Played Donjon some this weekend. Here are some thoughts that came up, in no particular order other than the order in which they occur to me.

1) There are clearly certain things that every Donjon character should have. These should be discussed explicitly in the rules on character creation. For instance, we had a barbarian fighter who wore no armor, had no passive dodge skill, and no damage soaking skill. His survivability was very low. Every character who intends to be able to melee should have SOME form of damage soak ability (like armor, or "Tough as leather") or a very high passive defense skill. This shouldn't be a requirement but should be explicitly discussed so that players who choose not to take these do so knowing what they're getting in for.

Another type of skill that should be discussed are the story direction skills (like find traps, spot hidden, etc.) so that all players have the opportunity to give their characters some area of impact.

In fact, I'd suggest actually listing several types of skills and describing each. For example: Damage increasing skills, Passive Defense Skills, Damage soaking skills, story direction skills and some examples of each. These are skills that really drive Donjon mechanics and leaving players to discover their utility after being moshed in the first encounter is probably not the best presentation.

2) Similar attention needs to be given to the Saving throws -- especially the ones with the silly names I refuse to say ;-) Our players wasted an awful lot of character points jacking these up for no really good reason. I myself don't really understand why they are necessary at all. I suspect they are a hold over from the Krawl-parody days but am not convinced they serve a purpose. It seems to me that their function can be served equally well by simply rolling against attributes (especially since they start the game equal to attributes anyway). Note: I'm refering to the Petrification, Poison saves, not the Wealth / Provision saves which work great.

3) The rules are very specific that success on rolls can be used for facts -OR- Rolling over bonus dice. I recommend eliminating the "or". The "or" merely sets up a conflict between narrating cool effects or maximizing bonus dice application. Lets say I get 3 successes. I want my opponent to stumble and thus get a penelty to his next defense. Currently that would use 1 success for the Fact of stumbling, leaving 2 for the defense penelty. Why would I do this, when I could just throw all three into a defense penelty anyway? Coming up with a cool narrative fact shouldn't be viewed as a "waste" of a good bonus. Let me come up with three Facts "he stubbed his toe on a rock", "causing him to stumble and almost fall", "in regaining his balance he's left himself open for an attack" -AND- let those 3 successes provide a 3 penelty to defense.

4) Using successes as bonuses in combat requires some more rigorous rules attention. For example, I think its a great idea to allow successes to be used for temporary bonuses or penelties to "the next roll", but that also comes into some conflict with the basic damage rules. As it stands, why (from a purely gamey perspective) would I use my 3 successes to give an opponent a -3 penelty to his next defense, when I could use 1 of those successes to make a called shot and then give him a -2 penelty to adroitness which would apply to ALL of his defenses AND all of his attacks for the remainder of the encounter? Clearly the temporary effect isn't a very efficient use of bonuses, yet it is a really cool feature that should be encouraged.

I suggest for this specific example that all damage to attributes require 2 successes per point of damage, where temporary penelties to a single roll require only 1.

But there are other things that can be addressed as well. For example, there are many situations that could be described where a loss of Actions could be justified. But simply allowing successes to cancel actions is very disproportional.

5) There is a very high whiff factor in the game, one that I find very deprotagonizing. For instance, our barbarian character had several very solid hits that pinged off for no damage (I know you've mentioned this one before), but there were a fair number of other whiffs also, including magical whiffing. My character spent 3 Actions accumulating magic points and managed to get a fair number (7 or 8 as I recall) to power my spell. When it came time to roll for effect 13 dice against 3 or 4, however...big whiff. Mike (the GM) rolled good enough on 1 of his dice to limit my massive roll to a mere 1 success. Thus what could have been a very very cool scene involving our barbarian fighter getting bashed around by a demon, only to be saved by a spell from a 2 foot tall Leprechaun (me) turned into a big waste of 3 actions as I could only roll over 1 die into damage.

The root of the problem is the Sorcerer die mechanics which are not very good at generating the multiple successes needed to power Donjon. They work in Sorcerer because 1 success in Sorcerer is "success" and anything beyond that is extraordinary. In Donjon, however, 1 success is minor and inconsequential in many cases and multiple successes are required to be extraordinary (you need ALOT of successes for a sword swing to do devastating damage in a single blow, or for a spell to have more than cantrip level effects). The die mechanics however actively oppose multiple successes. If I roll 20 dice and you roll 4, one would expect that I'd absolutely crush you. But in reality those 4 dice are enough to ensure (most of the time) that you'll get at least 1 very high roll which will limit my successes to a mere 2 or 3 tops.

I have to say that a more target number oriented way of generating successes would probably work better, but knowing how enamored you are of this mechanic I'll offer a couple of suggestions that will help a little.

a) don't use d20s. Go to a smaller die type. A smaller die type will encourage more tied rolls, and tied rolls favor the side with more dice to back up the tie. You will have fewer (only slightly, but every bit helps) occassions of 1 die being high enough to stop a 20 die roll, because the odds of that one die being tied and discarded will be higher with a smaller die type. I'd use d10s (d6s might be too small given the number of dice typically being rolled, and d8s being too uncommon). A good solid die pool even for beginning characters is going to be 7-10 dice. That puts a strong likely hood of a good pool getting at least one 10. This provides protection against that 1 in 10 chance of a 1 die pool getting a 10, as such a lucky roll would be most likely tied and discarded.

b) Make the tied dice part of the winner's over all success. This I believe was suggested before. For example: take a roll of 8, 6, 6, 5 vs 8, 6, 5. Currently the 8 and first 6 would tie and cancel, the second six would beat the 5 resulting in 1 success. However by keeping ties, the tied 8 and first 6 would add back in and give 3 successes. 3-4 successes are really required before a roll can be considered to have decent success.

c) reduce the default difficulty. Instead of easy being 3 dice and increasing in increments of 3 there after, make easy 2 dice increasing in increments of 2 there after. This will help in the story direction skill rolls as well as in mana gathering rolls.



Anyway, those are my thoughts.

Yep, I agree with everyhing that Ralph said.

In addition, please give more information on how to balance encounters better. I ignored the rule of thumb about not making foes more than a couple of levels higher than the characters thinking that at three to one odds, they could handle a foe four levels above them (in the first encounter, no less!). Well, they won, but they got pretty badly mangled in the process. Any notes to this effect would be helpful.

And I'd like to reemphasize the idea of making the need for certain skills explicit. As the GM I should have mentioned this, but got too absorbed in the creativity of the process. I just wanted to see what the players came up with. Had I known the nature of the characters selected, I would have made the adventure less combat oriented.

Mike

Did you guys use the new version I sent to Valamir? It addressed most of your questions: the tie issue is fixed, and explicit skills are mentioned in character creation.

Mike - don't ignore the advice written in a game and say "this game needs better advice." You're a technical guy with the statistics - I could have told you (or you could have figured out) a foe four levels above the characters is going to kick their ass.

As my rule of thumb, I use: 2 levels below party - really easy, 1 level below party - easy, level equal to party - average, 1 level above party - hard, 2 levels above party - nearly suicidal in any numbers, and used only for "boss" enemies.

IIRC, character creation was the only part of your new version that was actually written up yet. I don't recall reading anything in there that addressed these items.

Unless I'm completely misremembering:

The saving throws are still there, the default difficulty is still 3, the dice used are still d20s, and rules still explicitly state Fact OR die bonus roll over.

Can you be more specific, because I don't remember seeing anything that addressed these issues in the new version.

Clinton R Nixon wrote:
Mike - don't ignore the advice written in a game and say "this game needs better advice."

After a day thinking on it, I've got these ideas based off you guys' experiences. They're listed in rank of radicalness, from lowest to highest.

1. More suggestions in the text regarding skills - this is the biggest one, and one I can accomplish easily. Check on that, and thanks for your suggestions.

2. Change character creation for PCs and enemies slightly - skills useful in combat cost 2 slots instead of one, making uber-enemies a little tougher to build. I admit, when I create enemies, I throw away a couple of slots on useless skills so as not to make the enemies as tough as I could.

3. Change damage. I will be doing this - I think it's broken that you can connect and not do damage. What will probably happen is that I will drop the damage roll and have damage be equal to your attack successes, plus a modifier for your weapon.

4. Change the dice system. This is the scary one, but one that might just work well. Instead of the number of dice you have higher than your opponent's being your successes, I might have to move to a "roll against a target number" sort of system. For example, if everything equal to or above 12 is a success, two people might roll:

Player (8 dice): 4, 7, 10, 11, 13, 15, 15, 19
GM (7 dice): 5, 7, 7, 9, 13, 15, 20

The player gets 4 successes and the GM 3 for 1 net success for the player.

Of course, I could go quite a bit further, and change the rules so that successes are always successes - that is, in the above situation, the player and GM both narrate facts (the more successes guy always goes first, with the two alternating.) Let's say in the case above, the player was searching for a door:

Player: I find a door! (Fact 1)
GM: The door has a lock on it. (Fact 1)
Player: But the key we found before fits in the lock and opens it. (Fact 2)
GM: You hear growling behind the door. (Fact 2)
Player: It sounds like wolves (which the player has a skill with). (Fact 3)
GM: There's Orkish voices, too. (Fact 3, done).
Player: Orkish voices in pain from wolves attacking! (Fact 4, done)

Hmm. I kind of like this. In combat, the enemy would have taken 4 damage, and the player 3 damage.

1. I think that will cover a lot of ground effectively.

2. Hmmm, I'd be tempted to do it easier and just make what you normally do the rule. Monsters only get 3 slots for combat skills and 2 for other skills (like hunt for food). I wouldn't reduce the number of slots available for the player characters though...you'd get fewer non combat skills that way I think.

3. Not a bad idea, but then you have 3 effects: a) as Mike has pointed out elsewhere you lose the ability to ever just scratch someone with the giant axe of cleaving, b) you lose the effect of virility adding to damage. c) you then have to add armor to the defense roll, which might make hitting that much harder.

What if instead you keep the damage roll as you have now, but instead of adding the to-hit successes into the roll, you add the to-hit successes in as automatic successes in addition to those generated by the roll. For instance I hit you with 2 successes. I then roll weapon + Vir vs your Whe + Armor and get 1 success. That's a total of 3 damage, 1 for the weapon, 2 from the hit. Even if your armor wins the damage roll, I still get 2 from the hit.

If you want to get a little more simmy, you could say that if the defender wins the armor roll, then the damage done from the to-hit can only be used for flesh wounds, where if the defender fails the armor roll you can do damage against attributes.


4. Well, I never thought to hear you even suggest this, but I do think that your system cries out for a die mechanic that generates more successes over all and more reliable ratio of successes for big pools vs small pools. If you were to go for this, it would likely help with 3 above also.

Part II of 4, is a very interesting idea. Mike and I tried to do this with Universalis Complications in one of the versions. We couldn't get it to work smoothly, but its certainly worth playtesting.

In combat you could do alot of interesting things with it. The player got 4 successes, that could be 4 towards damage. The GM as the defender might then say that the players attack left him open and thus he has a -3 to his next defense (instead of doing damage out right), or some other variety of side effects.

FWIW, if you did go with the d20, 12+ for success, then we sneaky types with a lot of d10s could roll them instead for 6+ with the same effect.

Mike