Mapping the Trail

This Cyclopean textwall is a review of the Trail of Cthulhu RPG which got way, way out of hand. I considered breaking this into several parts, but then you’d get the thing where people start commenting and responding to an earlier part when they’ve not yet read and digested the later parts, so you’re getting the whole epic in one big post.

Disdain For Derlethians

My favoured flavour of Lovecraftian RPG is and always has been Call of Cthulhu, which may partly be down to my familiarity with the system and the sheer amount of material out there for it but I think also comes down to the strength of the original design (the lack of major revisions from early editions to 6th Edition is testament to this) and the way that 7th Edition has made genuinely useful improvements to the system (along with optional systems like luck spends or pushing rolls which help dial back the swinginess of the system).

Some of the significant improvements to 7th Edition seem to be a reaction to or refinement of ideas from Trail of Cthulhu from Pelgrane Press. Trail has carved out a niche for itself as perhaps the most significant of the surprising number of “it’s Call of Cthulhu, but with a different system” games out there, and I think you can track this pre-eminence to three important factors. The first is that Pelgrane have gave Trail it a fairly substantial support line right out of the gate, whilst much of Trail‘s early run has coincided with the old regime at Chaosium being in a bit of a decline and therefore not producing so many Cthulhu products in their own right (though in fact Trail is made by arrangement with Chaosium, so they probably get their cut out of this). The second factor which made Trail stand out from the crowd comes from it being written by Ken Hite, who’s well-versed both in Lovecraftiana and in horror in general – his Nightmares of Mine is still the definitive text on horror RPGs as far as I and many others are concerned. The third factor which put Trail on the map comes from it being a Lovecraftian implementation of the GUMSHOE system by Robin Laws, which unlike most systems people try to convert Call of Cthulhu to is designed from the ground up to support investigative RPG play.

That said, I resisted trying out Trail for a long time. There is an irrational part of me which largely rejected it because it’s named after August Derleth’s absolute worst Cthulhu Mythos story, an incredibly repetitive “novel” lashed together from a set of short stories which are outright mutually contradictory – and not contradictory in a cool, evocative cosmic horror sort of way so much as a “this is a massive display of basic authorial incompetence” sort of way. Hite seems to have this enjoyment of Derleth which is weirdly uncharacteristic of someone who is even remotely discerning in terms of their reading material – tastes do vary, but there is such a thing as objectively bad writing and Derleth’s Trail is living proof of that – though Hite at least admits that his is not the majority opinion.

This Trail of Cthulhu is bad and should feel bad.

My allergy to Derleth aside, however, I also had more rational reasons to be disinterested. Whilst Trail coming out when Chaosium were experiencing a slowdown made sense and was a smart move, at the same time it still felt slightly redundant to me – and in particular, having fallen in love with the way Call of Cthulhu handles specifically Lovecraftian-themed investigations, I think I would have been more open to looking at GUMSHOE‘s differing take on investigative RPGs in the context of a different setting. In fact, whilst on checking I realise that The Esoterrorists, the first GUMSHOE RPG, actually came out two years before Trail, I think Trail must have been the first release that came onto my radar; for a long time I was under the impression that it was the first GUMSHOE game, and my impression of the system may have been coloured by that.

It’s not that Call of Cthulhu was perfect at the time Trail came out – I think 7th Edition has in fact been improved by the developers seeing what Trail has been doing and responding to it accordingly – but because I had played Call of Cthulhu a lot, run it a lot, and thought about it a lot long before Trail came out, I already had Quite Developed Ideas as to how a Mythos investigation should work and what best practice is in refereeing such investigations, so I didn’t feel the need to have Ken Hite design an entire system based around his own differing ideas about best practice, but I would have been much more willing to listen to his ideas about how investigations in other settings would work.

Misgivings About the Trail

The basic conceit of the GUMSHOE system is that investigative games are not about gathering clues, but interpreting the clues you have obtained. This argument is based on an analogy to non-interactive media, which immediately makes me want to send Ken Hite and Robin Laws to the naughty step. They are correct that few detective stories really make the challenge the process of finding the clues and put more of the emphasis on interpreting them – however, I would point out police procedurals as a common exception to this, and I would particularly note that in an interactive medium where you want to feel like you, personally, are the detective, glossing over an important aspect of the detective’s work in this fashion is counterproductive. Conan rarely fails to hit enemies; this doesn’t mean characters should automatically win combats in Dungeons & Dragons.

Moreover, in interactive media you could arguably point to an entire genre where obtaining the clues can be as important as interpreting them – namely, point-and-click adventures. Imagine a situation where a billionaire happens to have a creepy occult tome displayed in their study under a glass case. In a point-and-click adventure, obtaining the information in the book is the subject of a puzzle. In GUMSHOE – or, at least, an extreme interpretation of GUMSHOE which I do not think that the authors actually intend, but which I worry the presentation here might suggest – obtaining the book ought to be automatic and isn’t really a challenge at all.

That said, this may be a semantic issue. Although in a point-and-click adventure obtaining the clues in the first place may be the challenge, a good point-and-click adventure won’t rely on you sweeping each and every screen to find every interactable object: they will at least let you know that the clue exists to be found. In the example I give above “the billionaire has a book in his study in a glass case” is a clue in and of itself, and in the context of the adventure game using your inventory items and game commands to obtain the book is an example of the successful interpretation of other clues; likewise, in the context of the RPG, working out how to get that book may be seen as an example of correct interpretation of the information you have in the game world and a correct utilisation of that in conjunction with the skills the player characters possess.

In other words, there is not – much as Hite and Laws would like there to be – a crisp, clear distinction between “finding a clue” and “correctly interpreting a clue”; sometimes a clue will only be apparent or accessible when other clues are correctly interpreted, and “not finding a clue (yet)” may sometimes be the inevitable and unavoidable outcome of making a serious error in your interpretation of your existing clues.

Nonetheless, I can at least agree with them that interpreting information is largely where the fun is, and one of the bad habits of sloppy refereeing is to insist that a particular clue must be accessed in a particular way; in the book’s advice to Keepers they are exhorted to remember that a clue should be provided to the players if any of them come up with a credible route of getting to it, even if that isn’t the route the scenario expected. Hite correctly recognises that this is important not just to avoiding investigations from being unduly cut off early, but also as a means of preventing railroading: insisting that clues must be obtained in a particular order in a particular way is perhaps the most infuriating way you can railroad players in an investigative game.

(I have more issues with the way that Hite talks about the crucial clue, the one you have to give the players, being the one which will get them to “the next scene”. A next scene, certainly. But if you truly want to avoid railroading you must never have a firm idea in your mind of what the next scene is. If you do, then you’re already laying down the rails – which is no bad thing if that is what your group signed up for, but it is a bad thing if you are doing it but still bullshitting yourself that you aren’t railroading. Incidentally, I find the use of the term “scenes”, an infuriating terminology which is over-reliant on drawing a comparison between interactive and non-interactive mediums. In particular, I don’t see anything wrong in a tabletop context with significant clues being obtained not through staged, roleplayed-through scenes, but through downtime actions, though to be fair I may be interpreting “scene” as meaning something a bit more involved and specific than the way they are using it here.)

Another point where I tend to part ways with the GUMSHOE philosophy for investigation-focused games is the assumption that failure to find clues at a particular point in time is necessarily a bad thing. I agree that the clues should eventually come into the hands of the players in order to avoid a frustrating experience, but I don’t agree that successful use of their abilities needs to be the only way they get those clues.

Here is one of those “best practice in scenario design” points. In general, if a situation can only possibly change or develop or move forward as a result of player character action, then I would say that’s a bad scenario – it’s a dull, static situation which risks devolving into a dull, static stalemate. Best practice is to have a plan for what happens next if the player characters fail to make headway. The classic example of this is the serial killer investigation: if you don’t piece together the clues, the killer kills again, both generating a set of new clues relating to the new killing and providing another opportunity to obtain the earlier clues needed to suss out their identity.

Likewise, you can use “failing forward” to good advantage here. Take the example of that book in the glass case in the dodgy billionaire’s study: let’s say the player characters have failed really badly in an attempt to break in and obtain it, and the billionaire has been spooked. In principle, one outcome of this could be that the billionaire is so shaken by the possibility of being caught that they turtle up completely – but that’s not an interesting outcome, and nor is it the only outcome which could possibly come about, so a skilled referee would avoid this. Instead, if it turns out the billionaire is a cultist, they may then break out the book and start using it against the interlopers – especially dramatic if they’ve taken someone prisoner! – so at least the PCs can start inferring what is in there from the horribleness that is called up on their tail. Alternatively, if they are genuinely innocent, the billionaire may decide to hand the book over to some other institution who can protect it better – giving the player characters a potential opportunity to get access from another angle.

Call of Cthulhu 7th Edition’s revision of the Idea roll, by the way, is a genius way of both working with this “failing forward” idea and providing, in one simple mechanic, an elegant solution to the “investigation stops due to running out of clues” problem that the whole edifice of GUMSHOE is erected to correct (and, to my eyes, risks overcorrecting). It works like this: if the player group decides it is well and truly stuck, it can ask to make an Idea roll. The difficulty of the roll depends on how effectively they have made use of the information they have already obtained; if they have outright just ignored a fact that has been placed in their lap, it will be harder than if they have genuinely exhausted all the avenues open to them.

So long as the lightbulb goes on, it doesn’t actually matter what it’s illuminating.

If they pass the Idea roll, they get a clue to keep things moving, with the referee deciding what fortuitous circumstances deliver them the clue. (“OK, so you all get the evidence together and sit down and plot it on a map in your living room, and you realise that all of the mysterious deaths happened within 10 miles of the Miller farm – which was one of the addresses in the dead Department of Agriculture inspector’s notebook.”) If they fail the Idea roll, they still get the clue – but the referee has licence to come up with whatever awful circumstances leads them to it they can dream up. (“OK, so you all get the evidence together and sit down and plot it on a map in your living room… and suddenly the front windows are shattered to pieces under a hail of shotgun blasts! You see Old Man Miller’s pickup truck out in the front, with several of his farmhands pumping their shotguns and aiming them at you…”)

I particularly like this because it effectively allows the players to mash a “We Wait For Further Developments” button – it acknowledges that player action doesn’t need to be the only thing driving the game world forwards, and that in particular once player avenues for action are exhausted, waiting and seeing what happens next is a legitimate strategy – although a bit of a gamble.

Now, of course I agree that a situation where the player characters fail to find any of the clues, or enough of them to stave off frustration, is bad. But that doesn’t mean they need to find all of the clues – just enough to suit the particular play group’s tastes. Call of Cthulhu 7th Edition’s Luck-spending mechanic is quite good for this, in that it means that players can succeed at a roll if they really, absolutely decide that it is important to them that they do so, so if that mechanic is in play but the PCs still failed to find any useful clues then that’s probably because the players are fine with failure (because if they weren’t fine, they’d have spent some damn Luck to earn some damn clues).

To be fair, GUMSHOE does require the players to get themselves into a position where they could conceivably obtain the clue before they get the clue, which does reiterate the important point that players should be given full credit for their correct conclusions: if they decide Ricki Enpeecee is worth talking to in order to establish their relevance, and Ricki Enpeecee is genuinely relevant, they ought to derive some benefit from that.

On the other hand, if the players decide to go talk to Ricki Enpeecee because they think Ricki is a friendly ally, and it’s actually the case that Ricki is out to deceive them, then they’ve specifically not got it right – they are going to see Ricki for the wrong reasons, and I think it would be entirely fair for Ricki to at least have a chance to successfully deceive them, denying them the clues Ricki holds (at least for that scene) because they’ve come at it from the wrong angle. Here’s where the central axiom of GUMSHOE contradicts itself: like I said earlier, sometimes the acquisition of future clues may not be reliant simply on innate player character ability, but through the correct interpretation of prior clues, and that if you have a game where success or failure depends on the correct interpretation of clues, losing out on an opportunity to get clues or having the PCs’ adversaries successfully trick them is in my eyes a perfectly legitimate result of a failure to correctly interpret prior clues.

“But wait,” you may say, “GUMSHOE does at least require the PCs to use an appropriate investigative ability at the right time in order to get the clue, so if the players choose not to use an appropriate ability to test what Ricki is saying then isn’t that on them?” It is true that GUMSHOE requires players to invoke a relevant ability to gather clues – except when it doesn’t. Specifically, there’s a rule that if a clue is “Innocuous” then there is no need to declare ability use – whoever’s got the best relevant ability spots it. I suspect, then, that my ability to get along with GUMSHOE highly depends on where the referee sets the bar for innocuousness; set narrowly enough, Ricki Enpeecee can definitely trick player characters who think Ricki is a trustworthy friend if none of them bother to use their resources to use an appropriate ability in assessing Ricki’s motives or fact-checking their story, and my concerns are thus settled (since then even GUMSHOE denies you clues if you don’t look for them at the right time), but set too broadly and it can be impossible for Ricki to fool the PCs because the GM will decide that “Ricki is with the bad guys” is Innocuous enough to qualify.

There does seem to be some awareness in the system design that it’s sometimes useful for NPCs to be able to lie to players. Given the genres that GUMSHOE has been used for – mainly horror and supernatural espionage – it would arguably be actively corrosive to an awful lot of what people may wish to do with those genres to give the PCs an infallible lie detector power! However, what GUMSHOE does, at least in Trail, is that it gives the PCs an infallible lie detector power and then breaks it in a way which flies in the face of the underlying philosophy of the system. The power in question is the Assess Honesty investigative ability, which like any other investigative ability just automatically works when invoked (though you can spend points from its associated pool to get extra bits). As described in the core book, the simple, baseline use of the ability lets you suss out when someone is lying to you – except when GM fiat decides otherwise.

On the one hand, that does allow for NPCs to effectively lie to the PCs. Great! On the other hand, the fiat thing seems to me to completely undermine the point of GUMSHOE. GUMSHOE works on the basis that your investigative abilities straight-up work when they are used in an applicable situation. With Assess Honesty, though, your ability can end up failing to work, even in a situation where you actually correctly conclude that it would be a good idea to use it, because it isn’t convenient for the plot for it to work at that time. (In fact, by definition every time it fails to work will be an instance when using it was the right call.)

And let’s be clear, we aren’t talking about some obscure skill which most characters will never even bother to invest in here, we’re talking about a major tentpole of the investigative process, and one which interfaces directly with a major feature of the sort of genres GUMSHOE tends to get used for. If you are playing a horror game or a spy game and you aren’t constantly using Assess Honesty, then the game has utterly failed to establish the air of trust-no-one paranoia that those genres absolutely scream out for.

It seems to me that there is a simple solution here, which is to say that “this NPC is lying to us” is not a clue but a conclusion you reach from the correct interpretation of other clues – remember, that’s the bit which is supposed to be the challenging part of the investigative process in GUMSHOE, right? This would fit the fiction much better: if you don’t have any clues to support the idea that someone is lying, and you don’t manage to use your knowledge-based investigative abilities to spot inconsistencies in what they are saying, then you don’t really have any basis to firmly conclude that they are lying. All you have is a suspicion based on their behaviour, and suspicion is not evidence.

Assess Honesty would make most episodes of this unworkable.

Another way in which the game seems to undermine its own philosophy is in point spends from investigative abilities to get additional information. This feels like an attempt to make them feel more game-y by adding a resource management aspect to them, except it’s completely toothless because even if you completely exhaust every pool all your investigative abilities will still function to provide you with the essential information you require.

On top of that, it feels like locking some clues away behind a point spend seems to sit awkwardly with the basic philosophy of the system. If these clues are actually important, you ought to be getting them for free, but if they’re not that important and just provide added flavour, why make you spend points for them? If the real challenge is in the interpretation of the clues, then paying points for extra pointers and clarification rather seems to short-circuit that challenge. It feels like Laws really wanted to just make the investigative abilities a checklist where you either have the ability or you don’t, but chickened out because he decided gamers needed numbers on their character sheets, and then had to scramble to find something to do with the numbers.

Furthermore, after going all-out to make a system that avoids the blunder of locking the players out of progressing because they didn’t find a clue, this subsystem seems to let that risk in again through the back door. Sure, you’re supposed to make the essential clues free, but from a referee’s eye view it can be hard to see what clues are truly essential and what can be happily missed. Perhaps the next step looks obvious to you without the extra detail, but from the players’ perspective the “essential” clues may be completely oblique.

If nothing is hidden behind a point spend, this is largely mitigated. If something is and people don’t actually have the points to spend on it, this may lead to precisely the failure state GUMSHOE is set up to avoid.

To be fair, almost all of the above issues with GUMSHOE are mitigated by the fact that a GUMSHOE game is going to be run by a human referee who can be reasoned with, and who – if skilled – will doubtless avoid absurd, game-wrecking, or boring outcomes. In the book Hite proclaims how GUMSHOE isn’t that railroady because all of the playtesters were given the same scenario but ended up with radically different outcomes, but I would say that those outcomes are enabled precisely because referees are a) not robots and b) likely to be experienced gamers who, by dint of such, will interpret and/or ignore the GUMSHOE philosophy accordingly in order to give a satisfying play experience, rather than implementing it to a fundamentalist extent (though I think the rhetoric that GUMSHOE presents itself with may inadvertently push people to a fundamentalist interpretation, especially if they have less prior experience).

In short, GUMSHOE as implemented in Trail of Cthulhu offers an awful lot of good ideas, but I feel it is an overcorrection – a fundamentalist adoption of an opposing position to the sloppy habits of the past, which despite lending itself nicely to some types of investigative games is not suitable or appropriate for all of them. At the heart of it is a distaste for the whole “one missed roll derails the whole investigation” thing, but I don’t actually think that is a system problem – that’s a refereeing problem. Insisting that a clue may only be acquired through a particular direction, or that a scenario can only progress if a particular clue is found at a particular point in time, is just plain poor refereeing. (It is also poor scenario design, but a good referee will recognise that and work around the limitations of a designed scenario; a bad one won’t.)

And the thing is, I don’t think you solve refereeing problems with system measures, you solve them by educating referees and encouraging best practice – which, to be fair, Trail is actually pretty good at in the advice it provides. The reason I say this is that the traditional tabletop RPG format is sufficiently referee-dependent that a poor enough referee will sooner or later drop the ball no matter how good the system they are using is, whereas a good referee will either work magic with an average or poor system or, at the very least, be sensible enough to realise that the system is a non-starter and switch to something better.

In my view, if you are a sufficiently incompetent referee that you make a habit of pulling the “oh no, you didn’t find this one specific clue in this one specific way, I guess that means the investigation completely stalls and there’s nothing we can do about it” thing, then you are probably a clumsy enough referee that you’ll find some way to wreck GUMSHOE with it. Conversely, if you have enough wit to realise that a tabletop RPG isn’t a point-and-click adventure and you don’t have to insist that the players click on the correct pixel in order to progress, then you will probably run a decent enough game. Even if you are not entirely comfortable with the improvisation involved in allowing players to advance an investigation through means and avenues you didn’t plan, by simply allowing them to do it in the first place you are already streets ahead of the straw Keeper complained of by Trail – not to mention legions of lacklustre published scenarios that fall into One Correct Route fallacy – and by practicing these improvisation skills you will rapidly get better at them.

Not only will this allow you to avoid the pitfall most of the time, but also it will allow you to deliver the actual unique selling point of tabletop RPGs – the flexibility of a human referee compared to a computer that can only accommodate a limited set of player decisions. And if you’re not interested in providing that, I question whether the tabletop RPG format is even the right format for whatever it is you want to accomplish. Wanting to referee without wanting to improvise is like wanting to learn a foreign language without picking up any vocabulary – you’re abandoning the one skill which means there’s any point to the exercise in the first place.

Actually Walking the Trail

So, therein sits the summation of my doubts about GUMSHOE before I sat down to play some Trail in late February. As I and Dan discovered, the road bumps we actually encountered were not the ones we were expecting.

For one thing, GUMSHOE as implemented in Trail doesn’t quite attain the clear distinction between investigative abilities and non-investigative abilities that GUMSHOE‘s underlying philosophy seems to require. In fact, looking at the character sheet I note that abilities are tagged as being usable both as general abilities and as investigative abilities, which raises a number of problems. For one thing, because the nature of the system means that investigative abilities and non-investigative abilities tend to be bought on different scales (4 points is quite a lot to spend on an investigative ability and is probably excessive, whereas for a general ability it’s at about a third of the ability cap for “Purist” games), these particular abilities are likely to have an excess of points in their pool compared to other investigative abilities. At the same time, because they can be used in two different ways, these abilities actually have two separate demands on their pools, so if you want to be competent at using them as general abilities but also make use of them as investigative abilities you need to buy them up more. Worse, you have the weird effect where if you do an investigative point-spend on the ability you suddenly become worse at using it as a general ability, because your reserves of competence have somehow been drained by you spotting something.

(Incidentally, Laws seems to have missed the point about the differing scales when designing the experience system for GUMSHOE. 1 point in a brand new investigative ability is much, much more valuable than 1 point in a brand new general ability, but they cost the same number of experience points to buy – despite the fact that the character creation rules imply that 1 point in an investigative ability is worth 3 in a general ability.)

The whole spending-points-from-pools thing is, we found, genuinely tricky to find an in-character rationale for. (You can do it, but the rationales we found felt like grasping at straws.) Run out of points in a pool and suddenly you are no better at the matter at hand than someone who has only an absolutely minimal, baseline investment in the ability in question. This makes somewhat more sense if you look at it more as a spotlight-rationing measure, but then again it clearly isn’t exclusively this – your health, sanity, and stability pools all have implications that go beyond “this is how much of this stuff you can reliably take the lead on before it’s time to let someone else shine”.

Another thing about point spends: the way general abilities work is that the referee sets a difficulty number for a task, which will generally be 2 to 8, and you roll a D6 and try to equal or beat that. You can spend points from your pool to add to the roll, but GUMSHOE – at least as implemented in Trail – does the infuriating thing that Mahna Mahna does of forcing you to declare how many points you are spending before you roll, and as Dan showed for Mahna Mahna this means that for a good proportion of the time your point spend will actually be partially or entirely wasted.

Of course, this is trivial to house rule, but I’m not talking about house rules here, I’m talking about the actual words in the book. I genuinely don’t get why Trail does this – if anything, the mild competence boost that arises from allowing people to choose a point spend after the fact would seem perfect for Pulp-oriented games. When we were discussing this Dan pointed out that some people really don’t like systems where you can effectively make a metagame decision to pass a roll, but with pay-before you still can do exactly that if your pool has enough points remaining. Pay-before is still a pay-to-succeed system, it’s just a markedly less efficient one than pay-after and a much, much more irritating one.

Another issue we noticed during character generation was the weird way Trail handles Credit Rating. To be fair, this is a historically weak aspect of Call of Cthulhu which Trail has inherited – it has been improved and clarified in 7th Edition, particularly by borrowing Trail‘s idea of specifying a particular range that Credit Rating should fall in for each occupation and better defining what it actually represents. Nonetheless, several of us found it odd that Credit Rating is an investigative ability.

I can see the thinking behind it – that way, it lets you use it to unlock clues that people wouldn’t share with discreditable sorts. However, this sits oddly with the way GUMSHOE actually works, because whilst I can see knowledge-based stuff or sensory-based stuff working in a binary way (either you know something/spot something or you don’t), and whilst I can even extend that to conversational approaches (either your banter or interrogation or hard sell or persuasion works or it doesn’t), I don’t see Credit Rating as working that way, and in particular I don’t see that it’s the sort of thing where, so long as you have a bit of it, you can quite happily succeed at getting information through it whenever it is important for you to do so. In terms of handing out clues, if Lord Snooty has a clue which is supposed to be obtainable through the use of Credit Rating, it should be just as possible for a party of Credit Rating 1 characters who are all barely a rung up from the dregs of society to get it as it is for a party of high Credit Rating aristocrats. That’s not how social class works.

A humble Credit Rating 1 clerk is just as able to get a basic Credit Rating-related clue as a billionaire.

Credit Rating in Trail would, conversely, work much better as a general ability with the general focus of “buying expensive stuff” or “persuading people that you belong somewhere swanky”. Both of those things are not clue-obtaining activities intrinsically (the clue-gathering functions of Credit Rating could quite happily be covered by Flattery or something), both are the sort of thing we should probably be rolling on.

(Indeed, with the exception of Cop Talk and Interrogation the social skills all seem to be on one level or another something which perhaps ought to be general abilities capable of being used as investigative abilities, because with things like Flattery and Intimidation and Credit Rating the thing you are trying to do may well be not “get information out of someone” so much as “get someone to do something”, which falls well outside the category of “getting clues” where the auto-successes are supposed to lie and well into the category of “non-investigative stuff” which is meant to be where all the dice-rolling is.)

Speaking of the sort of thing we are meant to be rolling on, one thing which jumps out at you when you look at the Trail character sheet is just how many options for investigative abilities there are compared to how few general abilities there are. (Incidentally, the character sheet design is one of the more minor gripes I have with Trail; for one thing, I think it would better communicate to new players the strong distinction between investigative and non-investigative abilities by having them in different boxes, rather than the general abilities looking like just another category on the list, and for another it would better communicate the idea of stat pools to new players if the skill entries had separate little boxes for the stat and the current pool total. These are small things, but I genuinely think a well-designed character sheet is an incredibly useful tool for teaching a game to someone or figuring a game out for yourself and these shortcomings impede that.)

But the thing about the choice of investigative abilities and non-investigative abilities is this: the philosophy of GUMSHOE is that it’s not OK to fail to progress in an investigation because you missed some clues, so you make investigative abilities work on an auto-success basis. The logical conclusion of that thought process is that it’s fine for the characters to fail at stuff that has general abilities associated with it.

Taken to a ridiculous extreme that only a truly awful referee would contemplate, this would imply that it’s fine for an investigation to come to a literal crashing halt when someone messes up a Piloting roll and the party dies in a plane crash. Less absurdly, it seems that GUMSHOE is comfortable with the possibility of a TPK happening because the characters failed to evade or run away from the danger and were bested in combat. Fine if that’s a climactic battle, but what if they got really unlucky in a scuffle with cultists in session 1? Why shouldn’t you get to use Fleeing as a quasi-investigative ability to book it out of there if a fight clearly isn’t going your way? It doesn’t give you a clue, but it may be essential to you reaching the next scene.

(For that matter, why is Sense Trouble a general ability? Isn’t “shit is about to go down” an important clue which leads naturally into the next scene, in which shit actually goes down?)

In short, why is it not OK for an investigation to end or fail in one sudden, arbitrary, and/or anticlimactic way, but OK for it to end or fail in a swathe of different sudden, arbitrary, and/or anticlimactic ways?

You could, of course, very reasonably defend this design decision by pointing out that just because such a potential game-ending situation can arise as a result of following the system thoughtlessly, most skilled referees will be perfectly capable of avoiding this by either fudging the situation or by coming up with some alternate failure scenario which doesn’t amount to a full-blooded “game over”. That’s very true! But then, if it’s OK to soften and manage the outcome of failed general ability rolls like this, why not just do exactly the same with failed investigation rolls in non-GUMSHOE systems? The bottom line is that once an RPG system starts partitioning off one sphere of activity as Too Big To Fail, it becomes much harder to justify not similarly partitioning off other spheres of activity in exactly the same way.

Beauty Spots On the Trail

I’m criticising GUMSHOE a lot, but I actually like it as a system – provided you don’t adopt it solely as a solution to the problem it claims to solve, and instead use it with an eye to the benefits it does provide and the problems it actually does solve quite elegantly. Although I think the GUMSHOE approach of giving out clues automatically is overkill for the problem it claims it’s solving, I do think it’s a very useful way to make sure a game keeps up a brisk pace. In some investigative games I quite enjoy taking the time to smell the roses, as it were – whether this is in the form of the characters taking multiple angles of attack before they finally get the clue they were aiming for, or having uptime sequences (or “scenes”, if you are bestial enough to use that terminology) which have nothing to do with acquiring a particular clue or following up on a different clue and focus more on establishing character or atmosphere or exploring character relationships.

In other circumstances, I’d prefer to just get on with things, and GUMSHOE prompts you to do just that – you must provide the core clue in a scene, that clue must lead on to the next scene, and that’s all there is to it. This is particularly good for games where you want to have an investigative element, but you don’t want to spending an awful lot of time on investigation – for instance, an action-packed spy game where a certain amount of investigation is necessary to track down Blofeld but the main thrust of the scenario is playing through invading his base and shutting down his missile silo. The non-investigative component of the system is also nicely compact; if you want riveting tactical combat with a richness approaching D&D 4E or Savage Worlds, it won’t deliver, but if you want a quick system to provide simple success/fail results it’s fine on that account.

The upshot of that is that a GUMSHOE game will progress at a good clip, which is perfect for the context we’re using it in – the group in question meets once a month at most so a session spent zooming off on some random sideshow or doing pure character development would be more frustrating than it would be in a game that ran more regularly.

Another benefit of the system was pointed out by Dan, and is an aspect which is arguably a genuine improvement over Call of Cthulhu which is that it avoids the issue where you paid a lot of points to an expert in some oblique skill, and the one time it comes up in the campaign you fail at it – and perhaps, to add insult to injury, some dork who only has it at the base level succeeds at the roll. Some might argue that it’s the player’s problem to find uses for their skills, and if they invest a lot in something which clearly won’t come up very much then that is the rod they chose for their own back, but this is something to perhaps discuss beforehand so if a player like Dan would be really annoyed by failing at an obscure skill they invested in, they can adapt appropriately. (For instance, you could simply avoid buying niche skills and go for a character whose skills come up all the time, so even if you have the odd failure you’ve probably got a success coming your way sooner or later.)

Trail neatly makes this a non-issue by making most of the “obscure bits of knowledge” skills investigative skills – so when the thing you have invested in being an expert on comes up, it’s your time to shine and you can count on being effectual. As Dan has pointed out to me, one of the nice things about the split between investigative and non-investigative skills is that the system seems to understand that calling up expert knowledge is a different process from swinging a punch, and isn’t the sort of succeed/fail binary system its often modelled as. More broadly, the investigative skills system, whilst I don’t think it’s necessary​ for solving the problem it claims to want to solve, is quite good at ensuring that investigative spotlight time is distributed among party members, so everyone gets to feel like they contributed to the gathering of clues.

Hang On, I’m Just Going To Improve GUMSHOE 100% Then I’ll Get Back To the Review

Having all these different skills is appropriate for Trail – both in terms of feeling like Call of Cthulhu and for the purpose of imitating that “piecing together of dissociated knowledge” that Lovecraft promised us would yield only sunshine and rainbows and kittens. At the same time, for non-Lovecraftian purposes – especially in a game where the characters are all spies or police officers or something, where their investigative skills are likely to be overlap a lot rather than being spread out disparately, I think you can radically simplify GUMSHOE simply by junking distinct investigative skills altogether.

Instead, have an Investigative Pool which starts out at, say, 10 points. To obtain a basic clue, you pay a point and give a plausible-sounding explanation as to why your character might spot a clue in light of the established facts about them and the scenario (what Dungeon World calls “the fiction”). Bear in mind that under most circumstances you don’t need to be an expert in a subject to spot a clue – just conversant enough with it to notice when something is unusual. (Remember, the stated GUMSHOE philosophy is that you auto-succeed at finding clues, you very specifically do not succeed automatically at interpreting them, so spotting that something’s weird without being able to give an expert diagnosis of exactly what is weird about it and what that may imply is fine for clue-gathering purposes.) Extra point spends can be used as in GUMSHOE to get additional information.

If the GM needs to drop an Inconspicuous clue in the lap of the players, they give it to whichever player character who has the highest current Pool total out of those characters who could, conceivably, spot the clue.

When your pool hits zero, you can spot basic clues for free, but you can only put yourself forwards to do so if everyone who still has points in their pool has passed on the clue.

If nobody in the party can conceivably spot the crucial clue (because the referee fucked up and didn’t pay attention to the character backgrounds when seeding the clues, or didn’t pay attention to the crucial clues when approving character backgrounds), anyone whose pool is not at zero can immediately spend all their remaining points in it to establish a new fact about their character (a hitherto-undeclared hobby, past career, acquaintance or whatever) that allows them to get the clue. If multiple people volunteer then whoever has the most points in their pool gets to do it, ties resolved by die roll.

Your pools refresh to 10 at the start of a new game session, or when none of the PCs are left with a pool of above 3 points.

Viola, I just radically simplified your character sheet, saved you a whole bunch of time in character generation, and got the referee out of the trap of having to assign clues to appropriate investigative skill types and keeping track of who has what skill, all with a system which still guarantees critical clue delivery and still ensures that the spotlight gets nicely spread around. You’re welcome.

Stuff That’s Actually Nothing To Do With the Investigative Skills Thing, Honest

Another thing I like about GUMSHOE is the perennial investigative RPG problem it actually does solve – namely, the “why the fuck am I not leaving this to the professionals?” issue. This comes up in games where the player characters are ordinary people drawn into an investigation, rather than representatives of law enforcement or espionage agencies or other people who actually have the duty and the authority to look into criminal or dangerous matters – thus, it often comes up in Call of Cthulhu and other such games. There will come a time in an investigator’s career that a purely rational actor would either just walk away from an investigation or try to fob off all responsibility for continuing it to the authorities. (“The authorities won’t believe you” only goes so far, especially if the PCs, for example, have obtained actual proof that cultists are involved in blatant criminal activity.) It is essential for the campaign to continue – or at least, for it to continue with that set of PCs rather than shifting focus to some freshly-generated police officer characters – that characters not succumb to this onslaught of common sense; it’s fine for them to try and enlist the aid of the cops, but they can’t just walk away from the investigation entirely.

Now, of course players are likely fully aware of this, so as a player you can bite the bullet and just have your character keep pressing on with the investigation even if they have no motive to do so, but this is an outcome I think most would be unhappy with on some level. If you like to look at RPGs as stories, it’s narratively unsatisfying; if like me you prefer to see them as immersive explorations of scenarios, it’s immersion-shattering.

It’s also, unlike the clue thing, not something which is easy to avoid simply by tightening up your skills as a referee or player. Whereas clue bottlenecking ultimately comes down to a bad decision on the part of the referee (since they have to either plan the scenario that way or run the published adventure unaltered and unimaginatively), here the issue could arise either from the player, or the referee, or from the interaction between the two. It could be that the player goofed and designed a character with no good reason to investigate stuff even when doing so has detrimental consequences for them (of course, the referee goofed somewhat here by not pressing them on the subject, but perhaps the referee trusted the player to make sensible decisions in character design). It could be that the player absolutely provided a suitable motive, and they communicated that to the referee who understood where they were coming from perfectly well, but the referee goofed and didn’t engage with that motive in the slightest, so the character is simply never provided with the stimulus they were designed to respond to. Or it could be that the player has one understanding of the motivation and the referee has a different, so the referee does try in good faith to work it in but the player doesn’t recognise that as relevant to their character’s interests.

This Trail of Cthulhu is good and should feel good.

GUMSHOE solves this neatly by bringing in the concept of Drives. At character creation you pick a Drive, a reason for your character to keep pushing an investigation when more reasonable people would have backed down. In Trail,  at least (and maybe in other iterations of the system) this plays into the sanity mechanics by giving you a little Stability bonus when you indulge your Drive, and by enforcing a hit to your Stability if you flat-out refuse to rise to the bait.

This doesn’t completely solve the issue, of course, but it goes a long way towards doing so. For one thing, by making your Drive an explicit part of character creation, the system ensures that the player will provide a suitable motivation for their character, and by basing it off a series of options on a list the player and referee have a common definition of the Drive they can look to, which won’t eliminate miscommunications about it but can help dial them back. The onus is still on the referee to remember and enforce that Drive, but that’s not so much of a burden, especially if several player characters have the same Drive. In fact, Drive seems to me to be an area where enforcement of niche protection and diversification amongst the party isn’t that necessary – and in fact, there can be advantages to having all the PCs possess the same Drive

(On the subject of sanity, the way Trail divides Sanity – the extent to which you are able to engage with and invest importance in ordinary human concerns – and Stability, which is your ability to retain self-control and resist panic in the moment, is quite nice.)

In terms of Trail’s own discussion of the Mythos, it’s pretty good as you would expect from Ken Hite; he is perverse enough to actually enjoy August Derleth’s work, though like I said he has sufficient self-awareness (and, hopefully, shame) to note that this is a minority opinion, and he doesn’t push Derleth’s One True Interpretation as canon. In fact, he doesn’t push anything as canon, offering several possible interpretations for each of the Great Old Ones and alien races detailed here, ranging from well-aired interpretations to excitingly novel ones. One thing he doesn’t do is offer stats for the Great Old Ones, which I kind of miss – if you want to take the view that they are transcendently powerful entities who will squish everyone by default if they show up, it is easy to just ignore provided stats, but if you want to have a very materialistic, pessimistic take on them and assume that they are big aliens who may be physically powerful but are no more important to the universe than mere humans, having stats is a nice way to make that point.

One thing Hite has a good understanding of is that there are “Pulp” and “Purist” strands in the fandom, with the former not being so keen on their horror gaming being actually horrific and keen on having characters who are made of sterner stuff than the average Joe save the day by punching Nazis on the jaw, and the former preferring actual horror and the dread certainty that the day cannot be saved, only postponed. If it sounds like I caricature a little, it’s because these are, as Hite points out, extremes, and most people’s tastes tend to land somewhere in the middle.

Trail offers pointers on how to adapt the rules here and there to take them in a more Purist or Pulp direction as desired, with Hite showing real inventiveness in finding small changes that can have a huge effect on a game. For instance, he suggests that a Pulpy option would be to let the players know the target numbers for their skill rolls, whilst a Purist one would be to keep them in the dark – in the latter case, of course, not only are you likely to waste more pool points because you don’t know the target bumber, but you may never realise you wasted them, which might be interesting or might be infuruating depending on your temperament.

The big problem with the implementation of this is in the little logos for different play styles. Hite here is following the lead of Paranoia XP , which described “Straight”, “Classic” and “Zap” approaches to the game and used appropriate logos for each – a pair of smoking boots for the trigger-happy Zap approach, the iconic Computer-screen-with-eye-on-it for the Classic style, and a somewhat creepy disembodied Computer eye for the ha-ha-only-serious Straight style. These icons worked because they very quickly flagged what approach they denoted.

For Trail, however, the icons chosen are completely abstract designs which do not denote either Pulp or Purist or really any intelligible concept whatsoever. As bits of weird art, they’re good, as signals of play style they’re rubbish. Just use a gun to denote Pulp and a dusty old book for Purist, Ken, that instantly gets across the distinctions.

This issue aside, the Purist/Pulp implementation here is pretty good. Whilst Call of Cthulhu 7th has taken the general idea onboard in its discussion of approaches to the game, at the same timr I think the baseline Call system tends to cater better to Purist play; the options wheeled out in the Pulp Cthulhu supplement do a great job of reskinning the system for Pulp play, but then you’re running with a core rulebook plus a supplement, whereas Trail is more flexible in this regard right out of the core book.

One area where Trail is less adaptable is in time period; whereas 7th Edition Call offers support for the 1920s and modern day straight out of the rulebook, Trail by default only caters to the 1930s, though in principle it’d be simple enough to tweak the skills and equipment lists appropriately for other eras, especially if you have a GUMSHOE game set in the era you want already. Where the 1930s setting really shines through is in the gorgeous sepia-toned presentation of the rulebook, which is so nice that I think Chaosium’s recent tightening-up of its production standards must have been partly inspired by comparing their material to Trail and noting its shortcomings.

So, how is it to actually play? There’s certainly a bit of a learning curve involved, especially if you are more used to Call; our Keeper sometimes called for rolls on investigative skills by mistake, though in part I think this comes down to the choice of what is and what isn’t an investigative skill not being entirely intuitive, and as players we tended to be a little unsure of whether or not we wanted to spend ability pool points in any particular situation, but otherwise our first session was reasonably smooth and I expect things will get smoother once we are all more used to the system. We are playing the Dreamhounds of Paris campaign, which casts the player characters as part of the original Surrealist movement getting entangled in Dreamlands nonsense, and I think the system adapts itself well to that sort of off-beat concept.

On that subject, interestingly different ideas for published campaigns seems to be a strength of the Trail line – there’s also Bookhounds of London, in which the PCs work in the antique book trade and get into buying and selling Mythos tomes – and whilst Pelgrane are nice enough to offer Call of Cthulhu stats for many of their published campaigns, if you are using a prewritten thing it’s probably as a labour-saving approach, so why give yourself extra work if you can just use the Trail system instead?

Ceci n’est pas un masque de Nyarlathotep.

Indeed, Trail to me seems strongest and most attractive when it’s used for something a bit offbeat and different from the usual Call of Cthulhu campaign; I listened to the Innsmouth House Players’ playthrough of Eternal Lies, which is designed as a globetrotting campaign in the vein of Masks of Nyarlathotep for Trail; the players actually used Call of Cthulhu for it rather than Trail, but they did note in their autopsy of the campaign at the end that there were several incidents in it which, whilst entirely suitable for a story in a passive medium like a novel or a film, are just kind of annoying and frustrating in an interactive medium, and it seems to me that they’d be similarly annoying in Trail.

When Should You Follow the Trail?

The rhetoric that GUMSHOE surrounds itself with rubs me the wrong way, and I think that as a solution to the problem it claims to solve it is clumsy overkill. That said, as a system it really ain’t that bad – it’s fast, simple, and it does solve some other problem, a few of which it apparently doesn’t even realise it’s solved.

I would be inclined to use it for a game where a rapid, fast pace is wanted, so the fact that you have to give the players the clue about Ricki Enpeecee even if they go to Ricki for completely the wrong reasons isn’t so much of an issue (because simply giving up the clue at least helps keep the pace up, which is a fair sacrifice). In this sense it’s actually a good fit for the group I’m going to be playing with, because we tend to only meet up once a month (if that) and so keeping things moving will be very useful. I’m also aware of someone who uses it for short, intense weekenders, where an entire mini-campaign is played out over a couple of days, and I can see how it would be very useful for that.

I also think it’s useful for a game based around what the Forge called Participationism, in which the game is to one extent or another railroaded and is overtly flagged as such, and everyone is 100% cool with the fact that one way or another​ you are going to hit the plot points and reach the climax that the referee planned in advance. As I said earlier, I think Hite doesn’t recognise that if you plan on the basis that “you must give the PCs the clue that leads to the next scene” then that is a form of railroading, because by definition you will have decided what the next scene is going to be, but I also think he protests too much about the subject of railroading to begin with. Railroading is not, at the end of the day, actually a problem: unwanted railroading is.

If your players want to get on the train, then obstructing that because someone told you Thou Shalt Not Railroad and you were fool enough to take it as holy writ makes you just as much of a game-wrecking asshole as someone who absolutely insists on railroading despite their players expressing their clear disdain for it. If your players want the train ride, GUMSHOE will certainly do a good job of keeping it on the track. (Unless, of course, they fail at interpreting the clues you have provided them! At which point you either have to let them fail or throw the GUMSHOE philosophy out of the window entirely and let them auto-succeed at that point of interpretation in order to allow the game to progress.)

Conversely, in a game where a slower and more careful investigative process is acceptable, or a sandbox campaign where “Well, we’ve failed to get to the bottom of this, so let’s go and chase up one of the other cases we’ve been pursuing and see if any further developments happen here later on” is an acceptable outcome of a strand of sessions, or a game of conspiracy and paranoia where neither players or characters necessarily know whether the outcome of an investigation was the “real” truth or the result of them getting successfully duped by others and there isn’t necessarily a singular truth to discover beneath a morass of clues in the first place… well, there I think GUMSHOE is less appropriate. It is not the only model of investigative game, though it is a useful enough one that I am glad it has been added to the ecosystem.

As for Cthulhu-related purposes, in general, I think that the 7th edition of Call of Cthulhu – as I outlined above – actually cleverly implements some of the lessons of GUMSHOE without adopting its extremes, so when refereeing it will still be my go-to resource, though Pelgrane are decent enough to not just provide conversion notes here but also seem to put out BRP stats for a lot of their supplementary materials, wisely realising that there’s no compelling reason not to sell to those of us not sold on Trail as a replacement.

For the purposes of playing a Purist-type game – which would tend to be my preference – I tend to prefer the way Call of Cthulhu handles things than the proposals for Purist play in Trail. (In particular, the combination of having to declare pool spends before rolling and, in the Purist system variant, not knowing your difficulty numbers would really annoy me over time in Trail.) Some of the suggestions for Purist play in Trail are entertainingly bleak, but others I am not entirely sold on, and to be honest I think the most Purist thing at all is to make failure possible in any particular field of endeavour; when I am feeling Purist I tend to feel that a “no victory, no moral, no explanation” outcome is an acceptable potential resolution of the scenario, provided that people were excited and creeped out by their descent into impenetrable Thomas Ligotti-esque hopelessness, so the central conceit of GUMSHOE itself seems to undermine Purism.

For Pulp purposes, however, things are somewhat different. “Pulp” is one of those terms which people throw around a lot but often don’t see eye-to-eye on, with the result that the sort of Pulp approach that Trail offers has a distinct and different flavour to the sort that Chaosium are presenting in supplements like Pulp Cthulhu. The latter is a more colourful take, willing to toss realism out of the window to the point of making player characters near-superhuman in their capabilities so that they can go around socking Deep Ones on the jaw. Trail‘s Pulp angle, conversely, is a bit more understated – less influenced by the likes of Doc Savage, more influenced by material like Robert E. Howard’s Skull-Face and his Detective Steve Harrison “weird menace” stories, where characters don’t so much have capabilities extending beyond those of their Purist counterparts beyond being somewhat more likely to succeed at actions they turn their hand to, and being somewhat more likely to be in situations where two-fisted action can usefully resolve matters. In general, I prefer Trail‘s style of Pulp to Pulp Cthulhu, so if in the mood to run such a campaign, Trail would definitely be a system I’d consider using. (Especially if you houserule it so that you can declare pool spends after rolling.)

On balance, it is a good thing for Trail that it has so much to offer aside from a heavy-handed solution to the missed clue problem, because otherwise innovations like the new Idea roll would have rendered Trail of Cthulhu largely redundant as a specifically Cthulhu-themed game line. As it stands, I’d gladly play in a Trail campaign run by someone else, use it to run a published campaign designed with Trail in mind, or use it to run a game using Trail‘s particular take on Pulp, and I’d also be interested in exploring both playing and running other GUMSHOE games – not because of the central declared aim of the GUMSHOE system, but because of all the myriad little good ideas which come along in its wake.

In addition, I am very glad that Trail exists as a testbed for system experiments which have both yielded interesting-looking subsequent GUMSHOE games and useful concepts that can be fed back into Call of Cthulhu itself. I note that recent supplements have made good use of the Purist and Pulp terminology, and I am interested to see if Pelgrane’s new Cthulhu Confidential book – providing a tweak of Trail of Cthulhu optimised for single player/single referee games – yields a similar offering for Call of Cthulhu.