Game Design: Good average chances of success for all situations?

[GMMichael]

If this zombie is going to unearth itself (just kidding, it's not that old), shall we discuss something a little more targeted?

What should be the typical chance of a PC making a successful attack?

I ask because the OP says to me, "should there be odds for things? Happening? What are they?" As GM, my PCs will usually have poor odds of doing difficult things, and good odds of doing easy things (difficult and easy from a typical human perspective). What they attempt to do is up to them, therefore, their "average chances of success" are determined by their own decisions - not mine, and not the game's.

 

[dragoner]

For Kosmic, I look at the average roll being 2d6+2, which any dice gives a 72% chance of success, good enough. Players can by action, change that, stacking other DM's, advantage/disadvantage; though for the basic odds, I think that is fine.

 

[Kannik]

If this zombie is going to unearth itself (just kidding, it's not that old), shall we discuss something a little more targeted?

What should be the typical chance of a PC making a successful attack?

For me I would still say it 'should' be between that 66%-75% range chance of success for 'typical' challenges for a character of a particular tier or experience. (Or is expressing it in fractions better? 2/3 to 3/4?) There'll be easier challenges and opponents, with adjustments accordingly, and opponents that are typical now will become easier as the character is more experienced and vice versa. But overall that, IME and IMHO, is the sweet spot to allow for uncertainty and the tension of failed rolls (and the fun of dealing with the outcomes of such failures) without feeling like the character is incompetent and becoming frustrated.

(Note that this value can still be used in 'harder' games; for example if good positioning and tactics are meant to be important, then set the difficulty such that, with a reasonable amount of said positioning and tactics, the chance of hit success is around 70%. That way it will facilitate and support the players to seek out those advantages.)

I used this for own game system (the Aurora RPG Engine), with the baseline competent roll equaling just a smidge over 75%. As a happy accident, it's also easy to pick a slightly different baseline which equates to 66%.

 

[overgeeked]

It depends on the context, but I think PCs should have about a 50/50 shot of success in most basic cases. If they have more than say a 75% chance to succeed, skip the roll...they just succeed. If they have less than say a 25% chance to succeed, skip the roll...they just fail. It's only interesting to make a roll if there's a real chance of success or failure and the results are interesting regardless of the outcome. If it's purely binary, yes/no, then that's too boring. You're rolling to see if the story progresses, that sucks. The story should always progress regardless of the outcome. Things like "no, and" "no, but" "yes, but" "yes, and" should be the results. If you have to have a clean binary, then add in boring yes/no into the mix. How the story progresses, that's interesting.

 

[pemerton]

What should be the typical chance of a PC making a successful attack?

What are the consequences of making a successful attack in the system? And how does the system structure the resolution of the attack?

In Burning Wheel and Torchbearer, and unopposed strike/attack will always achieve something; if it is opposed by a defence-type action, then the chance of achieving something depends on the resolution of the versus test. And if the action chosen is a feint rather than a strike/attack, then typically it will not achieve anything unless the opposing action is a defence-type one, in which case the feint negates the defence and counts the same as an unopposed strike/attack.

This is pretty different from the "roll to hit, roll for damage" of D&D and Traveller.

But there are plenty of other different models too. Eg in RM the likelihood of inflicting serious injury with an attack depends to a significant extent on the opponent's armour and the amount of their "combat pool" they have committed to defence. In Prince Valiant, combat is resolved via opposed checks, either a single one (eg for a joust) or with margin of success reducing the other character's dice pool, until one pool is reduced to zero and that character is taken out of the action.

So I think talking about chance of making a successful attack isn't a very meaningful question until we know what sort of resolution framework we are talking about.

 

[Kannik]

It depends on the context, but I think PCs should have about a 50/50 shot of success in most basic cases. If they have more than say a 75% chance to succeed, skip the roll...they just succeed. If they have less than say a 25% chance to succeed, skip the roll...they just fail. It's only interesting to make a roll if there's a real chance of success or failure and the results are interesting regardless of the outcome. If it's purely binary, yes/no, then that's too boring. You're rolling to see if the story progresses, that sucks. The story should always progress regardless of the outcome. Things like "no, and" "no, but" "yes, but" "yes, and" should be the results. If you have to have a clean binary, then add in boring yes/no into the mix. How the story progresses, that's interesting.

While I differ on the percentages, that is a good point to remember that skill tests are best only called for if there's a chance of narratively interesting failure (which can include mechanical effects), or if the degree of success/failure is important (for games that have Margin of Success/Margin of Failure built into their resolution systems).

 

[kenada]

What should be the typical chance of a PC making a successful attack?

Attackers in my homebrew system roll 2d6 + <proficiency> + <attribute> versus the target’s Armor defense, which has a base of 8. A PC with absolutely no modifiers to the roll at all (so rank +1 and a −1 attribute), has a ~41.7% chance of hitting (before mitigation, etc). More realistically, you’re going to be hitting more often than that. PCs are going to have +0 or +1 (up to +3) in the relevant attribute. Monster defenses scale by rank (= ⌊level ÷ 3⌋) with some adjustments per monster depending on their nature.

The worst PC in my campaign has a +1. Tama (the cleric) has not invested beyond her initial proficiencies and has Strength +0. The best is Deirdre (the barbarian), who is +7 due to investing and getting to her group rank (+2) due to being a warrior. Dingo (the thief) is +5. The characters are 5th level, so an at-level creature would have a base of 10 + adjustments. Deirdre is going to destroy at-level monsters, but she’s a warrior, so she’s supposed to do that.

 

[GMMichael]

So I think talking about chance of making a successful attack isn't a very meaningful question until we know what sort of resolution framework we are talking about.

Truth - but it's more meaningful than the Original Premise.

For what it's worth:
Attacks in Modos RPG always cause damage (combat progress) 100% chance of success . . . unless the opponent defends. Then your odds drop down to 50%, and start adjusting for physical power, skill, battlefield conditions, shields, and even heroism. What's the average? Only a heavy data file could tell you that. But the player has significant influence over her odds.

 

[MacDhomnuill]

I love me some necro threads!
I love DGP-CT/MT task resolution, it was simple but with enough granularity to account for time pressure (or lack there of) and presents PCs as competent. My biggest pet peeve in rpgs is the incompetent PC. If that the character concept the player wants its fine but if I am playing a PC that is supposed to have a college degree or have been in the military I would expect them to succeed at most tasks that fall in their experience areas.

It is one of my few complaints with BRP that a skill level of 40 is considered a professional.

 

[Kariotis]

You can be sure I'll unearth this one myself from time to time, because the question what the average chance of success and failure (or how spanners get thrown in the works) for any given situation in a game should be to maximize player engagement and fun has kept me reading game design resources for years now and I'm still returning to pondering it all the time. It seems so central to this hobby yet so rarely directly addressed.