Wednesday, June 30, 2010

The Isle Of Dr. FlawedMath

A couple weeks ago my wife and I picked up a cooperative card game called "The Isle of Dr. Necreaux". The concept behind the game is a good one - it simulates a team of spies infiltrating an evil scientist's island lair, and escaping before the bomb goes off. But the more I play it, the more I feel like there's a mathematical flaw at the core of the game.

What it comes down to is this: You'll like the game if you're not particularly fond of, nor frightened by, math, and you're okay with games with really wordy text-dense cards.

If you really like math, you're going to enjoy your first several plays of the game, and then tire of the fact that what is presented as multiple strategic paths is actually just one solution and an array of sub-optimal choices that deceptively look mildly appealing at first blush. The more you analyze the game, the less it holds up to the scrutiny.

It's not a bad game. It's actually a fun game, but it's got elements that are sloppy and annoying. You look at it, and think "this could have been so good, if they'd just spent another month in blind playtesting followed by a month in development followed by one more thorough editing pass." Gaming companies function on small margins, though, so I understand why that didn't happen. But it makes me sad.

My main gripe has to do with the Speed mechanic which lies pretty close to the core of the game.

The vast majority of the cards in the game assume you'll be traveling with a Speed rating somewhere in the range of 2 to 7. There's several traps that are triggered if you roll less than your speed rating on a d6, and others that trigger if you roll more than your speed on a d6. There's a few character cards that let you roll two dice for these sorts of things and use the better die, or let you add or subtract +/-1 to/from your effective speed for such things. Throughout the game is the implication that every turn you should be adjusting your speed a little bit to take into account these various cards. It implies you should move slower when the party is injured or low on resources, and make a bigger run when you're relatively unhurt. At least half the cards in the deck provide some sort of interesting tactical decision if your speed could only be between, say, 1 and 8, with an average of 5 or 6 being ideal.

Problem is, the game expressly states that there is no upper limit to how fast you can move. It's extremely clear about this - you can set a speed of 20 (or 5,000) if you'd like. Your speed determines how many bad cards you flip over in a turn, so at first blush it seems like a large number is suicidal. After playing about 4 or 5 games without a single win, my wife and I just on a lark tried a turn at speed 20. It went off without a hitch. So we did it again. We've played another half-dozen games since then, and found that as long as we move at speed 20 or higher as often as possible, we always win.

We either sit at speed zero for the rest bonuses, or jump ahead at a breakneck speed of 20 or more. It's come down to essentially a binary decision, rest or run as fast as we can. Which means those cards that care about your speed rating don't have any decision-making weight or randomness they initially seemed to have. All the cards that involve "if you roll less than your Speed rating, bla bla bla..." trigger automatically, and the ones that say "if you roll greater than your Speed" never trigger at all. There's no reason to ever want to move at a speed low enough that you'd ever agonize over, be surprised by, or even contemplate such cards. They just happen, beyond your control, and you can mostly laugh off the results.

The game has several cards that hose you if you move to fast - but they hose you equally if you're traveling at speed 8 or speed 80. Having a higher speed means you're more likely to draw these cards, and more of them in a turn, so it can get a little dangerous. In general, though, you can always survive any one card, and the retreat rules make it relatively simple to back off if a turn starts to go against you. In the default 2-player game you're trying to get through 65 cards in 11 turns, so Speed 6 sounds good. But you'll want to rest a few times, so you go speed 0 one turn and speed 12 the next. That just opens the doors to speed 20, or 30, or 50, any of which is vastly superior to a bunch of turns at speed 6. The benefits of setting an arbitrarily large Speed number significantly outweigh any minor benefit of potentially missing a trap or two by going much slower on some turns. It seems like such an inescapable conclusion, I've hunted through the rulebook and BoardGameGeek forums trying to find something we could possibly be doing wrong. Everything I've read in the rules or forums supports my conclusions.

Which would all be just fine, except for the fact that some of the character cards seem predicated on helping you only if you move in tiny little baby steps. The Security Expert, the Infiltrator, the Rocketeer, the Rogue, etc, all give boosts that would only be helpful if speeds 2 to 7 were viable, and those speeds just aren't.

Worse yet, the game presents this Speed-based risk management as one of it's core concepts. In theory, it should really resonate with people who like to number-crunch and analyze, or at least play the odds intelligently. But speaking as one of those people, I find I'm taunted by the fact that there's a better game with more meaningful decisions lurking beneath the surface if only there were some sort of speed limit and/or a smaller deck to go through. As it is, going infinitely fast is the way to win, and the deck is big enough you can't really hope to win if you don't do that on at least one turn. It just feels like they missed an opportunity in game design here. It's like they're aware of what makes the game fun, but didn't realize that the dominant strategy specifically downplays the fun aspects.

Anyhow, the game has a couple other flaws. It's fun, but the whole time you feel like something's amiss:
  • The pulpy retro-futuristic vaguely Buck Rogers or Flash Gordon theme is kinda neat, but if you went by the description on-line or on the back of the box you probably thought you were buying more of a vaguely James Bond themed product. It's grown on me, but it was a big shock when I opened the box.
  • The cards are insufferably wordy. Look back up at this post (and most of my other posts on this blog). You'll see that I'm a man who likes words, and I'm not the least bit terse or laconic. I'm certainly not phobic about words. I don't find text distasteful as a general rule. But these cards are too damn wordy for me. They need to implement a few more keywords. Example: There's 4 cards called "Brain Spiker", "Entropocyte", "Psychic Leech" and "Tangler". Each has 9 lines of text in tiny little print. Of those 9 lines, 8 lines are identical on all four cards (except for the repeated instances of the name of the card you're looking at). The one line that's different can be found in the middle of the text, making it hard to locate or parse. That's a huge pain in the visual cortex.
  • The rules are sloppy and contradictory. Example: You win by ending your turn with the "Escape Shuttle" card in play. The rules twice mention that the turn must be completed in order for you to win. But the card can (and indeed must) be put into play as soon as it's drawn, and the card says "You may end the game at any time". So you really don't have to end the turn, you can just end the game as soon as you play it. I think. There's a few other places where the cards disagree with the rules, and not in a fun "M:tG" way. Instead, it just seems like sloppy editing. There's no timing rules, and several instances where they'd be very helpful in clearing things up.
  • In this game, ninja's operate better on a team then alone. Everyone knows the opposite is true. If you're watching a movie and there's one Ninja after the main character, he's a badass. If there's a dozen ninja standing in the main character's way, they're just mooks and it will take 3 seconds to drop the whole bunch. In this game, however, a ninja on his own is ineffectual, and a ninja on a team is a powerhouse.
  • The sliding wall that splits up the party is actually good for you. If you are not willing (for whatever reason) to take turns at Speed 20, then your only hope to win is for the party to be split up into separate groups by the sliding wall. If you think you can handle 20 cards together, you can handle twice that easily if you split up. The sliding wall is the single-most beneficial card to the party, yet it's labeled a "trap" in big letters. Makes me scratch my head.
So, lots of little things to complain about, but overall it's still fun if you can manage to look past those flaws. Fun can make up for inelegant design, but only for so long. It's a game in desperate need of a revised edition.

Monday, June 28, 2010

Modding a Feral Beast

A while back I mentioned the Humble Indy Bundle, which was a set of 5 or 6 downloadable computer games. Very cool little titles, of which the cream of the crop is a beautiful game called Aquaria. My wife and I have logged a lot of time playing Aquaria. It's one of very few computer games I've ever felt the compulsion to truly play all the way through. After about 50 hours of play, we not only reached the end but also amassed nearly all the in-game achievements.

Then we broke out all the mods we could find for the game. Most of them were pretty short, though, so Sarah and I decided we'd take a stab at modding it ourselves.

Download Link
Our mod is called Feral Beast. What's available from that download is just a preview; enough to keep you busy playing for an hour or so. The mysterious beginings of the story, and the first 5 map levels we got finished. There's a few hints towards the eventual plotline, but it's mainly just an exploration mod at the moment. Poke around with it and have some fun. You'll need Aquaria to play, of course. You should carefully read the installation section of the readme.txt file included in the download, as the new recipes don't install automatically, and must be manually placed inside the game folders.

Additional chapters will be added as we create them. The actual level editing stuff is fast and easy, but there's some significant time investment in designing new creatures and coding anything particularly interesting, so it'll be a few weeks before you see an update.

Friday, June 18, 2010

You got your Everway in my Warhammer!

This is an article about how I folded two very different game systems into one another, and found the place where they intersect to be more immediately enjoyable than either on it's own.

First, some context about the two games:
The new(-ish) Warhammer FRP has some really awesome groundbreaking elements like group initiative, abstract movement, multiple-axis of success, and wickedly cool dice. However, on the whole it's a lot more fiddly than I'd like it to be, and it involves a boatload of components that hog the table-space. It's a strange (but mostly enjoyable) mix of complicated bean-counting and elegant time-saving innovations. It's a game that can't quite decide on its target audience. Too fiddly for fans of most indy games, too revolutionary and abstract for those who approach combat as a comfortable tactical challenge.

Everway is a cool older game that was one of the first really light and abstract game systems. But it's in practice, too light. Combat is entirely narrative, and things like damage, healing, initiative, and even who wins any given conflict is almost entirely a matter of GM fiat. With the right group, and the right GM, a good dose of trust, and plenty of time to get the balance of the non-mechanics fine-tuned, it can be incredible. I've had some fun with Everway before, and gotten a lot of joy from Amber, another game with roughly the same level of GM fiat and rules non-existence. While I like that style of gaming, I know they're not for everyone. If you're playing infrequently, don't know your play group well, or want a little more "game" in your game, Everway is not a good choice.

Hauling out my old Crunchometer, I'd say Everway is an umistakeable c4. It's light and simple, the rules could fit on a single page. Character creation is entirely open-ended, with the exception of four very basic and flexible attributes. There's hardly any math involved in character creation, and far less while you're actually playing.

Warhammer 3rd, on the other hand, occupies no single niche, and instead is spread across a hazy band near the other end of the spectrum. It's a c20 that plays more like a c12 (and gets to that lower rating by incorporating some features usually found in a c6).

The resulting hybrid was about a c8, which for me is pretty much the sweet spot where the gaming is really good. I was very happy with the way it turned out.
I'm not sure exactly how I realized that Warhammer had exactly what Everway was missing, but after that thought popped into my head last week, I started noticing other similarities. The real breakthrough came when I realized that in both games, nearly all character attributes will be in the range of 3 to 6 (the extreme apex of the range is higher in Everway, but in practice I've never seen a PC go above 7 and most don't go above 5). Which meant the basics of Warhammer's intriguing dice system could be ported over into Everway without having to make major changes to the character creation system of the lighter game. From there, it all fell into place organically.

Friday night, after arriving in Portland, I taught the basics of the system to my three players (two of which had never played either Everway or Warhammer), and they created characters. On Saturday, we played all day long (about 10 hours of play, IIRC), and everyone had a blast. It was probably the best gaming I'd ever had with this particular group.

So here's how "EverHammer" works. Character creation is identical to Everway. All the differences are in game play, not character creation.

The four elemental attributes use blue warhammer dice to actively do things. The difficulty is a number of dice equal to the defender's relevant stat, of which the first die is purple and rest are black. So if an archer with Air 5 attacks a target with Fire 4, they'd roll 5 blue dice, 1 purple and 3 black (about a 57% success rate, and the counter-attack with the stats flipped would be about a 37% success rate).

Powers added either white or yellow dice to active rolls. As I ran it, I had 0-point powers add a white die, and powers that cost a point or more add a yellow. I'm not 100% certain that was the right break point. The way I did it worked, but it might not have been optimal. I could see reserving the yellow die for powers whose points come from the "Major" quality, instead, for example.

If a power was defensive in nature, it added a purple die to the roll.

Additional white and black dice could be (and frequently were) applied as improvised situational and terrain modifiers.

In all cases, dice pools flow from the character creation system, and causality comes from the shared game world and goes into the dice. Never the other way around. In other words, you can't buy a yellow die on it's own, or upgrade a power to Major just to get a die.
Likewise, if a power is defined as "immunity to fire" and that character is attacked with normal mundane fire they cannot take damage. We still roll, to see if there's any fatigue changes or fortune effects (see below) but damage is off the table. If they were attacked by magical fire, or dragon breath, or some complicated mix of fire and something else (such as a volcanic explosion) then damage might be a possibility but they'd get the bonus purple die for defending with the fire immunity.
I ditched the stance system from Warhammer. I'm not sure I had to, there's probably some reasonably elegant way to work in red and green dice (possibly Fire & Air = Red, Earth & Water = Green), and finding meaning for the delay icons despite not having action cards (more about that later). I was a little pinched for time (having gotten the idea for this system less than 72 hours before we were going to be playing it), and didn't want to try to figure out how to include that extra layer of complexity that comes from stance.

You roll the dice, do all the cancellations, and if there's at least 1 success left, you hit for normal damage. If you score 3 successes, you hit for double damage. Normal damage is 1 card. I made some wound cards to handle this, they were basically the equivalent of a critical wound from Warhammer. I'll probably devote an article to discussing them in depth at some future date, because I think there's some non-obvious considerations on what types of criticals enhance a fight scene, and what just ruin the game. If you suffer more wounds than you have points of Earth, you pass out. Any hit after that point could be a Coup de grĂ¢ce.

For purely mental attacks, I ginned up a "disorder deck" that's basically criticals for mind powers. It's got a bit of overlap with various critical hits, insanity cards, and party tension meter effects from Warhammer. Again, when I get around to writing the article on the wound cards, I'll revisit this topic in depth.

Boons and banes were handled more simply than in Warhammer as well, eliminating all the complicated effects of various action cards. I kept fatigue, but rolled stress into it so there was just one type of fatigue point to worry about. I ran it with the existing Warhammer rule of 2+ banes = 1 fatigue, but in hindsight, I think I'll run that as 1+ bane = 1 fatigue next time. There were a lot of rolls of 1 boon or 1 bane, and not enough fatigue going around (An Everway character is less likely than a Warhammer character to have a weak-spot that fatigue penalizes.) If fatigue is higher than the active attribute, each additional point of fatigue adds a black die. Passing out sucks, and the wound system already emphasized Earth enough, so I set the pass-out to 10 instead of something attribute-based. One could argue the other way around (that wounds should be a set high number and fatigue based on Earth), and I wouldn't have a good counter-argument. I did what I did, and it seemed to work.

In theory, I still kept the door open for location cards to have more elaborate / non-fatigue-related boon or bane effects. I wasn't quite organized enough to implement that to the extent I would have liked this time out, but I'll be ready with it next time we play.

The biggest changes to the dice mechanics involved the Sigmar's Comet and Chaos Star results. In Warhammer, there's a lot of options when one of these is rolled, and deciding which option to invoke really slows down the process as either can change the other dice in the pool. Comets can, on a Warhammer attack roll, be used as a success, used as a boon, trigger a comet-line of an action, or add a critical hit. It's a lot to contemplate, and the chaos stars aren't any easier. I got rid of all the dice/symbol switching options, and just made these special symbols trigger a card draw. You'd flip over the top card of Everway's Fortune Deck, and if you rolled a Comet the GM would interpret that card in a beneficial way. If you rolled a Chaos symbol, the GM would interpret the card in a negative way. In theory this could have been as severe as a wound, but in general it was narrative effects and bonuses or penalties on future rolls. For next time, I may draft up a list of suggested/default effects by card type or severity.

I used the "group initiative" system from Warhammer 3rd. Initiative rolls were made with the water attribute, because it's linked to senses and awareness, and because Water was looking to be the least-used attribute otherwise (since Earth determines health and Fire and Air determine most attack rolls).

We used abstract movement rules and manoeuvre system per Warhammer 3rd, and the associated fatigue costs. Late in character creation after everyone was done with the Everway Vision Cards, I sorted out all the pictures of places to use as Warhammer Location Cards. I definitely need to supplement this deck, but it worked well enough for 10 hours of mostly-improvised adventure.

Pretty much everything else was pure Everway.

The hybrid system worked well. It had enough going on to spark some interesting challenges during the fights, including an interesting exchange with 2 PCs trying to draw the attention of the bad guys to protect a third that resulted in a very fluid rolling battle. Outside of combat, it was non-intrusive and simple enough that no role-playing scenes were disrupted by needing to consult rules. I think we were all very pleased with the results. There's a few things we'll do a little differently next time, but in general it exceeded all my expectations. Good gaming all around.

Thursday, June 10, 2010

From Peak To Valley

In which I admit that I am not always the world's greatest GM...

Last week's Continuum session was awesome, things went way better than expected. But, unfortunately, when you're standing at a peak, there's nowhere to go but down.

Last night's Continuum session was horrid. We had technical issues on both Skype and MeBeam, that made it painful trying to communicate with the players. I didn't handle that frustration well, and made some poor GMing decisions and kneejerk reactions. The plot was okay, I guess, but the presentation and pacing were dreadfull. Other than the tech issues, it was all my fault. I should have summarized instead of trying to play something out that was practically guaranteed to be mostly passive on the players part. I could have just said "You try that for more than an hour, and it proves a fruitless avenue of pursuit. You'll have to try something else." Instead, I basically narrated and played out in full detail the efforts that were certain to get them nowhere. We were on a tight clock because one of the players (my wife) had a union meeting in the morning, and my meandering presentation of the scene and poor topic-control squandered the time we had. So, to my players, I apologize, and promise next week will be much better. It certainly couldn't be much worse.

Saturday, June 5, 2010

Springing the Trap

This past Wednesday, I finally sprung a trap that had been set in a game since December. In my Continuum game, the PCs had been carrying around for quite some time a set of tickets to a stage show for a particular one-night-only performance by a mentalist (hypnotist/magician). They were fairly confident that their elder selves had been to the show. They knew for certain that an allied NPC had attended it, and that sometime thereafter that NPC had been fragged very badly.

They'd done a lot of fruning trying to find out exactly what had happened to that NPC, but essentially asked the wrong questions, and missed one really important clue because they focused on one particular "crime scene" at the expense of another one. His fragging had happened at a second location, and most of the PCs inquiries were about that other location and what had fragged him. What they never quite put together until it was too late was that he only ended up at that location because of hypnotic suggestions given him at the mentalist's show. They had solved all sorts of other mysteries tangential to and growing out from this, but hadn't quite figured out what the catalyst for the whole string of events was.

There was a long scene at the lobby of the theatre, and then they went in and sat down. The house lights went down, and the next thing they knew, they were at a third location. The entire audience were at a restaurant, and no one could remember getting there. Nor could anyone really remember anything about the mentalist's show beyond that it was really good.



Clearly, they'd been hypnotized and lead elsewhere, and made to forget the intervening hours. For a spanner, this is especially bad. Time-travelers in the Continuum setting lead regimented lives, with detailed "span logs" of where and when they've been, so they don't frag themselves. The rest of the session was spent trying to piece together what had happened, and the next session will involve a bit more of that. Amongst the great moments were:
  • One of the players had taken a really bizarre "hobby" skill of "Memorizing lists of words/names" for his PC. At the start of the restaurant scene, the only thing anyone could remember about the show was a list of memorized titles that had no context. They didn't know why he felt it was important to memorize this list, but he'd done so, and they pieced together some pretty big reveals out of it.
  • The revelation that the big bad they'd been investigating all along was probably in attendance at the event, and that they'd noticed her, mis-identified her, and chosen not to interact or interfere.
  • One of the PCs had brought along earplugs, in case there was a run-in with a hypnotist after the show, and chosen not to wear them at what was, unbeknown to her, the point of no return. Just a tiny bit more paranoid caution would have saved her.
  • The big-bad had hired a bunch of off-duty cops as extra security for the show. The PCs talked with the theatre manager, and found out how he was upset and flustered by the unexpected arrival of uniformed security. This tipped them off that something was amiss, but they failed to realize (until it was too late) the point of the cops was to provide witnesses with guns and security cameras so that it would be really difficult for time-travelers to sneak around and undo the events of the evening. It's going to be really tough to fix this now.
  • The scene at the restaurant afterwards was hilarious. The PCs were surrounded by other audience members at the show, in various stages of suggestibility, and only the PCs were suspicious enough to figure it out. They got into some pretty hilarious arguments with bit part NPCs that were adamant that they could remember the show, but could only regurgitate whatever little snippets the PCs threw at them.
It was a great session, with the players repeatedly exclaiming "if we'd only just done things a little differently!" They were really close to escaping the trap at two different points during the night, where a single simple action they almost took would have prevented all harm to them. To my player's credit, they did a great job of rolling with the punches, and accepting the consequences of the situation. I'm blessed to have such great players. You guys rock.

As George Peppard used to say, "I love it when a plan comes together."

Thursday, June 3, 2010

We aren't far from Swarm Prime afterall

My apologies for using the narcissist term "Swarm Prime" to refer to the default Continuum universe. I know it's in poor taste, and I hope my fellow spanners will forgive me.

The Continuum RPG has a list in the GM's section about events that happen in the official timeline of the Aquarian Era (2001 to 2400 CE). It's full of crazy fun stuff, like the birthdate of the first human born on Mars, and the implementation date of one world government here on earth, etc. Since there's a couple of shadowy upper Aquarians appearing as NPCs in my campaign, I decided to add that list to my GM's notes yesterday, which I keep on a secure wiki that I can access during sessions.

Typing it up, I decided I'd skip the entry for "2006: Turkey reasserts leadership of the Muslim world." That obviously didn't happen in the real world, and I was sure it wasn't likely to happen in the near future. Then I read this CNN headline this morning: Turkey emerges as Middle East leader. Shows how little I know about world politics... or perhaps just how weird the political landscape is in the wake of the Gaza flotilla raid.

Correcting for that 4-year discrepancy, and assuming the LHC's problems were some sort of cover story to keep word about the first successful temporal relocation experiment quiet, we're now perfectly in line for the Continuum's future to actually happen. Nothing is likely to conflict with that timeline again until 2019. Happy spanning!

Tuesday, June 1, 2010

A Touch Of Elders

I was thinking about the board game "A Touch Of Evil" the other day, and trying to figure out the math of the big showdown at the end of the game. In particular, I wanted to know how many dice should really be applied to any given Evil Elder, so I started looking at the math behind it all.

The conclusion I've drawn from this is that what the game presents as a strategic decision is actually more of a sucker's bet. It gives you the option of rolling very large numbers of dice against the Elders, but doing so is very likely to do exactly the opposite of what you expect it to. The whole point of targeting Elders is to reduce the damage you're taking, but if you put more than a die or two against them, you'll run a good chance of doing exactly the opposite.

Those pie charts show the results and percentages of rolling various numbers of dice against a single Evil Elder.
  • The green section represents the odds of rolling a single success - exactly the amount you want to roll.
  • The red section represents the chance of rolling two or more successes. You'd much rather roll 1 success than 2, since the extra successes are just wasted. That same die, if applied to the main Villain instead of his Evil Elder helpers, would have been 1 more step towards winning the game.
  • The black section represents the chance of total failure - not killing the Elder.
So how many dice are the right number to roll? The answer is almost always 2, with only a few exceptions. If you're feeling cautious, dropping it down to 1 die is perfectly fine.

Silver Shot is one of the main exceptions. If you're using Silver Shot, you should only roll against the Villain, and never worry about the Evil Elders at all. Silver Shot doubles the damage of 6s rolled, making even a single die on an elder has nearly a 17% chance of wasting a hit, and you're as likely to waste a hit as make a normal one. Better to pour all your Silver Shot into the Big Bad, then to waste those juicy 6's blasting a single Elder into mincemeat.

Without Silver Shot, you can always afford to put one die against each Elder with no chance of it causing any side-effects or wasted resources. One die per Elder can only help your situation, or break even, it can never hurt you.

The winning strategy is to eliminate Elders quickly, but do so while minimizing the number of wasted extra hits. Every wasted success has the potential to make the fight last one extra round - and in that extra round the Villain will roll at least 5 dice, possibly 8 or more. On average, the Villain will do more damage in one round than any single Elder does in the whole battle. As helpful as it is to pick off Elders and reduce the damage you're taking, you should never lose sight of the fact that killing the Villain is the real goal. There's no better way to reduce the damage you're taking than to just win the game.

Wasting one or two hits total during the whole fight is probably not going to cost you an extra round of fighting, however, because you don't often win by rolling exactly the number of successes needed in the final round. This is especially true in the Cooperative game, where all the players get to roll in each round, and thus as a group average somewhat higher damage per round. If a group of heroes is averaging 4 or more damage per round (and they should be) then they can afford to waste one or two points without it usually dragging the fight out into extra rounds... most of the time.

Wasting three or more hits over the course of the battle, however, runs a big risk of extending the fight by a whole extra round. So, if you were facing just 1 Evil Elder then it might be worth rolling three dice (as a single roll of 3 dice can only ever waste 2 hits). But if there's 2 or more Evil Elders, it's very risky to apply more than two dice against each of them. Resist the temptation to pile dice up on an Elder "to make sure" you take them out. You can never completely eliminate the possibility of failing against an Elder in a given roll. Both 2-die and 3-die rolls have the exact same odds (44.44%) of scoring precisely 1 hit, the only boost that's gained by rolling the 3rd die all comes in the form of wasted successes, which you don't really want. With four dice, you're actually more likely to score 2 or more successes than to score only 1. Personally, I would never roll 4 or more dice against an Elder, as the risk-to-benefit ratio is too high. Those dice are just better if applied to the Villain.