Here you are an example of Bofors L/70 inaccuracy:

I'm not trying to be snarky. The point is, some of the data on which we are working I may not disclose unless there is some other open source to which I can point, otherwise I might be in breach of contract. So I have to walk a fine line here. What we fixed was a conversion error, not the source data themselves.
Also, I admit that I sometimes ask "control questions" to find out if a poster actually knows what he's talking about or if it is just gut feeling. While the associated mathematics aren't exactly a rocket science, many people still don't have a very precise idea of the properties of a Gaussian standarddeviation.

The mathematics itself are essentially undisputed. What it boils down to is the veracity of the parameters involved. These can be obtained from some firing tables (if they do contain them, which isn't always the case), or from manufacturer claims (which are often deliberately obfuscating the relevant parts, like "dispersion: less than .2 mil" (great... is that the standarddeviation? The 50% percentile? Is it the diameter, or the radius?)).

So you see, I'm going with the parameters that I know but can't tell, or with parameters that are public but leave room for interpretation, or sometimes I have to estimate parameters from what I consider "comparable" rounds. Depending on how firm the foundation of our parameters are and depending on the credibility of a disputing source, I may come to the conclusion that a change is justified.

It seems that you are alluding to that old Challenger ammunition debate. That was a case of a source that I can't disclose but which I believe to be very credible, a very unfortunate combination for public debate. In the case of the 40mm rounds, I'm working on the basis of what I've been told by the Swedish Army as a part of the development contract (they may or may not have told us the truth). Coming from a development contract means that I must not disclose the exact parameters as a condition of the contract, unless of course someone else - it could be you - points to an alternative source; at that point the NDA regulates that the data are no longer to be treated as "secret". At that point I could also try and weigh the credibility of the new source against what we were told at the time of the CV90/40 development, and eventually reassess our parameter estimate.
Apparently the Swedish Army was happy with the results, which suggests that either the spread in reality is exactly like this, or that it is different but not so much that our model negatively affects the training, or that it was all an elaborate hoax to deceive the world about the performance of the Bofors gun. Take your pick.




Yes. Eventually we will come to that, too.

Well, your explanations are a bit...chaotic.

It is obvious that almost all current weapon data you had to implement in SBPro PE are kept secret. Anyway you stated before that you use open-source data. Yet later you claimed there are also some secret data that military customers disclosed to eSim Games. But now you tell us such data for 40 mm Bofors L/70 gun accuracy can be a hoax. So why do you so strongly believe in L27A1 penetration data? They also can be a hoax. So everything may be a hoax...

What I cannot find in your way of thinking is lack of being RATIONALE. Look at movie I made during firing Bofors L/70 gun. Only two of eight rounds were able to hit stationary target at 2 km range which means 25% "accuracy". Anyway BMP-2 was not destroyed and as a result it destroyed CV9040 with ATGM. In sum in your sim CV9040 is most often defenseless against very weak opponent like BMP-2 which can destroy it without problems launching ATGM (what they do always at longer ranges) which flies about five times slower than 40 mm rounds. It is obvious such firepower would have been unacceptable to the Swedish Army and CV9040 would have been kicked-off in a matter of minutes after achieving such poor firing results during real tests. Yet Swedish Army accepted CV9040 into service and procured about five hundred of those vehicles as the sole IFV platform for decades to come. That clearly proves CV9040 firepower is much better than that found in SBPro PE. Most probably CV9040 needs only 1-3 shots to kill BMP.

What can we do now? Basically the only viable solution is to replace CV9040 and CV9035 by CV9030FIN (which has decent fire accuracy and large ammo loadout) in all SBPro PE scenarios. Sorry, but I cannot worry out anything else...

PS. Is it possible to replace Finnish writings found in CV9030FIN's GPS by English ones? There are English writings in CV9035 and CV9040. Moreover not many people speak Finnish.

There is no 'proof' of accuracy from a sample of 8 shots - except that it falls below 100% and exceeds 0 %.
These points are given by the presence of both misses and hits, but there is no solid grounds for a statistical claim from such a small sample size.

Under the assumption that the interpretation error is such that a conversion factor was applied to take a 1SD interval (0.68+) to a 50 centile value for each of the two axes. This leads to a 25% hit rate, rather than the 48% hit rate of the 1SD box..

I ran a batch of tests, each of 10 groups of 100 samples for the 1SDx1SD and 50%x50% boxes:
Summary of the results:
Mean hits For 50%x50% box: 28.6 per group, minimum 17 in a group, maximum 37 in a group.
Mean hits for the larger 1SD x 1SD box: 45.5 per group, minimum 1 in a group, maximum 63 in a group.

Note that the '1' is an outlier, and would fall outside the 3SD interval.

Also note that despite 1000 samples, this is still insufficient to be confident of the mean values produced - an additional 3000 samples would greatly enhance reliability.

As for 2000m being 'typical', I would consider that somewhat unlikely - against a point target, something around 1000-1500m seems far more probable in typical European terrain - this could be even shorter in Scandanavian terrain, being very built up, wooded and without substantial relief allowing elevated sight-lines over terrain obstacles...

Where we could, we documented the source of our data in the various documents that can be found in the Steel Beasts installation, partially also as chapters in the user's manual.
That we are bound by NDAs can be no surprise to anyone who has ever worked in a commercial/industrial environment, be it software development or something else. You can hold that against us, but then please do so against any other software developer who makes any kind of simulation of products that exist in the real world.


Yes - where they are actually available (especially dispersion figures are particularly hard to come by), and we use open sources in the absence of anything better. Naturally, if we accept a development contract from an army and that army provides us with sources like firing tables to allow us to develop the parameters for our ballistic model, you surely do not expect us to withhold the results from the Personal Edition.


Of course if someone wanted to believe in conspiracy theories, I have no actual proof that all these armies do not perform elaborate schemes to feed us with false data. I don't believe it, but the only way how we could possibly be sure would be to do our own ballistic test firings - something for which we do not have the opportunity, the time, the need, or the money (and I'm pretty sure that you as the customer wouldn't want to finance the costs of, say, firing 2000 rounds of 120mm tank cannon ammunition to establish a dispersion parameter estimate with mediocre confidence ... assuming that this would actually be permitted (which would never be the case, as much fun as it might be).
All that I did was to point out the rather academic (or paranoid) possibility that we are just led to believe certain dispersion parameters. I do not want to suggest that I really believe that these parameters aren't to be trusted. I think that they are correct, and I think that SB Pro does a good job to visualize the actual effect, and I think that you simply overestimate the accuracy of these autocannons - but could I prove all this in court?
No.
Is there "reasonable doubt" about our models? I don't think so, apparently you do, and I let the rest of our readers here draw their own conclusions.


I can't tell, and I told you so on several occasions before. You want to hear it again?

I can't tell you. Once more?

I can't tell you.

You either trust me on this one, or you don't. I don't care what you want to believe. I've been about as open and as forthcoming about the fundamentals of our work as I can be. Either you accept that there are limits to what I can tell, or you don't. If you want to score easy rhetorical victories over me in internet fora - fine, you win. You are aware however that this is like beating a one-legged opponent in an ass-kicking contest. If that makes you happy, keep going.

If that event ever kicks off, can I volunteer as gunner...mind you that we would wear out a few barrels for this fun :-D

Regarding dispersion:

We considered a gun "good" if the MAJORITY of rounds where inside a PC-size target on the guns CORE range (NOT the maximum range).
I admit that our caliber was a good bit below 40mm though.

If you expect "surgical" precision with an auto-cannon at 2000m you will get disapointed 8-)

A BMP2 has a very low and compact siluate with EXTREME front slope. Hit at the wrong angle, and you're #%&*$#.
Allthough at hit with a MK just below front deck, should have a really high chance of a mobility-kill(i.e. damage/destroy the engine).

Spearfish seems to underestimate the BMP2 a wee bit.(Well, in SB it is allways 100% maintained and has an expert crew that never fliches ;-) ) But the AT-5 missle is quiete a potent weapon, esp. against an IFV. 580mm of penetration will suffice for most of 'em, and I don't see anaything wrong about that model.
Engaging it frontally at long ranges (1800 and above) has a good chance you'll come out 2nd. Getting it front the flank or at shorter ranges, the odds are on your side.
In short: I don't see anything wrong with that model and the way it is actually helps training.

It seems we cannot agree, so goodbye Ssnake!




No! The funny thing is now we have this surgical precision in SBPro PE when firing MK 30-2 and Bushmaster-I/II guns! I can score 70-90% hits firing them against BMP-2 at 2 km. Moreover BMP-2 is almost always quickly destroyed thanks to their high rate of fire in salvo mode, especially true for MK 30-2 gun, and large ammo loadout. I often can destroy three BMP-2s at 2 km before their ATMGs reach my IFV! And that precisely is what I expect from Bushmaster-III and Bofors L/70 guns - unfortunately without success so far.

What can I see now? Both those guns can hit targets at 0.5-1 km distance with good accuracy. That's derision...



No. Today BMP-1 and BMP-2 are basically an utter crap. In real life they are outclassed by modern Western IFVs mainly due to ridiculous armor protection and potent western autocannons firing APFSDS rounds. Please check if this "extremely sloped frontal armor" was penetrated by 12.7-14.5 mm AP rounds in the past. Those cannons can shatter BMPs at 1-2 km range in all possible scenarios. Moreover today's FCS nullifies BMP's small silhouette advantage.

Also note ATGMs is anti-tank weapon. Now outdated AT-5A missile cannot destroy modern Western tanks, and what? Is that the last remaining true BMP's weapon because Soviet 30 mm and 73 mm guns are useless? Come on, that's another proof BMP is finished as a serious combat platform in XXI century. It remains in service in CIS and Third World countries, sometimes in large numbers, but only due to long-term economic crisis following the USSR's collapse. Only recently Russian Army started to research completely new IFV concept. Maybe in 5-10 years new Russian IFV will be introduced.

Spearfish, did you actually read my post, or you just quoted it?

Again, to your points: Agreed the 25mm and 30mm cannons. They seem to be more acurate as the 40 /35 mm guns.
Yet I have no data or experience how EXACT FIGURES should be.

If you are on a flat ground and both sides agree to a 19th style shootout, the a FCS an nullifiy a small frame.
In most other aspects this statement is rediculous.
Small frame makes you harder to spot and you can use the terrain more efficiently. These are abilities that can in fact
nullifies even the most advanced FCS.
Yes the armour CAN be penetrated by 12.7 and 14.5mm guns...just not at ALL angles. And if you hit the front and penetrate
there is the engine/transmission block and another plate of armour behind that. You may very well get a mobility kill, but there is still an chance for the crew. And as the SB-crew are essentially "fearless", they may very well still guide the AT-5 to your nose(doubt they would do that in RL)

I don't really get the intend of your last §, but to pick on some of the points:

The 73mm low presure gun of the BMP1 did some nasty damage to US CAV units in GW 1. You just got to read BDA from that time.
Yes the ATGM is an "anti tank weapon"...so is there any reason not to use it against IFV? Which they can take out quiete reliably...

In 5-10 year? The BMP3 is alive and kicking since the 1990ies.
What new concepts do you mean?

PS.:
About the Bushmaster MK

Just tried 40 rounds(APDS) on a BMP-2 at 2000m...scored 22 hits.
If you get to 70-90% hit rate, you're better than me with that gun.