Tuesday, July 31, 2012

Gods and Spies

The expansion for Civ V has been out for awhile but what with a minor addiction to D3 I never picked it up. Since my interest in D3 has waned very dramatically recently I figured I would buy Gods and Kings and see what changes came to CiV.  I must say I am pleased with the changes they have brought in at first glance.  Many of the things that I fixed or tried to fix in my CiV mod have been implemented in G&K like reducing the power of Great Scientists and improving little used military units; I can't really take credit since many other mods did the same thing but I do like it when things I did end up in the game.

The addition of religion to the game works pretty well mechanically I think.  You can still play the game pretty much as you always have if you really want to but there are a lot of new options and styles available when you dive into religion.  The fact that I could play the way I used to but dip into religion carefully when it seemed useful is a great way to introduce a new mechanic; it means that players won't be overwhelmed with new choices and that the balance that has been achieved is easier to maintain.  When a new expansion adds completely new concepts to the core of the game it often ends up being either incredibly badly balanced or very confusing to current players or both.

Unfortunately with all of the really good changes there are still some things that stand out as not working well to my mind.  There are still buildings that are fantastically bad and which will never, ever be built by a human player playing optimally.  Having buildings that are mediocre but which have situational abilities is fine since we do want players to make meaningful decisions but I don't like it when you make a decision precisely once and then never revisit it regardless of what is going on in the game.  I am still going to build a mod and try to correct the flaws I see in the game but I suspect that it will be a much lighter touch this time as I make minor tweaks instead of major overhauls.

If you liked CiV but haven't picked up G&K then I would recommend you do so.  It gives a whole lot of new life to the game and makes it feel very much new and shiny again.  The complexity level has gone up a notch or two but I think that this is a good thing; at this point in CiV's life cycle it is mostly the hardcore players who are still going on it and they like a little extra meat on the bone.

Thursday, July 26, 2012


A few days ago I was introduced to a new game called Pigs.  It is a pretty simple party game full of big swings and opportunities for cursing but still a lot of fun.  It is vaguely similar to Can't Stop in that each turn you are building up points and you have to decide whether to bank those points and end your turn or keep rolling hoping to get even more points.  Instead of dice though you are rolling two plastic pigs onto a table and you get points based on how the pigs come to rest - on their sides, back, front, or snout.  It is fantastic because the equipment required is extremely minimal and you can fit the entire game into a pocket but it does suffer from extremely high levels of randomness.  You may or may not consider that last point a good thing!

I found it pretty amusing to see the different responses to the game from various players.  Most people decided to bank their points once they accumulated somewhere between 13 and 20 points but occasionally they got into a gambling mood and went for more; this was almost always disastrous.  Of course when somebody else is way ahead it makes sense to gamble more since you are almost certain to lose with cautious play but generally it seemed that the cautious players were more successful.  People generally thought the game was fun but nobody thought much beyond that.

My response was "I need to build a gigantic spreadsheet."

First I need to roll the pigs ten thousand times or so to get a read on how they roll on average.  Then I can do some really simple math to figure out the spreads on the various point totals and determine how many points you should stop at to maximize your point acquisition.  The more interesting part of the analysis is to figure out how that optimal number of points to bank changes based on the point totals of your opponents.  If we imagine that banking 19 points is right then it is reasonable to just bank 19 or more when everybody is at 0 but if one person is 5 points from victory then you should keep going until you either win the game or lose it all.  There are a very large number of scenarios to run of course so I would probably limit my analysis to a few simple ranges to try to figure out the best possible strategies.

After I sat there musing about analyzing the game and building a bunch of code to figure out my strategies everyone looked at me like I was a wee bit insane.  Their looks said "It is a silly game where you roll plastic pigs until somebody wins... you are building a spreadsheet WHY?"  Well, I am building a spreadsheet so I can just put my sheet down on the table and tell you all to roll pigs for me and make decisions based on the sheet.  This way I can go play other games and the Pigs players can just tell me if I won the game once they finish up.

I understand that there are people that don't feel the intense desire to build a spreadsheet whenever they play a silly party game.  They are crazy of course, but since they insist on spending time playing games instead of doing game analysis I should be accepting.  People are entitled to their idiosyncrasies.  

Thursday, July 19, 2012

Mass Effect endings: Round 2

Bioware recently put out new endings for Mass Effect 3.  It is pretty much universally agreed that the new endings are a big improvement over the default endings but of course the internet's opinion on how good the new endings are ranges from abysmal to amazing.  The structure that Bioware used for the end of the game was a strange one to my mind.  They chose to have the player talk their way through a ton of explanation and backstory and then offered four choices:  Destroy your enemies with massive collateral damage to friendlies, assume control of the enemy, turn every being in the galaxy into a synthetic/organic hybrid (guaranteeing peace for some reason) and Bite Me, I don't want to choose any of the above.

The reason I think it is bizarre is that there really wasn't any leadup to this final choice.  It wasn't foreshadowed that Shepard would have to alter the face of the galaxy completely with a single choice and your character's choices prior to this point had no bearing on the choice made at the end.  It isn't that there should have been more, better choices but rather that there shouldn't have been a choice at all.  The entire series led up to the moment where Shepard would hit the Big Red Button and destroy the Reaper menace so why not just do that?  It wouldn't require any sort of Deux Ex Machina ending nor introduce bizarre new characters but rather just finish things off in the way we always thought they would finish.  The fantastic part about doing it this way is that there could be a big variety of cutscenes afterwards that would display how Shepard's choices throughout the series played out in the long run.  It could show what happened with the Krogan (cured or not), the quarians/geth (destroyed or allied), and how the various individuals and races dealt with the end of the war.  

The trouble with the four choices at the end is that all the choices leading up to them are pretty much irrelevant.  What sort of character Shepard is hardly matters at all to the ending sequences because they are based around a single bizarre choice instead of a multitude of events over a long time frame.  This design decision is a hard one to make because there is always the temptation to try to top everything that has come before with something even crazier and more impactful so that the final moments of the story blow people away.  Writers naturally want to have a big reveal that leaves people surprised and amazed at how it all fits together but that works badly when the story isn't set.  You can't do a big reveal properly when you don't know what the character is going to do or what they have done before!  In a roleplaying game (which is what Mass Effect is trying to be) you simply have to let the ending flow naturally from the story rather than railroading them into The Big Twist Ending.

I will give credit where credit is due though:  If you are going to do a big Twist Ending and railroad the player into just a couple of options then it was done pretty decently this time.  The new endings give a solid sense of closure and feel properly done.  It isn't the best possible way to end a story but it is a damn sight better than their first try.

Wednesday, July 18, 2012

Overpowered villains

Last night I played the last encounter of the Pathfinder campaign I have been involved in for almost a year.  I really enjoyed my time playing with the group and had a grand time with the game but I felt pretty let down by the grand finale.  The final session was going to be pretty much exclusively a final battle with a lich that our party had been battling indirectly for so long but it didn't work out very well.  The trouble was that the lich in question should have handily mopped the floor with us and completed her takeover of the world; the only reason we won is because the DM very deliberately played her like a dunce.  I don't mind some characters doing things stupidly because being powerful in combat is no guarantee of being a master strategist but this opponent in particular was ancient and had absolutely massive Intelligence and Wisdom scores as well as plenty of time to prepare.

The fight should have gone something like this:  Lich teleports next to our group with Improved Invisibility on while we are asleep.  Lich leads off the surprise round with a Maximized Fireball while its undead dragon friend starts munching on us.  Lich hucks another Maximized Fireball and a Quickened Fireball next round and the dragon continues to munch.  This probably kills 3 characters, likely before all of them have even gotten a turn.  There is no reason to think that more than 1 of us would be alive by round 3 and that remaining character would be utterly certain to die by the end of round 3.  Instead the dragon attacked us by itself until we killed it and then the lich spent most of her time casting spells that would inconvenience us instead of ones that would flat out kill us.  Note that no advanced strategy is required here; surprising us and then casting Fireballs while invisible was more than enough to guarantee victory.  Advanced strategy includes things like having the dragon hover over our heads with Fly and have an active Silence spell on it so that it could utterly negate all of our party's spellcasting - letting me loose with a high level spellcaster is a dangerous thing.  :)

The biggest issue was the omniscience of our enemy.  The lich could scry on us from far away and teleport onto us any time she wished.  This guaranteed that the fight would take place where and when she wanted it to and that advantage is insurmountable.  She could wait until some of us walked a few hundred meters away to go to a shop, talk to a friend or do anything else and immediately teleport onto the smaller group and handily annihilate them.  She could attack us in our sleep or while we were fighting something else if she wanted to and we had no recourse.  Simply put there was no practical way for us to survive the encounter.  The galling thing was that the lich could have done this to us for a large part of the campaign; she was aware of us for some time and had every reason to wipe us out earlier but chose not to for some reason.  Knowing that the lich *should* have wiped us out a thousand times but didn't just because otherwise the campaign would end makes much of what we did feel pointless.

Having overly powerful enemies that could easily smash the main characters is a real problem in many games.  In the Mass Effect series it was the critical problem with the game that created endless plot holes and the excessively Deus Ex Machina ending (the new endings are released now and they are drastically better).  In ME it was always very strange to think that a person who is good at shooting people with a rifle would somehow matter in a battle against gigantic invincible alien spaceships and in the end that was unrecoverable - Bioware resorted to "ummmm, I guess you win by Magic, yeah....".

For the story to feel satisfying players do want to face superior enemies but the superiority needs to be limited.  It is very tempting to create a super awesome invincible bad guy and find a clever way for the heroes to defeat them but it usually becomes massively problematic that the bad guy doesn't just flat out *win* right away.  It is fun to outsmart clever but overconfident people but it isn't fun for the players to conclude that they only have a chance because their opponent is a moron.

Thursday, July 12, 2012

Magic Items and DnD Next

I remember in the second Lord of the Rings movie when Frodo and Sam went to the magic shop; it was such a great scene.

"Frodo, I know that Sting is really special and glows when orcs are around and such but we gained a few levels and we really need to upgrade it to something with a higher bonus."

"But Sam, Sting was a gift!  It is a marvellous magical item created using long lost techniques by some master magician..."

"Frodo, get real.  At our level we need better gear.  Vendor that old thing and come look at the selection of +3 weapons this guy has for sale.  Given that we have an appropriate amount of money for our level we *have* to get better stuff or we will die to the next challenging-but-winnable encounter!"

"Ok Sam, you are right.  I will trade Sting for 500 pieces of gold, though it wounds me to do so.  Shall we stare deeply and emotionally into each other's eyes for a minute before doing so?"

"I'm going to buy me a +2 dirty cotton shirt.  YES!"

No scene like the above one ever occurs in a fantasy novel.  Heroes wielding mighty magical weapons don't trade them in and they sure don't go to the store.  Just as much they don't accumulate staggering sums of money such that they could purchase entire kingdoms for the cost of a single upgrade.  Instead they find cool stuff and use it to defeat evil at just the right time.  They also find somewhere between 0 and 3 magical items during their entire adventure.

In DnD though we have gotten used to characters finding reams of magical items everywhere.  The games assume that characters will go to shops to buy astounding equipment and constantly sell off the stacks of gear they find to vendors who conveniently have millions of gold pieces just sitting around.  In 4th edition in particular the game was balanced around a very precise progression of magical gear from start to finish.  The wonder and awe at finding something new was pretty much entirely removed.  It also began to feel like characters at high level were just vehicles for their equipment - a high level fighter would lose 19 Armor Class without their normal armour equipped and that is *way* too much when random events are rolled on a d20!

I always thought that the magical item progression in DnD was ludicrous and the 4th edition progression in particular was overly gamey and led to ridiculous worlds.  Thankfully the designers of DnD Next appear to agree with me.  Gone is the reliance on +1 weapons and removed is the assumption of constant acquisition.  Their plan is to make magic items unique and to avoid giving them flat combat bonuses.  Huzzah!  Whether they will succeed or not is another thing entirely but it comforts me greatly to know that DnD is taking a step back towards feeling like a fantasy novel instead of Adventures In Shopping.  If I want a fantasy economics simulator I will just play Diablo 3.

Tuesday, July 10, 2012

Who is the sucker?

When you sit down at the poker table you had better look around and be able to spot the sucker.  If you can't, *you* are the sucker.

I have been playing a game of 'spot the sucker' in Diablo 3 lately and I can't properly determine who it is.  Lately there has been a surge of selling unidentified rare items for large sums in general chat.  People with lots of money buy gear for 200k gold and hope that they score big but usually lose everything.  Who is the sucker, the with 200k or the one with the item?  Since I identify my own items I apparently play as if the sellers are the suckers.

You can't math this one out.  Figuring out the market value for any given item is difficult enough but figuring out the total spread of stats that could be on items and the value of each as well as the probability of each stat spread appearing is not possible.  By the time you gathered enough data to be able to make a firm prediction the economy will have shifted so much that most of your data is useless anyway.  Since statistics can't help us spot the sucker we need to rely on psychology.

My reasoned bet is on the buyers being suckers.  People love to gamble and they love to imagine how great that next item will be.  I do it too!  I daydream about Ceremonial Knives with 1100 dps, 100% + crit damage, a socket and 900 Life on Hit.  I won't ever see one of course but it is nice to dream.  I know that back in Diablo 2 I made a fortune selling UNID rares to suckers for 1 SOJ apiece.  I had a regular list of customers and they came back again and again to pay an SOJ for another shot at the best bow/lance/mace in existence and never got anything of value as far as I know.  I do know that I bought myself an absolutely wicked bow once I had saved up 20 SOJs though!

There is, of course, a possibility that the price actually ends up being so close to true value that both the buyers and the sellers are suckers and the only winners are the flippers and bankers who spend their time cruising for deals on both sides of the fence.  Playing the AH looking for items to flip is a profitable business, or at least it *can* be, and playing merchant with UNID items is likely to be the same - you take advantage of people who are uninformed, impatient, or flush with cash purchased from gold selling companies.

By selling your UNID items you give up all chances of a big score, the gigantic single find that will net you a fortune.  Mostly you *don't* find those items, of course, but they are out there - Salacha recently sold a pair of gloves he found for 80M gold so it is clear that there are people out there with outrageous sums of money to spend if you have exactly what they want.  I think I am willing to miss out on the big find and I will happily settle for a constant stream of money.  The AH has all the items I will ever desire - all I need to do is get a large enough cash flow to be able to make all my dreams of glittering shinies come true.

Sunday, July 8, 2012

Lethal encounters and realism

This past week I spent some time working on the first adventure of my new Pathfinder campaign.  I am working on making combat encounters (my favourite part of building a campaign!) and ran into some roadblocks.  The biggest problem with monsters is that their stats are so interconnected:  A monster's Hit Dice determine its saving throws, attack bonus, and HP as well as immunities to various effects.  Its Constitution determines its HP and Fortitude saving throw.  So if I want to make sure my monster has enough HP to survive a few rounds against the players I need to jack up its Hit Dice or its Con or both.  Either way though I end up giving it a massive Fortitude saving throw which means that any spell the players throw at it with a Fortitude save cannot succeed.  This is a real mess.  I don't want giving a monster high HP to mean that it is immune to many classes of spells!  The other problem with stacking on more Hit Dice is that my monster would also have a extremely high attack bonus, quite probably making the character's Armor Classes irrelevant.  Ack!  If all of the monster's stats weren't so interconnected I could simply assign reasonable values to each stat and it would be grand but with the current system that isn't possible.

I can kludge things, of course.  Normally monsters get to roll a single 8 sided die for HP for every Hit Die they have.  Instead of a d8 though I can simply give my monster a d40 and it will suddenly have plenty of HP... but the players will be very confused when their opponents' stats don't make any sense.  Eventually it will become clear that the monsters aren't built on anything like the same rules system as the players and people tend to dislike that.  Quite understandably they like immersion and feeling like the playing field is level and randomly kludging stats can ruin that.  When given a choice between using the default monsters which have awful stat layouts which generate stupid fights and kludging I will kludge every time but I wish I was not put in that position.

The second issue I have had is trying to build a fight that has ~8 normal humans in it.  If the players decide to wipe out the little village they are going to visit then they will run into some opposition but the nature of the opposition is limited by the rules surrounding weapon damage.  A battleaxe or longsword hits for d8 damage which means that any scrub with a weapon has to hit for 4.5 damage on average.  The difficulty with this is that I can't have the players fight 8 dorks - if the dorks go first they stand a good chance of wiping the party out before any of the party members gets an action!  On the other hand if the players go first the dorks have so few HP that the players could easily kill nearly all the dorks before the dorks get an action!  It is difficult to get around this dilemma because aside from having all of the enemies have terrible Strength scores (your whole militia is a bunch of wimps?  wut?) or having all the enemies use daggers (your militia is armed with daggers?  wut?) you just can't get base damage down low enough to make having a large number of opponents reasonable.  Again, it 'makes sense' that a longsword does a particular amount of damage but it means that I really can't make particular fights work the way I want.  I don't want to kludge in a rule that these particular militia folk have a -50% Sky Is Cheating damage penalty but that might be the only way to make it work right.

Pathfinder does have the old DnD tradition of having lots of things constraining monster design because they 'make sense' but this is a real problem when trying to design really interesting, intense fights.  When fights get hard and design gets elegant the clumsiness and ugliness of the base systems really gets brought to the forefront; we know this from any number of games that had awful design at the start and had to clean up their systems when hard content was added in.  WOW is the best example I know of but Magic: The Gathering, Diablo 2&3 and others show us that when you want to add high level PvP or serious challenges you often end up wishing the base systems supported those properly instead of being 'realistic'.  

Monday, July 2, 2012

Levelling up the fun way.

I have been levelling alts in Diablo 3.  When I first levelled up I picked my skills ahead of time, tested most things out really quickly, and stayed with a static build for most of my time playing.  I never really tested many of the runes and ended up learning a lot about how things work at level 60 on my Witch Doctor.  I power levelled my Demon Hunter like crazy and did the same thing; hell, I hardly know how the class works aside from the 9 or so skills I used which were obviously the best.  With my last two classes, Monk and Wizard, I am using a different levelling system.  Each time I level up and get new abilities or runes I use those abilities for the entire level.  It has been a blast trying out all kinds of new builds and having a different set of abilities every time I level.  For those who stuck with a single build and found it boring I recommend this 'strategy' - it really keeps things new and interesting as you go through the game.  Note that I don't just put awful abilities on my bar and refuse to cast them; I beat the enemies with those awful abilities. 

Just so we are clear, some of the abilities are really awful.

On the Wizard it appears that Blizzard underestimated the power of actually delivering damage to target.  Most of the Wizard abilities are reasonable but there are a few that do appropriate damage but are so incredibly hard to actually hit the enemies with that they are laughably bad.  Explosive Blast, Meteor, Arcane Torrent, and Energy Twister are interesting designs that can only do enough damage in the instance that you attack immobile enemies who are not attacking back.  They probably look solid against a test dummy but even a couple minutes of playing the game through Normal trying to use them made it painfully obvious how often they completely miss and do no damage.  The most over the top version I have seen so far is Arcane Torrent - Death Blossom.  This targets a *random* spot on the screen for 670% weapon damage.  That's a lot of damage.... but to a *random* spot?  The first time I cast it I laughed at the spell for a full minute!

Now I think it is okay to have spells that are trickier to use and ones that are more dangerous to use.  I accept that many spells will have no place in the harder parts of the game - that's fine.  My benchmark is that a spell should be a reasonable choice at some point with some build.  Meteor might make that cut since I know bizarre tank Wizards use it but the others are just wretched.  If a spell is really hard to hit anything with it needs to hit like a truck full of anvils when it does hit.  The current damage numbers on Wizard spells really don't take this into account enough and that is why these spells that are very hard to hit with end up being the useless ones.  I am sure that Arcane Torrent could use a buff but even +50% damage would be sufficient to make it something you would consider using.  Explosive Blast and Energy Twister definitely need buffs in the +150% range though.  Explosive Blast is a small radius, zero range AOE with a cooldown!  It needs to absolutely mutilate the enemies to be remotely worthwhile.

Not everything needs to be good.  Not everything needs to be viable in endgame.  However, if you are going to go to the trouble of building all these cool animations it seems a good idea to make them good enough that people who like them thematically or for some other random reason can be reasonably successful using them.  There are plenty of ways to play Diablo 3 badly - having a few more ways to play it well would be great.