software where the developers basically demand blind trust.

 

You can make a game focus on entertainment, and still do a decent job with the underlying models and parameters. Our focus is on crew procedures, tactical education, and of course we'd like to take pride in good correlation between our simulation results and reality when it comes to damage mode, AI behavior, etc.

 

 

At the end of the day, claiming to be "the most realistic" combat simulation (without further qualifications) is first and foremost clamoring from the marketing department. Being a mouse warrior has little in common with actual combat, certain aspect can't be adequately modeled. Some epistemiological humility would become, well, the entire internet, I guess.

[Captain_Colossus]

watching videos of world of tanks / war thunder / armored warfare  / my impression is that even if they used the same figures as steel beasts, they would still generate very different results because their assumptions about what happens behind the armor are a bit different- hits exchanged between same tier vehicles are usually critical and kill outright. the graphic window they use to visualize the results serve the purpose to titilate the player, but the results are nearly always the same. subjectively, two of the toughest vehicles to ko in steel beasts are the MRAPS and the pirahana based 90 and 30 mm gun carriers, not because of their armor values, but because perforations do not do much to them- as a result i have seen tanks deplete their ammunition loads while they still remain in play; every imprortant component might be damaged, and they might be immobilized, but the attempt is hopeless . still, with the mission editor, i often work in a destroy if condition attached to all vehicles in order to approximate crews bailing or incapacitated and so on, so that opposing units do not spend all their efforts on disabled but still live targets which refuse to die

Edited February 6 by Captain_Colossus

[Ssnake]

What happens after armor perforation it anyone's guess. There simply are no good data available; the best being the early 1960s Conqueror trials which have been declassified a few years ago. I think it's a safe assumption that the basic principles still apply. At the same time engineers learned a lot how to protect crews even after penetration (e.g. the introduction of spall liners, which were completely unknown at the time of the Conqueror trials - as well as insensitive munitions, flame-retardant paints and clothing, quick-reaction fire suppression systems, and probably more).

In SB Pro we have a high spatial resolution as far as component location is concerned. We have some idea how resistant some components are to spall/fragmentation damage. But can anyone say with certainty that in case of an armor perforation in a certain part of the hull or turret the radio will be damaged with a 5% likelihood? 6%? 4%? 15%? 30%?

At some point you'll go with whatever you find "plausible" - which is a function of one's expectations (IOW, a bias), and then you try to set up consistent rules between vehicles of a similar class or design era, and simply roll with it.

 

In the end, our damage model serves certain purposes that are subordinate to the purpose of the simulation as a tool for crew training and education. We probably overestimate component failures at least in vehicles that have crew positions, simply because our customers want their crews to correctly diagnose their faulty systems and apply corrective measures. We may slightly underestimate the lethality of armor perforations (by how much is anyone's guess) simply because you don't learn much from a "you're dead" end screen. And overall, our software shall still give a realistic impression (not a prediction!) about combat outcomes in a given scenario where crew training is not the focus.

 

Someone working in a classified facility explained to me once why he purchased SB Pro for his work. Yes, they have "better tools" and better data to predict the result of specific armor/ammunition interactions. These are, to use a photography analogy, a macro lens to take pictures of pieces of flowers, like pollen sticking to the legs of a bee. As awesome as these tools are, they do not give you an impression of what a garden is. SB Pro, to him, is the impressionistic painting of a garden. It's fuzzy in the details, but it gives a unique perspective of the bigger picture. And then there's constructive simulation tools that are the equivalent of an aerial picture of a garden colony or a suburb. Lots of gardens, but you can't make out the individual flowers anymore.

 

Other armor games might not be an impressionistic painting of a garden. But maybe they still count as paintings in a different art style, if you want to take the picture analogy to an extreme. The real danger lies in mistaking image quality of a render engine as a proxy for simulation accuracy. Even people who should know better make that mistake all the time.

 

"Beauty is truth, and Truth is beauty"

[Arch]

4 hours ago, Ssnake said:

The real danger lies in mistaking image quality of a render engine as a proxy for simulation accuracy. Even people who should know better make that mistake all the time.

The real old heads will do the opposite, and assume the title with flashier graphics can't possibly be a better sim, because all of the current leading consumer sims are a generation or two behind in 3D rendering, and it's been that way as long as I remember.

Edited February 7 by Arch

[Ssnake]

Visual and simulation fidelity are orthogonal to each other. That's the takeaway.

[thewood]

9 hours ago, Arch said:

The real old heads will do the opposite, and assume the title with flashier graphics can't possibly be a better sim, because all of the current leading consumer sims are a generation or two behind in 3D rendering, and it's been that way as long as I remember.

Just a few years ago, there was ample justification for that line of thinking.  A dev's resources are a finite pool and can be very restrictive in the planning of the game/sim.  Devs typically chose good graphics vs barely passible simulation or barely passible graphics vs plausible simulation.  And the curve on that is constantly shifting based on tool kits/APIs greeting released.

 

There is a third factor and that's motivation.  Investor- or market-driven devs are generally pushed for cutting edge graphics with simulation deprioritized to be good enough on launch day.  Client-driven devs tend to build more sim-oriented, depending on clients or niche market.

[Captain_Colossus]

dcs and microsoft  flight simulator have the graphics and the meat of 'simulation' and which i think disprove the argument that users are forced to compromise one or the other. there are console action flight combat simulators developed in the same generation (like call of duty are more similar to interactive movies than 'games'), but even these no longer have any advantage in graphics- if they ever did at all. flight simulators always had the most resources in graphics and presentation relative to their generation; if you go back to the simple flight simulators from 40 years ago, as crude as they were, there were never anything close on anything on a console, and flight action games scaled roughly similar in terms of visual results- only the most advanced arcade boards and home computers could do anything like that and were quite similar to one another visually with vector or flat shaded polygon models.  for my time however, dcs and ms flight simulator just are not practical to get into. and so there is the niche that MMO tank games fill- though i have to wonder about the effect that  the "free to play, pay to win" schema psychologically manipulates the player base to grind and invest more money and time into them. the game experience without that is actually quite shallow, it is the fact that players are psychologically invested into it because a carrot is constantly dangled in front of them with the vague hope of 'winning' which is also the meat of the experience. you could replace the tanks with elves and dragons and fantasy creatures to get a similar effect

[Ssnake]

3 hours ago, Captain_Colossus said:

dcs and microsoft  flight simulator have the graphics and the meat of 'simulation' and which i think disprove the argument that users are forced to compromise one or the other.

That's hardly translatable outside of the flight sim sector. Flight sims are immensely more popular. Note that even then MS did abandon its flight sim for several years before coming back with their latest version.

[Captain_Colossus]

i do not dispute that at all- rather it is implied that the relationship between graphics and sim-like fidelity is not antagonistic, but converge with several factors in play, one of which as you say depends on the audience and popularity of the subject matter to begin with, certainly you see that in other categories, say the resources committed to racing simulations vs golf simulations. but certainly better graphics does contribute to better simulation results if all the other ingredients are also present. we can compare the barren environments of 1980s DOS games to see just how poor they were in modelling the subject matter that they purported to represent. i think in large part player's imaginations filled in the gaps, and probably explain the strange nostalgia players still have for games like M1 tank platoon I & II when they clearly are obsolete

[Arch]

4 hours ago, Captain_Colossus said:

dcs and microsoft  flight simulator have the graphics and the meat of 'simulation' and which i think disprove the argument that users are forced to compromise one or the other. there are console action flight combat simulators developed in the same generation (like call of duty are more similar to interactive movies than 'games'), but even these no longer have any advantage in graphics- if they ever did at all. flight simulators always had the most resources in graphics and presentation relative to their generation; if you go back to the simple flight simulators from 40 years ago, as crude as they were, there were never anything close on anything on a console, and flight action games scaled roughly similar in terms of visual results- only the most advanced arcade boards and home computers could do anything like that and were quite similar to one another visually with vector or flat shaded polygon models.  for my time however, dcs and ms flight simulator just are not practical to get into. and so there is the niche that MMO tank games fill- though i have to wonder about the effect that  the "free to play, pay to win" schema psychologically manipulates the player base to grind and invest more money and time into them. the game experience without that is actually quite shallow, it is the fact that players are psychologically invested into it because a carrot is constantly dangled in front of them with the vague hope of 'winning' which is also the meat of the experience. you could replace the tanks with elves and dragons and fantasy creatures to get a similar effect

Yet the best consumer flight simulations have 90s tier graphics with some modern features slapped on top.

[Captain_Colossus]

but if you are trying to make the argument that graphics is unrelated either to fun and entertaining or 'realistic' or is somehow at odds with these things, i fail to see it-  and which looks like to me some kind of over used fallacy.  of course you certainly can compromise one or the other with limited budgets and resources, and if you can creatively get around this sometimes it works for the better (take for example the early star wars films- the low budget practical effects, set designs and look of the costumes and rubber muppets are superior to the lackluster computer generated scenes of later films- the designers used props which they bought in hardware stores and possibly apparel stores modified for the look). i am not really a huge fan of flight sims, but falcon 3.0, one of the few i spent any time with was in my view something special even by today's standards. but i returned to it just a few years ago and it does not quite capture the same experience as it once did. however the point i make is that if anyone argues that graphics is somehow icing on the cake at best, then why does esim periodically update its render engine- what purpose does it serve if not also to achieve more parity with the world it attempts to simulate? if that were not the case you could replace all the models with rasterized sprites or perhaps every vehicle with a 3D wireframe block with the words 'tank' or 'pc' inscribed on them, and render the ground terrain as a few shaded terrain tiles and a few tree models every square kilometer or so matched with the few sterotypical pyramids like you saw in those older programs representing hills  and have no different experience than what you have now- if that is what you mean

Edited February 8 by Captain_Colossus

[Ssnake]

To answer your question, our philosophy is that graphics are what psychologists call a "hygiene factor". You don't want your product to look ugly. But it doesn't have to look world-class either. Creating AAA graphics still requires disproportional allocation of scarce resources to artwork, and artwork alone. Dollars spent on artwork development can't be spent on better AI, or whatnot.

As long as you have a limited budget - eSim definitely does, though it has become much better over the past fifteen years - you must prioritize. We chose to emphasize tactical depth and fidelity of simulation outcomes in our development efforts. So we have three times as many programmers as we have artists. Usually the ratio in game development is five artists to one programmer.

I disagree that better lighting effects result in better simulation fidelity. They may certainly improve immersion, they may increase the fun factor - all that is undisputed. But the decision whether a certain vehicle gets destroyed by a certain round and at a certain angle and distance does not depend on the quality of the smoke that it will emit afterwards. I would tolerate the argument with respect to our high-resolution terrain engine. It certainly looks better, but its undeniable contribution to better simulation results is what's far more important; it offers more opportunity for combatants to seek cover, hence lifts survivability of units in battle to more realistic levels. I will admit that we chose to work on the high resolution terrain for the wrong reasons - better looks. I'm grateful that it was a benign error that reaped unexpected benefits. But the new terrain engine was far more than just some shader effect or other visual sleigh of hand.

[Captain_Colossus]

5 hours ago, Ssnake said:

To answer your question, our philosophy is that graphics are what psychologists call a "hygiene factor". You don't want your product to look ugly. But it doesn't have to look world-class either. Creating AAA graphics still requires disproportional allocation of scarce resources to artwork, and artwork alone. Dollars spent on artwork development can't be spent on better AI, or whatnot.

As long as you have a limited budget - eSim definitely does, though it has become much better over the past fifteen years - you must prioritize. We chose to emphasize tactical depth and fidelity of simulation outcomes in our development efforts. So we have three times as many programmers as we have artists. Usually the ratio in game development is five artists to one programmer.

I disagree that better lighting effects result in better simulation fidelity. They may certainly improve immersion, they may increase the fun factor - all that is undisputed. But the decision whether a certain vehicle gets destroyed by a certain round and at a certain angle and distance does not depend on the quality of the smoke that it will emit afterwards. I would tolerate the argument with respect to our high-resolution terrain engine. It certainly looks better, but its undeniable contribution to better simulation results is what's far more important; it offers more opportunity for combatants to seek cover, hence lifts survivability of units in battle to more realistic levels. I will admit that we chose to work on the high resolution terrain for the wrong reasons - better looks. I'm grateful that it was a benign error that reaped unexpected benefits. But the new terrain engine was far more than just some shader effect or other visual sleigh of hand.

 

i admit my question is rhetorical and that i am already certain of what i intend already, nor do i think i have made myself clear. i will try again- when i say that graphics matter, i do not necessarily mean a particular thing like lighting (although in a game like doom back in the 1990s that was extremely immersive and important to the gameplay- the long dark corridors with malfunctioning lighting flickering and this sort of thing with the monsters breathing somewhere around the corner was what that was all about- not per se the violence and the killing which would have been just another wolfenstein 3D, but it was the environmental lighting and thoughtfulness of the game design which was different from anything else at the time. clearly it was not a detailed simulation of shooting, the gameplay is rather shallow- so doom was a very rare case where graphics and sound were more important than the gameplay or the accuracy of the ballistics or whatever. no one complained about that in doom and so of course it went on to fame and fortune).

 

i mean by graphics as they relate to a 'simulator' just the overall impression of what you see on the monitor somewhat resembles reality. i would argue that in the 1980s and 1990s this wasn't really possible, and the kinds of home computers available to consumers at that time outside of research institutes or military simulators could scarcely if at all be used as a training aid or tool- some may point to microsoft flight simulator was actually used by flight students for instrument navigation and familiarity but still- flight sims in those days could not render environments where a user could discern the difference between 500 meters or 5000 meters above the ground- hence why landing and take off or low level flight would be entirely more difficult than it should be in real life, nor could night flying really be trained in this manner. from a ground level perspective, you could not render an environment which would not look anything like reality- so a game like electronic arts' seal team released in the 1990s could not render thick forest jungles and all the information content needed to replicate special forces operations in vietnam, that is, enemy parties could clearly be seen at any distance and there simply wasn't enough cover or concealment that personal computers at the time could render to make that a more realistic experience. so graphics are related to function in that respect- better graphics mean the potential for a better or more functional experience for lack of a better description, if of course there is the rest of the content to go along with it.

 

and so there is a false dichotomy which often arises when someone objects and points out the such and game is pretty but it is a shallow game or whatever- and of course they are right, but that does not mean that graphics simply do not matter. again because if graphics do not matter, then you could render your entire vehicle fleet in 50 triangles and then your m2 bradleys will probably look no different than an m1113 and then when games in those days did look like that when their objects were almost indiscernible looking masses, they often had to use artificial cues like labels of the vehicle type somewhere on the screen or in the HUD or whatever to inform users.

 

so again this isn't to say that graphics alone matter, i will also point out plenty of cases where mainstream games are completely shallow and boring at least in my view which are really just graphical experiences (which is in my opinion what games people play on their smart phones are) so that's not what i am saying. there are two extremes: graphics either don't matter, or only graphics matters, and either of which i think are wrong. now of course you will often see visitors on steel beasts channels on youtube complain about the graphics and so on, and then other users bring attention to esim's purpose and business model being military client driven first and then graphics second, but this doesn't mean however the extreme case that therefore graphics do not matter to a simulator to steel beasts at all- so for example more terrain detail naturally lends itself to simulation results rather than simply looking better. they coincide. and so of course the more detailed terrain in steel beasts as it is now is clearly better looking and more functional than steel beasts version 1

Edited February 8 by Captain_Colossus

[Captain_Colossus]

and so i will add, and this not meant to discredit the idea that substance does matter, but what i am saying is that graphics do matter, and it is largely subconscious that they matter, you do not have a choice of it. if you were not born blind, your brain is attempting to map the world around you by estimating what it thinks is out there based on the information coming in through your eyes- therefore the external world you believe that you see is an experience inside your skull. your mental map of what is 'out there' is inside your brain. and what you see is your brain attempting to make sense of the world visually, combined with your sense of hearing and smell and balance and other tactile simulation.

 

i had not played many computer games or simulations outside of steel beasts for the last 20+ years. as such i did not realize how far computer games graphics had come since then, because i had not been paying that much attention.

 

then i played dcs and skyrim for the first time and i was quite astonished what i was seeing.

 

flight sims without special cockpit simulators cannot replicate g effects and the feel of the aircraft and the feel of wind currents buffeting and things like this to really get an appreciation of what real flying is like- however what i saw in dcs and microsoft flight simulator comes as close as they can just with graphics techniques however they are doing that- without the aforementioned aids in the real world to orient your sense of movement in the environment that you would have in real life- nevertheless say what you will about the simulation or whether it is fun or not, i found how far they come with a sense of scale and distance which i never saw before. take the water in the below screenshots- you intuitively sense how close or how far away it is visually at different altitudes and distances- whether it is the lighting or texture results or rendering techniques at far distances or whatever it is- it looks like real water at the distances and altitudes they are trying to represent. this is much different than i remember where in the past the water really looked the same at any distance but for perhaps at a few simulated meters distance from the ground level, you may see some pixels or a few lines flash by to try to give the impression of the surface beneath the player's point of view.

 

so this is not to put steel beasts down at all, because i will argue that the steel beasts terrain engine has improved and come far so that there is more plausible representations of an actual environment than what tank sims used to be doing.20 or 30 years ago. just look at m1 tank platoon 1 and compare it with any game which is released now and see immediately how utterly artificial those environments were and therefore could not generate results that would look anything like reality

 

 

Edited February 8 by Captain_Colossus

[Sabrewolf]

On 2/5/2024 at 9:22 PM, Volcano said:

So the person thought that AW's armor or ammo data was more realistic than SB. That's a good one! 🤣

 

Jokes aside, what it is really about is how confident does a developer feel with their numbers? We feel very confident in ours, and in the few areas we don't we revise periodically. If the AW developer feels confident in their values, then good for them.

One on the really HUGE problems I had with them is their modeling of the M1A2C Abrams. They gave her VERY VERY weak side armor. A Hunter AFV, or any vehicle, with a 30mm cannon will vaporize the Abrams in seconds, however, when the Abrams fires back, it will give a result of NO EFFECT. Somehow these people think that all American equipment is garbage, yet the Russian armor is superior and can't be hurt.

[Ssnake]

In all fairness, side hulls of Leo 2 and M1 are rather thin. They are no wonder weapons. And Russian tank designs aren't quite as bad as many people would have you think in the mid 1990s. The question is maybe not so much whether the numbers they are working with are very different, but what the model is when a perforation occurs.

[Fat Otaku]

On 2/6/2024 at 3:22 PM, Volcano said:

So the person thought that AW's armor or ammo data was more realistic than SB. That's a good one! 🤣

 

Jokes aside, what it is really about is how confident does a developer feel with their numbers? We feel very confident in ours, and in the few areas we don't we revise periodically. If the AW developer feels confident in their values, then good for them.

Hey, just wondering how confident you guys are with armor of M60A3's turret? Will you guys consider revising it now that we have people actually measured the turret armor?

[Ssnake]

What kind of outcome would you expect?

[iamfritz]

My teen son plays War Thunder, and the hit/damage philosophy there is , in a word, completely and totally ridiculous. OK, more than a word. I play some times too, only to scratch my World War II itch, but I never last long. Last time I played an M-42 Duster jumped my M-4A3E8 and blasted away at my armor as I put 3 x 76mm rounds into its front hull from 8 inches away. War Thunder style, I died in seconds.
Now, he's very involved in the game and he knows how they calculate impact and all the involved math, which he tried to explain to me. But we often have such conversations, and we always wrap up these convos by simultaneously saying, "Russian bias".
The way weapon effectiveness, armor effectiveness, track maneuverability, etc., etc. are always changing I have no idea how/why he still plays. I know he enjoys the work/reward of earning new tanks and planes, but he is often frustrated when he tries a well established move to kill an enemy, only to die and then remember "they" changed the values for that certain kind of attack.
I ramble, but my point is it's very similar with Armored Warfare and World of Tanks. Values and qualities are always changing. The changes often favor Eastern powers and deny Western weapon fans their gaming truth.
It's nice to have a sim where "This is what it is, thank you very much, no we're not changing it."
Ah well. </rant>

[Ssnake]

Well, we are changing things, just not in pursuit of a particular game mechanic or with the purpose of shifting the balance in favor of a certain vehicle. New information comes to light, we revise numbers. After a thorough check. Both other game developers and us do commercial software development. But given the different nature of revenue streams, we can afford to maintain independence from the preference of a certain user group, or investors, and it turns out that for professional usage purposes, trying to remain neutral and realistic is the best overall strategy.

Of course, we're still operating in an environment where a good part of some information is classified, so we have to offer our best guess. But then again we don't claim that you'll get The Truth from Steel Beasts, just something that is sufficiently plausible and fit for the purpose.

[iamfritz]

Yes, and I think we all get that.
But you don't change the numbers to quiet down the squelch from a majority of players who cry out loud "That this tank is too powerful, my favorite tank that's less powerful in the real world is now not fun to play because it blows up too much!"
You change it because you receive new data from an actual battlefield where new data can be gleaned.

And by the way I just made a killer taco gulosh.