Tuesday, July 1, 2014

Venture City Problems

Hey, welcome back to my blog. Here's a really depressing blog post.

("Oh great,  JD is going to talk about math again.")

My buddy Ron Blessing just ran a couple of sessions of Venture City Stories, a Fate Core game, with me, my wife, his wife Vern, and a couple of others.  I really like both Fate Core and Venture City, for reasons I'll describe later, but for the purposes of this blog post, what's more important is that it is incredibly popular.  There are a stunning 19 reviews for it at the time of this post, the overwhelming majority of which are five stars. There are 35 ratings, 25 of which are a perfect 5 stars. So this blog post is less an indictment of the game than an indictment of gamers since there is a major, glaring mathematical error in one of the core mechanics of Venture City.

It's disappointing as hell. Even in a very mathematically simple and solid system like Fate Core, RPGs continue to have major problems with simple probability and expected outcomes. And it seems like nobody cares. The designers don't care because the community doesn't care and it's time-consuming (and therefore expensive) to thoroughly evaluate the probabilities of a game system and come up with the answer to the question "is this what we want our system to do?"

Venture City Stories bills itself as a superhero game - as with most superhero games, the player characters are formidable and iconic. So a mechanic that has the expected result of a TPK when in conflict with some very basic street gang members is a mechanic that should leap out at people. But it didn't. It hasn't. And nobody quite seems to care.

So What's The Problem?

Before we get started, a quick overview of the Fate Core system. Skills of characters are rated from +0 to +4. You get one skill rated at +4, two at +3, three at +2, and so on. This is called the "skill pyramid". You get stunts which can be things like using a skill in an unusual way, or receiving a small boost to a skill in broad circumstances or a large boost to a skil in narrow circumstances.

To use a skill, you roll four "Fate Dice", which are six-sided dice each with two "+" sides, two "-" sides, and two blank sides. These represent +1, -1 and 0 respectively. So your results may be +,+,-, blank, and that adds up to a +1 result. You add that result to your skill and compare it either to a target number set by a GM or generated by an opponent using their skill.

Here is the basic probability breakdown of Fate dice. For Fate Core, it's the "4dF" entry.

As you can see, the expected total outcome of rolling 4 Fate dice is 0, this happens around 24% of the time. There is a 63% chance that the outcome will be -1, 0 or +1. There is a 90% chance that the outcome will be between -2 and +2. It's a simple, easy-to-calculate normal distribution, and importantly, it's solid game design.  

If I have a skill of +3 in Fate Core, and I'm up against someone with +0, the chance that I will tie-or-win is around:

  • 38% (the chance that I'll roll a +1 or more, in which case my success is guaranteed) +
  • 23% (the chance that I'll roll a 0 and that my opponent won't total a +4) + 
  • 18% (the chance that I'll roll a -1 and my opponent won't total a +3 or +4) + 
  • 10% (the chance that I'll roll a -2 and my opponent won't total a +2, +3, or +4) + 
  • 3% (the chance that I'll roll a -3 and my opponent won't total a +1, +2, +3 or +4) + 
  • ~1% (the chance that I'll roll a -4 and my opponent won't total a 0, +1, +2, +3 or +4) =
Around 93%.

This means that in a contest between someone who is ranked as "Good" in a skill and someone who is "Mediocre", the "Good" character wins or ties 93% of the time. And that makes sense in context of a game where you're only going to have two "Good" skills and only one "Great" Skill (It is impossible for the "Mediocre" person to beat someone with a "Great" or +4 skill, and even tying has a probability of around a hundredth of one percent.) Your "Good" and "Great" skills are areas where your character dominates the situation.  

It's a good, solid feeling that you're going to win or tie 93% of the time over someone with a skill level 3 less. The same probability distribution applies to someone with a +4 skill competing with someone with a +1 skill. I leave it as an exercise for the reader.

There are some ways this distribution can be shifted. For example, Aspects can be used to give one side or the other a +2 bonus. As you can see from the analysis above it doesn't much matter if you give yourself a bonus or the other side a penalty in a contested situation, since the probability of success relies on the difference between the two skill rankings, not their absolute value.

Fate Core is a pretty mathematically sound system. There are a few quirks here and there but nothing that leaps out and screams "this is a problem!" And you can read it at the Fate SRD for free (or download it for pay-what-you-want at the DrivethruRPG link above.)

Nameless NPCs

In Fate Core, there are rules for Nameless NPCs.  These are thugs or minions that pose more of an obstacle than a threat to action heroes like the player character.  As with the main rolling mechanic, the Nameless NPC rules are good because they permit GMs to resolve the actions of many such NPCs with a single roll, multiply their effectiveness while not bogging the game down and not inviting an "outlier" roll by making many, many die rolls.

"Average" nameless NPCs have a few skills at +1, and are taken out if they lose any physical contest with a player character.  

They only are likely to pose a significant threat if there are many of them: in Fate, if someone has a +1 or better in a skill, they may "mob" with someone else and provide an additional +1 to the "lead actor". So if 6 "Fair" nameless NPCs are firing their guns at the gargoyle Batman is perched on, instead of six rolls at +1 (versus, say, Batman's +4 Athletics), they can team up to do one joint +6 attack. This has its upside for Batman as well, since he can use limited Aspects and Stunts only once in response to the attack instead of quickly running out his resources, using them 6 times.

If, instead of 6 Average Nameless NPCs, there are 5, plus Harley Quinn with a comically oversized revolver and a +3 Guns skill, Batman could be in sincere danger: Harley leads with +3 and the five goons each add a +1 for a total of +8. Or the GM could have Harley attack on her own with her +3 and the other five combine for a +5. Or Harley could try to use her skill in other ways, to Create an Advantage to help her boys out (say, shooting at the guy wires that hold those catwalks in place that Batman is always escaping to.)

Average nameless NPCs have a few skills at +1 and can't take any physical stress. Slightly better, "Fair", nameless NPCs have a skill at +2 and can take 1 physical stress without going down. That means that if a PC attacks them and only beats them by 1, the NPC can keep fighting until that happens again. And "Good" nameless NPCs have a skill at +3 and have 2 stress boxes. It's emphasized in the Fate Core rules that Good nameless NPCs can get very dangerous very fast, and should be used sparingly.

I think you can see how this concept really adds to Fate Core and how flexible it is. Again, it's mathematically sound - there's no weird outliers because the NPCs are mapped closely to the same skill tree as characters. "Good" Nameless NPCs are going to succeed against "Mediocre" PCs 93% of the time, as we learned above - that's pretty formidable - and that's what the rules say they are.

So what happened in Venture City Stories?


In Venture City Stories, there are several factions listed. These are groups that have several aspects in common and work together. The statistics associated with a faction are to help GMs understand and express what the capabilities and focuses of these groups are. Well and good. 

But here's what Venture CIty Stories says specifically about the nameless NPCs that belong to them:

All right, sounds good.  But then when they actually get to the faction of non-powered street toughs, The Crew, their Violence rating is +6...twice what the most formidable nameless NPC is in Fate Core, along with an Espionage rating of +3, which would normally be at the very peak of the most formidable nameless NPC's capabilities.  Now, the new faction rules don't say if their ratings apply to stress boxes in the same way as they do in Fate Core, but based on the Nameless NPC rules, I would guess that each of The Crew has 4 physical stress boxes...and that's not even counting the superpowered ones.

In one potential confrontation, there may be "four or five" with even more powerful Violence ratings due to pyrokinesis or super-strength, adding a further +2 to that +6 Violence rating for a total of +8.  

So how is this going to go? Let's lowball it and say there are four nameless Crew NPCs with a Violence rating of +8, two super-strong and two pyrokinetics, and four "normal" Crew nameless NPCs with a Violence rating of +6 (though the text seems to suggest there might be many more).  Up against them let's just use the sample characters in the Venture City supplement itself: The Brick, The Psychic, The Speedster, The Flamer.

Before we get into the physical confrontation, let's make it clear that there are other options in the scenario like sneaking around or trying to parley with the Crew - however, the scenario does indicate that physical confrontation is possible and depending on the skill mix of the player characters, even likely.  When we actually played the scenario, we had a super-hacker, a mega-telekinetic, a shapeshifter and a super-martial-artist, so a fairly good mix.  Nevertheless we couldn't avoid all physical confrontation with the Crew, and, given that our mission involved disrupting their drug dealing operation, didn't really want to.

So let's start the confrontation.  We'll assume that something has happened and the Crew as described above: two Super-Strong characters, two Pyrokinetic characters (all with Violence ratings of +8) and four non-powered Crew (with Violence ratings of +6) become aware of the 4 player characters all at once.

Turn Order

The Crew has an Espionage of +3, which is the closest thing to Notice. The sample characters don't have a full skill pyramid for each one, but Notice is a "Suggested Other Skill" for the Psychic and the Speedster. Let's say they both have it at +2 - it's not one of their 3 "peak" skills (of which they have one at +4 and two at +3), but it's a significant investment for them.  And let's say that the Flamer has Notice at +1 and the Brick has it at +0.  

Since the Crew, the Psychic and the Speedster all have +3 to Notice, we look to Athletics or other appropriate skill to see who goes first. The Crew has a Violence rating of +6 (we won't give the bonus to the pyrokinetic or super-strong people for this since their powers don't really affect their ability to respond quickly), but the Speedster has Athletics as a peak skill (let's say it's her one +4 skill) and a power that gives her a +6 to Athletics when it comes to moving quickly, for a total of +10. Nice.

Our Turn Order is:

  1. The Speedster
  2. The Crew
  3. The Psychic
  4. The Flamer
  5. The Brick

The Speedster goes first. This is a great example of a character who is optimized to take out normal Nameless NPCs.  She gets a +2 when fighting in close combat (something with her Athletics and power she will be able to do virtually all the time).  If she gets more than 3 excess shifts in fighting someone, she can move those shifts into performing other skills as if the skill was at a +1, and she can use Fight as an Area Attack, hitting everyone in a zone (but they gain a +2 bonus to defend.)

But because of the way that the Crew is not just more powerful than most nameless NPCs, but twice as powerful as the most powerful ones listed, the Speedster is extremely unlikely to be able to do that Area Attack. Let's run the numbers and see why.

Right now when the Speedster zooms into a Zone and attacks exactly one Crew member, she's rolling 4 Fate dice, +3 for her Fight (remember that Athletics is her only +4 skill), +2 for the up-close-and-personal bonus, for a total of +5.  The Crew member, trying to dive/block/whatever and not be injured, is at +6 (their Violence rating).

To do any physical stress to the Crew member whatsoever, the Speedster must beat, not tie, the Crew member's total.  In other words, since the Crew member starts "up 1" on the Speedster, the Speedster has to beat the Crew member's Fate dice total by 2 in order to do any damage at all.  The probability that the Speedster will do any damage at all is:
  • ~1% (the chance the Crew member will total a -4 and Speedster totals a -2 or better) +
  • 4% (the chance the Crew member will total a -3 and Speedster totals a -1 or better) +
  • ~8% (the chance the Crew member will total a -2 and Speedster totals a 0 or better) +
  • 7% (the chance the Crew member will total a -1 and Speedster totals a 1 or better) +
  • 4% (the chance the Crew member will total a 0 and Speedster totals a 2 or better) +
  • ~1% (the chance the Crew member will total a 1 and Speedster totals a 3 or better) +
  • ~.1% (the chance the Crew member will total a 2 and Speedster totals 4) =
Around 25%.  There's a 3 in 4 chance that the Speedster, who is described in her description as "You don't want to get into a fistfight with her" will do absolutely no damage whatsoever to the person she hits.

Now, it's true that the Speedster has Aspects that could be used to boost her total. But, then again, so does the Crew, so that's a wash.  The real question of whether a matchup in Fate is fair or not is where the core rolls shake out.

What's worse is that if we use the normal rules for stress tracks for nameless NPCs, each member of the Crew has 4 stress boxes since their applicable skill (Violence) is higher than +5.  This means that to take out a member of the Crew in one hit, the attacker must beat their result by 5.  The lowest result the Crew can generate is 2, meaning the lowest result that can accomplish this is a 7.  The Speedster will generate a 7 or better only 18 percent of the time, meaning that even if the Crew does as horribly as they possibly can with their roll, getting a result that only happens less than 2 percent of the time, the Speedster will be unable to take advantage of their crap defense a solid 4 times out of 5.  I won't run the other numbers - suffice to say that even if the Speedster succeeds in hitting, it's extremely unlikely she'll do more than one or two stress to the target.

Why is it so important to take a Crew member out instead of just whittling them down? Well, it's their turn. Let's see.

Allow Us To Riposte

There are eight of them and four player characters, so let's keep things simple for ourselves and say that they simply divide up two per person. They also mob up as suggested in the rules.  We'll say that the super-strong people attack the Speedster (she is "up close" to get her bonus, after all), the pyrokinetics attack the Psychic, and the remaining four non-powered Crew members divide themselves between the Brick and the Flamer.

The super-strong Crew member starts with a +8 - mobbing up with their super-strong pal brings it to a +9.  The Speedster's Athletics is her peak skill at +4.  You can already see that the super-strong goons are guaranteed to inflict some physical stress on the Speedster.  But how likely is it that they can one-shot her?  Let's find out.

The Speedster's Physique is an important but not peak skill, so let's say it's at +2.  As a result she has one bonus physical Stress capacity for a total of 3, meaning that to take her out with one blow, an attacker must beat her defense roll by 4 or more.

The chance the Crew will one-shot the Speedster is approximately:

  • 1% (the chance the Crew will total a 4) +
  • 5% (the chance the Crew will total a 3) +
  • 12% (the chance the Crew will total a 2 and the Speedster rolls a 3 or lower) +
  • 18% (the chance the Crew will total a 1 and the Speedster rolls a 2 or lower) +
  • 19% (the chance the Crew will total a 0 and the Speedster rolls a 1 or lower) +
  • 12% (the chance the Crew will total a -1 and the Speedster rolls a 0 or lower) +
  • 5% (the chance the Crew will total a -2 and the Speedster rolls a -1 or lower) +
  • ~1 (all the rest barely add up to this) = 
73%.  Now, the Speedster can stay in the game because she (unlike the nameless NPCs) can take consequences. These are the "death spiral" of Fate - injuries or other disadvantageous things that make it harder for her to make it through the next round, in exchange for staying in this round.  But it seems sort of insane to be taking immediate consequences from a physical confrontation between nameless NPCs and a character that is expressly designed for going up against nameless NPCs.  

The expected outcome for this fighting specialist against nameless goons should not be "I have a 1/4 chance of chipping away at them a little bit, and they have a 3/4 chance of breaking my leg or taking me out in the first round".

And as you will see, it will get much worse for the other characters who are not so targeted in their design.

The Pyrokinetics vs the Psychic

Let's assume for the moment that the Psychic can use their Will to defend against the pyrokinetic attack. (It's not clear from the text whether the pyrokinesis is considered a "physical attack" or not but let's just say that it is.)  That's great, it means that, like the Speedster, the Psychic is defending with his peak skill, we'll say it's a +4.  However, the calculation is even worse for the Psychic than it is for the Speedster because the Psychic doesn't have Physique as a skill they're focused in.  Like a lot of comic book Psychics, they can't take much physical punishment before they're knocked out.  With no (or low) Physique, the Psychic only has two physical Stress boxes.  An attacker only has to overcome them by 3 in order to take them out, not 4 as with the Speedster.

So the Psychic is defending with +4 against an attack of +9 (+6 Violence for the faction, +2 for the pyrokinesis, +1 for a second pyrokinetic mobbing up), and we want to know how often the attacker beats the defender by 3 or more:

  • 19% (the chance the Crew totals a 2 or more) +
  • 19% (the chance the Crew totals a 1 and the Psychic totals a 3 or lower) +
  • 22% (the chance the Crew totals a 0 and the Psychic totals a 2 or lower) +
  • 16% (the chance the Crew totals a -1 and the Psychic totals a 1 or lower) +
  • 7% (the chance the Crew totals a -2 and the Psychic totals a 0 or lower) +
  • 2% (the chance the Crew totals a -3 and the Psychic totals a -1 or lower) +
  • ~1% (the chance the Crew totals a -4 and the Psychic totals a -2 or lower) =

It's around an 86% chance that the pyrokinetics will one-shot the Psychic even if we permit the Psychic to use Will to defend.

The remaining conflicts I leave as an exercise for the reader.  Just to let you know, they're even more mind boggling.

What The Hell Happened?  (And why am I the only one who seems to notice it?)

My suspicion is that the Faction ratings are left over from an earlier draft, where +6 (or even +8 for the super-powered members of the faction) was supposed to be the maximum that a gang of the faction could throw down. In other words, it makes sense to me to say "Hey, GMs, when you have a bunch of Crew characters, they should mob up, or have superpowers, or some combination of those, that gets them approximately to a +6 when they try to do damage to the PCs."  In fact, that's a good piece of guidance that the original nameless NPC rules can use!  +6 is a nice solid danger - as you can see, the player characters that are good at defending defend at +4, so that means they can be in peril of taking some stress or even some consequences, but with teamwork and some creativity they can overcome.  (This level of danger also gives characters more "breathing room" to establish those creative decisions, and if that's where the Crew starts at, their danger can be lessened as characters take out various baddies.)

But to just say "hey, the nameless NPCs in this setting are between 3 and 4 times as powerful as the nameless NPCs in Fate Core" without identifying any other changes to how they're taken out, if they mob up the same as other nameless NPCs or how player characters might exploit their situation seems to fly in the face of the superhero setup promised in both Fate Core and Venture City, where the player characters are highly effective and exceptional.

Venture City Stories is a brilliant game, generally speaking.  The setting raises real-life issues in a literalized way, in the way the best comic books do.  It takes our anxieties about wealth inequality and corporate power and says "yeah, so not only are they rich and influential, they can also throw a car".  Unlike many generic superhero settings it comes with a clear ethos, a moral presentation that reflects the best that the superhero genre has to offer.  

And the superpower system for fully fledged characters is really good!  Note that if you put any of the sample characters up against (say) Splitstream, the head of the Crew, it's a much more evenly matched bout!  This is a game where it is literally easier for you to be effective against the main villain than against nameless goons because the nameless goons have that huge faction bonus and the ability to mob up. (Spoiler alert: actually all of the factions are far overpowered from what the nameless NPC rules allow.)

I would suggest just using the nameless NPCs from the main Fate Core book.  If you want some of them to have powers, make them Supporting characters and decrease the number of goons accordingly.

As for why nobody else, including the obviously very talented design team seems to have noticed this...all I can say is:

  • Either people aren't playing it by the rules and so don't notice that the rules lead them down this path, or
  • They aren't playing it at all, just reading and eyeballing it and not thinking about the consequences of the mathematics of the system.

And both of these seem disappointing to me.  Like, someone should have developed a "Factions FAQ" or a "hey, here's how you fix Factions" in your game that gets posted all the time when people are talking about Venture City Stories.  But rpg.net had a 52 page thread on the game and not a single person brought it up.  52 pages - that's 22 pages longer than Venture City Stories itself! - and nobody noticed that one of the rules introduced in the game straight up broke one of the key elements of superhero stories, the nameless minions. 

No wonder mistakes like these get past designers and editors - the target audience just doesn't really care that much about these types of things.  It takes effort to run the numbers and think about what they mean, and that means money, and margins are tight. A designer who is prioritizing how they spend their time would be nuts, I guess, to spend it on making sure all their mechanics work properly, since gamers overwhelmingly don't care if they do or not.

I'm disappointed in this because it devalues the good mechanics that are well designed.  For example, in Venture City Stories itself, characters have Special Effects on their powers, which activate when they succeed "with style" (significantly) over opposition. I already mentioned some of the Special Effects that might activate in the Speedster section above.  This seems really cool - it makes you want to succeed more vividly - it makes your big successes seem more special than just "oh yeah, you did more damage".  And it fits with the superhero genre.  I read it and pumped my fist, woot, that's a cool mechanic.  It even meshes will with Fate Core's "succeed with style" mechanics, just amping it up to reflect how awesome supernormal people are.

But looking at the Faction rules makes me wonder - if the Special Effect mechanics had been straight up broken, would we even care or have noticed? Why make Special Effect mechanics good when it doesn't matter to the target audience if they're good or bad?

Anyway, I didn't mean to come back to the blog with a long depressing rant, but it just really bothered me and I don't see any way past it.

Edit: As a result of some very kind re-shares of and commentary on this article the problem has been somewhat fixed by adjusting the numbers assigned to the factions. They still are more powerful than what I think they should be (again, just use the standard rules!), but they're a lot closer to the norm now.