Dogs are an expensive weapon that takes up the ally slot. Having just 2 dogs out compares compares quite favorably to the .45 Thompson, albeit one that can only shoot once per round. The movement effect is mostly icing — situationally useful, but not what you get the dogs for.

Being able to fight only once per round is bad for a primary fighter, but being such a heavy investment, they're too expensive to run as an off-class combat tool if you don't expect to use them often. Usually events like Occult Invocation and Backstab are favored for those purposes because Seekers and Rogues usually have better things to do with their ally slots. Playing down 2-3 dogs just to use them twice or thrice a scenario is an extremely poor use of resources and actions. The same is true for Mystics have their combat spells, and Guardians just have better weapons. That includes you, Leo Anderson.

Which leaves us with Survivors, who can actually use the dogs quite decently. Survivors have good draw and econ now, but notably poor action compression which the dogs help with. Calvin for example likes the soak the dogs give and doesn't have to worry about managing horror from the Meat Cleaver. If they miss, they have Live and Learn and Oops! (2) to compensate for the miss the same way they would use an Old Hunting Rifle. Tommy can also borrow those tools if he wants to be creative, but like other guardians he's got better weapons when he gets XP.

The other candidate is Lily Chen, who lacks good low-level weapon options due to her Firearms restriction. Instead of playing several mediocre spells just to power Dragon Pole, why not bring a couple of dogs along?

This card is absolutely wonderful, even on hard, if you’re willing to sit on it a bit. Unless you’re Lola, you already control one class card, your investigator card. And with Edge of the Earth, there are a few of triple class cards that get you to 3. Double class cards are available, even for level zero. Also useful: having teammates who can play things for you (like Bob).

My favorite option is the investigator who can play cards from any class, as long as they heal horror. With Peter Silvestre, illicit alcohol, and the scroll of secrets(0), you’re there. (There’s also Shrewd Analysis, which is vital if you’re doing the ancient Stones trick. Easy!)

With your tableau set up, you can swing a +5/6 damage brutalizer for the extremely reasonable price of three resources. Overpower or promise of power can get you to ridiculous numbers. In duo, this can kill tough enemies. In a larger group, it can supplement your dynamite as a boss killer. And late game is when bosses typically show up!

Fear actresses and therapists, because they have many friends.

Monterey Jack wants to go fast. Which means being away from the party tank. This weapon solves a lot of problems in an elegant way:

*You don’t need to play it until an enemy shows up.

*It doesn’t exhaust until want it to. Kill a one HP foe, then exhaust a boss. Or get it enchanted to do a two damage evade.

*You can use it on massive enemies or people engaged with others, with a low risk of damaging them.

It’s also extremely cheap for what it does, and can trigger any number of skill cards, like watch this or manual dexterity, which are harder to get from your seeker cards.

The most powerful encounter-cancel card in AHLCG. If you know the encounter card, you can achieve Ward of Protection (5) with no resource and sanity cost. Moreover, you could play it for another investigator at your location. More important thing is that you can cancel weakness. Unlike other cards, there are NO weakness restriction. Additionally, it works for Peril encounter, I believe (since play timming is before you know the card is peril).

However, it's hard to use this by itself. You should know the encounter/player deck to play it. It means that you need to use Foresight with some supporting cards. However, it's not hard work when you firstly see this card.

• Scrying/Alyssa Graham: Scrying/Alyssa is the best way to look cards for mystics. If you're Gloria, I think you already use this. One problem is Arcane/Ally slot to handle it. If you have enough slot, this card will be good way.
• First Watch: Here is the best solution, I think. Play First Watch. Look before dealing cards. Then, naming what you just looked. It seems that the investigator do not draw the encounter card from the deck when resolving the First Watch according to the FAQ 2.0, so that it does not work.
• Parallel Fates: 0-level is not robust, but 2-level is robust. I don't use those yet, but it will be helpful.
• Norman Withers: Norman Withers usually reveal the weakness before drawing it. It means he cancel his weakness by Foresight. However, it's not true for The Harbinger. The solution is simple. You may use replacement set :)
• Mysterious Chanting: Some encounter make you search encounter deck and draw what you want. If you're not first-run, you may know which is in encounter so that naming it.
• Search Cards: I'm not sure I can play Foresight AFTER search BEFORE draw a card. Currently, my party plays Foresight as a ruling that I should name a card before search deck. Even if this cases, searching can increase the probability. If you're Agnes, you may play Word of Command to search Dark Memory and discard by Foresight without effect.

Lets talk about this card, with dummy probabilities. Because numbers are tricky and I think I can explain it simply.

First off, this is a "cantrip", it's a cheap/free card that draws another card. Why is that very good? Most arkham decks have a few cards that you cannot go without, your most important cards, for example a weapon for a fighting , spells. Having a cantrip like Tempt Fate means that when youre sifting through your 30 card deck to try and find those key cards, you're actually sifting through a 28 card deck.

But, it is'nt "free", it comes with the bonus / penalty of and tokens.

Without going into exactly how, / tokens are inversely more powerful depending on how many tokens on a chaos bag net you success or failure. When the chance of success is low, is more powerful, the inverse is true for .

I.E. If you're doing a lot of tests that you know you'll win just barely (most of the time you'll be avoiding to do tests if your chance to beat it are negative!) then curses really hurt you, similarly if you do a lot of tests you know you'll just barely loose, then the blesses will help more then the curses hurt. This effectively means that Tempt Fate hurts consistent characters with good baseline stats, but helps risky characters with a glaring weakspot (Like many folks who find themselves doing tests from low starting points.) Incidentally, this means that strategies can be very helpful to middling characters like Roland Banks, who find themselves throwing repeated Investigate checks at shroud 2 or 3 locations.

To explain it more detailed: You can calculate the most common token draw from a bag pretty easily, many // tokens are net -1's or -2's on standard, and on standard -1's and -2's are common. If you test such a chaos bag with a baseline of +2, and then add blesses, your chance at success increases only by the amount of tokens in the bag that a bless might help with (a -3 or -4), but because those are less common than the opposite, the blesses help less then the negative effect had by a similar number of curses, because there are more tokens that a curse can turn into failure, than there are tokens that a bless can turn into success.

For example: If you have five -0's in a bag, and three -2's (and nothing else), and you're testing from a baseline of +1 over the difficulty, then if you add a bless into the bag, there are 3 tokens where bless + token replaces failure with success, if you added a curse instead, there are 5 tokens where curse + token replaces success with failure. I.E, in this case, adding a curse hurts more then the help gained by adding a bless.