The Inevitable Bugfix Release: v0.510

As I should have expected, as soon as I released GearHead Caramel v0.500 last week a whole lot of bugs that had been hiding for months suddenly made themselves known. So this week I present to you v0.510, a release with far fewer bugs but one extra crab mecha. You can download it from the github releases page. The precompiled release packages don’t require you to install Python3 or any other dependencies; just unzip to a convenient place and start playing immediately.


    • AmkG on April 20, 2020 at 3:46 am
    • Reply

    Charging is fun. I put a spear-type weapon on each melee mecha, plus a normal melee weapon (usually a Heat Axe but I might try the Wrecking Ball). Charge into battle then slash with the normal melee weapon. Or just charge twice: first at the front line, second at the support/artillery mechas, make the Claymores run!

    Speaking of which, charging means that melee mecha are actually feasible now, whereas in previous games you’d strip off every melee weapon and stock up on PHS-8 Phase Cannons. As long as there isn’t too much power creep in the ranged weapons as you add more of them, Charging makes a melee mecha build possible. In case power creep does come in for ranged weapons, maybe Charge +1, Charge +2 etc. weapons can be added which not only increase accuracy, but also damage and allowed distance (so you can charge further than you can move, for example, because your drill (with Charge+1) will pierce the heavens). It also makes wheels more useful — skim is less slot-efficient than wheels in terms of dpr, and you want high dpr for greater possible Charge distance. When water gets implemented I’ll put skim on support/artillery mecha and wheels on melee mecha, have the support/artillery skim on water, and then use the faster wheels to charge my melee mechs at the enemy support/artillery before *they* can skim over water.

    Ion Lances are great, maybe even a bit OP. You can remove either HAYWIRE or OVERLOAD (just keep one of them) and I’d still love it. You do have to kill a Dielancer for one, so if that’s the only way to get Ion Lances in the future, I suppose having both HAYWIRE and OVERLOAD would still be fine.

    The changed combat rules (every mecha gets two moves in a turn) means no more heavily-modded flying Savins with 6 Phase Cannons sniping at everything on their very first initiative turn. I think that’s a good thing, I personally felt it was ridiculous loading up on phase cannons, there was no elegance to such a design.

    There was also that weird mechanic in previous GearHeads where installing weapons in the mecha frame directly was more encumbering (reduced MV/TR more) than having them mounted on mount points, which, at least by my imagination, did not make sense (you’re putting them into the hull directly FFS surely you can adjust the weight distribution or whatever better there!), but was necessary for game balance due to the ability to add more stuff on things than their slots could fit — GearHead Caramel taking away that ability and imposing strict limits on what you can install on any module means that the total weight is the only thing you have to consider for your mobility score. This makes a more elegant balancing mechanic: you can choose to try to fit a weapon into the hull directly,, or you can add a 1-slot mounting point (at 0.5 tons added weight) and put other things into the hull.

    Spellcasting (Prescience, Sensor Lock, Spot Weakness, Scrambler) are still a bit marginal. You are sacrificing a move for this, and you only get two moves per mech. With the power of the weapons available you are usually better off either trying to get to the optimum range and shooting, or just shooting twice instead of spellcasting and then shooting. Maybe reducing the damage of weapons slightly may tip the balance in favor of spellcasting more. Sensor Lock in particular seems to be very weak still, even with the improved chances of casting it (from the laughable base of 10 to a chuckleable base of 50). I did get some good use for Prescience when fighting the final challenges, but that is because casting it is 100% certain.

    Part of the weakness of spellcasting might be due to the sheer number of enemies you have versus the size of your own forces. With a bit of renown you are getting 1.5x to 2x the number of mechs against you. Casting Sensor Lock or Spot Weakness on an enemy mech is less useful than casting Prescience on your own mech — Sensor Lock and Spot Weakness have less than 100% chance of success, and if you have four Mecha against your enemy’s eight, your choice is between casting Prescience once improving 25% of your own forces, versus*successfully* casting Sensor Lock or Spot Weakness once worsening 12.5% of the enemy forces, versus shooting a Heavy Gauss Rifle for a chance to cripple or even eliminate 12.5% of the enemy forces . With that, the choice is usually the weapon, and *maybe* Prescience for really tough combat, but almost never Sensor Lock or Spot Weakness.

    That’s a bit hard to balance — if you improve Sensor Lock or Spot Weakness or Scrambler, then the large number of enemy forces means they can cast it on the fewer player forces, which impacts the player forces much more than the player forces can impact the larger number of player forces with the same spells. For Scrambler, maybe you can change it instead so that it affects at most 25%, or at least one, hostile units in a large area around the caster (put them in an array, shuffle, if the array is 7 or less affect the first unit, if the array is 8 or more take 1/4 the length rounded down and then truncate the array to that number and affect those units – rolling 25% per unit makes things too luck-based, imagine a very unlucky roll where the enemy casts Scrambler once and it affects all the player units, an occurrence that will happen 1/256 of the time). This is plausible as you broadcasting a scrambling signal that you coordinate with your allies so they aren’t affected, but has a chance of scrambling hostiles. Not sure if you can do the same to Spot Weakness or Sensor Lock — maybe having more opponents means the user of the skill is more likely to successfully spot *some* weakness or get *some* lock.

    • AmkG on April 21, 2020 at 4:55 pm
    • Reply

    Spellcasting ideas:

    * Sniper Mode – Cast on an ally or yourself. While in Sniper Mode you gain multiplicative defense bonuses from cover, and enemy defense bonuses from cover is reduced subtractively (i.e. if you have a Sniper Mode of particular level, a Cover: -10 might be reduced to 0 but a Cover: -20 might be reduced to -10). It is permanent until combat ends, OR until the unit moves (from normal movement or Charging). Since moving a unit also improves its defense, the defense bonus of Sniper Mode in good cover exists to compensate against removing the option to move while in Sniper Mode. The reduction in the cover defense of enemies is important to give the Sniper an advantage, since cover is symmetrical. As an active skill that adds a UI-visible status, it should be more discoverable by the player as well. It might be made more vulnerable to melee by removing all cover effects at melee range (a non-sniper in forest has Cover even at melee, but a sniper in forest in melee range has 0 cover and would not reduce the cover of an enemy melee unit).

    Note that while the typical trope for a sniper is a sharpshooter, technically a sniper is someone who is also good at camouflage and spotting targets. This is due to the original snipe, the bird, which has good camouflage, is very skittish and panics easily making it take flight, and has an erratic flight pattern that is hard to shoot: a good snipe-hunter (sniper) was someone who could spot targets easily despite them being in cover (to spot camouflaged snipes), zemself was very good at camouflage (to keep snipes from also seeing them and then escaping), and was a good marksman (to shoot down snipes even if they took flight). Playing around with the cover system and making it asymmetric in Sniper Mode seems to most closely emulate this, and gives interesting map-based strategic choices as well. In flat spaces Sniper is useless, so you might, on a Logistics level, be willing to pass up on Sniper skill in a lancemate if you expect to mostly fight in large open spaces (though you would have to make the story missions more selectable by the player whether the player will fight in an open dessert or plains, or in a forest or underground or city with lots of possible cover; Charge is a lot more useful in open spaces, so those are strategic options the player can select: sniper for tight spaces, charge for open spaces without water).

    * Pathfinding – cast to affect all allies in a nearby large circle. Units with Pathfinding active get a % dpr bonus. Times out after a while. Could be added as an additional action of high-level Scouting skill. Makes Charge and melee mechs even more powerful, which is important to get better combined-arms instead of just stocking up on Phase Cannons. Must affect the entire lance since you are sacrificing a move to cast it, meaning the caster has to stop in its tracks when it could have instead moved.

    * Mecha Analysis Program – Eletronic Warfare program with two actions, which when cast affects allies in a nearby large circle. Mecha Movement Analysis, which boosts weapons accuracy of affected allies, and Mecha Armor Analysis, which boosts weapons penetration of all affected allies. Alternately, they could effectively % reduce the mobility or armor stats of enemy mecha that the unit targets. Crucially, casting one will cancel the other. So you would need to consider which mode is more important, depending on whether your enemy’s armor or mobility is better, and consider as well your available weapons (thus making them % reductions of the enemy’s stats might be better, so that casting Mecha Movement Analysis on low-mobility enemies is less effective). This could have been how Target Analysis worked instead, which might fix the problems with Spot Weakness and Sensor Lock (maybe replace Spot Weakness with Pathfinding in the Scouting skill?).

    At the Logistics level a good combined-arms mecha designer might design two multi-role mecha variants: a heavily-armored lancer/sniper and a lightweight mobile lancer/sniper. If the enemy casts Mecha Movement Analysis the armored mecha act as lancers while the mobile mech act as snipers. If the enemy casts Mecha Armor Analysis the mobile mecha act as lancers while the armored mecha act as snipers.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.