Sunday, November 17, 2019

MTG Arena: Efficient Card Acquisition

'Free to play' games have evolved over time to have more and more obscure monetization schemes. Humans are bad at making many decisions on a row and the developers have figured this out, so they tend to make things take more effort to figure out. The less clear it is what you're doing, the more likely you are to make a purchase that you wouldn't actually make with a clear head and perfect information.

You'll never find an actual price tag on a thing you want to buy. In League of Legends, for example, the new champion costs 7800 blue essence or 975 riot points. What does that even mean? Blue essence you get over time for playing the game, riot points you can buy with cash. 490 points for $5, with 'bonuses' if you buy in bulk. So Senna is priced at $9.95, but you can't spend exactly that amount. In this case the next highest amount you can actually spend is $10, so really close, but often that is not the case. They make you buy extra points and then you're committed to the platform by having more currency to spend later. But not too much more! You'll need to top up if you want to buy something else. The idea is you spend more now, so if you don't come back they get more than they would have with an actual price. And if you do come back you're hooked and more likely to spend more. And by having to look at all this math your brain gets a little overwhelmed, and if you have money to spare it's easy to just spend rather than work it out. They've also worked anchoring in there by giving you a base price for points but then giving you even more if you buy in bulk. Lots of marketing trickery going on.

MTG Arena has lots of this happening. Not necessarily maliciously evil, but it's certainly using things that worked in the past for other people and those things work by taking advantage of people. MTG Arena also seems to have extra systems making it more convoluted but I suspect this was more no one had an actual vision for how to do things so more people tacked things on that they thought was needed. It certainly has kept me from really figuring out what I should be doing to build a collection properly. But I've qualified for some sort of tournament so I need to play a wider variety of decks. So I need more cards. So I need to work this all out.

MTG Arena doesn't use a disenchanting system like Hearthstone does, so you never need to make a decision about keeping cards you've opened. You always keep them, you can't destroy them, and if you get more than the max amount you get a very minor boost elsewhere in your collection. I think this is an actively good decision. I've hated that I've had to blow up Hearthstone cards to build decks in the past because it actively hurts my collection to make short term progress and that's a bad feeling no matter which side you're on.

There are a bunch of different currencies going on. You've got gold, which you earn by doing daily quests and as tournament prizes. You've got gems, which you buy with money and earn as tournament prizes. You've got XP, which you earn by doing daily quests. You've got wildcards, which you earn by opening packs. You've got packs, which you can buy with gold, or gems, or cash, or earn as prizes, or earn from XP. And you have the individual cards that you can get by drafting, or playing sealed, or opening packs, or doing dailies, or as prizes, or from earning XP. They all switch back and forth in different ratios depending on different tournament formats!

But for now, all I really care about is getting more cards. How do I make best use of the different currencies in order to get more cards? Gold and gems are the two hard ones to look at, so I'll do the rest first and hope that makes it easier to look at those.

XP - There's a cap on how much XP you can earn and not really anything to spend it on. You can get at most 750 XP per day (first 10 wins and the daily quest) and every 1000 XP gets you a level on their mastery system. You do also get 3750 XP per week (first 15 wins) for a max of 9000 XP per week. There are 110 levels with rewards and 112 days in the season for a max of 144k xp. So you do have some extra to spare, but you really need to go hard to get all the rewards. (For every 1000 XP above the cap you get an extra uncommon for whatever that is worth.) You need to spend gems to get access to all the rewards, with the following being available.

everyone - 42 packs
old players - 6 packs, 10 uncommons
3400 gems - 20 packs, 2000 gems, 10k gold, 10 mythics, 3 rares

wildcards - These are how you actually make specific cards that you want. There are wildcards for each of the different types of cards (common, uncommon, rare, mythic) and nothing to do with them except trade them for cards. You earn more wildcards in one of three ways. You can open them in packs as a replacement for an actual card (1:3 packs for commons, 1:5 for uncommon, 1:24 for rare/mythic). For every 6 packs you open you get an uncommon wildcard and a rare wildcard. (Every 5 rare wildcards in this way are mythic instead.) And you can get wildcards out of 'the vault' which turns extra commons and uncommons into 3 uncommon, 2 rare, and 1 mythic wildcard. It takes 1000 extra commons to get this reward, with extra uncommons counting as 3 commons. You get these extras by opening packs and doing drafts/sealed events.

packs - Opening a pack gets you 5 commons, 2 uncommons, and a rare. (1 in 8 rares will be mythics instead) Getting your 5th copy of a common or uncommon gets you vault progress. A 5th copy of a rare gets converted into a different rare from the same set. If you own every rare you get 20 gems. Same with mythics except you get 40 gems when full. You also get 1/6th of an uncommon wildcard and 4/30ths of a rare wildcard and 1/30th of a mythic wildcard each time you open a pack.

individual card rewards - Daily reward uncommons have a 9/80 chance to become a rare and a 1/80 chance to become a mythic. If you already have 4 of the assigned card you get vault progress, 20 gems, or 40 gems.

gold - 1000 gold gets you a pack

gems - 200 gems gets you a pack

Ranked draft events cost 750 gems or 5000 gold. You draft 3 packs and keep the cards, with 5+ copies replaced by vault progress or 20/40 gems. You do not get wildcards from the packs. You do not get wildcard progress for opening packs. You do get 14 cards per pack instead of 8. You also earn prizes based on how well you do in the event, which are packs and gems. If you have a 30% win rate then you rate to get 1.23 packs and 153 gems. A 50% win rate is worth 1.33 packs and 347 gems. A 70% win rate is worth 1.62 packs and 672 gems.

Traditional draft costs 1500 gems with the same cards drafted. It's best of 3 instead of best of 1, and the prizes are spikier. A 30% match win rate is worth 1.85 packs and 244 gems. 50% gets you 4.4 packs and 1350 gems. 70% match win rate is 6.9 packs and 2361 gems.

Sealed aren't up right now, but the info for the last sealed shows it was 2000 gems to join and you got the cards from 6 packs. You don't get to rare draft them. You always won 3 packs and then scaling gems based on your wins. 30%-524, 50%-1002, 70%-1672

Standard events cost 95 gems or 500 gold. The rewards are gold and three uncommon cards. The cards have upgrade chances based on how many wins you get. The gold return is 230 for 30%, 410 for 50%, and 715 for 70%. As for card upgrades, you have a 7% chance of an upgrade with 4 or fewer wins,. A guaranteed upgrade (and 6% more) at 5 wins. And another guaranteed upgrade (and 5% more) at 6 or 7 wins.

Traditional standard is 190 gems or 1000 gold, with a best of 3 format. Prizes again are 3 uncommons and some gold. Gold returns are again spikier, with only 415 for 30%, 1470 for 50%, and 2396 for 70%. The card upgrades are better here, but not twice as good so you definitely lose out here compared to the best of 1 event.

Ok, so, let's assume I'll hit the 70% win rates for everything except ranked draft, where 50% is more reasonable. (The problem with ranked draft is it pairs by rating, not by record in the draft, so I expect to only play good players the whole time. The other events aren't on the ranking system so I expect to get a decent mix of everyone interested in playing events. Constructed I'd need an actual good deck to hit that, unfortunately, but let's pretend we get one.) For building a collection the main thing that matters is rares. Rare wildcards, in fact, since there are a lot of junk rares out there, but any rare is decent. For now I'm going to assert I get enough mythics, uncommons, and commons. I'm going to count a random rare as half as good as a rare wild card. Well, mythic wildcards I'll count as rare wildcards even though they're worse because if they're worth nothing I won't have enough and it makes things a little easier for my spreadsheet.

An opened pack is worth .7385 rares. Constructed uncommon rewards are actually worth .64 rares if you win 70% of the time. The best of 3 constructed uncommons are worth 1.16 rares. A limited pack is worth .557 rares, but you do get some selection in a draft and can choose to rare draft the shock lands and not the garbage rares. And if you do end up with most of the rares you can get small gem rebates.

So a ranked draft is 5000 gold or 750 gems for 2.65 rares and 347 gems. This is a better use of gems than buying packs (you spend the gems that would buy 2 packs to get 2.65 rares). It's actually a bad use of gold compared to buying packs.

A traditional draft is 1500 gems for 6.77 rares and 2361 gems. If you could actually win 70% of your matches this would be an absurd rate of return. I feel like this can't actually be a fair potential win rate. Looking at a 50% win rate is 4.94 rares and 1350 gems which is still really, really good.

Sealed is 5.56 rares and 1672 of your 2000 gems back. Or 1002 of your 2000 gems back at 50% win rate. So just worse than the traditional draft, but also probably an easier place to get a high win rate.

Standard events are both essentially infinite ways to grind up small amounts of value. They don't actually make wildcards is the big problem, so while you can still use the random rares they both seem good. They do also let you build up gold to spend on something else. That something else could be packs, but probably it should be drafts so that you can launder the gold into gems in order to do more drafts.

It's really not much clearer now than it was before, but I think what I've learned going forward is that sealed events are actually a pretty good deal. They only show up right after a set is released so saving up some gems to do some for the next set makes sense. Beyond that if you have a good constructed deck you should grind cards in constructed tournaments. Drafting should be used to convert gold into gems, and then the gems should be spent drafting or sealeding.

New cards likely get banned tomorrow, so it seems like I need to see what happens there, then build a new deck and play tournaments for uncommon upgrades. I should also spend most of my uncommon wild cards to make it more likely I get vault progress out of drafting.

Saturday, November 24, 2018

Blood Bowl: Fireball The Orcs!

I played a game of Blood Bowl yesterday as Lizards against an Orc team that set itself up on the opening kick-off with 8 orcs in a little box in the backfield. Getting 8 coin flips to get a mighty blow hit on an orc seemed too good to pass up, so I threw my wizard at him. It ended up knocking 3 guys over and removing none of them, and ultimately the game ended up as a draw where he barely scored and I likely win the game 87% of the time if I'd kept the wizard for the lightning bolt on the ball carrier. It got me thinking about if I got played on the kickoff... Did I get unlucky to not hurt anyone or are orcs just tough enough that it was strong play from him to bait me into using the fireball.

I decided to do a spreadsheet to figure out the numbers and then discovered Excel just straight up has a function that does it for me. Hello BINOMDIST, thank you very much. Here are the odds against any number of 9AV targets (columns) of getting any number of KO+ removals (rows).



So I was 55% to remove 0 orcs when I shot a fireball at 8 of them. Maybe this play would make sense if I was desperate to get lucky, but I didn't need that at all. I'm convinced the attrition fireball was a mistake, and I definitely got baited.

On the other hand, if 7AV elves were to make the same offer we'd get the following:



Now we're 75% to get rid of at least one and have pretty good chances of removing multiples. They're also better at recovering from the lightning bolt, so if an elf team were to give me that opportunity I would jump all over it.

Saturday, August 11, 2018

September Humble Bundle

The September Humble Monthly prepurchase games have been announced and they've added a new twist in this month. I'm not really sure why exactly, but they're giving you the choice between three games or a single game that was in a previous Humble Monthly. I'm guessing maybe the distributors for Rise of the Tomb Raider wanted to be featured again but Humble really didn't want to give away the same game to someone who's been a subscriber of their system forever. At any rate, you can get that single game ($70) or you can get Sniper Elite 4 ($67, but currently on sale on Steam for $10), Tales of Berseria ($65), and Staxel ($22). I've decided I intend to stream Humble Monthly games on the weekend going forward to shake things up from always playing Blood Bowl and decided to go with the three game option primarily because my sister keeps trying to get me to play a Tales game.

The three games are a sniper shooter game, a jRPG, and a farming simulator. A good variety that probably hits a lot of games with at least one of the games and seems like it should really hit me with two of them. I remain bad at aiming but jRPGs are ostensibly my favourite game genre and I really like Stardew Valley. I say ostensibly because I was thinking about it earlier and haven't actually played a new jRPG since probably Final Fantasy XIII in 2009 and I didn't even finish that one. So, well, I have high hopes but maybe my mental image of the genres I like doesn't line up with who I am today? Hard to say.

I'm going to stream Tales of Berseria tomorrow and see how it goes!

Tuesday, July 31, 2018

World of Warcraft: Expansion Additions

There's a new World of Warcraft expansion coming out in two weeks and I thought I'd take a look and see what stuff they're adding in this time around. This was probably a mistake because it turns out it sounds pretty sweet and now I'm probably going to end up wanting to play it again. They're adding two new game modes that sound pretty fun and they added one in the last expansion that I didn't really play that they're continuing on that really seemed up my alley. Thematically it sounds like this expansion is trying to return to making the Alliance and the Horde be fighting each other again instead of teaming up to fight the big bad. I don't much care about this one way or the other, but it does mean the mechanics are more PvP focused.

The first new mode is a Warcraft II mode to take over control of a zone from the other faction. Oddly it isn't PvP at all; it seems they actively want the zone to flip back and forth and if it was PvP then on most servers it wouldn't flip at all. Instead it's like a 20 person PvE fight where you build resource gathering buildings and barracks and stuff and then lead dudes into battle. It sounds neat, but probably not something you're going to want to do a ton of. You could just go play WCIII (which got a patch recently to revive that game) instead. But it should be fun.

The second new mode is a 3v3 race to score the most points by doing WoW type things. Kill monsters, do quests, open treasures, etc. The map is covered by a fog of war so you need to explore around and adapt your plan based on what actually spawned in this particular instance. There are a bunch of different islands to create ground rules but the location of things is procedurally generated so you need to adapt on the fly. Blizzard claims to have designed good bots to play the other 3 person team for the easier difficulty levels but the hardest mode requires you to have a premade group of 3 and pits you against another 3 person team. Spread out and get treasures? Gank squad to kill the other team? Someone with stealth and a lot of CC? Healers? Lots of options and it should be fun. Small scale skirmishes are the sort of PvP I can get behind, so hopefully this mode is done well.

The last mode, which existed already, is mythic+ dungeons. It's a system that's been evolving for a while, and has stolen some things from Path of Exile, and seems to be in a pretty good spot now. The basic idea is you do a hard 5 man dungeon with a time limit. Finish under the time limit and you get an item to let you run another harder 5 man dungeon. Beat that one in time and you get an even harder one. As you get harder and harder ones you start getting affixes on the dungeons that change the way things work. More enemy damage, less healing, weird dots... All sorts of options, and they're adding new ones. The loot is good and I think people pushing the really high mythic+s is a fairly competitive thing so I'd like to get in on the ground floor for it this expansion and see what I can do. It takes 4 other people and I don't know people who play anymore, but I imagine my old guild will at least have a few people interested at the start. I'm pretty sure Death Knights rate to be really good tanks for mythic dungeons at any rate, so I have a niche to fill! And who knows, maybe streaming mythic dungeons is a thing people are interested in watching?

Saturday, July 07, 2018

August Humble Monthly Bundle

Humble Monthly is a subscription service run by the Humble Bundle people where they sell you a fixed bunch of games for a low price. The bundles tend to be 'worth' about $200 and it costs $12 per month but they are using a non-discounted MSRP to determine that value and you don't get to pick the games you get so chances are you're going to get some games you're simply not interested in due to genres or whatnot.

The trick they use to get people to sign up is they announce some of the key games for the next month shortly into the current month. Typically this will be three games, but when they get a particularly big game they have it as the only headliner. Civilization VI and Destiny 2 are examples of those sorts of games from previous months.

My brother got me a year long subscription for Christmas last year and it's been pretty sweet thus far. I spend too much time playing a lot of individual games (like Blood Bowl in recent times) to play all of them but I have played some of them, that I never would have played on my own, which were great. Quantum Break was a movie/FPS hybrid that I would have never considered getting on my own but it was very cool.

I have recently signed up to be a Humble Partner. Basically this means I advertise for them and if people buy things through my link I get a cut. I did a little bit of this for the game Cultist Simulator which I was given a copy of to stream and generate hype for them prior to release, but I could and should do more.

I think what I should start doing is streaming the preview games for the Humble Monthly with my chat bot periodically spamming the link. Humble really wants their partners to sell the Monthly bundle in that I actually get most of the payment for someone who signs up for the first time. It will also give me a little structure for playing different games, which is always a good thing for me.

The advance games for August were just announced and they are A Hat in Time, The Escapists 2, and Conan Exiles. A cute platformer, a cool looking prison break simulator, and some sort of open world crafting game. The first two were both on my wishlist from browsing through various Steam sales, the third I'd marked uninterested. But it's one of the top sellers of the year, and has a $50 base price, so I'm going to give it a try. Letting people know they can get it (and some other stuff) for $12 could be a useful thing!

Sunday, August 27, 2017

Path of Exile: Ultimate Party Buffs?

Path of Exile is a game that can be played in a group of up to 6 players and has a massive number of different options for how one character can do things to help the other characters in their party. Group buffs, enemy debuffs, specialized roles... The world is your grouping oyster!

And yet right now I feel like I get very little benefit from playing with my friends right now. We've all built characters in order to be able to successfully play alone which means they're all self sufficient. They haven't really put any effort into buffing a group (since it tends to be more expensive skill point wise to buff a group than to just find something to buff yourself) and the efforts put into debuffing the enemies will often come into conflict with each other (curses overwriting each other, elemental equilibriuming away an ally's damage, knockbacks, non-magic finders killing things).

So we started mulling over the idea of creating characters actively designed to work well together. If a character is built knowing they're going to get buffs from friends then they can make tradeoffs as they build their character. For example, if I know someone is going to make some really good curses to put on the enemy then I know I don't have to put any skill points into making my curses better. I don't have to waste skill gems on a curse setup. If I know someone is going to focus on doing a lot of damage then I don't have to worry about taking any damage nodes in the tree at all! Everyone can focus on being tough and doing a couple of good things for the group and actually play together without feeling bad about it.

I'm a little hyped, but now I want to make sure I pick up all the ways to help that I can! So I wanted to go through the list of skills and uniques and stuff to see what the different ways of helping out would be. It's important to not try to take too many helping things and end up squishy but it's entirely possible there are some fairly trivial buffs that people could pick up as long as they know about them. So I want to have a list of broad 'roles' and then a list of extra add-ons that could be picked up.

Roles:
- AoE damage dealer
- single target damage dealer (for bosses)
- aurabot(s)
- multi-curser
- magic finder
- super tank

Specifics:
- elemental equilibrium (would need to coordinate to use the right element)
- conduit to share power, frenzy, and/or endurance charges
- consecrated ground
- no stun after blocking from Guardian
- speed buff after warcry from Guardian
- armour, energy shield, physical reduction, life regen from Guardian auras
- attack/cast speed, damage, resists from Necro auras
- maim
- blind
- culling strike (@20% from Slayer)
- taunts (double reduction from Champion)
- damage and move speed from Champion
- bleeds buff Bloodlust support gem
- knockback (reverse with unique gloves?)
- stuns/freezes
- hinder
- warcries?
- corpse consumption?
- vaal auras
- frost bomb (cold resist reduction)
- spirited response jewel (rallying cry -> mana regen for party)
- the vigil jewel (vigilant strike fortifies the party)
- dying breath weapon (curse, aura, damage amp) {animate guardian?}
- shaper's seed amulet (health/mana regen aura)
- leer cast hat (damage amp) {animate guardian?}


Enemy damage increases
- Inquisitor (nearby 16% elemental damage)
- Champion (taunted 20% damage)
- intimidate (Champion)
- Warchief (near totem 16% phys/fire damage)
- maim support gem (10-14% physical damage)
- shocks!
- wither (140% chaos damage)

Tuesday, August 22, 2017

Path of Exile: Perma-Freeze Bosses

I have been playing a lot of Path of Exile lately. My current character is running the freezing pulse skill gem because they added a unique jewel that buffs it and I had fun with it years ago. The goal of the build is to do a bunch of cold damage and inflict the 'freeze' debuff on all of the enemies. This means they can't take actions, which means they can't do damage, which means I get to be invincible! Mwahaha!

The theory turns out to be better than the practice, however. I can absolutely crush trash monsters but I've been having trouble with bosses. I've been trying to straddle a line of being able to still deal with bosses that aren't frozen so I've ended up with a mishmash of a build where I have some survivability and some freezability but not really enough of either. I will be getting more of both with more levels (both in terms of getting more passive points in the tree and in terms of leveling up my skill gems) but I'm starting to feel like I need to really go all in on one or the other. Running a bunch of curses and instant leech and the like is fine if enemies aren't frozen but if they are then it's not really doing anything for me and maybe I'll be better off without 4 item slots dedicated to doing all my cursing.

Alternatively maybe I'm better off doing all the cursing and leeching without worrying about freezing at all. If I'm building up to being able to leech tank bosses I can do the same for all the trash too and I'd be better off without all the freeze related points in the tree. At that point I may even be better off just using a different spell entirely!

So I think the way to proceed is to work out what I could do to go all in on freezing. See if I think that'd be good enough to freeze some bosses. If so, great! If not, well, probably start a new dude.


Ok, so, freezing... What are the specific mechanics behind freezing enemies? For our purposes only cold damage can freeze and a hit of cold damage will freeze if two conditions are met:

- The hit is a crit OR an inate 'chance to freeze' roll succeeds.
- The hit does enough damage to freeze the enemy for .3 seconds, which at base means doing 5% of the enemy's max health in cold damage.

The first condition is relatively easy to meet. In my current setup I have a 19% chance to crit, 20% chance to freeze, 20% chance to freeze a chilled enemy, 13% chance to freeze a frostbiten enemy, and somewhere from 0-25% chance to freeze based on distance from my character. So while not every hit will meet this check a significant number of them will. If bosses were getting frozen some but not all of the time I would worry more about working on this condition, but they aren't, so I won't.

The second condition is the one I'm getting stuck on. Having to hit a boss for 5% of their health in order to freeze them actually seems a little silly. If I only need to hit them 20 times in order to kill them I don't think I care that they're not frozen. I can kite and/or use flasks for the 7 seconds it'll take to burn them down! Fortunately damage isn't the only way to work on this condition. You can also work to extend the duration of your freezes so that smaller amounts of damage are good enough to hit the .3 second threshold. If you had 100% freeze duration you'd only need to hit them for 2.5% of their max health. 400% freeze duration would mean you only need to hit them for 1% of their max health. I'm not quite sure that would be good enough, and I really doubt I could get that much freeze duration anyway, but I might as well look into it.

Currently I have 60% increased freeze duration, which is every freeze duration node in the tree. There is one 10% ailment duration node which probably also applies, but I don't have. Other potential sources of freeze duration would be:

- jewels which can add up to 16% freeze duration each
- 20%+ freeze duration from quality on a frostbite gem (and maybe more if my curse nodes impact the quality boost)
- elemental proliferation support gem would be up to ~29% freeze duration (and 20% chance to freeze)
- The Halcyon unique amulet has 30% freeze duration (and 10% chance to freeze, and potentially 80% cold damage)
- Southbound unique gloves have 25% freeze duration
- Rashkaldor's Patience unique amulet had 20% elemental ailment duration

And that is all. The last two unique items don't feel very good while The Halcyon runs over 150 chaos to buy. I do have that much, and I suspect I can flip the amulet back if I don't like it, so possibly that is what I need to do if I'm going forward. I can probably grab jewel sockets to get 3 of those jewels, and I can easily get a quality frostbite gem. I currently need my amulet for life and resists, so switching to The Halcyon hurts my health to the point where I probably need to give up on the Kitava's Thirst 4-curse plan. I also _might_ be able to stretch my tree to get that last 10% ailment node at the assassin start location.

All told I can probably get my freeze duration up to 164% without elemental proliferation and 203% with it. That's almost twice what I'm currently sitting at (303 compared to 160) which almost certainly gets me a little closer to freezing some bosses. That's enough that it's probably worth working on...

On top of that, I can also work on doing more cold damage. With all the freeze duration I can manage I only need to do enough damage to hit for 1.7% of the enemy's max health. I'm still not sure if that is good enough... That's 60 hits to kill them, and I attack almost 3 times per second, so it's probably trivializing a fight that would be lasting 25 seconds otherwise? That's actually a lot of dodging... I don't think I can facetank a lot of these fights for 25 seconds with my current setup, so it is probably worth trying. But let's max out our damage too...

Initially my thinking was that if I'm doing a perma-freeze plan I don't want to be relying on the swings involved in crits. So I've taken the node that makes me do 40% more damage if I've crit in the last 8 seconds. This is way better for consistent damage and it saves me taking crit nodes but it's entirely possible I could be freezing some of the time with crits instead of the current none of the time. At this point I'm really committed to this line I think (I'd want a different bandit reward and probably an entirely different class for critting) but maybe I want to create a different character to do crit freezing pulse...

Anyway, with crits totally abstracted away there are really 4 ways to boost the damage of my spell. I can increase the base cold damage of the spell with flat damage (including leveling the gem itself), I can add to the 'increased damage' multiplier in a variety of ways, I can get extra 'more damage' multipliers, and I can lower the enemy cold resistance.

There are so many ways to change all of those things so I went and built a spreadsheet. It was a little surprising to me in some ways, in particular that a level 4 empower gem isn't really worth using. I suspect that's because of how much flat damage I'm adding to the spell via herald of ice? It means there isn't a red gem worth using which makes me sad... I'm going to have to spend a bunch of chromatics to change my 6 link away from BBGGGR into BBBGGG. (I only have a level 2 empower gem anyway...)

At any rate, here's the support gem ordering for different criteria:

30 cold resist (standard boss), freeze duration - controlled destruction > hypothermia > ice bite > slower projectiles > empower (4) > cold penetration > unbounded ailments > added cold damage > pierce > efficacy > empower (3) > elemental proliferation > iron will > empower (2) > faster projectiles > faster casting > spell echo

30 cold resist (standard boss), DPS - spell echo > controlled destruction > ice bite > hypothermia > faster casting > slower projectiles > empower (4) > cold penetration > added cold damage > pierce > efficacy > empower (3) > iron will > faster projectiles > empower (2) > elemental proliferation = unbounded ailments

85 cold resist (cold boss), freeze duration - controlled destruction > hypothermia > cold penetration > ice bite > slower projectiles > empower (4) > unbounded ailments > added cold damage > pierce > efficacy > empower (3) > elemental proliferation > iron will > empower (2) > faster projectiles > faster casting > spell echo

85 cold resist (cold boss), DPS - spell echo > controlled destruction > ice bite > hypothermia > cold penetration > faster casting > slower projectiles > empower (4) > added cold damage > pierce > efficacy > empower (3) > iron will > faster projectiles > empower (2) > elemental proliferation = unbounded ailments


Basically it seems I have enough spell penetration that unless a boss is overcapped it barely isn't worth using the cold pen gem. This is only true if frost bomb is up, and that's not a given.

The spreadsheet wants me to use slower projectiles, but the spreadsheet don't understand that freezing pulse loses damage the longer it travels so slower projectiles has an invisible less damage multiplier in it. This depends on how close I get to the enemies, and it really hurts my trash clear because the size of my attack is based on how fast it travels. Similarly the faster projectiles damage is better than it seems here (but it still seems quite bad).

Spell echo is the best for doing damage, but it's actively bad at freezing a boss. This means I may want to carry one around to swap in on bosses I can't freeze? I suspect I'd rather have faster casting instead for mobility reasons if I can't freeze the boss? Hard to say.

Ice bite looks good, but that's because it's a way to get frenzy charges. If I can get the charges up then it is worth using. If I can't then it's trash. You get them for killing frozen enemies, which will happen all the time on trash and on boss fights with tons of adds but never on a standalone boss.

I think what I'm ultimately going to want to run is controlled destruction, hypothermia, ice bite, cold penetration, and faster casting. It feels like a good setup for trash with a reasonable shot at freezing most bosses. I can swap in a slower projectiles for the ice bite when I can't keep up the frenzy charges and just get right up in the single boss' face. Either way I do want my colours to be BBBGGG so I should get on that.

Friday, June 30, 2017

Twitch Subscriptions for Affiliates

On Wednesday of this week Twitch rolled out the third stage of their program for getting more income streams for small streamers. The first was the bit system, the second was the game selling system (but the games I play aren't part of that scheme so I haven't really looked into it), and the third is opening up the subscription service, which is arguably the most important option of them all.

Subscriptions allow someone to pay $5, $10, or $25 each month in order to get a few bonuses for chatting on Twitch. Twitch takes a pretty big chunk of that (I had one person sub to me yesterday and my dashboard says I made $1.74, so I guess Twitch takes $3.25, but it could be that Twitch Prime subs are worth less? I don't know!) but it's still a way to set up ostensibly recurring income. Donations (direct through Paypal or via the bit system) are super nice and all but they're very erratic. Someone who is looking to be able to pay rent off of streaming is going to appreciate the consistency of subscription revenue.

There's also the weirdness around Amazon Prime which automatically comes with Twitch Prime which lets the person subscribe for free to one streamer. So it's like your Amazon Prime subscription actually comes with a $1.74 rebate, you just have to have the rebate mailed to someone else. Someone else like me!

As for what the person gets, they get a few things that have never seemed terribly interesting to me but there are a lot of people who get really into them. The first is you get an icon beside your name when chatting in that stream's chat that shows you're a subscriber. Partners can customize that icon and can have different ones for different numbers of consecutive months subscribed; affiliates just get a default star. But it's a way to show in chat that you're someone ponying up to support the streamer, and some people like that, so that's cool. The other thing they get is access to a chat emoticon designed explicitly for the stream. They can use it in any stream chat anywhere on Twitch so if you make a really cool emote then it becomes a form of advertising for your channel. (I've never watched AdmiralBahroo stream, for example, but he has some really sweet emotes that people use in my chat all the time.)

Partners get lots of emotes based on how many subscribers they have. Affiliates get exactly 3. One for each tier of subscription. (But if you have enough subscribers to be eligible for tons of emotes I'm pretty sure you can become a partner too so it's not that big a restriction.)

I didn't do a good job of planning ahead (they did say this was coming in the near future) so I don't even have any emote ideas let alone anything made. You can make changes, though it'll take a couple weeks to get things in or changed, so I really need to keep in mind that it doesn't need to be perfect. Getting something decent implemented quickly is more important than getting the optimal thing done in a year.

So I need to come up with some ideas, and then I need to get 28x28, 56x56, and 112x112 PNGs for those ideas.