BV-J: My custom BV system

Not related to MegaMek/MegaMekNET/MekWars

Moderator: Moderators

Jackal
Posts: 1410
Joined: Thu Jan 21, 2010 3:02 pm

BV-J: My custom BV system

Post by Jackal »

BV-J: My Custom Battle Value System

Hey everyone. Those of you who have played on this server for a while know that I'm sort of a numbers guy. In my professional life I do a lot of financial modeling, and crunching numbers is sort of a hobby of mine. One project that I've been working on for quite a while is a new (unofficial for now) battle value system for Battletech. Thanks to COVID, I've had a lot more more free time (no more commuting to work) so I've made quite a lot of headway. I call this system BV-J (for Jackal, naturally 8)).

This system works similarly to BV2 and I've actually set up my most recent builds to return numbers that are about in the same range, so that it is possible to make direct comparisons between BV2 and BV-J, and—importantly—to see how individual units differ from one system to the other.

Here's a sneak peak. Note columns G and I, which are the BV-J values and BV2 values, respectively. You can also see the % change in # change in columns J and K. Note also the BV per ton in column L, which I've found is a good objective measure of quality (click image to enlarge):

Image


It is important to note that what you are seeing above is a dynamic spreadsheet, meaning everything you see here is being calculated from formulas. Basically you plug in a unit's stats and it spits out a number, just like BV2.

Major Effects of BV-J

Like I said, BV-J works similarly to BV2, but I have tweaked a lot of the values and also added many new mechanics. The differences in 'how the sausage is made' between BV-J and BV2 are both too numerous and too subtle to explain without having to write a small book on the subject (which I may do at some point). What's much more interesting is what *effects* BV-J has on the estimated value of different units. Here's a quick rundown:
  1. Pilot and crew skills have less overall impact on BV growth than in BV2
    • on average, 1 gunnery level adds ~15.6% (compare to 20% for BV2 and 15% in BV1)
    • on average, 1 piloting level adds ~6.4% (compare to 15% for BV2 and 5% in BV1)
  2. Gunnery levels have different effects on different kinds of units.
    • close combat mechs gain less BV from gunnery than longer range units
      Image
      • fire support mechs gain more BV from gunnery than shorter ranged units
        Image
        • melee-oriented mechs gain barely anything from gunnery (check out the Charger!)
          Image
          • mechs with lots of weapons with to-hit bonuses—particularly pulse lasers—gain less BV from gunnery levels (pilot levels don't add as much because the weapons already have low THN's)
            Image
        • Piloting levels also have different effects on different kinds of units.
          • close combat mechs gain an average amount of BV from piloting levels
            Image
            • melee-oriented mechs gain a lot more than other mechs from piloting levels
              Image
            • Speed, weapons and armor combine in different ways than in BV2.
              • Speed has a weaker impact on gunnery than in BV2
              • Speed has a stronger impact on defensive value than in BV2
              • Speed also has a very strong impact on the new BV component, physical value (PV)
              • Overall these factors combine in an interesting pattern:
                • "Sentries" (2/3 movers like UM-R60's) go down slightly in BV
                • "Walkers" (3/5 and 4/6 movers) mostly have the same relative BV (but loadout has a big impact, see #4-6, below)
                • "Runners" (5/8 and 6/9 movers) mostly go down in relative BV, and
                • "Sprinters" (7/11+ movers) mostly go up in relative BV
                • "Jumpers" (6/9/6 and below) mostly go down in relative BV
                • "Extreme Jumpers" (7/11/7 and higher) hold steady or go up in BV
            • Each unit's critical hit chart is evaluated for critical hit vulnerability.
              • "Zombie" mechs (lots of inert critical hit locations) tend to go up in relative BV
              • "Explodey" mechs (lots of ammo or ammo-only hit locations) tend to go down in relative BV
              I actually created another dynamic spreadsheet that looks a mech's critical hit chart to determine how vulnerable it is to critical hits. Here's what it looks like (note: this is rated on a 0-1 scale with a lower number indicating a lower level of crit vulnerability):

              Image
              • Heat efficiency matters more.
                • Units with double heat sinks mostly see big BV increases
                • Units with inadequate heat sinking capability (RFL-4D, etc.) mostly go down in relative BV
                • Units with efficient salvos at multiple ranges (STK-3F, TDR-5S) perform slightly better than in BV2
              • Long range units start with lower BV's than in BV2, but gain BV faster with levels compared to short range units.

                Here are the BV's and BV growth for two well-known short and long range mechs in BV2:
                • 4/5 ON1-VA – 1328 (BV2)
                • 3/5 ON1-VA – 1594 (BV2) [+20%]
                • 4/5 ARC-2R – 1477 (BV2)
                • 3/5 ARC-2R – 1772 (BV2) [+20%]
                Here are their BV's under BV-J:
                • 4/5 ON1-VA – 1381 (BV-J) << starts higher than BV2
                • 3/5 ON1-VA – 1556 (BV-J) [+12.7%] << gains less % than the long range unit
                • 4/5 ARC-2R – 1422 (BV-J) << starts lower than BV2
                • 3/5 ARC-2R – 1664 (BV-J) [+17.0%] << gains more % than the short range unit

              Progress To Date
              So far I have done all the data entry and 148 introductory tech mechs and 100 standard tech mechs. But because the system just calculates the BV from a unit's base stats potentially any unit could be calculated.

              So far, here's what my system supports:

              Completed:
              • All introductory tech mechs
              • All Inner Sphere standard tech mechs (except for Narc, TAG, and C3)
              Not completed:
              • Vehicles – I already have a partially working model for vehicles, but I've not fine-tuned it to the degree that I have for mechs. Perfecting this model really is an issue of testing an iteration.
              • Clan standard tech – The roadmap for building-out Clan tech already exists, so implementing this is just a matter of time and effort. It should work just like IS tech (the values are just bigger).
              • Narc, TAG and C3 - These items are different from most standard tech gear in that they boost a gang rather than just the unit on which they are mounted. I do have some general ideas about how to approach these, I just haven't dug into them very deeply yet.
              Not Planned (At least for now):
              • Infantry and BA - I haven't done any work at all on infantry or battle armor, but since BV-J numbers are roughly in the same range and BV2 numbers, they will probably work as-is. I might eventually take a hard look at these, but I currently have no plans to do so.
              • Aero – I actually like playing aero games, but I feel comparing aerospace fighters to ground units with a single system is an apples-to-oranges comparison. If I were to ever to tackle this, I would probably come up with a separate 'aero value' that is just for aerospace combat
              • Advanced, Experimental, and Primitive tech – I don't have plans to do any of these tech levels anytime in the near future, simply because they aren't used as widely as the other tech levels. Theoretically, they would work something like Intro and Standard Tech, but maybe with some new dynamics.
              Last edited by Jackal on Tue Apr 13, 2021 2:34 pm, edited 7 times in total.

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              Hey Jackal, as i've already told you, I love it.

              BV-J captures many of the nuances that earlier editions of BV completely miss.

              Image

              But i was wondering why all the melee mechs went up in BV. I do think melee is a particularly difficult one to model. What were your considerations for that? A slow melee mech should be worth much less than a fast one, regardless of the strength of its kick

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              Good question, Pigeon.

              So there's something that's important to understand about this model—it's not the first model that I've made. This particular one is v8.9.X10. I've made literally dozens of them.

              Each of these models (and BV2 itself) are each based on slightly different assumptions about how much a point of armor is worth, what the base-to-hit of weapons and physical attacks should be, how much speed should count for, etc., etc. When you change just one of those variables it changes the BV of every unit in the game... and that can have a lot of unintended consequences, which makes this very tricky to do.

              What I think is different between BV2 and v8.9.X10 is I've put my model through a pretty rigorous testing process (which I'll elaborate on later, I don't have time right now).

              In the beginning of building BV-J most of the variables were based on hunches, personal experience, and what 'looked right.' Later, when I had a pretty developed testing process, improving my models became more about making the model fit the data.


              v8.9.X10 is simply my best-fitting model to date. I expect if I do more testing—particularly with more human players—the model will get better (and maybe melee-oriented mechs will go down in BV, we'll see).






              Edit: Oh, and I missed your last statement about slow melee units. Yes, you're right. Speed is an important factor in melee combat. You'll notice the mechs in that "Melee" breakout are mostly fast. The BNC-3E is in there simply because the formulas in the spreadsheet identifies it as having a high "Melee Priority," meaning that it NEEDS to be trying to use physical attacks.

              The way the spreadsheet determines a mech's melee priority is to compare its "Physical Value" (a measure of its physical combat prowess that I invented for this model) to the total of its gunnery values (SRGV + LRGV).

              If a mech's physical value is more than half of the total gunnery value, it is considered to have a High Melee Priority.

              High Melee Priority: PV ≥ (SRGV + LRGV)*0.5

              If a mech's physical value is more than the total gunnery value (as is true on the CGR-1A1 and OTT-7J) this mech is considered Primary Melee.

              Primary Melee: PV ≥ (SRGV + LRGV)


              And, as you know, melee combat is an important aspect of a BNC-3E. If you're not using the mech's physicality (which includes THREATENING to melee, not just using melee attacks), you're not getting the most you can out of the mech.

              Finally, when you look at the BNC-3E on a BV per ton basis (which I've learned is a good objective measure of quality) it is just 15.1. The average BV/ton for intro tech mechs is 19.0 so the BNC-3E is a pretty low quality mech. The fact it goes up 1% doesn't change much.

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              I was thinking about the test game we played. Four 3/5 jumpers against four 4/5 walkers.

              I don't think its a good test. If you want to sanity check the levels, you should do a match up something like four 3/5 walkers against four 4/5 walkers. If you want to test jumpers vs walkers, then they should be the same the same level.

              in other words, hold everything constant except for the variable you are testing for.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              Yes, I agree. That matchup involves 2-3 points of difference–levels, speed, and jump jets. The reason I went with that particular matchup is just that it was one of the few human-vs-human tests I had ready to go.

              It's not hard to generate single variable tests. It takes a bit of time.

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              1) describe from your point of view how you think the battle unfolded

              I had 3/5 WVR-R, 3/5 PXH-1, 3/5 JVN-10F and 3/5 JVN-10N (jumpers). You had 4/5 BJ-1, 2 * 4/5 VND-1R, 4/5 HBK-4J (walkers). This battle could only go one way - the jumpers had to charge the walkers, simply because the walker lance has much much greater 'pull'. Even though the jumpers were all 3 gunners, the fact that they only had 2 medium range weapons (LL and AC5) meant the amount of 'pull' the lance had was still not good.

              Since only one of my mechs had pull (the PXH), the plan was to keep the PXH slightly apart from the rest and use it to snipe or flank if possible. The plan for the 2*JVN and the WVR was to charge in together as a brawling wrecking ball and hopefully take down one of yours before the JVN-10N got raped by PPC fire.

              You had to climb elevation which meant that you didn't have a lot of backing up room. The only way you could disengage if you were chased to the edge was to use Jump Jets.

              Can't quite remember the details but it didnt go well the jumpers. i think the JVN-10N got ammo'ed and the WVR-6R was knocked down and then we had to call it quits. I think i was close to taking down one of the VND but nothing of yours was down yet.

              2) I'd also be interested in knowing how threatening my army seemed from your POV

              It was threatening mainly because i had a mech (JVN-10N) that could not take a PPC hit without a high risk of blowing up. you had a lot of medium and long range fire and i had no way to respond. There wasn't really any range bracket where i felt i had the firepower advantage. It felt like at long range i was f**ked. At short range, i at best had a slight advantage. hmm actually no. HBK-4J and BJ-1 have quite a number of MLs and my brawling element would likely have to take a PPC or two at short range as well. I guess the fact that i had 3 gunners meant that i should be hitting on better numbers at short range, even though i may not have had the same weight of fire.

              It felt like given the match up I could not create a situation where the advantage was strongly on my side. This could come about if i won init a few times in a row. Then my faster speed would mean that I could position my guys in safer positions and get back shots.


              3) finally after describing how the battle unfolded etc. now offer your thoughts on the relative balance of the matchup (feel free to add qualifiers–slightly very not at all etc.)


              This kind of pure jumper brawler type lance is something I would typically never put together. When all you have is a hammer, every problem looks like a nail. To me the match up felt difficult probably because I had a very limited toolset to take apart your lance. To me, this is not so much a bv problem, rather its an issue with the lance construction itself. A lance of 2 walkers + 2 jumpers is much stronger than a lance with 4 jumpers - even if the bv works out to be the same simply because the diverse lance has options.

              Also, in my experience - firepower trumps both armor and speed. The best defense is a good offense. This game is won by killing your opponent faster than he kills you. So perhaps a pure jumper / flanker lance is not my playstyle.

              Lastly, i think a lot of this game is identifying at which range bracket you have a firepower advantage over your opponent, and using positioning to engage mostly at that range. In the match-up, it didnt feel like there was any range bracket where i had a strong advantage
              Last edited by obese pigeon on Sat Apr 17, 2021 3:58 am, edited 2 times in total.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              So here was how the battle looked to me (an interesting contrast).


              1) describe from your point of view how you think the battle unfolded
              So to me, in the opening phases of the battle it looked like you were going to obliterate me. It seemed like you landed a zillion medium laser and large laser hits from a 5 or 6 hex range, and it felt like none of my shots were hitting.

              But then the battle shifted to a close-range brawl in a relatively narrow raised area that I had moved to at the start of the fight. I was actually pretty apprehensive about this because I thought my VND-1R's were going to have a lot of trouble in close quarters, but I kept on being able to get just far enough away from your mechs to keep my minimum range penalty to +1 or lower.

              I had thought you were going to jump around a wear me down from range, but in a battle with limited mobility the advantage seems to shift to me and, though all but one of my mechs had pretty worn-down armor, I killed the JVN-10N, did a ton of damage to the WVR-6R and even opened a side torso on the JVN-10F.

              That said, I still felt like there was going to be a tough fight ahead. On the turn we stopped, my HBK-4H and BJ-1's armors were both shredded, and though your JVN-10F had an open side torso, it was otherwise unhurt and your PXH-1 wasn't damaged at all, so not really sure where that battle would have gone had we continued.

              2) I'd also be interested in knowing how threatening [the opposing army] seemed from your POV
              Particularly with how the battle opened (and also our aborted battle before that), I had a pretty healthy fear about your army's firepower. If I recall correctly, my BJ-1's armor bar was already deeply into yellow territory by the end of the 2nd round and I basically had to retreat it. Since that was one of my key close range mechs, I was kind of running scared. Kinda interesting you saw it the other way.

              3) finally after describing how the battle unfolded etc. now offer your thoughts on the relative balance of the matchup (feel free to add qualifiers–slightly very not at all etc.)
              So I'm not sure yet what I think about the relative balance of this battle. What I will say is that BV2 seems to get it wrong. In BV2 your army supposedly has +5.4% BV, which should be a pretty significant (but not insurmountable) edge in this game:
              • Leveled Jumper Army: 4285 BV (+221 BV, +5.4% advantage)
              • Unleveled Walker Army: 4064 BV

              v8.9.X10 has this army looking this way:
              • Leveled Jumper Army: 3918 BV (-39 BV -0.9% disadvantage)
              • Unleveled Walker Army: 3959 BV

              I actually felt like your army was a little stronger and mine a little weaker than v8.9.X10 suggests but, in any case, it seems closer than BV2 does.

              Of course, this was just one battle, so I think it's premature to draw any hard conclusions, but was definitely interesting. Thanks for helping me test this out!
              Last edited by Jackal on Tue Apr 13, 2021 12:07 pm, edited 2 times in total.

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              Jackal wrote:
              Fri Apr 02, 2021 5:40 pm

              2) I'd also be interested in knowing how threatening [the opposing army] seemed from your POV
              Particularly with how the battle opened (and also our aborted battle before that), I had a pretty healthy fear about your army's firepower. If I recall correctly, my BJ-1's armor bar was already deeply into yellow territory by the end of the 2nd round and I basically had to retreat it. Since that was one of my key close range mechs, I was kind of running scared. Kinda interesting you saw it the other way.

              3) finally after describing how the battle unfolded etc. now offer your thoughts on the relative balance of the matchup (feel free to add qualifiers–slightly very not at all etc.)
              So I'm not sure yet what I think about the relative balance of this battle. What I will say is that BV2 seems to get it wrong. In BV2 your army supposedly has +5.4% BV, which should be a pretty significant (but not insurmountable) edge in this game:
              • Leveled Jumper Army: 4285 BV (+221 BV, +5.4% advantage)
              • Unleveled Walker Army: 4064 BV

              v8.9.X10 has this army looking this way:
              • Leveled Jumper Army: 3918 BV (-39 BV +0.9% disadvantage)
              • Unleveled Walker Army: 3959 BV
              Noted that in BV2 the bv of the jumpers is actually 5.4% higher. It got me thinking how the skill level of the players involved actually could affect this perception of balance. I think we are both relatively skilled and experienced players who know how to counter speed, mobility and terrain. I guess in this match-up between two skilled players, jumpers would have their advantages easily nullified by positioning

              But... if this game were played between two players of lesser experience, it could very well seem that the jumper lance is 5% stronger. It certainly is better at jumping into terrain (which is something baby seals love to do - since they tend to overfocus on defense mods). Also, since newer players may have difficulty getting the best to-hit numbers, having a lance entirely consisting of levelled gunners definitely helps out as well.

              So perhaps bv2 (which seems to more highly value jumpjets) is a more accurate representation of balance when played by less experienced players

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              So in my experience it is the opposite. Less experienced players tend to do better with walkers than with jumpers/mobility units. You even see this with the bot. If the bot is has walkers I can beat about 25% more BV. If the bot is has jumpers or runners I can beat 50%+ more BV.

              You NEED to be skilled if your defense relies on making good moves. Walkers are a lot more forgiving.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              But for the record, it was my general feeling that the leveled jumping army's BV was a little too low. I just don't think it is *5.4%* too low. Based on my testing, a BV difference that big is enough to make balanced fight into one where one side usually wins (and in the battle we just played the jumpers were losing not winning). I don't think the battle was quite that slanted.


              But in any case, we're trying to draw a lot of conclusions from just 1 1/2 battles. It's way too early to make those calls. Give it some time and some more battles before drawing any hard conclusions.


              And finally, thank you. I'm very happy you're putting a lot of thought into this. It's taken a lot of time for me to develop this thing, so I appreciate that you're taking it seriously.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              So going back to the idea of single variable tests, here are the things that need to feel this thing out (and this is a reduced list from the things I test vs. the bot).
              1. baseline test: some 4/5 walkers vs. some different 4/5 walkers (in standardized test terrain)
              2. gunnery levels (vet): 3/5 walkers vs. 4/5 walkers (in standardized test terrain)
              3. piloting levels (vet): 4/4 walkers vs. 4/5 walkers (in standardized test terrain)
              4. gunnery levels (green): 5/5 walkers vs. 4/5 walkers (in standardized test terrain)
              5. piloting levels (green): 4/6 walkers vs. 4/5 walkers (in standardized test terrain)
              6. mobility: 4/5 runners vs. 4/5 walkers (in standardized test terrain)
              7. jumpers: 4/5 jumpers (5+) vs. 4/5 walkers (in standardized test terrain)

              After that you would get into mutlivariate testng. Leveled jumpers vs. unleveled walkers, etc.


              There are lots and lots of things one could potentially test. I'm not asking for that kind of time though. I think right now we just want to roll through some of these scenarios and get a 'general sense.'

              I can come up with the specific scenarios, or we can work together on them. I'm flex.
              Last edited by Jackal on Tue Apr 06, 2021 2:19 pm, edited 2 times in total.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              1. baseline test: some 4/5 walkers vs. some different 4/5 walkers (in standardized test terrain)

              So here's the first human vs. human test I've designed for the series above—six 4/5 walkers that all go down in BV[/b] vs. six 4/5 walkers that all go up in BV. I test with six vs. six units because it makes each army better able to recover from unlucky losses/cripplings.

              Once again the light yellow column is the BV-J value, the white column with the italics letters are the BV2 values.

              Image

              As you can see, the BV2 column shows this matchup as 8620 vs 8200 BV, in which the top army supposedly has a significant +5.1% advantage, something which I feel is questionable (the bottom army looks very strong to me).

              With v8.9.X10 this matchup is instead 8458 vs 8488 BV, where top army actually has a very slight -0.3% disadvantage which is essentially an even matchup (which seems plausible).

              Obese Pigeon and I are going to give these armies a whirl, each of us playing both sides at least once. I'll report our findings here.

              User avatar
              Windschreiter
              Posts: 112
              Joined: Sat Feb 16, 2008 1:07 am

              Re: BV-J: My custom BV system

              Post by Windschreiter »

              wow - just wow to all the effort and thought put in there. Great work!

              Would it be possible to adjust BV according to lance composition? I know that is hard to tackle and mostly personal play style, but I think it has a major impact.

              For example a RFL is better off in a light lance than in an assault lance.

              Or in the last matchup example you gave: the WHM(MAD to some degree too) feels underarmored for the BV, that would be my point why I too think the T-Bolt/Onion force has the edge: all Mechs are equally sturdy and none seem to be a weak link, but the WHM(MAD) has significantly less armor than the other Mechs in that force.
              so (relative) low armor + high firepower => prime target [which can be offset/or even used by experienced players]

              This is represented in the lower BV values to a degree, though there might be an additional effect.

              Maybe something like BV Spread might be a way to rate this. The second army has a BV spread of 98 vs. a spread of 288 in the first army. Maybe checking for outliers could be a way. Just writing up thoughts here now.

              Need to think this through and maybe I can come up with something I feel comfortable with.


              If you need more testpilots I'd be willing to volunteer - though this would add another variable to the mix.

              Regards
              Windschreiter

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              Windschreiter wrote:
              Wed Apr 07, 2021 7:48 am


              Or in the last matchup example you gave: the WHM(MAD to some degree too) feels underarmored for the BV, that would be my point why I too think the T-Bolt/Onion force has the edge: all Mechs are equally sturdy and none seem to be a weak link, but the WHM(MAD) has significantly less armor than the other Mechs in that force.
              so (relative) low armor + high firepower => prime target [which can be offset/or even used by experienced players]

              I highlighted exactly the same thing to Jackal, that the first lance feels weaker because it has two soft spots.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              Windschreiter: So, ya, the goal in this battle was just to show that BV2 overestimates the toughness of the top army. Yes, the MAD-3R and WHM-6R are weak points, but that's why the BV is lower in BV-J.

              If there were better armored units in the place of the MAD and WHM, they'd also have significantly less firepower. I guess we could try putting Hoplites in those places or something.

              I think it would be worth doing a number of walker vs. walker tests.

              And @Windschreiter! You're totally welcome to join us for the testing if you want! The nice thing about these games is you can stop a game in the middle and pick it up at another time. (Good for those of us who are a bit time challenged.)

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              baseline test: some 4/5 walkers vs. some different 4/5 walkers (in standardized test terrain)

              I played army 1: ARC-2R (4/5), BNC-3M (4/5), BL-7-KNT (4/5), BL-7-KNT (4/5), MAD-3R (4/5), WHM-6R (4/5)
              Jackal played army 2: TDR-5S (4/5), TDR-5S (4/5), TDR-5S (4/5), TDR-SS (4/5), ON1-K (4/5), ON1-K (4/5)

              1) describe from your point of view how you think the battle unfolded

              I had the option of a high ground deploy. Given my stronger long and mid range firepower, I decided to take it. I saw that Jackal had also made a rather big strategic mistake in splitting his lance into two, with one half significantly closer to my position than the other half.

              I did a general advance in turn 1 so as not to clue Jackal in as to what i was planning, then in turn two turned and moved aggressively towards the closer half of his lance. Fortunately for Jackal he realized the strategic error by turn four and basically turned and ran to open up the distance. I was able to do a fair bit of damage, however, TDRs are tough mechs.

              I had taken less damage in the early game exchange, but at the same time, my units were softer.
              I could not really press hard because I didnt have init throughout the early part of the fight, else i would have tried to press half of Jackal's force to the north map edge with 2 black knights and the banshee.

              Meanwhile the other half of Jackal's force is also engaging, and he rushes in a single thunderbolt to confront my camping ARC and WHM (probably to take attention away from another thunderbolt that was taking quite a beating at range). Doesn't go well for the aggressive Tbolt as it gets knocked down every turn (Jackal failed a lot of PSRs). In the last round of the fight, Jackal had lost init and the aggressive tbolt was on the floor. I had init, and got WHM behind it at point blank range. WHM's back was exposed to an ON1 at medium range, but i was ok with taking that risk. In the end, TDR took a pounding (40+ damage), but WHM took a rear torso hit from AC10 and sustained an ammo explosion.

              At the point, state of my lance was as such. One black knight was badly damaged, having lost its right torso. MAD was moderately damaged.

              Jackal had a few moderately damaged TDRs (but still very much fighting fit) and was pressing a counterattack. Also his ON1 had sustained a 5 pt head hit.

              2) I'd also be interested in knowing how threatening [the opposing army] seemed from your POV

              Not very threatening, especially since i had the high ground and Jackal's lance was badly deployed. I was pretty sure I could hurt his lance in the early game before he closed the distance.

              3) finally after describing how the battle unfolded etc. now offer your thoughts on the relative balance of the matchup (feel free to add qualifiers–slightly very not at all etc.)
              I think army 1 is weaker despite the much higher BV per BV2. However, this is a due to composition rather than bv. MAD and WHM are too soft to survive well at 8k bv. Once these two go down, i expect army 1 to get steamrolled by army 2. Army 1 has better long and mid range engagement capabilities but will melt once army 2 gets stuck in.
              Last edited by obese pigeon on Thu Apr 08, 2021 11:22 am, edited 1 time in total.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              1) baseline test: some 4/5 walkers vs. some different 4/5 walkers (in standardized test terrain)

              So here's how the battle unfolded from my perspective. As Pigeon noted, these were the armies:
              • Obese Pigeon: ARC-2R (4/5), BNC-3M (4/5), BL-7-KNT (4/5), BL-7-KNT (4/5), MAD-3R (4/5), WHM-6R (4/5)
              • Jackal: TDR-5S (4/5), TDR-5S (4/5), TDR-5S (4/5), TDR-SS (4/5), ON1-K (4/5), ON1-K (4/5)
              Here, also, is a snapshot of the minimap that shows our starting positions (Pigeon is red, I'm blue):

              Image

              This, BTW, is the "standardized testing terrain" that I referred to in the above posts. It's sort of a rolling hills map with scattered woods. I use this map for baseline testing because it basically has something to offer any unit (you can fight in the flats, you can fight in the hills, there are decent lines of site, but plenty of obstacles, vees can use roads, etc., etc.).


              1) describe from your point of view how you think the battle unfolded
              I could see during deployment that Pigeon was placing his units on top of the hill. The hill had a lot places where he could get partial cover, so my concern about a frontal assault was that he going to fire from behind cover as I approached, and keep falling back to more partial cover, dragging my forces around the hill. I was also concerned that his more vulnerable WHM and MAD would just end up in his rear in this situation and I would have to make it through all of his better armored units to get to them. Driven by those concerns, I decided instead to split up my units into two packs and attack from multiple sides. (This may or may not have been a tactical mistake. I'm undecided.)

              In the first couple rounds of the battle, things actually were going well in terms of the BV race. I was landing more hits on him with LRM15's, LL's and AC10's than I was taking in return fire. But starting about round 2 or 3, I could see that Pigeon was shifting all of his forces to engage my northwestern group of mechs (the 'NW Group') which consisted of a TDR-5SS, a TDR-5S and a ON1-K.

              Realizing I was in danger of getting pinned in the NW corner by his entire army, I ran these units to the north, scrupulously staying beyond medium range of all his units except for the particular mech that I was firing at—a BL-7-KNT that was further ahead of the pack. I also moved my SW group (which consisted of two TDR-5S's and an ON1-K) to the hill's southern slope and started working my way up, hoping this would take pressure off my other mechs.

              As I did this, Pigeon adapted. His WHM, ARC and MAD (which were, in fact, in the rear of his army) were now the closest to my SW Group which was scaling the hill. Pigeon tried to back his MAD up the hill, but failed a PSR and fell. His WHM and ARC meanwhile took up positions behind some small 1-hex rises that offered partial cover and started unloading on the SW Group. He basted one of my TDR-5S's to the ground, so I retreated it to the SW. I also diverted the ON1-K in this group the the W, so I could shoot around the PC. My remaining TDR-5S readied itself for a charge up the hill...

              That same round my NW Group was still trying to back off to the NW. I managed to get one of the TDR's into some woods where it took up a firing position. It tried to back my ON1-K up onto a 1 level plateau in the NW but it fell. I backed off the TDR-5SS (which was now showing yellow on its armor bar) to just beyond the plateau's edge so I didn't fall down with it too.

              At this point I'm not sure what Pigeon was thinking (had to get back and protect his southern flank?) but he started backing off the BL-7's that were attacking my NW group. I'd also done a fair amount of damage to the BL-7 nearest me, and if I recall correctly I did a lot of damage to it this round (even with my ON1-K grounded). The next round once all of my NW Group were on their feet, I surged forward and destroyed the BL-7's right torso.

              Back on the hill, things weren't going so great. I charged in with the TDR-5S that was furthest up the hill and, even though I made it all the way up to the PC that the WHM was hiding behind, I got knocked to the ground. My ON1-K found some light woods to snipe from an angle, and my other TDR (the one that had to retreat early) also fired away with LRM's and LL's from the base of the hill. Then next round my TDR at the top of the hill fell trying to stand and, having no place to go, took cover behind the PC again. Because of my first failure to stand, I couldn't get out of the BNC-3M's firing arc, and it plus the other mechs there unloaded on me, knocking it to the ground yet again.

              By now my NW group was rushing in to help having forced one of the BL-7's to retreat. But at a full run, and still at medium range, they didn't do much damage.

              The final round we played, I lost init. Still not really having anyplace to go with my TDR-5S that had fallen up on the hill, I again stood up behind PC and just hoped for the best. As Pigeon said, he ran his WHM around behind me, but left his rear open to my stationary ON1-K that was a medium range with its LRM15 and AC10 (THN's of 7 on both). The AC10 hit him square n the back, found the ammo, and destroyed the mech.

              And amazingly the TDR-5S stayed upright for once, taking 40 damage, but suffering no breaches (impressive considering it had been taking concentrated fire at close range for 3 rounds in a row).


              2) how threatening did the the opposing army seem from your POV?
              Definitely threatening, but I also had confidence in my mechs. I've known Pigeon for a very long time and know him to be a very skilled player so that also upped threat rating. His position on top of the hill also complicated things.

              But I also felt my mechs had very good armor and—if I could keep getting him to switch targets—I could get enough of them up the hill that I would be able to overcome him at close range.

              3) finally after describing how the battle unfolded etc. now offer your thoughts on the relative balance of the matchup (feel free to add qualifiers–slightly very not at all etc.)
              So I'm always reluctant to make a call on something like this after just one battle, but I do think this battle ended up being about even. I 'won' largely because of the lucky ammo crit. But I also had a LOT of bad luck with falls, so maybe this balances out.

              What I will say though is none of my mechs had any major breaches by the time we stopped our battle. By contrast, even without the WHM loss, some of Pigeon's mechs were looking a little beat-up (namely the side-torso destroyed BL-7, the MAD and the WHM).

              I also tracked the stats on this battle and we each won roughly half of the rounds, something I've found in previous testing to be an indicator of a close battle.

              I look forward to playing this battle in reverse. Will be interesting to see how that goes.
              Last edited by Jackal on Wed Apr 07, 2021 9:09 pm, edited 1 time in total.

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              Jackal: ARC-2R (4/5), BNC-3M (4/5), BL-7-KNT (4/5), BL-7-KNT (4/5), MAD-3R (4/5), WHM-6R (4/5)
              obese pigeon: TDR-5S (4/5), TDR-5S (4/5), TDR-5S (4/5), TDR-SS (4/5), ON1-K (4/5), ON1-K (4/5)

              1) describe from your point of view how you think the battle unfolded
              So i play the armoured deathball lance this time. We start on opposite hilltops. Given that Jackal has range superiority and softer mechs, I have to close in to brutalize him.

              I saw that Jackal again had a bad deploy, splitting up his force, and i rushed aggressively to try to exploit the 1 or 2 turns he was out of position.

              Unfortunately, i lost a TDR-5S to a PPC cockpit crit in turn 3 - a minor setback. If I could close the distance, I should still be able to win out through superior weight of fire.

              The next turn (or the turn after), Jackal shot extremely well. Hitting with almost every shot (lowest he was firing at was an 8 ). The TDR-5SS got hit by something like 5 PPCs and 2 LRM 20s, took 80+ damage, lost its PPC and was knocked down. I think i did 60+ damage to his Warhammer but it stayed standing.

              I had great difficulty closing the distance subsequently because of the terrain, and the fact that i was failing a PSR every turn from this point on. I managed to get 1 ON1-K onto his hill, which did good damage to his black knight, but failed to stay standing in return.

              Because I had a downed ON1-K on his hill dangerously exposed, I had to take pressure off it and could not afford to back up with my other units (despite losing init). I continued pushing forward, with the knowledge that his banshee could get a kick in on some of my undamaged units from higher ground. I put a few units in position to shoot at the banshee if it came charging in.

              Banshee did come running in for the kick and i was able to get a decent volley on it (40+ damage). However, Jackal hadnt failed a piloting roll yet at this point and he wasn't about to start now. The subsequent kick connected and took the head off a largely undamaged Orion.

              At this stage of the game, I had lost 1 TDR-5S to early cockpit crit, 1 ON1 to a head kick, TDR-5SS was badly damaged and missing its PPC. A second ON1 had little armour left and was probably 1 turn away from an ammo explosion.

              Jackal's MAD and WHM had thin armor but no breaches. One of the black knights had an open leg with 10 pts internal structure left but hadnt taken any actuator damage

              2) how threatening did the the opposing army seem from your POV?
              Not very threatening - I was confident the deathball could brawl Jackal's lance to death if i could get it up close and personal


              3) finally after describing how the battle unfolded etc. now offer your thoughts on the relative balance of the matchup (feel free to add qualifiers–slightly very not at all etc.)

              I think the fight is balanced. Despite this fight being completely one sided, I think this was more a function of poor RNG as well as unfavourable terrain making it hard for the deathball to get stuck in.

              I think if the terrain were favourable or RNG not bad, it would definitely have been a more even fight.

              I still feel the deathball is a bit more powerful actually.
              Last edited by obese pigeon on Sun Apr 11, 2021 12:52 am, edited 1 time in total.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              So here's my sense of the battle. Again the forces were:

              Jackal: ARC-2R (4/5), BNC-3M (4/5), BL-7-KNT (4/5), BL-7-KNT (4/5), MAD-3R (4/5), WHM-6R (4/5)
              obese pigeon: TDR-5S (4/5), TDR-5S (4/5), TDR-5S (4/5), TDR-SS (4/5), ON1-K (4/5), ON1-K (4/5)

              1) describe from your point of view how you think the battle unfolded
              As Pigeon noted, I made a mistake in my deployment. I failed initially to notice a good defensive position—a small hill where he would have to cross a canyon to get to me—and I had already deployed half my force before I decided to switch. The BNC-3M, MAD-3R and WHM-6R end up in the defensive zone, but my brawling BL-7-KNT's and ARC were both a few turns run away, which created a situation where I had to close ranks quickly before his units could charge in.

              As he approached (turns out it was only the 2nd round), my snipers stood and delivered several PPC shots at medium range (THN's of 8 ). I got an extremely lucky hit, and inflicted a head open/cockpit crit on his leading TDR-5S.

              Thereafter I regrouped my forces and prepared for Pigeon's onslaught. I put my BNC-3M and both BL-7-KNT's in between my snipers and his forces on a low broad hill right above the canyon. The canyon force Pigeon to approach more carefully than he probably would have otherwise and it slowed him down just enough that I was able to concentrate my attacks on one mech at a time.

              To make a long story short, I think I just outlucked Pigeon for the next couple of rounds. More of my shots hit and my units mostly stayed upright where his units kept getting knocked down.


              2) how threatening did the the opposing army seem from your POV?
              So Pigeon and I briefly played this same match once before but had to quit because the DED wasn't loaded with standard MMNet settings (of particular importance, extreme range was turned off in that battle which really reduced the value of my PPC's). In that battle we just decided to rush in an duke it out and my army got rolled! So, ya, I had a healthy respect for his "Lyran death ball" after that. :)

              3) finally after describing how the battle unfolded etc. now offer your thoughts on the relative balance of the matchup (feel free to add qualifiers–slightly very not at all etc.)
              I agree with Pigeon. Although this battle ended in a decisive win for me, had it not been for the early cockpit kill it would have been much closer.

              The battle definitely seems to be in the ballpark of balanced and, possibly, even balanced slightly in the TDR/ON1 army's favor. (And since BV2 actually has the WHM/MAD army as being stronger, it does seem that BV-J has balanced this battle better.)

              So @obese pigeon, should we call this particular matchup +1 for BV-J?

              obese pigeon
              Posts: 910
              Joined: Sat Mar 19, 2011 10:51 am

              Re: BV-J: My custom BV system

              Post by obese pigeon »

              Jackal wrote:
              Sat Apr 10, 2021 6:15 pm


              The battle definitely seems to be in the ballpark of balanced and, possibly, even balanced slightly in the TDR/ON1 army's favor. (And since BV2 actually has the WHM/MAD army as being stronger, it does seem that BV-J has balanced this battle better.)

              So @obese pigeon, should we call this particular matchup +1 for BV-J?
              Sure. BV-J does seem to balance this matchup better than BV-2

              Tuco
              MegamekNET Campaign Operator
              Posts: 2872
              Joined: Wed Feb 13, 2008 10:23 pm

              Re: BV-J: My custom BV system

              Post by Tuco »

              The BV System is made for all tech eras. How does this work out outside the 3025 cosmos?
              When the injustice is great enough, Justice will lend me the strength needed to correct it.
              None may stand against it.
              It will shatter every barrier, sunder any shield, tear through any enchantment and lend its servant the power to pass sentence.
              Know this: there is nothing on all the Planes that can stay the hand of justice when it is brought against them.
              It may unmake armies. It may sunder the thrones of gods.
              Know that for ALL who betray Justice, I am their fate... and fate carries an Executioner's Axe.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              Tuco wrote:
              Sun Apr 11, 2021 7:19 am
              The BV System is made for all tech eras. How does this work out outside the 3025 cosmos?
              Hi Tuco. So as I noted in the intro, the system currently works for both Intro Tech (which is 3025 era and parts of Age of War) and Inner Sphere Standard Tech (which is later Age of War, Star League Era, and Clan Invasion Era).

              Here is a sample:
              Image
              Note: The examples in this image are all pre-Succession War standard tech mechs (so LosTech). Mechs in Purple are SLDF Royal and Blue are just other mechs from the era (click for larger view).



              Basically the system works everything all the way up to TRO 3050 except for Clan mechs. I haven't done those because I would have to expand my calculator quite a bit. It could be done though. Just a matter of time and effort.


              The key thing though is all of these values are based on testing. I've tested a huge range of Intro Tech mechs and have tested a fair number of the Standard Tech mechs (principally the ones with big BV shifts because I wanted to confirm those).

              User avatar
              Windschreiter
              Posts: 112
              Joined: Sat Feb 16, 2008 1:07 am

              Re: BV-J: My custom BV system

              Post by Windschreiter »

              Following up on my post regarding some modifiers for outliers in lance BV:

              I don't think it is workable to account for lance composition in BV calculation, as this could be easily abused. Plus it would add to complexity and could mislead/irritate inexperienced players when 4 units don't simply add up their BV, but give a different lance BV.

              So I retract my suggestion. ;)

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              Windschreiter wrote:
              Mon Apr 12, 2021 10:40 am
              Following up on my post regarding some modifiers for outliers in lance BV:

              I don't think it is workable to account for lance composition in BV calculation, as this could be easily abused. Plus it would add to complexity and could mislead/irritate inexperienced players when 4 units don't simply add up their BV, but give a different lance BV.

              So I retract my suggestion. ;)
              Was worth thinking about anyway Windscheiter. Believe it or not, I actually developed a system along the same lines as what you are describing (it takes a look at a whole army's BV instead of just summing the individual units), but I am undecided about its value.

              On one hand, it might get closer to the true combat worthiness of some armies. On the other, it requires extra math (just adding up individual units is much easier), which probably isn't worthwhile in a tabletop game. Maybe it would be useful in a computer version of CBT where the math is done for you. Maybe.

              In any case, I've set it up in my model so its something that can be toggled on an off. We'll probably do some testing of it in human vs. human trials.

              Cheers,
              Jackal
              Last edited by Jackal on Tue Apr 13, 2021 3:31 pm, edited 1 time in total.

              Jackal
              Posts: 1410
              Joined: Thu Jan 21, 2010 3:02 pm

              Re: BV-J: My custom BV system

              Post by Jackal »

              I just revised the original post to better highlight the effects of BV-J. In particular I added a lot of screen grabs showing how BV-J handles pilot skills.


              Short version: in BV-J, pilot skills have a lower overall effect on BV and are differentiated by unit. This makes pilot skills more desirable and also reduces the problem of units getting the 'wrong kind' of level up (like a 3/5 pilot in a CGR-1A1 instead of a 4/4 pilot), because the BV increase is tailored to that specific mech.

              Post Reply