Tumgik
#the weird thing is it's always the same 12 kobolds
probablybadrpgideas · 2 months
Text
Need to punch up the difficulty of a combat encounter? Just add 12 kobolds regardless of any other context!
Attacked by wolves? 12 kobolds show up and help the wolves! Exploring an ancient tomb that hasn't seen the light of day in 1000 years? Somehow, there's 12 kobolds at the end alongside the final boss.
The system doesn't even matter: offend the Prince of the city and he sends a batch of vampire thugs to make you not a problem? 12 kobolds!
768 notes · View notes
Text
A bunch of ramblings about personal stuff with regards to my own identity as 'human' or not follows. Wanted to get in on paper somewhere and was like "Hey, I have a blog," so. It's personal, but it's my (our) blog I get to post rambling personal posts if I wanna.
Today marks my (Quinn) second run-in w/ Something hitting the rat choir of yearning (good post) (technically like the umpteenth but in this case second) that made me go "Hm, am I a therian," stuff. We're no strangers to non-/alter-humans as a system, there's a wolf, a pair of dragons, and a couple other 'unclassified' folks, but y'know.
I've always kinda been the "token human"? In the past I've (semi-jokingly) used the phrase 'species nonconforming', I've just kinda bounced around from 'sona to 'sona pretty regularly (with a rough base in puppyesque vibes for a lot of reasons), but it's always been a role to play or mask to wear for fun. Y'know, typical furry stuff.
Then I read Taxxon's HRT fic (some of the others had read it before, but I ran into it myself later) like 2 months ago and that slapped the aforementioned rat choir into singing their familiar tune, and uh. I got stuck as a dog for like, a week.
So that was fun.
By which I mean terrifying, actually! For a lot of reasons! I'm really bad at introspection actually! But I tried to settle in as best as I could and when it passed I kinda breathed a sigh of relief and moved on. Went back to the, y'know, "have fun with it" vibe and kept going. Canid-specific 'sona's and roles were kinda poisoned a little by the experience, unfortunately, but it happens. Kobolds are where it's at anyway.
And then Last Night. A very good piece of art by ayviearttv here on Tumblr was passed to me (it's a series, go to their blog it's REALLY good). And uh. Oh Boy the Rat Choir. The night soured for unrelated reasons but a solid amount of it was "Why won't these tiny cheesebrains stop singing for like 5 minutes."
Like, they sing a lot is the thing. Not like, constantly, but a lot, in response to a lot of things that I won't detail. A few include like, specific depictions of androids, or organics becoming androids, etc., messy TF (ie; semi-realistic like in the case of Taxxon and Ayvie's pieces) both organic and mechanical, and so on. We/I have also been into dragons since, like, a super young age (raised on Dragonology, fantasized about being one, raised on Animorphs which also explains a lot, never read Pern but did read Eragon, etc.) so, y'know.
...I'm saying this like I'm trying to justify it. I guess I have to, to myself, a little. Not..."have to" but...feel the need to? I woke up this morning (like 12 hours ago) feeling like I had phantom limbs, they're still around when I'm not otherwise distracted (ADHD makes a lot of things go away when I'm distracted) and have been...odd to deal with. I've never been more aware of how dirty floors can be when it feels like a part of me is dragging on them, ha.
It doesn't feel scary. I mean, I'm anxious, a little. Moreso earlier. I'm anxious in the same way, the imposter way, the faking for attention way, so on, the ways that are usually externally motivated (or at least pretend to be). But unlike before I'm not...bone-deep terrified of it. It feels nice. Wings at my back, tail balancing me out, the strange feeling of horns and crest above, it's...I'unno. It's nice.
On the other hand, I miss being able to lay on my back without feeling weird, hah. Also I was basically sprinting out of the car whenever we stopped for errands.
My proprioception's (the sense of where your body is in space) always been kinda fluid and easy to fool, even for a human who developed a fluid proprioception to handle tool use (that's why tools and/or vehicles often feel like a part of your body when you use them and you (generally) know where in space they are even without looking). Like, as a kid I found a tarot book in our grandmother's workspace that had advice for developing proprioceptive wings through meditation (not joking) and that worked pretty well for me, among other things. It's just how I am. So we'll see, uh, how long this sticks around, I guess.
If...it doesn't...no harm. Mm...maybe a little harm. I'll be sad. I think I'd miss it now that I know how it feels. If it does, work's gonna be awkward, hah. They don't make chairs for that. It'll be fine, just funny.
I don't need advice or anything, to be clear, just musing. I like hearing about the experiences of other therians in specific/alterhumans in general. It's neat. Plurality was neat too, and then turns out I was we. Regardless, it's neat, so I guess I'm just putting this out there to have on paper, and if anyone reads it and it makes them think a little about themselves or encourages anyone or whatever, that's an added bonus.
1 note · View note
malaismere · 3 years
Text
Campaign 3 Predictions - Compiled
So, as a fan of compiling statistics, I've been keeping track of race/class predictions for campaign 3 for the past...at least a year, from tumblr, reddit, and twitter. with EXU over, and my spreadsheet hitting 400 (?!), I figured I'd share the fandom's current predictions
Travis
Human (30), Dwarf (29), Elf, Shifter (11), Half-Elf (10)
Cleric (103), Bloodhunter (63), Wizard (50), Fighter (49), Druid (46)
Lycan bloodhunter (41), Forge cleric (26), War cleric (16), Eldritch Knight fighter, Bladesinger wizard (14)
Marisha
Elf (16), Genasi, Tiefling (14), Dwarf (13), Dragonborn (11)
Paladin (112), Rogue (53), Fighter (43), Cleric (40), Warlock (37)
Eldritch Knight fighter, Glory paladin (14), Artillerist and Armorer artificer, Battlemaster Fighter (8)
Liam
Dwarf (25), Halfling (18), Tiefling (13), Elf (12), Warforged (11)
Druid (77), Cleric (72), Bard (71), Fighter (59)
Stars druid (16), Dreams druid (10), Eloquence bard (8), Alchemist artificer, Whispers bard, Twilight cleric (7)
Sam
Dwarf (34), Kobold (24), Goliath (17), Warforged (15), Kenku (13)
Sorcerer (106), Cleric (98), Druid (63), Wizard (47)
Wild Magic sorcerer (61), Wild Soul barbarian (14), Twilight cleric (8), Life and Forge cleric, Wildfire druid, Divination wizard (7)
Laura
Elf (21), Human, Tabaxi (19), Genasi (15), Gnome, Aasimar (10)
Barbarian (75), Sorcerer (74), Bard (64), Warlock (60)
Wild Soul barbarian (14), Wild Magic sorcerer (12), Glamour bard (10), Shadow monk (9)
Taliesin
Warforged (16), Elf, Changeling (14), Gnome, Genasi, Tabaxi (8)
Sorcerer (68), Rogue (65), Warlock (55), Bard (45), Wizard (43)
Aberrant Mind sorcerer (18), Whispers bard, Phantom rogue, Soulknife rogue (11), Mastermind rogue, Clockwork sorcerer (9)
Ashley
Elf (26), Human (18), Tiefling (17), Half-Elf (12), Dwarf (11)
Rogue (95), Bard (91), Monk (43), Ranger (41), Warlock (38)
Swashbuckler rogue (21), Glamour bard (13), Mercy monk (9), Drunken monk, Wild Magic sorcerer (8)
I also (although less consistently) collected continent/setting predictions. Marquet was the top (49), then Issylra (29) and the Shattered Teeth (22). For non-continent settings, some form of Spelljammer was the top (19), followed by the Age of Arcanum (17), and Planescape/Planehopping (15). Underdark, Ravenloft, Blightshore, and a return to Tal'Dorei were also suggested multiple times.
Much longer and rambly discussion (and my own predictions) under the break.
Top predicted races were Dwarf, Elf, and Human (~100). Dwarf and Elf haven't been played before, so that tracks, and I don't think it's out there to assume we'll get at least one human again. Also, post the whole thing with Essek and long rests, people really started jumping on Elves (which, fair). Warforged, Dragonborn, Tabaxi, Genasi, Tieflings, and Changelings all are pretty prominent (~50).
Of the races not yet established as existing in Exandria, Warforged and Changeling were the most popular (Warforged now dubiously canon post-Aeor, and Changelings dubiously canon with the LoVM bartender), followed by Shifters, Leonin, Kalashtar, Fairies, Grung, Ravnica races (Loxodon, Simic Hybrid, Vedalken), Van Richten's Races (Dhampir, Reborn, Hexblood). Locathah and the other Feywild/Strixhaven races are the only officially published races at 0 suggestions. The lowest previously seen race is Gobins at 2, one of which was for Sam again, and the lowest PHB race was Half-Orc at 17.
Class wise, Sorcerer was actually the most predicted class (which kind of tracks, as it's the one that hasn't shown up even as multiclass), followed by Cleric (generally assumed as compulsory), Paladin (only as a multiclass), and Rogue (also assumed as compulsory, but way less so. Not surprisingly, Bloodhunter, Ranger, and Artificer were the lowest.
Wild Magic Sorcerer was far and away the most suggested subclass, the only one to break 50, although it hasn't hit 100 quite yet (I think it will by the time the final characters are announced though). EK Fighter, Lycan Blooodhunter, Forge Cleric, Swashbuckler Rogue, Wild Soul Barbarian, Stars Druid, Glamour Bard, Bladesinger Wizard, Eloquence Bard, and Echo Knight Fighter are the other top subclasses.
Every official subclass has been suggested except for Berserker Barbarian, Grave Cleric, and Transmutation Wizard (previously played), Battlerager Barbarian and Banneret/Purple Dragon Knight (SCAG subclasses, which are widely unpopular), and the dubiously-official Planeshift subclasses. Open Seas Paladin is the only Matt homebrew to not be suggested at least once. For dead UA, Satire Bard, Brute Fighter, Giant Soul and Stone Sorcerer, and Raven Queen Warlock have all been suggested, usually only once, although many of the suggestions were collected while classes were in UA for Tasha's, Van Richten's, and Fizban's which is technically still UA but announced so...
With Travis, the predictions bounce between two main ideas - a melee spellcaster (Forge/War/Tempest cleric, Bladesinger/War wizard), or going back to a melee class (Bloodhunter, Fighter) but with a bit more mechanical interest (Lycan, EK/Echo/Rune/Battlemaster). I think those are both solid predictions, and while I really, really doubt we'll see a Lycan bloodhunter or a Forge cleric, I think the general vibe is probably spot on.
My own prediction is one of the more out there, but still in line with the general thinking - Artillerist Artificer. Travis is definitely a very tactical player, and it would be cool to see him get a turret for the battlefield, plus all the general utility/versatility of the artificer. Alternatively, I really could see a rogue, although more like what Mastermind or Inquisitive is trying for as opposed to how they actually turned out, if that makes sense.
Race wise, the top guesses are fairly plain, outside of shifter (which is mostly tied into the "werewolf" vibe). None of them would shock me, but I don't have any predictions.
I think that everyone's right on the money with Marisha as a paladin. Her next character being high charisma seems spot on, and I think moving to a half-caster also tracks. EK/Echo/Rune/Psi fighters would also fit, although they don't lean towards high charisma, or a warlock, maybe a more melee one.
Rogue seems unlikely purely due to the fact she's played one before, kind of. Matt and Marisha have both talked some, but her first game wasn't Vox Machina, but a previous game Matt had run where she'd played an assassin. You can do non-assassiny rogues, but still.
(Other fun facts about this game because it's wild: apparently the session she sat in on before playing involved half the party getting eaten by ghouls. the party joined up with another half-tpk'd party (marisha and the replacement characters) to get the raven queen to bring their dead friends back, and a fate-touched rogue swore service to the Raven Queen in order to bring the last party member back.)
My prediction for Marisha is also paladin, although I don't have any thoughts on the subclass, with genie warlock as a second because they are fun. No real thoughts on race other than I too would love to see tiefling Marisha.
Most people are going with a support caster for Liam, which I totally buy. Caleb definitely leaned towards support caster, even if he usually did end up played as DPS. Druid has taken the top given the polymorph->wildshape vibe, although it's still very yclose with Bard and Cleric. Suggestions for fighter dropped after EXU, and while Liam does play a lot of fighters, I doubt we'll see it for C3.
Honestly, Liam is the one I have no predictions for outside of 'support caster'. I'd lean away from Cleric and towards Druid or Bard, but it's hard to say. I also think Artificer deserves to be in the running, as it seems like something Liam would really enjoy, but also...might not want to go Int-caster to Int-caster. My only real thought on race is that I want to see whether Marisha and Liam choose the same again.
Top guesses for Sam is, far and away, Wild Magic Sorcerer. This was also the top guess for C2. I do not think Sam will play a Wild Magic Sorcerer. In general, though, the vibe is going back to fullcaster - Sorcerer, Cleric, Druid, Wizard. I think full caster is probably right.
Sam is so hard to predict because it isn't what he'd choose, but what Liam chose for him. I think it's either something really standard or something really out there, and since I can't guess the really out there, I'll go for the standard - Elf Wizard or Dwarf Cleric, leaning towards Dwarf Cleric, due to the support class and the fact that Sam's mentioned never playing a religious character.
The main vibe for Laura is definitely "DPS" which is understandable. I don't know if I agree with it, but I understand it. Aside from Barbarian, the rest of the vibe is spellcaster - and I don't think we'll see a completely no magic character from her either.
Prediction wise...I understand barbarian, but I'd actually go with Ancestor or Beast over Wild Soul. I could actually see a Bloodhunter from her too, although leaning away from Vex vibes. I think I'd want to go with Wizard, though I'm not certain on that. I would bet Tabaxi but idk, I could see her avoiding that for Travis' sake.
Everyone always names Taliesin as the hardest to predict (he had the lowest count at 354, under even Ashley at 365, to everyone else's ~380/400) but I don't think he's harder to predict than Sam. The thing that makes him hard to predict is that he likes to build characters to fit the party, which he (probably) won't be doing, same as with Molly. The other main thing he tends towards is mechanical complexity in a way that suits his characters.
The main driving influence in the top suggestions is Eldritch Weirdness. Aberrant Sorc, Whispers Bard, Phantom Rogue, Warlock in general. I don't disagree with any of the subclasses, but I really don't think he'd go eldritch for eldritch sake, if for at the very least being...he has always been this weird and it's yet to be a driving force behind any of his characters before. Like the Taliesin-is-an-elder-god thing, I think this is mostly people who don't hang out around occultists. Look, I've had multiple people sell me their actual souls, and you don't see all my characters being warlocks.
That being said, I don't think I disagree with the top classes, just the subclasses. I definitely agree with Sorcerer as a good choice for him, although I'd actually go Clockwork, as I think it has a fuck-with-the-DM vibe. Taliesin is the most heavily suggested for dunamancy subclasses, which wouldn't surprise me, but I think he might avoid on the sole point of not wanting something too tied with the last campaign. A lot of people also name the psionic subclasses, which I'd be more likely to second if they had kept the weird mechanic from the UA, but don't disagree with, excepting my issue with Aberrant Mind.
My out there guess is that he's going to choose a multiclass build. He definitely enjoys playing around with weird builds (Owlbear, he did a non-CR oneshot as a monk/stars druid). On the one hand, a lot of these builds work best for oneshots or starting at higher levels, as they can take a bit of time to come online, but with such a large party, I think it will still function.
(my actual prediction for Taliesin is that his character is weirdly reminiscent of either the aasimar echo knight or the elf blood cleric from the exandria game I'm running.)
Ashley is being predicted as a Dex/Cha build, and I'm totally here for it. Pre-Fearne, I was leaning Ranger, especially Fey Wanderer for a fey build, but post-Fearne, I'm going Rogue, especially Swashbuckler. I agree that seeing a high Cha Ashley would be great, especially to let her be more center-focused than Yasha had been, and swashbucklers are just...really fun. Also, the whole Aeor arc really left me wanting to see Ashley as the go-ahead-and-scout character, just to watch her push buttons.
For continents...I understand why people are guessing Marquet, since it's currently the most explored. I think that if they're going to do Marquet, then Matt will sit down with a cultural consultant. I say will over should, because I won't make any value judgements, but I think it's in line with what Matt and CR would do in that situation.
I can't really tell whether this is a prediction or what I'd like to see (the two are distinct but often difficult to untangle) but I'd actually go with Issylra, and specifically playing up the (at least initial) set up of explorers and adventurers heading out into the wilds. I will also place my bets on them having some sort of more steady home base, and my hopes on that they get an airship. My wildest out there guess is that the plot will move towards either planescape/spelljammer in the upper levels, tying into some of the seeds from the end of C2.
I have seen a handful of people predicting table seating order, which is both very minor and also the thing that I may be most interested in. A while back, someone made a post pointing out that the main romantic relationships were all cross-table, while the strongest platonic relationships were same-table or side by side. Because I am the sort of person that I am, I did statistical analysis on ao3 fics....and it's statistically significant. So I am trying to see whether or not, based purely on C3E1, I'll be able to predict what the top ships for the campaign will be.
This rambling has mostly gotten out of hand because I don't have much opportunity to talk about this, but, you know. If you send me predictions I will give you the current odds gambling style, so that you'd know how much you'd win if you'd place a bet, because I did the tables up as a joke for something else and now I kind of want them to be used for something.
59 notes · View notes
dapperkobold · 6 years
Text
Hypotheticals of the Storm: Sombra
“Kobold, no,” you say as you read the title, “No more Overwatch heroes, we just got two and they’re both flipping weird, don’t go asking for more.”
It’s okay! I’ve chosen the single overwatch hero LEAST likely to get in! I just think that the mechanics would be fun!
“Fun?” you say, “Sombra’s mechanics wouldn’t work in a MOBA at all!”
“Yeah!” I say, nodding, “That’s what I said! Fun!”
Who is this?
The hacker from Overwatch, Sombra runs around with a SMG, a hacker rig, some camouflage, a teleporter, and that’s it. She is hard to pick up, hard to use, only situationally useful, and when done well a giant pain in the enemy’s hind quarters. While called a offence hero, she has no damage-dealing abilities and is instead reliant on her abilities to be fast, unpredictable, and control health packs to be a monkey wrench in the opponents’ plans. 
She’s generally a wonky character who always seems to have people arguing over whether she’s any good.
Basic Ideas
In Overwatch, Sombra has a good gun, three utility powers, and a very big utility ultimate. She can get away with being an Offense hero in Overwatch, but I don’t think she’ll cut it as a HotS Assassin. She isn’t tough, so she’s not a Warrior, and she doesn’t heal at all so she’s not a Support. Welcome to the other category, Sombra: you’re a Specialist. With stealth, a movement bonus while in stealth, and teleportation, she may be one of the most mobile Specialists in the game.
Core Mechanic Ideas
I think she’d still work on Mana (lacking a different system to use instead) but it wouldn’t be an immediate concern for her: the Mana would more work as a pace limiter than a steep limit, more akin to Artanis than Li-Ming. Chances are she’s going to be considering pulling back for health before Mana, but even so there’s no reason to let her run around indefinitely if even Raynor needs to rest up for Mana on occasion.
She’d be a fragile character, with a solid basic attack similar to dedicated sustained damage Assassins. This is pretty strictly needed, for the same reason that she does so much damage using her gun in Overwatch: It’s the ONLY damage output she has in a game all about killing things. She doesn’t get any abilities (ANY!) that do damage, so killing things is entirely dependent on her base attack damage output. However, I don’t think being able to attack while moving is needed, mobility on the attack isn’t that integral to her character idea. 
Other than that, even at a thought it’s clear that she’s going to be leaning a lot on Hack to do things. In Overwatch it kept enemies from using abilities (that would be silence in HotS) but it could also target Torbjorn’s turret (but few other player-based structures, which I think is a shame) and most importantly health packs. Seeing as there are no Health Packs in Heroes of the Storm (I don’t think Regen Globes can be targeted by abilities) she needs a different edge.
The obvious conclusion is that her Hack ability should also be able to target enemy buildings! It could keep Turrets from firing and the like, that’s the beginning of a solid core mechanic for her.
Problems Already
What do you call a hero who goes around stunning structures?
Sylvanas.
Sylvanas can stun any mercenary, structure, and minion with both her basic attacks and all her basic abilities, and it’s easy to keep two things stunned at a time. A hero who can 1. only stun structures and 2. really only stun one at a time while 3. having less combat flexibility than almost any other character in the entire game is just a strictly inferior Sylvanas. Having high mobility, stealth, and a good escape option isn’t enough to counteract that. Sylvanas already even has a good, reliable escape option! 
This is really why I’ve been thinking about this: on paper, Sombra (a hit-and-run opportunistic character) and Sylvanas (a lane-push master) should have different playstyles, but a quick and easy Sombra translation puts her square and plain in Sylvanas’ claimed territory. How do you translate Sombra’s hit-and-run backline-threatening supply-interfering playstyle into something that works in HotS?
Let’s roll back a little bit.
Core Mechanic Ideas
Sombra’s gameplay is going to focus very strongly on her mobility, as that’s the focus of her powers. She’s going to be able to engage well, and disengage well, she just needs good options for what to do while engaged. She either needs raw damage output that permits her to score kills that her enemies don’t expect her to be capable of, or some other utility whose effect doesn’t stop just because she leaves the area.
That’s the big deal: since she’s so mobile, what she does needs to persist even though she likely won’t be in the area very long. Her playstyle should encourage going somewhere, being a right pain to the area, and then running off as enemies arrive to go be a pain somewhere else. Sombra is not going to be participating in long trades or pushing a lane, she doesn’t want to be in the middle of a team fight. She wants to be off to one side, taking opportunities, and then vanishing when things get dicey so she can cause trouble somewhere else.
The core theory behind her gameplay in Overwatch seems to be that she is opportunistic: she takes opportunities that present themselves. Her passive is even called opportunist. As a result, it’s important that she be able to both A. identify opportunities and B. be able to capitalize on them.
Trait: Opportunist enemy units at 50% health or below are visible within a (large) area around Sombra. This would not allow her to target them with abilities, and it would not put them on the minimap: only Sombra can see them on her screen. This would work really well with the recently changed camera. It would not allow her to see enemies in stealth, but it would allow her to see enemies on the far side of walls. So if she beats an enemy around and they vanish into a bush, sombra can watch and see where they run when they exit the bush. Note also this is ALL enemy units, including enemy minions, mercenaries, and maybe even neutral mercenary camps and monsters. Under the right circumstances, she’d be given information that no other character would have, like the fact that a minion wave is just passing by outside the bush, or that a Mercenary camp was started but for whatever reason not finished.
It could also possibly point out hack targets if any of the more complicated hack suggestions are implemented.
Hack, possible targets: Hack will obviously act as a silence against heroic targets, but I think that’s just the beginning of its possible uses. Consider:
1. Stunning Turrets, Forts, and Keeps for an extended time: Not just one second, or however long the silence acts, but longer than Hack’s cooldown, ideally a lot longer. 15 seconds or something else similarly absurd. Long enough that even just leaving it unmanaged against a minion wave would cause damage.
2. Opening Enemy Gates: letting a trapped ally escape from behind the turret wall, or maybe letting an ally enter it? This would not make it stop being a target, minions and mercenaries would still attack it and it would still be hid by any ability that would hit a building.
3. Blocking Healing Fountains: Not messing with any character’s cooldown, but making the fountain itself unusable by enemies. Maybe even usable by allies, but I don’t see that getting a lot of use since the time between a healing fountain being accessible by the enemy and its destruction is generally pretty short. This would have to last a long time to be relevant at all, 45 seconds or something.
4. Regen Globes: If there IS code in the game to let them be targeted by abilities, it would be a nice touch for a targeted regen globe to be set to your team and re-set its timer, but it never switches to neutral.
5. Mercenary Camps: NOT to take them. Instead to make them harder to take for the enemy, give them 15 armor vs the enemy team or something.
6. Enemy siege minions: Like the Catapults, stunning them for a time. Also possible to apply to enemy mercenaries, but not boss mercenaries.
7. Enemy hero-made buildings: Like Gazlowe’s turret, or Probius’s turret. It would stun these buildings. Of course, the list of heroes that make buildings that could be stunned is pretty short. Maybe Zagara’s Creep Tumor, maybe Abathur’s Toxic Nest? Would hacking a creep tumor or nest just disable it, or destroy it?
Hack, not allowed: Some things are too important and should NOT be viable targets for hack.
1. Any Map Objective: No, just no. Even vehicles, that said I think all Vehicles are immune to Silence in any case.
2. Cores: Maybe a rank 20 Talent that slows their attack rate briefly or something. Maybe. More likely nothing.
3. Boss Mercenaries: If hacking mercenaries in any form is possible, Boss mercenaries in any form deserve enough respect to ignore it.
Basic Ability Ideas
Okay, here’s where I port the powers from Sombra in Overwatch over to HotS.
Ability 1: Hack The musings above are well and good, but what’s the end result? It’s a click ability with a short channel time. You can move while channeling but not attack or use other abilities, and it can be interrupted by stuns or silences. It silences the enemy for a long time, 3 seconds or something, leaves them revealed (unable to enter stealth), and keeps track of them as if they’re visible. They do not give visibility, they instead show up even outside of any hero’s vision range and on the minimap for a while (8 seconds or something). Buildings hacked (let’s say turrets, hero-made buildings, gates, forts, and keeps) are stunned for a time (12 seconds?) in which time they cannot attack, give vision, or do their other essential building roles (Probius plyons don’t have an energy field, gates are stuck open.) The duration on heroic buildings might be lower due to the fact that they are attached to heroes (8 seconds?)
For the example times above, I’m assuming a power cooldown of 8 seconds: this means that Sombra could (potentially) keep a single hero visible indefinitely, and stun time on buildings would overlap some. Hack cast on a target that’s already hacked would just re-fresh the duration.
Talents could include normal things like improving the range, removing the channel time, adding a slow, but there could also be options to unlock other hack options I listed above: The ability to hack a healing fountain for a long time (but only one fountain at a time) or hack mercenary camps to bias them against the enemy team.
Ability 2: Thermoptic Camo A stealth power, giving sombra stealth and a move speed increase for a duration. Personally I figure a move speed increase of 30% or more. Yes, as fast as a mounted hero. In Overwatch the speed increase from this power is just shy of 47%, making her temporarily the fastest footspeed hero in the game. Giving her mounted speed would leave her proportionately slower than she is in Overwatch. Obviously, the move speed bonus should vary alongside the ability’s duration; a longer duration should have a slower speed bonus, and a shorter duration a higher one.
If you care, mounting would still be faster for long distances since the ability would need to go on cooldown.
Traits would, admittedly, look a lot like Samuro’s Wind Walk Traits. I feel that a shorter duration, faster move speed, and shorter cooldown time should differentiate the two abilities if needed, Sombra will be leaning on her own stealth more than Samuro leans on his.
Ability 3: Translocator Activate once to throw the translocator, activate again to teleport to it. This needs a lot of considerations; in Overwatch Sombra has a 15 second window in which to use it, which seems a little short for my tastes in HotS. There’s a few possible changes for this:
1. Make the duration more like 30 seconds. This would really need testing to pin down a good duration.
2. Make it indefinite duration, with cooldown activating as soon as you place a Translocator and a different button to recall to it.
3. Make it so that Translocator does not go into the full cooldown if it expires. Maybe a very short cooldown, 4 seconds or so.
My bias would be for 3 to be the default, with one tier of Talents giving Sombra a choice between increased duration, indefinite duration (so you activate the Trait to teleport to the Translocator) but with increased cooldown, or no cooldown if the Translocator expires.
Heroic Ability Ideas
Heroic 1: EMP Scientific accuracy aside, this ability ports over from Overwatch pretty cleanly. Activate to silence heroes in an area around you, reduce shields to 0, and stun hackable buildings. None of the advanced hacking abilities would trigger, but I do feel that stunning turrets, forts, and keeps is important enough to include in the EMP.
Heroic 2: Trojan Okay, here’s a fun idea: the second heroic targets one building, Turret, Fort, or Keep. That building attacks your enemies for the duration. It’s still an enemy building, you can still attack it, but it acts as if it’s allied to you. Turn the fortification around on the enemies. I’m not sure how long this would last, certainly not as long as the Hack stun, but it needs to be long enough to have a solid impact on the fight instead of just a momentary event. Remember: the theme for every non-mobility Sombra ability is that there is an impact even if Sombra walks away.
Playstyle Ideas
We’ve already discussed this at length, but let’s look back over it now that we have some solid notes.
Opportunist lets Sombra track anything that’s hostile to her and below 50% health, giving her improved situational awareness and information about things that she normally wouldn’t be able to see. In addition, with a solid basic attack she has a chance to finish off the wounded she finds.
Hack is Sombra’s butter to go with the bread of her solid basic attack, and she can use it on enemy heroes for a long silence and then the ability to track them for a bit. A hero who is being tracked can’t enter stealth and does show up on the Minimap for her team, so it increases the situational awareness of the entire team. Hack also stuns buildings for an extended period of time, so even if she runs off someone else pushing the lane might benefit from it. She doesn’t really have waveclear so she won’t be doing a lot of pushing herself, but if she works with someone else who is pushing she can make a siege go much smoother.
Her mobility is a mix of Camo for stealth, moving fast into engagements, and a set up Translocator for escape. It needs some forethought and practice, but Sombra’s MO should be to appear from stealth, make trouble, and then teleport away before she can suffer the consequences. In addition, having such high mobility from two different powers means that she can apply them differently in different situaitons: throwing a translocator and then traveling to it immediately might on occasion be the best option you have, as might using Camo while running to get raw distance and protection from targeting.
More Problems
At this point the theory behind Sombra is pretty solid, I think. The overlap with Sylvanas has been addressed and a unique playstyle has been planned. The question suddenly becomes, ‘but is it any good?’
You see, this playstyle is different enough from anything else, and different enough from the core theory of the game, that it’s a big toss-up whether it will work at all. The theory is as sound as it’s going to get, I’d certainly try the character I have laid out here, but that doesn’t mean a thing if the character is bad and never used.
Sombra would need to be on knife-thin balancing. If she falls behind even a little, she would get tossed aside in favor of Gazlowe, Sylvanas, or Zagara; characters that can effectively tie up turrets while still dealing more damage. If she’s too good, however, she would be an insufferable menace. I don’t think ANY of us want to have to worry about Sombra every single game. There is, in theory, a balance point where she’d be a fine if gimmicky character. The issue is that finding it wouldn’t be a matter of theory and base mechanics, it’d be a matter of fine number tuning.
That’s part of why I chose Sombra as least likely Overwatch character to see HotS: while her theory can be translated over, the implementation would take such testing and fine-tuning for a character that a lot of people are never even going to consider playing because the mechanics are just THAT weird that even if the fans were clamoring for more Overwatch characters, or more stealth characters, or more siege characters, balancing Sombra would be too much effort in the first place. You need to balance stealth, teleporter ability, not having any damage powers, and a unique hack mechanic. Rory Swann would be easier to make into a specialist, or Mira Han, or Torbjorn, or Symmetra, or a heroic Swarm Host, or any number of things would be easier to make, balance, and sell than Sombra.
So, if they do decide to do it, I’ll be here to watch the carnage, but if they do decide to do it, most likely it would be with an all new kit.
Still, I’m willing to bet we’ll see Bastion before we see Sombra. He doesn’t have any huge, glaring overlap-Sergent Hammer DARNIT.
1 note · View note
ciathyzareposts · 5 years
Text
Game 322: Nemesis (1981)
Hey, it was a bare-bones era.
       Nemesis
United States
SuperSoft (developer and publisher)
Released in 1981 for CP/M
Date Started: 24 March 2019
Date Ended: 24 March 2019
Total Hours: 4 Difficulty: Moderate (3/5) Final Rating: (to come later) Ranking at time of posting: (to come later)
In 1977, the innovative first-person dungeon crawler Oubliette appeared on the PLATO mainframe system at the University of Illinois at Urbana-Champaign. Two students, aspiring programmers, became convinced of its commercial potential. Taking various elements from the game, they reprogrammed it for the microcomputer and released their version in 1981, offering no credit or acknowledgement to the Oubliette authors. The game was a smash hit and launched a dynasty of sequels and imitators, influencing the genre down to the present day.
The last sentence makes it clear that the above paragraph was about Wizardry, but take it out and you also have a description of Nemesis, one of a very small number of RPGs released for the CP/M operating system. The CP/M was a popular OS for Intel 8086 and 8088 computers in the 1970s, and based on most accounts, it would have been the OS of choice for the new IBM-PC if some issues hadn’t arisen over a non-disclosure agreement, leaving the door open for Microsoft to sell IBM on PC-DOS, which ironically took some of its elements from CP/M. If things had gone another way, Nemesis might have been one of the first RPGs for a booming OS rather than one that died the same year.          
A mix of D&D, Tolkien, and Donaldson in the race list was an early clue.
         Like Wizardry, Nemesis isn’t an exact copy, and has plenty of its own innovations, so we shouldn’t go too far in making accusations of plagiarism and such. In fact, in making their adaptation, the authors–Michael A. Pagels and Michael Q. Hiller–changed enough of the elements (in particular getting rid of the 3-D interface) that I might not have noticed the association. What tipped me off was the use of “ur-vile” (from Stephen Donaldson’s Thomas Covenant series) as a character class. I knew I’d seen that before, searched my blog, and came up with my entry on Oubliette. From there, I noticed that the list of races for the two games were exactly the same, in the same order, excepting the replacement of “Eldar Elf” with “Grey Elf.” Then I got hold of the game manual and noted that the address for SuperSoft was a post office box in Champaign.
Getting the game running was no picnic. The only reliable CP/M emulator that I could find (Simeon Cran’s MicroFast) was for DOS, which put me in the weird position of running an emulator within an emulator. The game then requires you to create a configuration file for the terminal you’re using before you can run it. It has configurations programmed for numerous terminals, but none of them seemed to overlap with the various options offered by MicroFast. Actually, one did–the D.E.C. VT-52–but I overlooked it for a while, wasted a lot of time trying to define my own terminal type, and nearly gave up before I figured it out.
Nemesis is necessarily dumbed-down from Oubliette. Microcomputers of 1981 had nothing like the resources of the PLATO mainframe. Oubliette‘s explorable “town” level with numerous shops, inns, and so forth was (like in Wizardry) turned into a menu town. Instead of a party, a single character adventures alone. Combat is rendered considerably easier as a consequence.
But the basic rules, logistics, and statistics come directly from Oubliette, which itself drew heavily from Dungeons & Dragons and a few other sources. Character creation has you choose first from 15 races: human, elf, dwarf, half-dwarf, half-elf, hobbit, orc, uruk-hai, ogre (misspelled “orge”), pixie, goblin, hobgoblin, kobold, ur-vile, and grey elf. The game then randomly rolls for your strength, intelligence, wisdom, charisma, constitution, dexterity, gold, and (weirdly enough) sex. The rolls are modified by your race choice. You also choose an alignment from lawful, neutral, or chaotic, which was also used by Oubliette but goes back to original D&D.        
Choosing a class after rolling attributes.
       You can re-roll as many times as you want before accepting the character, at which point you choose from a list of available classes, with those that don’t meet your minimum attributes filtered out. There are 15 classes, and they again match Oubliette’s list in names and order, except for the substitution of “rogue” for thief and “featheror” for courtesan. Nemesis‘s full list is cleric, demondim, featheror, hirebrand, mage, minstrel, ninja, paladin, raver, peasant, ranger, rogue, sage, samurai, and valkyrie. After a few false starts with boilerplate characters (e.g., an ogre hirebrand), I decided to aim for a “minstrel” because it amused me to think of an adventurer pratfalling his way through a dungeon while belting a tune about his mammy down in Alabamy.
The menu town has an armory for buying and selling weapons and armor, a hospital, an inn, and Archives. Hospitals and inns both let you restore hit points. Hospitals cost money but heal you a lot faster (in game days) than inns, which has implications for your longevity. The Archives is where you go to pay money to have unknown items identified. New characters have no equipment, but they also have so little gold that you usually can’t buy anything. Even a “pointed stick” costs over 100 gold pieces. So you enter the dungeon and take a chance with your hands.           
Visiting the store. I have no idea why I’d need a hanky, brick, or beenie.
Visiting the hospital after a rough dungeon trip.
           Gameplay consists mostly of wandering the 21 x 23 dungeon levels, picking up equipment as you find it, and killing monsters as they attack you. The items you find almost immediately outclass what’s available in the store, so you mostly use gold for healing and identifying items in the archives. You want to stay near the entrance until you gain a few levels and extra hit points; although combat is relatively easy on Level 1, the game still features permadeath, and you can always get unlucky.
The dungeon is rendered in roguelike fashion, with ASCII characters representing the walls and doors, rather than in the 3D graphical fashion of Oubliette. It’s possible that the developers were exposed to Rogue but equally possible that they came up with the idea independently. In addition to stairs, players can encounter chutes and pits to lower levels, teleporters, anti-magic rooms, anti-cleric rooms, “melee rooms” (every square has a combat), and special treasure rooms. You maneuver with the URLD keys.         
Making my way through the dungeon. The character is a flashing underscore, so don’t ask me to find it on this static shot.
          Combat is drawn largely unchanged from Oubliette. When you encounter an enemy party, you’re taken to a separate screen. Where you can see your own statistics and inventory and the enemy groups that you face. Your options are only (F)ight, (C)leric spell, (M)age spell, and (U)se a special item. There is no fleeing or parrying. Even worse, the point of most of the character classes is nullified, as the authors failed to adapt any of their special abilities. Courtesans/”featherors” and minstrels can no longer charm enemies; rogues and ninjas can’t hide; clerics cannot dispel undead; paladins cannot lay on hands. There is a suggestion that some of these abilities were intended for a sequel.         
I believe the primary party I’m fighting is orcs, but I caught this in the process of refreshing the screen. Two mediums (priestly classes) have joined the battle.
         Other monsters may appear to join a battle in progress. As you kill them, you see your experience and gold increase. Leveling happens when you leave the dungeon, and it’s accompanied by increases in maximum health and spell points.
Nemesis offers 55 different monster types, and all of them appear in Oubliette with a few exceptions, and those exceptions are all simple substitutions. For instance, Oubliette‘s giant spider and giant ant become “huge spider” and “large ant” in Nemesis. Oubliette has a lot more monsters than Nemesis; those that didn’t make the cut tend to be the higher-level monsters like dragons, medusas, and advanced spellcasters, and I suspect that the Nemesis authors didn’t know how, or didn’t have the space, to program those enemies’ special attacks.             
The game’s town. Oubliette had stores, hospitals, and inns, but I think the Archives are original to this game.
          There is some overlap in the games’ spells, but on the other hand, the 13 mage spells and 11 cleric spells offered by Nemesis are common enough that they could have come from anywhere. Nemesis doesn’t require you to know a spell code name to cast its spells. They are separated into travel spells (“Light,” “Protect,” “Levitate”) and combat spells (“Damage,” “Sleep,” “Fireball”), and each depletes a number of magic points from the character’s pool.
There’s no main quest or winning condition in Nemesis. The manual encourages you to set your own goals, such as a certain experience level or treasure level. Survival isn’t very hard if you can live past Level 0 and if you play conservatively, for instance returning to the surface when you’ve lost half your health. The game earns only a 12 in my GIMLET, with no element rising above a 2. It is particularly hurt by the lack of any backstory, NPCs, or quests (all 0).           
Ironically, one of the monster types that the game did not adapt from Oubliette was dragons.
         The manual indicates that Nemesis II was already under development when Nemesis shipped. The creators intended to bring multi-user capabilities to the sequel. Players were invited to join the “Nemesis User Group,” which met at Hiller’s residence, to test the new adventure. Alas, it was never finished.
Pagels, Hiller, and SuperSoft issued at least two other products: a multi-player science fiction game called StarJump and a dungeon level and character editor called Nemesis Dungeon Master. The latter came with the edition of Nemesis that I downloaded, but it must have been a late addition because the manual doesn’t mention it at all.            
The Nemesis Dungeon Master character editor.
         On a Google Group about a year ago, Pagels indicated that he and Hiller “had a great time writing this game, and it helped pay for grad school.” Neither continued in the gaming industry. I reached out to both for comments but didn’t get a response.
If we ever get hold of OrbQuest (1981), we may have a challenger, but until then, I’m willing to call Nemesis the best CRPG issued for the CP/M operating system. I’m glad we had a chance to check it out.
source http://reposts.ciathyza.com/game-322-nemesis-1981/
0 notes