Perception · ArkhamDB

If you take a closer look at the popular decks here on Arkhamdb, you may notice that still most of them use a complete set of Guts, Perception, Overpower and Manual Dexterity. I omit Unexpected Courage here, because it functions a bit different.

These cards' popularity is based on their marginal cost and the high versatility they give you. Each of them occupies one deck slot, but there is (on Standard difficulty) a huge chance of drawing another card, so it feels as if they cost nothing. They kind of "replace themselves instantly". Whatever strategy you are running, tests over all four attributes will meet you at every corner (or in every Mythos phase) and you don't have to commit these skills to a particular kind of skill test. The above-mentioned reasons, as a sum, virtually "reduce" your deck size and enable you to draw your key cards much faster. In fact, playing them includes two slight risks:

The test can fail and the skill does not replace itself.
The skill succeeeds and the card you draw is your Weakness. Especially, this risk can hurt if you Weakness happens to be Amnesia.

Pros

+2 on a test for (almost) no tempo hit is really, really good. If taking the test costs you an action, and you're even on the difficulty, then the resulting tempo bump is, on average, worth almost an entire action.
Can be committed to another player's skill test.
Zero play cost
Card text has no trigger restriction (as in Vicious Blow e.g.)
Combines perfectly with cards that have an additional effect if you overfulfill: List
"Ashcan" Pete gets more value out of the skill card than the other investigators, since the extra card you draw also works as smelling salts to wake his dog up.
On High/Expert difficulty you do not want to fail on skill tests, so the +2 bonus alone justifies the inclusion.

Cons

"Max 1 committed per skill test" across all players.
The skill succeeeds and the card you draw is a Weakness.

6/13/17

Synisill · 795

There is another weakness - which is that you are never forced to make a test of that type, and the card sits unplayed. This can happen for Manual Dexterity more than Guts, and for Perception/Overpower for people not searching/fighting. I think you always want Guts and 1 of Perception/Overpower though. — duke_loves_biscuits · 1256 6/28/18

Perception, Guts, Overpower, and Manual Dexterity are by far most cost efficient and overpowered non-asset cards in this game. Usually a good card is something that gives us an equivalent of getting 1.5 actions per 1 action (example: 1 action to draw Emergency Cache, 1 action to play it, and you get equivalent of drawing 3 resources). But Perception only requires to draw as an action, committing is not action, as an example we'll take The Gathering scenario standard difficulty which has 16 chaos tokens:

If Shroud 4 and Lore 3, we have 2 success tokens, Perception gives us +8 success tokens against 37.5% chance of not replacing Perception card, or 21,3 success tokens per spent card (10.67 per 1 card/action efficiency).
If Shroud 3 and lore 3, we have 6 success tokens, Perception gives us +7 success tokens against 18.75% chance of not replacing Perception card, or 37.3 success tokens per spent card (6.22 per 1 card/action efficiency).
If Shroud 2 and lore 3, we have 10 success tokens, Perception gives us +4 success tokens against 12.5% chance of not replacing Perception card, or 32 success tokens per spent card (3.2 per 1 card/action efficiency).

How i calculated numbers above ^, example: assuming no ghouls in your location, taking one plain action to retry Lore 3 against Shroud 2 test without committing resources would give us a chance of drawing 10 success chaos tokens from 16 total tokens, or 62.5% of success, vs. Perception committed it's +4 tokens. How much does committing Perception costs in cards/actions_to_draw_a_card? It's the chance to lose Perception, which is the amount of losing tokens (2) against total tokens (16) = 12.5%. Multiply 12.5% by 8, and you get one action you will spend to draw a card, also multiply 4 tokens by 8, and you get 32 success tokens for one draw card action vs. 10 success tokens of taking another plain research action. 32 vs. 10 = 3.2 as efficient.

And that's not even taking into account its deck thinning ability, 5-8 of these cards will give you the same or better deck thinning than Underworld Support, but 1) without limiting good cards to one copy, and 2) you can commit these to help other players.

The upgraded versions of these cards are not nearly as cost efficient.

I'll remove the review if my math is wrong.

10/25/24

Senji975310 · 1

10 definitely seems incorrect, if you wanted to count the resource gain it would be the chance to pass the test next action. So say you have a 50% chance to pass without perception then it would be worth half a resource, or a full resource if you rightfully say an action is worth more than the other resource types. Also underworld support does not cost an exp. — Spamamdorf · 4 10/25/24

You can't a success in test by spending resources, that's why i calculate cost in actions, actions to draw, actions to play card, actions to attack for example. Unfortunately action's worth is different each time, in first example to get one success without committing even more actions, you will need to spend on average 8 actions (12.5%). By spending Perception, you increate your chance from 2 tokens to 10 tokens (+400%) by just drawing Peception from your deck with an action, so 4 times the efficiency. If we didn't spend any action on this card, than action efficiency would be infinite. But every 37.5% of times you will have to spend ONE action to draw another +400% card (400 : 37.5 = 10.67). Perhaps it would be more believable if a card was worth 100% of one action and costed only 37.5% of action to draw, then 100 : 37.5 = 2.66 action efficiency. — Senji975310 · 1 10/25/24

This might just be me but if I've only got a 10% chance of passing a test I'm probably not supposed to be the one doing that test. Like if you have 3 book why are you the one investigating a shroud 4 location. That doesn't seem like a worthwhile calculation to be making. — Spamamdorf · 4 10/25/24

The problem with this math is you're not doing this by expectation, you're doing this by elasticity. You're taking into account the risk of losing the draw but comparing it to how much more likely it is you pass a test, which for our purposes I agree it's simple to say that failing a test is minus one action. What's clear is that there is a trade off: The higher up on the test you are, the more certain you will get back the card from success, BUT you are also less likely to have needed the card to pass. On the other hand, at lower odds of success, you moe greatly improve your chances of success (especially going from down 1 to up 1, or even to up 2) but are more likely to lose the card. The sweet spot, assuming 1 ghoul and an elder sign of +1, is when you are dead even with a test. I created the following profit function and plugged in the odds accordingly: PROFIT = — StyxTBeuford · 12985 10/25/24

PROFIT = E(Success with Perception) - E(Success without Perception) + 1*E(Success with Perception) - 1. — StyxTBeuford · 12985 10/25/24

You can combine terms but I'm writing it out this way for clarity. Here we're assuming passing the test is net 1 action, so you have to take the change in your likelihood of getting that action, so the difference in success between having and not having perception committed. Then we take the probability of success with it committed * 1 to show that that's the chance we draw a card from it. Then se subtract a fixed cost of 1 because we had to commit the card. You'll find your action profit is 1.375 at the peak of when you are even on a test. It goes down to 1.0625 when you're 2 up on a test, and when you're 4 up it would reach 1 - 1/16. Similarly, it drops off if you're well below a test: 1.125 if you're 1 below, 0.5 if you're 2 below, and 0.25 if you're 3 below. So definitely not infinite, but clearly quite good, and I should add, MUCH stronger if the value of passing the test is more than just an action to draw a card. . — StyxTBeuford · 12985 10/25/24

And to your point about efficiency: By this logic, we say Emergency Cache is 2 actions (a card and an action) for 3 actions, aka you net 0.5 actions per action spent. In the case of Perception, you are gaining 1.375 actions per action spent and probably more depending on the value of the test, but most optimally when testing at even on standard with the chaos bag given. So that means Perception/Overpower etc are probably going to be anywhere from twice as good as E Cache to more than 3 times as good. If you're a Rogue who loves succeed by stuff, it could be much higher. — StyxTBeuford · 12985 10/25/24

CORRECTION: All of those should be -1 because I forgot to enter the -1 into my excel formula. Which intuitively makes sense: perception isn't going to be more than 1 card and 1 action (assume a fast card that says "you automatically succeed" when you're otherwise certain to fail" so it can't be more than net 1. Still that gives us the following curve: -0.5 when down 2; 0.125 when down 1; 0.375 when even; 0.125 when up 1; 0.0625 when up 2. But again, this is assuming the value of a card and passing the test are the same, which they really never are. Passing the test is probably much more valuable since you choose which tests are worth committing to. And the common knowledge of committing to be up 2 on standard holds. — StyxTBeuford · 12985 10/25/24

@Spamamdorf correct, i just gave that example to demonstrate that it won't be cost ineffective in such test, in fact the cost effectiveness will rise dramatically. — Senji975310 · 1 10/26/24

@StyxTBeuford first message is correct, and it doesn't contradict what i wrote. But where is ***"-1"*** at the end in your formula comes from? You don't lose an action to **commit** a card. And this ***"1*E(Success with Perception)"*** is redundant (like 1*0.5=0.5). And how can you even calculate PROFIT if you did not even include the cost? Remember, we are calculating profit of Perception card so you need to include her cost, and we are not including cost of initial test action sinse we need to make it anyway, we only compare the cost of playing Perception card vs. addidional plain actions to research clues. ***"Here we're assuming passing the test is net 1 action"*** it never is, if it was, there would be no need whatsoever to commit anything to that action. That's where your formula is incorrect, you take already 100% succes on a test, and you add to it a chance to lose a card when no card was needed. — Senji975310 · 1 10/26/24

"so you have to take the change in your likelihood of getting that action" Phrase doesn't make any sense to me. Which ***"the chance"***?. And what is "likelihood of getting that action" there is no likelihood, we are already taking test action and do not account it in the formula sinse we have to take it anyway, the only thing we calculate is the benefit of committing Perception to that test action. "committed * 1"" you mean multiply 100% by 1? Ok, that's our card back 100% of the time. "Then se subtract a fixed cost of 1 because we had to commit the card." again, commit doesn't take an action. "You'll find your action profit is 1.375" By now i'm completely lost and i don't understand how you arrived at 1.375. From your words it was laready 100% success, how can you get 137.5% by commiting Perception? Already at 100% you will get infinite redraws. — Senji975310 · 1 10/26/24

In The Gathering scenario assuming lore 4 and shroud 3, the research action gives us 10 tokens to succeed, but takes one action, on the other hand Perception gives only 4 tokens to succeed, but costs only 12.5% of one action. We multiply 12.5% by 8 to get one action comparable to plain research action, and we do same thing with 4 tokens getting a total of 32 tokens which are not comparable at all to research action, 32 tokens vs. 10! I'm not math professor, so i can't create a formula out of it, but it's clearly correct. — Senji975310 · 1 10/26/24

And easier way to explain: If test difficulty is one higher than our skill, Perception gives us +400% chance of success and costs only 37.5% of one action (10.67 per 1 card efficiency). If test difficulty is same as our skill, Perception gives us +116.67% chance of success and costs only 18.75% of one action (6.22 per 1 card efficiency). If test difficulty is one lower than our skill, Perception gives us +40% chance of success and costs only 12.5% of one action (3.2 per 1 card efficiency). — Senji975310 · 1 10/26/24

In general, you should not express increases in odds in terms of "+%."It makes what your writing much harder to follow, especially because you don't even include the base percentage chance of what you are modifying. When trying to turn statistics into a narrative or actionable policy (which is what your doing), it would behoove you to say something like "You go from a 16% chance to pass to a 62% chance to pass, making you 4 times as likely to pass" which makes it much more clear what you are actually saying when you say "+400%." While I don't agree with your 'efficiency' calculations, when you put it like that it becomes much more clear why someone might want to say... put a Guts into a deck with 2 willpower. — dezzmont · 210 10/26/24

Also this review skips over the deckthinning aspect pretty hard. If the cantrips just read 'If you take a skill test your 3+ up in, discard this card and draw a card" they would still be essentially auto-includes in most decks that can hit those numbers. The fact they ALSO come with a really big probability swing on a test is often at most a 'side benefit' to effectively running a 28, 26, or 24 card deck, depending on how many Cantrips you can realistically force to pass. When you look at the Cantrip skills from the lens of "I am running these to effectively reduce decksize and thus play a much more streamlined deck' the upgrades become a lot more potent, because they are twice as good at that function and allow you to play a 26, 24, or 22 card deck, which combined with strong tutoring and good mulligans lets you run your combos with extreme consistency. — dezzmont · 210 10/26/24

@dezzmont I replaced percentages with success tokens to make it easier to understand. I also had to turn review into narrative to explain where the numbers come from, hopefully i did a good job. You don't agree with my calculations because your instincts tell you so? Understandable, it's was hard to believe for me too. xD Didn't skip deck thinning aspect, just didn't include a lot of exclamation marks, its importance is different for everyone. — Senji975310 · 1 10/26/24

@SeNJI The -1 in the formula is the cost. We're doing this in terms of 1 action allowing you to draw 1 card. Passing the action is similar to gaining an action because presumably fialing the action means you need to spend an action solving the problem in the future instead of drawing a card. I even mentioned in my formula that there is a redundancy, on purpose, for clarity. The left side of the formula is the expected gain of action, the right side is the expected loss of cards. I explained as such in my comment already. I agree with Dessmont, it doesn't make any sense to me to compare the increased chance of passing an action as efficiency. Going for 0% to pass to 0.1% to pass is an efficiency gain in this view of infinity. Clearly though, a card that does that in my deck at the cost of a card is not worth running. — StyxTBeuford · 12985 10/26/24

@ StyxTBeuford I see, 1 in your formula is the drawing of new card, except it's not the benefit of such action. Your formula is almost correct then, instead of -1(drawing a card) you need to insert the success of one action (it's different each time). Just because you could have used Investigate action instead of Draw Card action, it does not gurantee a sucess. In my third example sucess is 62.5%, so if in my example, the cost of losing a card is 12.5% (because i only calculated amount of success tokens per Perception card, 4 tokens per 12.5% or 32 tokens per 1 card), while in your example the success tokens of one action would be 10 or 62.5% . In the altered formula we would divide the worth of Perception by worth of a plain action: — Senji975310 · 1 10/28/24

PROFIT = (E(Success with Perception) - E(Success without Perception)) : (E(Chance to lose Perception))*(E(Success without Perception). — Senji975310 · 1 10/28/24

As for "Going for 0% to pass to 0.1%", i completely agree, except in my first example it would still be 10/16 success tokens, which is worth attempting if your seeker has to spend 1-2 actions to get to your location, and if he is in your location already, you can give this bonus to him, he gets a small bonus, and you get to replace a card and reach all your desired cards faster. — Senji975310 · 1 10/28/24

洞察力

FAQs

Reviews

Pros

Cons

洞察力

相关卡牌

FAQs

Reviews

Pros

Cons