Sunday, September 28, 2014

Competition

Diablo 3 is trying to be a competitive game.  Blizzard is making a go at setting up an environment where only the extremes of skill and effort will be successful.  The system they have gone for is Greater Rifts, where rifts are timed and become more and more difficult with each level.  The increments are multiplicative so no matter how good you are eventually the enemies will one shot you and you will be unable to beat them within the fifteen minutes alloted.  There has been quite the outcry recently because recently Blizzard changed the way people get access to high level GRifts.  In short, they made it much harder to jump immediately into the highest level GRifts and forced people to beat lower level ones and grind their way up.  This would have been fine except that people can't actually beat any of the GRifts near the top at all and now it is nearly impossible to advance.

The fundamental issue here is that GRift difficulty is hugely variable.  The way people were getting top scores is not be being better, but rather by dashing into GRifts to see if the monster composition was favourable and then usually dashing back out again in ten seconds to try again.  Even with a favourable composition the only way to get a really top time was to get lucky and get a conduit pylon.  People were literally unable to beat GRifts above say 38 consistently but because they could get into GRift 42 and just keep on trying until they got a super easy one they could win - rarely.  Now that you actually have to beat 38 and then 40 to get into 42 fishing for an easy 42 is nearly impossible.  This is not a good competitive environment, particularly when new players find the scores that are posted are effectively unbeatable due to mid season changes.

I am not entirely convinced that competitive GRifts are actually the thing to do but if Blizzard wants to do them what they need to go for is preset levels.  If every single level 42 looks identical then people can truly practice and theorycraft the best way to beat it and the game won't come down to seeing who gets the lucky 1 in 1000.  An actual competitive way to build this would let players simply select any level from 1 to 42 as long as they have beaten 41.  Go nuts and keep slamming your head into that wall if you want, and let those who build properly and practice diligently succeed.  With preset GRifts it is completely okay to have some levels be easier than expected - as long as 38 is super trivial for everyone it really doesn't hurt things overall.

A slightly less extreme way to accomplish the same end is to remove all Pylons from GRifts and group monsters in a less random way.  As long as there was always a mix of easy monsters (Zombies!) and brutal monsters (Anarchs!) then it should be fine, or at least fine enough.  Of course this further tinkering would completely shut down any new attempts to beat current records and season 1 would be effectively over, at least for those wanting to be anywhere near the top.  This would preserve some randomness in the system which would be nice for those like myself who aren't anywhere near the edge but still like bashing our way through GRifts for fun.

Something Blizzard has to learn to deal with though is how much precision and control is required to have a competitive environment.  You can't just wander in and make changes to the system and not completely upset the very delicate balance at the absolute bleeding edge.  If they want to patch in new abilities and alter systems on a semi regular basis there simply cannot be a really competitive environment.  Their current strategy is very much like the Tour De France deciding that bikes may not be made of a carbon and announcing that after three days of racing.  Nobody would take the competition seriously, and rightly so.

What I actually expect to see is a set of changes that flattens the randomness of GRifts very seriously but I don't think that will actually come in until the next season.  Both monster mixes and pylons will be addressed for the start of season 2, and not before.  I don't think Blizzard actually intended to bork the season up with the latest batch of changes as their intended effect of making things less punishing for melee actually worked out just fine.  What I do think is that they didn't anticipate the serious change to the top players their seemingly minor alteration made and they don't want to wreck things up further until a full reset is happening anyway.  I also suspect when this happens they will alter the scaling so that old records will fall easily to give people something to shoot for.  They would likely have to change it so that 40 then is roughly as hard as 35 is now to give people a reasonable shot at beating the old records, though it might even have to be more extreme than that.

Wednesday, September 24, 2014

Details!

I have been slugging away at the half formed idea I had for a camping themed cooperative game.  The basic mechanics are down and seem very good but as it is a game that requires a whole bunch of cards with wildly different effects on them it will take a while to actually flesh out.  This process is illustrating to me just how little the basic idea matters in game design.  People who have an idea for a game often get completely wrapped up in secrecy and worrying about somebody stealing it when that is a laughable proposition in nearly all cases.

The hard part about building a game is very rarely a simple, raw mechanic.  There is so much to do to polish it, to create all of the working bits, and to get everything to feel right that even if you tried to get people to take your game ideas they wouldn't bother.  That is all not to mention the real challenge, which is to build prototypes and get the game published and convince people to buy it.  Heck, even if you took a complete set of game rules and tried to get somebody else to go through the mess of getting that published it would be hard to find takers.

I am very excited about the build.  I look forward to seeing people trying to deal with bear attacks in the middle of the night and yelling at one another about who gets to control the marshmallow stick to produce delicious smores.  Obviously in a coop game everyone could play along nicely and such, but a little bit of conflict over what to do about the raccoons seems like it could be pretty amusing.  Kill them!  Bribe them with shiny things!  Just wait until they go away!  Feed them to the bears?  And when the Park Rangers show up, definitely just bribe them with beer.

The pictures for the cards are going to need to be cute.  I am thinking of a vibe similar to Munchkin with cards mocking the terrible things that happen on camping trips being portrayed in silly cartoon form.  I have no talent for drawing such things myself and art is getting way, way ahead of myself but I think having a vision for where I want to go at the end is a useful thing.

Hopefully over the next few weeks I can get documents up here so everyone can see what I am working on.  If you have suggestions for a name for a silly camping themed coop game, get them in now!

Friday, September 19, 2014

Do not believe

I have been having an entertaining time reading about review scores for games lately and comparing them to what actually came out the other end.  Tobold recently complained that previews are pretty much useless as they are just marketing from the game companies and that there are far more of them then there are reviews.  It is definitely true that the hype surrounding a game is an important predictor of sales but that promises contained in previews are pretty near worthless in terms of figuring out whether or not a game is worth buying.  I am reminded of the Bears, Bears, Bears speech promoting Warhammer Online that people got so excited about but which turned out to be complete rubbish (though I called it out as such at the time, go me!)

Certainly Tobold is right that previews are worthless in terms of promises, though you can likely glean some information from watching actual gameplay footage.  You just have to completely ignore anything anyone says and only look at what you see when someone is doing things in the game itself and you just can't get much out of that.  Sadly I don't think that reviews after the game is out are actually much better.  Unfortunately there is way too much of people who played a game for 100 hours in the first two weeks complaining that the game is the worst thing they have ever played and not worth the $50 they paid for it.  These are the same sorts of people who sit on forums for a game they have logged thousands of hours in and whine endlessly about how the game is garbage.

The real problem here is expectations.  When an indie game comes out with no expectations and little fanfare people who don't like it stop playing but they rarely go ballistic and post 1 star reviews.  Just another mediocre game, right?  But when it is a much anticipated big title things are very different and people invest huge amounts of their persona in being fans of that particular game.  Any disappointment in how the game turns out becomes a very personal attack against them and their judgement and they lash out.  It is much like interpersonal interactions in that if a stranger utters some nasty slur it is going to irritate me but it isn't going to wreck my day.  A friend who I have spent months telling my friends is a super sort of person who does the same thing is going to be a much bigger deal and my reaction will be much more vitriolic.

The only way to get useful reviews about a game is to only accept information from people who you trust to have similar tastes and judgement to your own.  Any sort of large scale system is going to be completely corrupted by marketing, whining from bittervets, and false hope.  Even if game quality were some sort of objectively measureable quality (Hint:  It isn't) then actually finding that information would still be a mess.  It seems like with the giant information collection device that the internet is that we could actually figure this stuff out but the systems that are in place right now are pretty lame at doing so.

Tuesday, September 16, 2014

Coop mode

Over the past little while I have been playing a fair number of coop board games, specifically Hanabi and Sentinels of the Multiverse.  I think these are two examples of very good coop board games because they avoid the two major issues coop games have:  One player playing the entire table or the game being very simple and random.  What I have been thinking is that my next board game creation endeavour should be to create a coop game that is superior to Sentinels of the Multiverse.  Hanabi is so simple and excellent that I don't think I can just sit down and build a better version.  Sentinels on the other hand does some things right but it also does a ton of things terribly wrong and I am confident I can build something similar but avoid many of the pitfalls of that game.

One of the things I wish Sentinels did better is limit game length.  While I have played quick games I find that things often begin to drag out in a fairly extreme way.  In the worst case you end up with a couple of players dead and the other players manage to stabilize and take hours to take down the villain a bit at a time.  That can be very frustrating especially from the perspective of a player who has been eliminated and thus has little to contribute.  Initially I liked the idea that a dead player could still do a single small thing from a short list but how it ends up working out is that the player does the exact same single thing each turn.  It ends up locking them into the game without actually letting them play properly, and though it tries to solve player elimination I think it solves that problem badly.  Keeping game length limited is a great idea both because it lets you actually play it when you don't have half a day to spare (are you listening Diplomacy?) and because people like to know what they are committing to when they sit down.

My idea is to make a game themed around going camping, which is certainly due to the challenges I experienced in my camping trip this weekend.  Of course the camping trip the players will experience is going to be much more complicated and tragic than the one I experienced myself.  I expect many bear attacks, hordes of mosquitoes, fights over the marshmallows, and tents that catch on fire.  The basic mechanic I envision is one where each player first either plays a card from their hand or chooses a basic role each turn, and then chooses a terrible event from a short list.  Something along the lines of "Okay, so I play a Marshmallow Stick which makes Marshmallows give +3 food.  Then I guess I have to choose between Raccoon Attack and Downpour."  I like the idea of players getting to choose the terrible things that happen because then they can plan around what is coming up and figure out what sorts of things they can handle at the moment and what simply has to be put off.

In theory the deck of terrible things can be used to limit game length while also providing the challenges the players have to overcome.  The theme can essentially be that the players have to deal with all or at least some very large subset of the problems that exist before the camping trip can be declared over and the score tallied.  I am not entirely certain if it should be possible for the players to lose entirely but I think Hanabi actually works very well in that regard because it usually goes right to the end but there is a chance of losing if you really bone it up.  I am leaning that way at the moment, but until I actually build things more concretely I won't know for sure.  These things have a way of getting away from me and following their own path.

Wednesday, September 10, 2014

The problem with proximity

I was talking earlier about issues with pets in Diablo 3.  They are the same issues other games face - pets either always live and are overpowered or always die and are useless.  People love pets though so I want to talk a bit about the real problems with pets and how we might go about fixing them.  The problem with pets is not that they do damage, or that they do too much damage.  Rather, it is that their toughness matters.  It is very easy to create a balanced class or ability based on a pet that beats up the monsters as you can simply balance it against whatever else you do to kill things.  The problem is that pets can completely invalidate the enemies' attacks and that wrecks everything.

One way to fix this is to have pets that are very short duration.  In this model pets are more like an extended nuke spell that run up to the enemies, bash for a bit, then vanish.  In theory this works just fine but in practice I think what players expect from a pet class is a bunch of pets that follow the character around.  If all the pets only appear when the player is spamming then a lot of the flavour is lost and flavour is really what we are after.  If we were after balance we would just shelve pets permanently because they are a neverending problem.

D3 is actually far worse in this regard that it has to be because pets are on long cooldowns.  If my Zombie Dogs required real resources to summon I could keep recasting them as they die and that would fix the all or nothing problem a lot.  I would have to choose between casting damage spells and resummoning pets and although my pets dying would suck at least I could continue to move forward.  With big cooldowns on pets though there is no middle ground.  Either the pets are immortal, or they die and I can't play and there is nothing in between.  I am not super familiar with Path of Exile but as I understand it that game did things a lot better because you can spend your time resummoning pets so you don't fall off the cliff quite so dramatically.

Really though if you want permanent pets I think the way to make them work better is to fix enemy aggro.  Right now in D3 and indeed most games of its ilk monsters operate purely on a proximity aggro system.  They happily stand and bash on the zombies and ignore the summoner in the background.  If the enemies ignored the pets and all went after the player then things would be much smoother since the player would have to be able to tank, kite, CC, or do whatever else to deal with the damage just like any other player would.  The pets could happily bash on the monsters and do the damage that they need to do without worrying much about their toughness.

This suggestion has the flavour issue that people really want a meatwall that makes them invincible.  I think though that the invincible meatwall desire intrinsically creates intractable problems.  In any case the meatwall is going to interfere with enemy movement and keep some of the enemies away from the player, it just won't be a complete defence.  With this sort of design pet toughness isn't much of an issue as long as they don't die to AOE effects but are still killable in extreme circumstances everything is fine.

In short, I think the best way to fix pets is to make them into weapons rather than armour.  You don't summon zombies to get punched, you summon them to do some punching!

Saturday, September 6, 2014

Falling off the cliff

Diablo 3 has just launched patch 2.1.  By and large the patch does a lot of good things but it is also illustrating some real problems that appear in nearly every RPG type game out there.  In particular there are always tremendous issues with pets and pet classes that constantly plague these sorts of games, mainly that either the enemies are capable of killing the pets in which case the pet class is nonfunctional or the enemies cannot kill the pets in which case the pet class is invincible.

Player toughness is not nearly so binary because players do things like crowd control enemies, position themselves so that only a few enemies can hit them, kite, use defensive buffs, etc.  Because of all the different tactics players can employ to trade off damage for survivability there is a wide range of monster damage that is workable.  Pets can't do any of those things.  They just stand there and get beat and either they are tough enough or they are not.

I am running into this issue with my Witch Doctor right now.  Generally speaking the changes in 2.1 were an outrageous buff to me - several bugs were fixed with the interaction between Fetish Sycophants and Fetish Army, pets occupy less space so they can swarm more easily, and Life On Hit was massively buffed which is insane because I was already stacking a lot of it.  I can safely say that my health and damage are both effectively doubled.  I can handle the highest base difficulty without much worry and I am collecting new gear and cash at a fantastic rate as all I need to do is follow around my swarm of evil and watch them maul the enemies.

However, when things go bad they go really bad indeed.  If the enemies can actually deliver enough damage to kill my dorks my army falls over very quickly and then I am left in a completely untenable position.  If I summon more dorks they die rapidly as they are outnumbered and building back up to a critical mass is extremely difficult.  I can potentially run back through the dungeon to hunt for weak enemies to farm to get my army back up to size but that isn't a very good strategy for a timed run, or indeed for any sort of efficiency.  Doing so also assumes I can survive a situation where I have to take hits in the face instead of relying on my meat shields.

It feels a bit like launch day of the original D3.  Pets were really excellent in the beginning, both tanking and beating down like champs.  As soon as the challenges ramped up a bit though they were utterly useless.  Right now pets are really good for the base game but as soon as I end up running through timed Greater Rifts which continually ramp up in challenge I am going to hit a pet wall where I simply can't keep them alive and no more progress can be made.  I haven't hit that wall yet but it is clear it is coming and there is nothing I can do about it.

I won't trash on the developers for this situation though as the model for pets that D3 uses can't avoid this problem.  When pets are expected to be very durable and not die this is inevitable.  To get around the pet problem what we need is an entirely different vision for pets.  Pets need to be something you constantly summon because they are primarily an offensive force and they only last a short time.  They should not be so durable and numerous that their boss can ignore all enemies most of the time.  A zombie that runs over to the enemies, vomits acid on them for 3 seconds, and then disintegrates would be a fine example.

This situation actually extends to melee classes too.  At some point in progression it will definitely be the case that melee classes cannot keep up as nothing can survive in melee range and only range classes played like a bullet hell game will be capable of playing.


Wednesday, September 3, 2014

Are gamers really that bad?

Gamer culture on the internet gets a pretty bad rap.  Ziggyny talked about his own personal experience with it recently and concludes that the crap he had to deal with was brutal so it must be truly a mess for women in similar situations.  I definitely agree that there are a lot of crappy places out there but I know for sure that there are also some really decent places but that maintaining that decency requires a ton of time and effort.  I was always impressed with how decent Elitist Jerks was because they have punitive moderation policies in place and the discussions are quite reasonable as the trolls get banhammered very quickly.  I even got a few infractions there for things like quoting multiple people in a single post - the worst of the internet sure doesn't survive under those conditions.  The great majority of the internet is a pit of despair though and we can't clean it all.

The thing I have been considering is the question of whether or not gamers are actually worse than other people when it comes to hurling abuse on one another.  After thinking about the many places and social situations I have been in my life I concluded that gamers truly aren't worse, but they do tend to exist in a place that enables bad behaviour.  When I worked at a gas station as a teenager the difference between how things were when a woman was in the shop and when they were not was staggering.  As soon as it was just the boys the abuse that was heaped on pretty much everyone was dialled up to eleven.

The same story holds true pretty much everywhere I have been.  Bad behaviour has little to do with the hobbies of people and everything to do with anonymity.  As soon as people think nobody who can hear them will call them on their shit they suddenly turn into ravening monsters.  The Greater Internet Fuckwad Theory was developed after watching gamers misbehave but it rightly doesn't specifically distinguish them as the problem.  Targetting gamers as the source of online harassment problems is like saying that camouflage clothes are the primary source of violent atrocities.  The problem is that people in camo clothes spend a lot of time in foreign countries while holding weapons; dressing them in fluorescent spandex would not prevent soldiers killing people.  The problem is the guns / anonymity, not the personal details of people involved.

However, given that gamers do end up being really abusive because they are constantly in spaces where they can get away with it all of us have a real opportunity to improve things.  It isn't easy to constantly push back against crappy behaviour as there is a seemingly endless supply of jerks but there is no other way.  No government regulation will step in to save us.  There is no greater ruling authority of gamers that will come to its senses and suddenly fix the internet.  The way things will improve is for people to consistently speak out against anonymous abuse and to support with our voices and our dollars those companies, moderators, and policies that aim to clean up our virtual spaces.