Slitherine

quote:

Originally posted by Paul Vebber: Sorry - I don't have much time to respond lately, but wanted to quickly chime in here. The M36 seems to have the wrong 90mm gun designated. THis will be fixed, but hopefully one can understand the problems trying to rate everyt weapon form every country to the level of detail we have tried to go here. 10's of thousands of data points, even if we are 99% accurate hundreds will have typos, like the M36 given weapon 64 instead of 65. I'm sure this will get fixed. To the specific point, Please read the designers notes and look back at some of the threads on how the weapons are rated. IT is not just by using "book" penetration data. One can't just take book numbers and compare them at ace value becasue every country had different (and sometimes multiple) definitions of "penetration" that differed in effect by 25% or more. The data I have from multiple sources indiates teh 88 had significantly better AP penetration than the M3 - an example is the www.wargamer.org/GvA data, several other sources corroborate this data. The figures used in the game are backed off form 30 degrees vertical slope at 500m to o slope at 0 meters using the same proceedure for all weapons. The result is consistant - but generally will not match any book data exactly (much of which compares apples to oranges based on differeing definitions of penetration). The game uses a database of "book values" and two different threoretical estimators to come up with penetration. Fire control and accuracy are guesstimates. There is simply insufficient data to reliably set accuracy numbers that aren't skewed by crew proficiency, so these are often based on the original values in the game extended. FOr every instance of a gun being "accurate" are anecdotes to the contray, so barrel length and muzzle velocity are used to baseline raw gun accuracy. Using www.wargamer.org/GvA data as an example - the 88 APCBC ammo MV=1000m/sec down an L71 barrel, while the M3 APBC MV was 808 m/sec down an L52 barrel. Given that data I fudged the M3 to the HIGH side becasue I felt it was more accurate than the raw numbers gave it credit for. There is no "ground truth" metric that can be objectively be debated, hence you have the editor to change them to your taste. But one needs to be careful to judge context when throwing quotes around, and in the context of rating scores of weapons for many of which there is little or no data, a schem of rating values based on info common to them all must be used. That means outliers will exist, and many of those have been argued vociferously (see tiger front turret debate:-) and we change things as deemed appropriate by the OOB working group. As to OOBs and formations one has to remeber that we are under some severe limits based on what teh game allows us to do. We can't just change EVERYTHING as teh bugs that crop back up after being stomped, and persistant "toughies" point out. SO we have to live with "inaccuracies" form strict history (which I would argue is in many ways moot becasue "book" TO&E went out the window by contact with the enemy...but we need some sort of baseline and "book" TO&E - or as close as we can come was chosen. We also have a body of scenarios with which combatibility must be maintained. I submit given the time expended and scope of work undertaken, that the current result is about as accurate as can be had, and we will continue to improve them - a process that could easily take years as data is found and reviewed and discussed. We welcome questions about where the stats come from, and strive to make the OOB's better, but they can never be "perfect" only interanlly more consistent... ...but I do take a bit of offense to being caste a Nazi-phile. I did all the weapon numbers and have gone out of my way sometimes to give credit to the US and UK equipment even when the raw data tended to indicate otherwise. Honest disagreements ABOUT DATA are welcome, but lets keep the snide remarks out of technical debate, eh Greg? [This message has been edited by Paul Vebber (edited October 17, 2000).]

quote:

Originally posted by Paul Vebber: Sorry - I don't have much time to respond lately, but wanted to quickly chime in here. The M36 seems to have the wrong 90mm gun designated. THis will be fixed, but hopefully one can understand the problems trying to rate everyt weapon form every country to the level of detail we have tried to go here. 10's of thousands of data points, even if we are 99% accurate hundreds will have typos, like the M36 given weapon 64 instead of 65. I'm sure this will get fixed. To the specific point, Please read the designers notes and look back at some of the threads on how the weapons are rated. IT is not just by using "book" penetration data. One can't just take book numbers and compare them at ace value becasue every country had different (and sometimes multiple) definitions of "penetration" that differed in effect by 25% or more. The data I have from multiple sources indiates teh 88 had significantly better AP penetration than the M3 - an example is the www.wargamer.org/GvA data, several other sources corroborate this data. The figures used in the game are backed off form 30 degrees vertical slope at 500m to o slope at 0 meters using the same proceedure for all weapons. The result is consistant - but generally will not match any book data exactly (much of which compares apples to oranges based on differeing definitions of penetration). The game uses a database of "book values" and two different threoretical estimators to come up with penetration. Fire control and accuracy are guesstimates. There is simply insufficient data to reliably set accuracy numbers that aren't skewed by crew proficiency, so these are often based on the original values in the game extended. FOr every instance of a gun being "accurate" are anecdotes to the contray, so barrel length and muzzle velocity are used to baseline raw gun accuracy. Using www.wargamer.org/GvA data as an example - the 88 APCBC ammo MV=1000m/sec down an L71 barrel, while the M3 APBC MV was 808 m/sec down an L52 barrel. Given that data I fudged the M3 to the HIGH side becasue I felt it was more accurate than the raw numbers gave it credit for. There is no "ground truth" metric that can be objectively be debated, hence you have the editor to change them to your taste. But one needs to be careful to judge context when throwing quotes around, and in the context of rating scores of weapons for many of which there is little or no data, a schem of rating values based on info common to them all must be used. That means outliers will exist, and many of those have been argued vociferously (see tiger front turret debate:-) and we change things as deemed appropriate by the OOB working group. As to OOBs and formations one has to remeber that we are under some severe limits based on what teh game allows us to do. We can't just change EVERYTHING as teh bugs that crop back up after being stomped, and persistant "toughies" point out. SO we have to live with "inaccuracies" form strict history (which I would argue is in many ways moot becasue "book" TO&E went out the window by contact with the enemy...but we need some sort of baseline and "book" TO&E - or as close as we can come was chosen. We also have a body of scenarios with which combatibility must be maintained. I submit given the time expended and scope of work undertaken, that the current result is about as accurate as can be had, and we will continue to improve them - a process that could easily take years as data is found and reviewed and discussed. We welcome questions about where the stats come from, and strive to make the OOB's better, but they can never be "perfect" only interanlly more consistent... ...but I do take a bit of offense to being caste a Nazi-phile. I did all the weapon numbers and have gone out of my way sometimes to give credit to the US and UK equipment even when the raw data tended to indicate otherwise. Honest disagreements ABOUT DATA are welcome, but lets keep the snide remarks out of technical debate, eh Greg? [This message has been edited by Paul Vebber (edited October 17, 2000).]

quote:

I have done that, using as a first example obvious error regarding the M3 gun as a first test of the good will of that request. I assumed it was not just a rhetorical proposal?

quote:

I was not providing data about penetration and plate angle test varacity or type of ammo or any other technical information or test results. I was simply stating that three sources agreed on the matching power of the 90mm and the 88. The penetration value technical assessment or non technical, was not anywhere in my text. The only statemnet regarding detailed and tested performance was regarding muzzle velocity. This came from tests performed by the US Army during the war. It did not measure penetration to any angle or accuracy or any other aspect of performance. It stated that the muzzle velocity of the 90 was slightly better than the 88. That is all.

quote:

Originally posted by hammerhead: If such issues cannot be discussed here, where a game displays an SS soldier [Nazi] on its opening screen, I must say that this also puzzles me.

quote:

Originally posted by pbear: So on the subject of the OOB errors to report, The French have a 25mm AT gun that has and icon of an M3 Halftrack, as does their Fuel Dump. Version 4.4 Campain Game (I forget the exact name) but the ivasion of NA by the Americans. The OOB is connected to the icon files is there a way to get a print out of the Icon files so that I can tell which number goes with which icon???

quote:

Originally posted by Warhorse: Yes, in the main directory of SPWAW, there should be a txt file called Icon data list Oct.26, or something to that nature. If not in the main, try documents folder in the game, this can be printed out. It has a lot of code numbers designating shp files and such, too much to go into here, but the number before the vehicle/weapon name, is the icon number entered in the oob at graphic slot.

quote:

Originally posted by Unhappy Camper: Personally I don't think the Matrix created scenarios are all that sacred. If they hadn't screwed up the Oob files in the first place, they wouldn't be faced with redoing all of the scenarios. I appreciate that a game like this is a lot or work, but if you're going to be releasing something like this to the public, even for free, it should be accurate, and work properly. There are some glaring errors in the army lists. Most notably is the armor ratings for a lot of the tanks. There is a definite leaning towards the Germans in this. The front turret on the Tiger is listed as 176mm, when all the historical info I've seen says 100mm. The Sherman M4A3 in the game is listed around the 75mm mark, I've seen at least four sources list the frontal hull armor as 100 to 110mm. And even the organizations are screwy. According to the TO&E included with the game a US Mech Inf platoon should have three squads of men, with MGs and Mortars and Half-Tracks. The game only gives them two squads, and adds an AT gun, but you can't fit everything into the HTs provided. They've contradicted themselves. And they've included many weapons that were never used operationally. The 128mm AT gun, the Maus tank, 105mm RCL. All of them German by the way. I understand that play must be balanced, but then don't claim you're going for historical accuracy. I've seen several messages arguing about the 50cal, guys, it was a big, heavy machine gun. I fired a heavy bullet a really long way. The Germans never, never developed a heavy machine gun. So the 50cal should have a huge range advantage. As to the 90mm AA guns for the US, later models were in fact equipped with sites for use against ground targets, and they were almost certainly more accurate than the German 88.

quote:

Originally posted by Unhappy Camper: According to the TO&E included with the game a US Mech Inf platoon should have three squads of men, with MGs and Mortars and Half-Tracks. The game only gives them two squads, and adds an AT gun, but you can't fit everything into the HTs provided.