Good video. Well thought out! Several years ago for Netmech we were doing similar testing because of the Gauss rifles not functioning properly at any range, but we got the damage values via hex editing. A guy named Highlander made a nice spreadsheet of it all. I'll have to go see if I have that saved somewhere or look into recreating it, but yes they definitely were not accurate 100% in mechlab. Regarding the Medium Pulse Lasers though, I definitely can attest to them being beasts through similar testing that I did. The damage gets even more convoluted when you do a side by side comparison between Titanium and DOS/95!!! If you compare the DPS of MPulse in Titanium vs DOS/95 you get even more DPS on the Titanium version. Side note: In testing online using 1.0 gravity er mediums cap out at about 310 meters, LRM 20's do around 800-850. These values are based upon where they work before they disappear. You definitely are correct about there is a maximum amount of missiles in the air as for a very long time it has been a tactic to chain your missiles so your opponent loses their missiles if they are using group fire; however, I am pretty sure that there is a maximum amount of lasers in the air count as well, but the number is significantly higher so that you would not notice unless you had four or more players with lasers playing at the same time.
Very nice! Though you may be interested to learn that, according to the strategy guide they released with the game, the pulse lasers work on very unusual logic. When the devs were interviewed for the guide, this is how they explained it: The game runs two checks every time a pulse laser hits a target, one for each pulse in a salvo. If the game detects that the first pulse hits a target, that pulse deals full damage, and the damage of the second pulse is ignored, even if it hits the target. If the game detects that the second pulse hits a target, it then checks to see if the first pulse in the same salvo has already hit. If not, then the second pulse gets to deal the full damage of the pulse laser.
@@Pixelmusement Lol, no kidding! I'm glad lasers behaved themselves in 3 onwards. According to the devs, the idea was that they wanted the lasers to obey tabletop rules as much as possible. So instead of spreading damage between two pulses (trading better chances to hit for more reliable DPS) they just tried to make it as easy as possible to deal the base damage value of the laser.
This viewer remembers figuring out that's how they behave, way back when. It started making a big difference when spraying at hilariously mobile targets! Did that interview include anything on how PPC worked? They seemed to cause targets to die earlier than expected, and it felt like they applied excessive crit hit rolls due to splash and DoT effects, since anything with an XL engine ior lots of volatile items would explode and die with one or two hits. PPC worked very well in versions of the game not afflicted by the collision with firer bug, especially up close, but one PPC in the centre torso was also great for picking off targets at long range. Low max active projectile count, slow travel and long range added up to harsh punishment for sloppy use, but it was great fun to run perpendicular to target heading and lead a pretty string of blue pearls into a slower-moving 'mech so far away it's invisible in the distance and unable to retaliate.
@@rainbowsprinkles4234 Would you believe that 1.0 versions of MW2 had PPCs using Gauss Rifle firing code and missile damage code? Guess what they forgot to adjust for in doing so. If your answer was "tabletop damage values," you would be correct, and it would explain why they were so absurdly lethal.
As Modern Vintage Gamer would say... "Mistakes were made." lolz I'm not much of a Battletech connoisseur, and I've played only a few games with less than five total hours of gameplay time overall, but even I can understand and appreciate just how precise given values should be. After all, such is quite important in RPG mechanics, whether it be a videogame or table-top game or whatever medium.
This was amazing. I always knew the weapons stats weren't entirely accurate but it's actually pretty awesome to see that the environment changed the ranges too.
I just remembered. IIRC, The Medium Pulse damage is incorrectly higher than what it's supposed to be in Mech2 and fixed in mw2:Mercs. Which is why at 24:48 you're doing more damage than even a LARGE pulse laser. If you look at the stat list for weapons, they're supposed to scale up normally. (3, 7, 10, damage per statcard. Likewise for mercs, it's 3,6,9 damage per statcard. In both cases, the medium should do LESS than a large.) If it's not a bug, you'd have to be a fool / noobtrap to use the large pulse (heavier, more heat, more slots, and less damage than medium lasers).
Youre certainly not the only one to doubt the mechlab stats! Thanks for the excellent testing! I imagine Mercenaries would be even harder to test as many weapons seem to take into account if you even have the target selected (especially with the "easy control mode")
+Skull Gun MW2:Mercs has BOTH Inner Sphere and Clan tech, so that alone doubles the workload, and on top of that there are discrepancies with the weapons depending on if you're running the DOS or Win95 versions of the game. Not to mention, the MechLab has bugs that can result in things not working out properly and the collision detection in the sim itself is buggy and sometimes allows shots to pass STRAIGHT THROUGH. Doing this same testing with Mercs would be a nightmare and nearly impossible because of all of this. :P
@@Pixelmusement OH btw im not sure you mentioned this in the video, but what "difficulty" did you have the game set to in options? From what I know this effects how much damage is done by the player and AI. (Not sure it effects things like bridges though)
@@Skull_Gun I had it set to "Hard" as I never played this game on anything less; I found it a bit too easy on the lower settings even as a kid, however the way the numbers were working out it may be that the damage values themselves aren't being adjusted, rather instead the amount of armour the enemies get is adjusted. :B
Just for reference: Think of fixed point as like an integer value N, but you treat it in your code as N / (some power of 2) so that conversion to a whole number is just one shift instruction away. For example if I chose bit 16 as the "decimal point", then 0x10000 = 65536 = 1.0 and 0x8000 = 32768 = 0.5. Of course this means with 32 bits you can only count from -32767 - (65535/65536) to 32768 + (65535/65536), but it means you can do finer calculations without an FPU. x86 processors did not integrate the FPU until the 486, and later the Pentium, though early 486s with faulty FPUs were sold as not having one, and non-Intel CPUs eventually integrated the FPU as well. Good examples of fixed point include DOOM and Wolfenstein 3D. Use of fixed point in DOOM for your position is the reason you can NOCLIP, walk backwards out of a map, and eventually hit a point where your 3D view goes wonky and then disappears, and when you turn around, you're on the opposite side of the map.
+TheGreatCodeholio I know how fixed point works; I dabbled in it briefly from the coding side of things but long after floating point had become lightning fast so abandoned my tinkering with it shortly after I started. Because of how much slower floating point is prior to Pentium CPUs, even with a math co-processor installed, most 3D engines are fixed point engines, but Mech2 came out after Pentiums debuted and also around the time the floating point bug with the early Pentiums was making news in tech circles, so the Mech2 docs make it VERY CLEAR that they're using fixed point math and that the game is not affected by said Pentium bug. :B
+Zontar82 There's actually a fairly stupid reason why we all keep going back to it: Because Mech3 and Mech4 don't like being run on modern systems. I've already spent two hours today trying to get Mech4 working on Win10 to no avail. :P
@@Pixelmusement Took me a while too but Mech4 is running fine now. You require dgvoodoo2 if you have too modern a video card, and you need to break the cd copy protection if you are running it from a disc, as it's incompatible. That's pretty much it. Mech3 will also run on dgvoodoo2, and you will need rivatuner to limit the frames to 60 fps for the physics to work properly.
Andy Yiu Actually, MW4 Mercs works absolutely fine. The Battletech reddit discovered it basically works if you delete a certain file in the game for modern systems.
Great research, very well-done video. Some reflections: 1) ranges in the Battletech tabletop game were not meant to be literal... they were intentionally made shorter than what was realistic in order to facilitate good tactical gameplay given movement speeds, 2) it sounds like the small laser was nerfed a little bit in order to make the small pulse laser more attractive... in the tabletop game, pulse lasers give a very important bonus to hit which is not really possible in the computer game, so the alternative would be to make pulse lasers shoot a little farther and hit a little harder than their regular laser counterparts. Again, this is in line with the design philosophy of the tabletop game (even when the stats end up differing from the tabletop), as these things have always been designed with gameplay in mind and simulation taking second seat. In any case, it's great to see the differences and put hard numbers to them!
I had the modding/"gamepiece" manual for MW2 a while ago. I don't recall doing much with it after some severe arguments resulted from my trying to get a copy of the various editions of the base game and NetMech to see if there were greater differences. -- trying to remember if it included weapon definitions or only those needed to add vehicles and missions.
What an interesting breakdown - it's been a while since I gave Mech2 a proper play, but I'd forgotten about the reticle lighting up outside of range. The convergence factor is new to me - though I'm sure I experienced it for sure.
I'm going through the same data-mining exercise for Mechwarrior 1 at the moment, with the intention of optimising speedruns. I can recognise a lot of the little problems and foibles you went through, because I've seen them all myself - which means both of us were on the right track in our own ways. The strategy I adopted was to assume the smallest weapon, the machine gun, did one unit of damage, and every other weapon would be an integer multiple of that, which seems to be borne out in my testing. DPS and range are a lot harder to figure out, as there's no way to get the game to run at a consistent speed, range to target is never explicitly displayed, and weapon ranges are listed as "short", "medium" and "long" (and there are variances between weapons within those categories). The best generalisation I've come up with is that the shorter the range for a given weapon, the more DPS it has, so if you can get a teammate to distract an enemy Battlemaster, you can sprint in with a Locust and blow it to pieces in seconds with machine guns with impunity. I've also seen the same smallest-weapon-does-one concept in another game, Fire Fight, which the developers got *very* adventurous with afterwards. The starting machine gun does one damage per hit, the swarmer missiles also do one per hit but launch four at a time, the plasma gun follows a pattern of 2, 2 and 1 damage (three shots doing 5 DMG and four doing 7 DMG absolutely did my head in until I figured it out), homing missiles do a flat 7 damage, the cannon does *up to* 12 depending on how close you are (again, no range display, so hard to quantify), and I think grenades do a flat 12 with a certain radius as well. I'd love to see more videos like this, about mining data from old games - it's a super interesting topic to the right people, and this one's given me a few ideas for new avenues to explore with Mechwarrior 1, so thanks for the effort!
I try to do at least one Pro video every couple of months and there will definitely be more such videos in the future where we try to determine the raw values of things in games! :)
I would have guessed that next week's game is Evasive Action. But you already did that one in episode 195, so I guess there's another game that allows Fokker Dr I vs spaceship battles.
Back in the day i used to play this using a "nice" flightstick that had throttle-lever. After remembering that, i need to find dedicated usb-throttle that fits at the end of keyboard.... :DDDD
+Wolde Munster I have a Thrustmaster HOTAS T.Flight X joystick. Inexpensive and features the full throttle and stick setup. That said, I find Mech2 and Mech3 work better with mouse and keyboard, whereas Mech4 definitely has the best joystick optimization of all the MechWarrior games! :B
Alright, update: you're absolutely right, CERLL do 8 dmg only. The algebra just wasn't adding up, contradicting equations ^^ That 40 armor DWF leg goes in 8 shots, not 7. My mistake. Poor counting on my part. Weird I made like 3 tests giving me 7 shots, but I guess I just was tired today and forgot how to count :D heh Very conclusive as a result. Good thing I made that mistake: I would never have found your video and would have been redoing your work. Only slower, on dires, on Goat Path, and without the fancy DPS testing you did. All in all, people like you, and MMPD Radick are exactly what make the MW2 community awesome. So, thanks again. MW2, you just got decrypted :P God, I'm such a kid ^^
That DPS chart actually checks out perfectly to me. I never liked the ER lasers, I always preferred the pulse lasers and I usually boated mediums or large depending on the mission. Pulse lasers and LRMs were usually my go to, didn't really like SRMs or any of the ballistic weapons. I ran the AC a few times and wasn't putting enough ammo in them because I was a dumb child so I just went back to pulse lasers and LRMs lol. Of course all of that changed with MW4 Mercs, I loooove me some AC cannons.
In MW3 the ACs are STUPIDLY powerful. Combo that with taking out a mech just by shooting out one leg and all of a sudden a single AC20 loaded with ammo can carry you through an entire mission. :P
PPCs also have a limit of how many can be on screen at once, at least in Mechwarrior 2 Mercenaries. I want to say it's the same in Mech2 and Ghost Bear's Legacy. I remember taking a Warhawk out where I removed everything except the 4 PPCs and added more heat sinks and found the limit pretty quick. I want to say it's 6 or 8, but I cannot recall right offhand.
You're awesome! And patient, too :) This is so useful, and IMHO fixes one of this game's rare flaws, by giving us the truth! This is especially important in the GBL expansion, which is sometimes brutally hard. The stats completely lie, and glorify CERLL, while underrating a devastating weapon in this game: the CERPPC. If you base you weapon choice on stats,( which you should), and trust the game's stats, (which you shouldn't), and play an unforgiving game like GBL (which you should), you get wowed, and not in a good way either lol. Soo, thanks a lot for all your work. This is more useful than you know. It's also funny how alike we both are. Both did tests, both piloted MAD IIC (cuz it's not an omni, so I don't feel guilty changing the engine), both used weaponless dires as enemies, both put 40 and 0 armor on their legs (I'm serious), sole difference: I stripped the armor off the left leg and you the right, I put 3 identical dires in the enemy team, and I kept their 300XL, cuz, lore, I'm infected with adhesion to lore, sry lol. Tried to make sense what the actual structure values are. Came up with the conclusion they're 50% of the max armor for that component. E.g. 21 for a DWF leg (since max armor for a DWF leg is 42). Definitely found out about CERSL being 3 dmg (instead of 5!!, which is a massive 40% less than promised) Finding the actual damage value is actually not that hard: pick the weakest weapon you can, aim at the toughest, clearly defined, immobile target you can. I used the CERSL as a weapon, and started hunting Dire Wolves in the desert on Goat Path. Trick is to leg the powered down mechs while looking at the number of times you shoot, while disabling the first enemy star by removing one leg from their Dires. Heh. Armorless DWF legs go with 7 CERSL shots, 40 armor fresh legs go with 21 shots. Difference in shots? 14. 40 hit points / 14 = 2,8571428571428571428571428571429 Rounded we get 3. What makes testing a pain is that sometimes, a laser shot will go right through a mech and deal it no damage (the blast isn't colliding with the target). If you're using Enhanced Imaging like I do, this doesn't show (or at least I don't see the difference when a laser connects). This so messes up stats. I thank you for the CERLL 8 dmg stat finding. Couldn't determine if it was 8 or 9. Will try to understand why a 40 armor dire leg (21+40 = 61) goes with only 7 shots (7*8=56). Heck, maybe I'm wrong with the structure thing, but I don't think so... Will see. Your testing is a sight for sore eyes, motivates me to do some more testing on my end lol. I read somewhere that pulse laser damage is dealt on the first pulse hit, and the second is only cosmetic. I cannot yet confirm or deny this but believe it to be true (would explain some of your results). I also read somewhere that there can be a limit of how many hits per second a target can take, and that boating lasers agressively can be inefficient. My testing shows this to be rarely true, but most of the time the game does everything correctly: group firing 10 CERSL does exacly the same damage as firing them chained at a slow rate. An important thing to know is that damage output depends on difficulty. Or maybe enemy armor becomes tougher then. There is a massive difficulty difference between easy and medium. I did all my testing on easy to see if the game actually buffs our weapons. It doesn't. It would appear Easy is the actual Lore friendly Realism mode, while Medium and Hard artificially gives them twice or thrice the armor and structure. Next testing I intend to do is try to destroy an armorless dire with MG crits to the CT., before the Enhanced Imaging display giving it a red (less than 25% health, armor and structure included). Finally, believe me, I'm very glad to see a fellow pilot enjoy the game of my childhood in 2020. Heck, I even have, believe it or not, dreams about the Mannedorf Trial of Grievance map, and nightmares about a Warhawk looking for me and trying to kill me! All this while not having played this game for years. Once I finally dropped in a Trial on Mannedorf and saw that Canyon, I immediately recognized the place from my dreams! It was awesome, I thought I had invented that map in my dreams, deep inside my head, after about 2 decades, my mind still hadn't forgotten it. Ah, this game... It might be a quarter a century old, yet still feels futuristic, exotic, mysterious, distant. It aged beautifully. Yeah, it's a legendary classic, glad to see poeple play it and watch this :)
Final testing for today: MG crits against engine. I destroyed 20 armorless, weaponless dires, going for the CT. 30% (6 kills) happened while the mech was on yellow damage (backstabbed, this can't be a accidental cockpit hit), 70% (14 kills) happened due to MG destroying the CT component (DWF was with a red torso on enhanced imaging display). Soo, critting engines seems to be a thing in this game. Not that I ever did that. Due to the game's rather simple damage indicators, you generally know it's time to crit only when the CT is red. At that time, you might as well finish it with raw firepower. I guess now it's time to test MGs during campaign missions to see if it is useful, or just a nice lore friendly thing that's not practical in actual combat (in MW2, of course).
I recently started playing the Mech 2 trilogy again, and I suspected there was some Tom-Foolery going on with the Mech lab stats. Mercenaries made some more changes. They definitely made the Auto-cannons and LRM’s weaker, and the pulse lasers even more powerful.
(12:29) A bit unrelated, but I've noticed some games are rather inconsistent with how they treat splash damage. In some games, if you hit a bad guy with a rocket launcher (or an equivalent), it'll blow the bad guy apart with nearby bad guys taking splash damage. In other games, when a bad guy takes a similar direct hit, he absorbs _all_ of the damage, leaving no splash damage to surrounding bad guys; if the hit is anything less than a direct hit (glancing), then splash damage is dealt normally (as far as I can tell).
+Christopher Noel Yeah, it really depends on the game. Generally speaking, you have "impact" damage and "splash" damage, the damage caused by the weapon making impact, and the splash caused to surrounding targets, although various other methods exist too, such as ONLY having splash but calculating the quantity of splash damage based on distance from the point of impact. However, here in Mech2, splash damage is not instant, it occurs over a couple seconds, and it's very minimal, so if you ONLY get hit with the splash from a weapon it does very little damage. :P
Final note: Lasers were dissapointing in this game, but that's just because Activision decided to nerf 'em. Since I'm a sniper this is kinda a big thing. You end up with a weapon that has a 1.5 heat/dmg ratio (CERLL, way too hot IMO), and no nice recycle. Learning to use the CERPPC and just outright spamming it at a range the enemy can't fire back from is verry rewarding. Just put plenty of CDHS to offset the heat and you're good to go. Now I will basically carry loads of CERPPC and Gauss Rifles.
*Lasers weren't hitting at max range* Hmm... I'm wondering how much you fired the weapon. I mean is it that the range stated is wrong or that at that distance it is hard to hit things? I didn't play the game back in the day but I watched other people play and I seem to recall them rolling dice to see if they hit or not.
There are no dice rolls in the weapon mechanics in Mech2. In fact, there's virtually no RNG at all. I was firing the weapons more than enough to determine that RNG was not a factor and for results that seemed unusual I triple-checked them. :B
This gets even more complex if you include Mech2 vs Mech2:Mercs,[A quick look shows LRMs have a range of 1000m vs 630m in Mw2. Though maybe those are all 'clan' weapons? Many other weapons are identical stats though.] as well as possible changes from patches.[Though I don't see any yet.] There were also bugs in various versions of Mech2, and Mech2:Mercs. One forum poster: > Shots from weapons phasing through target when fired in rapid succession, making weapons with quick fire times close to useles (e.g. Pulse Lasers, Ultra Autocannons) > Small Pulse Lasers having no hit detection due to bug in weapon's effective range If that forum post is true, it definitely changes how viable fast firing weapons are [if they bug out and don't count to your DPS]. That's also something that could be tested.
Those bugs are in reference to MW2:Mercs and were not a factor in the original MW2. The main bug in the original MW2 was that one side of the torso took WAY more damage than the other side, suggesting the overlapping collisions volumes were simply choosing one over the other when both were triggered, rather than trying to figure out which side was actually hit. :P
I thought the UAC in Mechwarrior 2 rapid fired when you held down the button to simulate the increased fire rate from tabletop. For those who don't know the UAC in tabletop can fire 2 shells per turn.
+OptimusShr Nah. Every weapon in the game will auto-fire if you hold the button down. However, it's coded in such a way that if you had let go of the fire button and press it back down while a weapon is still reloading/recharging it will NOT auto-fire once ready as this normally causes a skip-over to the next weapon in sequence.
The damage stats on the board game does not translate accurately in many cases to the computer game. The best example is using my favorite mech, the Nova, which is armed with 12 Extended Range Medium Lasers. In the Battletech board game a single alpha strike with all 12 lasers that hits the same spot can destroy an assault mech even if it is hit in the center torso, which is the most heavily armored part of the mech. In Mechwarrior 2 an alpha strike on a light or medium mech that hits its arm or leg which should have completely blown it off often doesn't.
I know, which was the entire reason for making this video because the stats are pulled straight from the tabletop game yet they never made sense here for whatever reason! :P
I typically only go through the effort of making tables and such available when I feel the information is worth referencing. Here, it didn't really seem worth going through the effort. :P
I don't understand how anybody could suffer through testing and calculating all this stuff. However, I think the results are really interesting... so go right ahead ;) Great video, in any case! :D
+R. F. R. For the same reason why grinding in games doesn't bother some people. It may be tedious, but it's ultimately not all that hard and so you just kinda fall into it and can escape from reality for a bit. That's what doing all this testing mostly felt like. :B
Good video. Well thought out!
Several years ago for Netmech we were doing similar testing because of the Gauss rifles not functioning properly at any range, but we got the damage values via hex editing. A guy named Highlander made a nice spreadsheet of it all. I'll have to go see if I have that saved somewhere or look into recreating it, but yes they definitely were not accurate 100% in mechlab.
Regarding the Medium Pulse Lasers though, I definitely can attest to them being beasts through similar testing that I did. The damage gets even more convoluted when you do a side by side comparison between Titanium and DOS/95!!! If you compare the DPS of MPulse in Titanium vs DOS/95 you get even more DPS on the Titanium version.
Side note: In testing online using 1.0 gravity er mediums cap out at about 310 meters, LRM 20's do around 800-850. These values are based upon where they work before they disappear. You definitely are correct about there is a maximum amount of missiles in the air as for a very long time it has been a tactic to chain your missiles so your opponent loses their missiles if they are using group fire; however, I am pretty sure that there is a maximum amount of lasers in the air count as well, but the number is significantly higher so that you would not notice unless you had four or more players with lasers playing at the same time.
Very nice! Though you may be interested to learn that, according to the strategy guide they released with the game, the pulse lasers work on very unusual logic. When the devs were interviewed for the guide, this is how they explained it:
The game runs two checks every time a pulse laser hits a target, one for each pulse in a salvo. If the game detects that the first pulse hits a target, that pulse deals full damage, and the damage of the second pulse is ignored, even if it hits the target.
If the game detects that the second pulse hits a target, it then checks to see if the first pulse in the same salvo has already hit. If not, then the second pulse gets to deal the full damage of the pulse laser.
+subtlewhatssubtle ...I'm REALLY glad future MechWarrior games actually made the lasers... you know... ACT like lasers? ;)
@@Pixelmusement Lol, no kidding! I'm glad lasers behaved themselves in 3 onwards.
According to the devs, the idea was that they wanted the lasers to obey tabletop rules as much as possible. So instead of spreading damage between two pulses (trading better chances to hit for more reliable DPS) they just tried to make it as easy as possible to deal the base damage value of the laser.
This viewer remembers figuring out that's how they behave, way back when. It started making a big difference when spraying at hilariously mobile targets!
Did that interview include anything on how PPC worked? They seemed to cause targets to die earlier than expected, and it felt like they applied excessive crit hit rolls due to splash and DoT effects, since anything with an XL engine ior lots of volatile items would explode and die with one or two hits.
PPC worked very well in versions of the game not afflicted by the collision with firer bug, especially up close, but one PPC in the centre torso was also great for picking off targets at long range. Low max active projectile count, slow travel and long range added up to harsh punishment for sloppy use, but it was great fun to run perpendicular to target heading and lead a pretty string of blue pearls into a slower-moving 'mech so far away it's invisible in the distance and unable to retaliate.
@@rainbowsprinkles4234 Would you believe that 1.0 versions of MW2 had PPCs using Gauss Rifle firing code and missile damage code? Guess what they forgot to adjust for in doing so. If your answer was "tabletop damage values," you would be correct, and it would explain why they were so absurdly lethal.
As Modern Vintage Gamer would say... "Mistakes were made." lolz
I'm not much of a Battletech connoisseur, and I've played only a few games with less than five total hours of gameplay time overall, but even I can understand and appreciate just how precise given values should be. After all, such is quite important in RPG mechanics, whether it be a videogame or table-top game or whatever medium.
This was amazing.
I always knew the weapons stats weren't entirely accurate but it's actually pretty awesome to see that the environment changed the ranges too.
This confirmed my childhood suspicions that it was always more fun to get as many high caliber Autocannons loaded, and watch mech limbs fly.
I just remembered. IIRC, The Medium Pulse damage is incorrectly higher than what it's supposed to be in Mech2 and fixed in mw2:Mercs. Which is why at 24:48 you're doing more damage than even a LARGE pulse laser. If you look at the stat list for weapons, they're supposed to scale up normally. (3, 7, 10, damage per statcard. Likewise for mercs, it's 3,6,9 damage per statcard. In both cases, the medium should do LESS than a large.) If it's not a bug, you'd have to be a fool / noobtrap to use the large pulse (heavier, more heat, more slots, and less damage than medium lasers).
Youre certainly not the only one to doubt the mechlab stats! Thanks for the excellent testing!
I imagine Mercenaries would be even harder to test as many weapons seem to take into account if you even have the target selected
(especially with the "easy control mode")
+Skull Gun MW2:Mercs has BOTH Inner Sphere and Clan tech, so that alone doubles the workload, and on top of that there are discrepancies with the weapons depending on if you're running the DOS or Win95 versions of the game. Not to mention, the MechLab has bugs that can result in things not working out properly and the collision detection in the sim itself is buggy and sometimes allows shots to pass STRAIGHT THROUGH. Doing this same testing with Mercs would be a nightmare and nearly impossible because of all of this. :P
@@Pixelmusement OH btw im not sure you mentioned this in the video, but what "difficulty" did you have the game set to in options?
From what I know this effects how much damage is done by the player and AI. (Not sure it effects things like bridges though)
@@Skull_Gun I had it set to "Hard" as I never played this game on anything less; I found it a bit too easy on the lower settings even as a kid, however the way the numbers were working out it may be that the damage values themselves aren't being adjusted, rather instead the amount of armour the enemies get is adjusted. :B
Interesting, wonderful to see a full breakdown of Mech 2's weapon stats.
Damn, it looks even better than I remembered. I really need another one for VR.
Just for reference: Think of fixed point as like an integer value N, but you treat it in your code as N / (some power of 2) so that conversion to a whole number is just one shift instruction away.
For example if I chose bit 16 as the "decimal point", then 0x10000 = 65536 = 1.0 and 0x8000 = 32768 = 0.5. Of course this means with 32 bits you can only count from -32767 - (65535/65536) to 32768 + (65535/65536), but it means you can do finer calculations without an FPU.
x86 processors did not integrate the FPU until the 486, and later the Pentium, though early 486s with faulty FPUs were sold as not having one, and non-Intel CPUs eventually integrated the FPU as well.
Good examples of fixed point include DOOM and Wolfenstein 3D. Use of fixed point in DOOM for your position is the reason you can NOCLIP, walk backwards out of a map, and eventually hit a point where your 3D view goes wonky and then disappears, and when you turn around, you're on the opposite side of the map.
+TheGreatCodeholio I know how fixed point works; I dabbled in it briefly from the coding side of things but long after floating point had become lightning fast so abandoned my tinkering with it shortly after I started. Because of how much slower floating point is prior to Pentium CPUs, even with a math co-processor installed, most 3D engines are fixed point engines, but Mech2 came out after Pentiums debuted and also around the time the floating point bug with the early Pentiums was making news in tech circles, so the Mech2 docs make it VERY CLEAR that they're using fixed point math and that the game is not affected by said Pentium bug. :B
TheGreatCodeholio 386s also had an FPU that could be purchased separately.
incredible how, with so many mechwarrior games made after this,we always go back playing this one . for me it's clearly the best
+Zontar82 There's actually a fairly stupid reason why we all keep going back to it: Because Mech3 and Mech4 don't like being run on modern systems. I've already spent two hours today trying to get Mech4 working on Win10 to no avail. :P
@@Pixelmusement Took me a while too but Mech4 is running fine now. You require dgvoodoo2 if you have too modern a video card, and you need to break the cd copy protection if you are running it from a disc, as it's incompatible. That's pretty much it. Mech3 will also run on dgvoodoo2, and you will need rivatuner to limit the frames to 60 fps for the physics to work properly.
@@Pixelmusement Vengeance and Black Knight might cause trouble, but the Mektek version of Mech 4 Mercenaries should work fine.
Andy Yiu Actually, MW4 Mercs works absolutely fine. The Battletech reddit discovered it basically works if you delete a certain file in the game for modern systems.
Pixelmusement The reddit discovered that a simple file deletion solved almost all the problems. Me and my mates play through gameranger weekly.
Great research, very well-done video. Some reflections: 1) ranges in the Battletech tabletop game were not meant to be literal... they were intentionally made shorter than what was realistic in order to facilitate good tactical gameplay given movement speeds, 2) it sounds like the small laser was nerfed a little bit in order to make the small pulse laser more attractive... in the tabletop game, pulse lasers give a very important bonus to hit which is not really possible in the computer game, so the alternative would be to make pulse lasers shoot a little farther and hit a little harder than their regular laser counterparts. Again, this is in line with the design philosophy of the tabletop game (even when the stats end up differing from the tabletop), as these things have always been designed with gameplay in mind and simulation taking second seat. In any case, it's great to see the differences and put hard numbers to them!
I had the modding/"gamepiece" manual for MW2 a while ago. I don't recall doing much with it after some severe arguments resulted from my trying to get a copy of the various editions of the base game and NetMech to see if there were greater differences. -- trying to remember if it included weapon definitions or only those needed to add vehicles and missions.
What an interesting breakdown - it's been a while since I gave Mech2 a proper play, but I'd forgotten about the reticle lighting up outside of range. The convergence factor is new to me - though I'm sure I experienced it for sure.
God, so nerdy! Love it.
I'm going through the same data-mining exercise for Mechwarrior 1 at the moment, with the intention of optimising speedruns. I can recognise a lot of the little problems and foibles you went through, because I've seen them all myself - which means both of us were on the right track in our own ways.
The strategy I adopted was to assume the smallest weapon, the machine gun, did one unit of damage, and every other weapon would be an integer multiple of that, which seems to be borne out in my testing. DPS and range are a lot harder to figure out, as there's no way to get the game to run at a consistent speed, range to target is never explicitly displayed, and weapon ranges are listed as "short", "medium" and "long" (and there are variances between weapons within those categories). The best generalisation I've come up with is that the shorter the range for a given weapon, the more DPS it has, so if you can get a teammate to distract an enemy Battlemaster, you can sprint in with a Locust and blow it to pieces in seconds with machine guns with impunity.
I've also seen the same smallest-weapon-does-one concept in another game, Fire Fight, which the developers got *very* adventurous with afterwards. The starting machine gun does one damage per hit, the swarmer missiles also do one per hit but launch four at a time, the plasma gun follows a pattern of 2, 2 and 1 damage (three shots doing 5 DMG and four doing 7 DMG absolutely did my head in until I figured it out), homing missiles do a flat 7 damage, the cannon does *up to* 12 depending on how close you are (again, no range display, so hard to quantify), and I think grenades do a flat 12 with a certain radius as well.
I'd love to see more videos like this, about mining data from old games - it's a super interesting topic to the right people, and this one's given me a few ideas for new avenues to explore with Mechwarrior 1, so thanks for the effort!
I try to do at least one Pro video every couple of months and there will definitely be more such videos in the future where we try to determine the raw values of things in games! :)
I would have guessed that next week's game is Evasive Action. But you already did that one in episode 195, so I guess there's another game that allows Fokker Dr I vs spaceship battles.
Back in the day i used to play this using a "nice" flightstick that had throttle-lever. After remembering that, i need to find dedicated usb-throttle that fits at the end of keyboard.... :DDDD
+Wolde Munster I have a Thrustmaster HOTAS T.Flight X joystick. Inexpensive and features the full throttle and stick setup. That said, I find Mech2 and Mech3 work better with mouse and keyboard, whereas Mech4 definitely has the best joystick optimization of all the MechWarrior games! :B
Alright, update: you're absolutely right, CERLL do 8 dmg only. The algebra just wasn't adding up, contradicting equations ^^ That 40 armor DWF leg goes in 8 shots, not 7. My mistake. Poor counting on my part.
Weird I made like 3 tests giving me 7 shots, but I guess I just was tired today and forgot how to count :D heh
Very conclusive as a result. Good thing I made that mistake: I would never have found your video and would have been redoing your work. Only slower, on dires, on Goat Path, and without the fancy DPS testing you did.
All in all, people like you, and MMPD Radick are exactly what make the MW2 community awesome.
So, thanks again. MW2, you just got decrypted :P
God, I'm such a kid ^^
That DPS chart actually checks out perfectly to me. I never liked the ER lasers, I always preferred the pulse lasers and I usually boated mediums or large depending on the mission. Pulse lasers and LRMs were usually my go to, didn't really like SRMs or any of the ballistic weapons. I ran the AC a few times and wasn't putting enough ammo in them because I was a dumb child so I just went back to pulse lasers and LRMs lol. Of course all of that changed with MW4 Mercs, I loooove me some AC cannons.
In MW3 the ACs are STUPIDLY powerful. Combo that with taking out a mech just by shooting out one leg and all of a sudden a single AC20 loaded with ammo can carry you through an entire mission. :P
really useful video for Clan Mechwarriors. marvelous.
PPCs also have a limit of how many can be on screen at once, at least in Mechwarrior 2 Mercenaries. I want to say it's the same in Mech2 and Ghost Bear's Legacy. I remember taking a Warhawk out where I removed everything except the 4 PPCs and added more heat sinks and found the limit pretty quick. I want to say it's 6 or 8, but I cannot recall right offhand.
It may be a per-mech thing... Even the missile limit may be a per-mech thing; I didn't take a closer look into that... :o
You're awesome! And patient, too :) This is so useful, and IMHO fixes one of this game's rare flaws, by giving us the truth!
This is especially important in the GBL expansion, which is sometimes brutally hard. The stats completely lie, and glorify CERLL, while underrating a devastating weapon in this game: the CERPPC. If you base you weapon choice on stats,( which you should), and trust the game's stats, (which you shouldn't), and play an unforgiving game like GBL (which you should), you get wowed, and not in a good way either lol.
Soo, thanks a lot for all your work. This is more useful than you know.
It's also funny how alike we both are. Both did tests, both piloted MAD IIC (cuz it's not an omni, so I don't feel guilty changing the engine), both used weaponless dires as enemies, both put 40 and 0 armor on their legs (I'm serious), sole difference: I stripped the armor off the left leg and you the right, I put 3 identical dires in the enemy team, and I kept their 300XL, cuz, lore, I'm infected with adhesion to lore, sry lol.
Tried to make sense what the actual structure values are. Came up with the conclusion they're 50% of the max armor for that component.
E.g. 21 for a DWF leg (since max armor for a DWF leg is 42).
Definitely found out about CERSL being 3 dmg (instead of 5!!, which is a massive 40% less than promised)
Finding the actual damage value is actually not that hard: pick the weakest weapon you can, aim at the toughest, clearly defined, immobile target you can.
I used the CERSL as a weapon, and started hunting Dire Wolves in the desert on Goat Path. Trick is to leg the powered down mechs while looking at the number of times you shoot, while disabling the first enemy star by removing one leg from their Dires. Heh.
Armorless DWF legs go with 7 CERSL shots, 40 armor fresh legs go with 21 shots.
Difference in shots? 14.
40 hit points / 14 = 2,8571428571428571428571428571429 Rounded we get 3.
What makes testing a pain is that sometimes, a laser shot will go right through a mech and deal it no damage (the blast isn't colliding with the target). If you're using Enhanced Imaging like I do, this doesn't show (or at least I don't see the difference when a laser connects). This so messes up stats.
I thank you for the CERLL 8 dmg stat finding. Couldn't determine if it was 8 or 9.
Will try to understand why a 40 armor dire leg (21+40 = 61) goes with only 7 shots (7*8=56). Heck, maybe I'm wrong with the structure thing, but I don't think so... Will see. Your testing is a sight for sore eyes, motivates me to do some more testing on my end lol.
I read somewhere that pulse laser damage is dealt on the first pulse hit, and the second is only cosmetic. I cannot yet confirm or deny this but believe it to be true (would explain some of your results).
I also read somewhere that there can be a limit of how many hits per second a target can take, and that boating lasers agressively can be inefficient.
My testing shows this to be rarely true, but most of the time the game does everything correctly: group firing 10 CERSL does exacly the same damage as firing them chained at a slow rate.
An important thing to know is that damage output depends on difficulty. Or maybe enemy armor becomes tougher then. There is a massive difficulty difference between easy and medium. I did all my testing on easy to see if the game actually buffs our weapons. It doesn't. It would appear Easy is the actual Lore friendly Realism mode, while Medium and Hard artificially gives them twice or thrice the armor and structure.
Next testing I intend to do is try to destroy an armorless dire with MG crits to the CT., before the Enhanced Imaging display giving it a red (less than
25% health, armor and structure included).
Finally, believe me, I'm very glad to see a fellow pilot enjoy the game of my childhood in 2020. Heck, I even have, believe it or not, dreams about the Mannedorf Trial of Grievance map, and nightmares about a Warhawk looking for me and trying to kill me! All this while not having played this game for years.
Once I finally dropped in a Trial on Mannedorf and saw that Canyon, I immediately recognized the place from my dreams!
It was awesome, I thought I had invented that map in my dreams, deep inside my head, after about 2 decades, my mind still hadn't forgotten it.
Ah, this game... It might be a quarter a century old, yet still feels futuristic, exotic, mysterious, distant. It aged beautifully. Yeah, it's a legendary classic, glad to see poeple play it and watch this :)
The million C-Bill question is, DOS or Titanium Trilogy?
+Lukas Stone Fully patched original DOS release. :B
Final testing for today: MG crits against engine.
I destroyed 20 armorless, weaponless dires, going for the CT.
30% (6 kills) happened while the mech was on yellow damage (backstabbed, this can't be a accidental cockpit hit), 70% (14 kills) happened due to MG destroying the CT component (DWF was with a red torso on enhanced imaging display).
Soo, critting engines seems to be a thing in this game. Not that I ever did that. Due to the game's rather simple damage indicators, you generally know it's time to crit only when the CT is red. At that time, you might as well finish it with raw firepower.
I guess now it's time to test MGs during campaign missions to see if it is useful, or just a nice lore friendly thing that's not practical in actual combat (in MW2, of course).
Wait, the plane of Gothos that belongs to fighter ace Baron Von Trelane????? :0 WHUUUUUUUUUUUUUUUUUUH Excited!!!!
I recently started playing the Mech 2 trilogy again, and I suspected there was some Tom-Foolery going on with the Mech lab stats.
Mercenaries made some more changes. They definitely made the Auto-cannons and LRM’s weaker, and the pulse lasers even more powerful.
who could have thought that a video on weapon stats in a 25 year old game (even if it's one of my favorite games) could be so interesting; thanks
(12:29) A bit unrelated, but I've noticed some games are rather inconsistent with how they treat splash damage. In some games, if you hit a bad guy with a rocket launcher (or an equivalent), it'll blow the bad guy apart with nearby bad guys taking splash damage. In other games, when a bad guy takes a similar direct hit, he absorbs _all_ of the damage, leaving no splash damage to surrounding bad guys; if the hit is anything less than a direct hit (glancing), then splash damage is dealt normally (as far as I can tell).
+Christopher Noel Yeah, it really depends on the game. Generally speaking, you have "impact" damage and "splash" damage, the damage caused by the weapon making impact, and the splash caused to surrounding targets, although various other methods exist too, such as ONLY having splash but calculating the quantity of splash damage based on distance from the point of impact. However, here in Mech2, splash damage is not instant, it occurs over a couple seconds, and it's very minimal, so if you ONLY get hit with the splash from a weapon it does very little damage. :P
17:40 Hmm... I'm wondering if armor is also factoring in here. Like the mechs, all likely have an armor value, whereas the pylons likely do not.
Armour in Battletech is simply treated like extra hit points. ;)
Final note: Lasers were dissapointing in this game, but that's just because Activision decided to nerf 'em. Since I'm a sniper this is kinda a big thing.
You end up with a weapon that has a 1.5 heat/dmg ratio (CERLL, way too hot IMO), and no nice recycle. Learning to use the CERPPC and just outright spamming it at a range the enemy can't fire back from is verry rewarding. Just put plenty of CDHS to offset the heat and you're good to go.
Now I will basically carry loads of CERPPC and Gauss Rifles.
*Lasers weren't hitting at max range*
Hmm... I'm wondering how much you fired the weapon. I mean is it that the range stated is wrong or that at that distance it is hard to hit things? I didn't play the game back in the day but I watched other people play and I seem to recall them rolling dice to see if they hit or not.
There are no dice rolls in the weapon mechanics in Mech2. In fact, there's virtually no RNG at all. I was firing the weapons more than enough to determine that RNG was not a factor and for results that seemed unusual I triple-checked them. :B
@@Pixelmusement Fair enough. ^_^
Great vid!
Well, I knew that the LB-X 2 has the highest range. But I am a nerd for the battletech Tabletop.
Very interesting. Also a cool game,
This gets even more complex if you include Mech2 vs Mech2:Mercs,[A quick look shows LRMs have a range of 1000m vs 630m in Mw2. Though maybe those are all 'clan' weapons? Many other weapons are identical stats though.] as well as possible changes from patches.[Though I don't see any yet.]
There were also bugs in various versions of Mech2, and Mech2:Mercs. One forum poster:
> Shots from weapons phasing through target when fired in rapid succession, making weapons with quick fire times close to useles (e.g. Pulse Lasers, Ultra Autocannons)
> Small Pulse Lasers having no hit detection due to bug in weapon's effective range
If that forum post is true, it definitely changes how viable fast firing weapons are [if they bug out and don't count to your DPS]. That's also something that could be tested.
Those bugs are in reference to MW2:Mercs and were not a factor in the original MW2. The main bug in the original MW2 was that one side of the torso took WAY more damage than the other side, suggesting the overlapping collisions volumes were simply choosing one over the other when both were triggered, rather than trying to figure out which side was actually hit. :P
Amazing analysis, but where the hell are your STDev?
Actually, for Mech2... there ISN'T any as there's no random chance anywhere in the simulation, best I can tell. :o
I thought the UAC in Mechwarrior 2 rapid fired when you held down the button to simulate the increased fire rate from tabletop. For those who don't know the UAC in tabletop can fire 2 shells per turn.
+OptimusShr Nah. Every weapon in the game will auto-fire if you hold the button down. However, it's coded in such a way that if you had let go of the fire button and press it back down while a weapon is still reloading/recharging it will NOT auto-fire once ready as this normally causes a skip-over to the next weapon in sequence.
The damage stats on the board game does not translate accurately in many cases to the computer game. The best example is using my favorite mech, the Nova, which is armed with 12 Extended Range Medium Lasers. In the Battletech board game a single alpha strike with all 12 lasers that hits the same spot can destroy an assault mech even if it is hit in the center torso, which is the most heavily armored part of the mech. In Mechwarrior 2 an alpha strike on a light or medium mech that hits its arm or leg which should have completely blown it off often doesn't.
I know, which was the entire reason for making this video because the stats are pulled straight from the tabletop game yet they never made sense here for whatever reason! :P
is the next game corncob deluxe?
Oh, I think Frankomatic played next week's game recently. ;)
Hm, that would imply Corncob 3D. There's also another contender: Evasive Action (working title "Air Duel") by Glyn Williams.
+Steven Don Already did Evasive Action! ruclips.net/video/fJMK8Tr2uvY/видео.html
I was thinking (ROT13) fgnegerx whqt zrag evgrf
Is there a table to download these values?
I typically only go through the effort of making tables and such available when I feel the information is worth referencing. Here, it didn't really seem worth going through the effort. :P
I don't understand how anybody could suffer through testing and calculating all this stuff. However, I think the results are really interesting... so go right ahead ;) Great video, in any case! :D
+R. F. R. For the same reason why grinding in games doesn't bother some people. It may be tedious, but it's ultimately not all that hard and so you just kinda fall into it and can escape from reality for a bit. That's what doing all this testing mostly felt like. :B
all great info, but holy shit, this is too much