So, another thread mentioned the use of skill or attribute "tests" in combat in lieu of attacking, and I wanted to discuss how more games are including these options (and why I find them generally unsatisfying).  Of course, the title of this thread frames the question as an absolute (primarily because there is a group of posters on here who just lose it every time someone posts an absolute... and it's amusing to watch the 'tards "Reeeee!"), but I'm interested in games that have created a viable non-attack option in combat... because I just haven't seen one yet.

To set the table, I want to make sure we're all on the same page as to my assumptions and assertions.  So I want to lay the following assumptions out before we look at the present iterations of skill tests:

• I am focusing on mechanical results, and not subjective or narrative results --- Sure, you can do a backflip to awe the enemy so that your partner might kill it, and the DM can rule that you impress the spectators enough that you boost your standing with the locals... but that's not a mechanical result and it's not doing anything to help with the actual combat
• I am focusing on the most efficient resolution of fights --- sure, there are any number of ways to use skills to prolong, delay, and change the nature of a fight.  But unless the tactic can result in an equal or less cost in resources, time, damage, etc., then it is considered "sub-optimal" for the purpose of this discussion (an additional note: "optimal" results don't mean some kind of nefarious min-maxing.  They actually make the most sense in character as well as in the game.  What character will prefer to lose half his hit points in a fight when the option was available to lose only one quarter?  No rational human is ever going to choose the tactic that costs the most of anything, except in rare enough circumstances that don't really address the main issue here)
• I am talking about situations where violence either has or will imminently break out and the only choice is to defeat the enemy, either by killing, wounding, capturing, or scaring off.
• I am assuming enemy combatants are equal or greater in number than the player characters --- this is a bit of a narrowing assumption, but just because it comes up very frequently in most of the RPGs I've played, and because a highly unbalanced action economy in the players' favor doesn't seem like it is a true test of utility.

OK, with that out of the way, my primary hypothesis is as follows: Any kind of non-attacking skill test or challenge will never be more effective than degrading the enemies' hit points via attacks, and will often be much less efficient.  So, how do I come to this conclusion?  Well, let's look at combat tests from probably two of the most popular game systems right now: D&D 5e and SWADE.  In 5e, a player may perform a "help" action, which gives the helped creature advantage on the next attack roll against the chosen creature.  In SWADE, a character may use the support action (resulting in a +1 or +2 bonus to a fighting roll for an ally) or perform a test, which results in either a distracted or vulnerable (-2 or +2 respectively to opponents' rolls or allies' rolls) with the possibility to make the target shaken on a raise.

Now, either of these options are almost always going to be worse than attacking the target.  I don't want to get too far in the weeds to figure out how much better your skill would have to be as opposed to your chance to hit in order to make the test viable (getting to do something, even if it's not really helping, as opposed to just missing every turn), so let's assume a relative parity in chances to hit and pass the test.

Now, I know that an argument against that is that tests allow a player to build a character that is not combat facing, but still useful in combat.  But, I think, based on the following points, that "useful" may not be accurate, and that this illusion helps to make unbalanced (and even min/maxed) parties more viable and likely because of the lip service paid to tests.

So, what ends a fight?  All enemies are one of the following: dead, incapacitated, or fled.  Just sticking to the mechanics related in the games, none of the skill tests create a chance of flight.  Sure, you can house rule it, but that's not really confirming the utility of the mechanics as they stand... since you're having to change it.  Incapacitation might be possible due to some kind of grappling, but you haven't removed the threat; you've just delayed having to remove it (and if there are more enemies than players, you still have active combatants that can kill you while you try to keep your target out of the fight).  But grappling is its own thing, so I'm not really talking about that in this case.  So that leaves killing/rendering unconscious due to damage as the primary goal.  And tests suck at this.

I'll note that a similar argument has been proffered previously by many folks when it comes to damage or healing.  Generally speaking, healing is less efficient than doing damage, as it spends resources without reducing the threat or capacities of the enemy.  The best healing is the healing that revives or keeps a character in the fight, as it preserves damage output and shortens the fight.

What most skill tests seem designed to do is allow one character to give up a chance to do damage in order to make another character slightly more likely to do the same damage that he would have done without the test.  To be viable (i.e. to actually reduce the time in combat), a test would need to guarantee a least as much damage as both characters would inflict on average.  Now, this will only happen when one character has a huge disparity in accuracy or damage over the other (an extremely unbalanced party), or in a case where one party member's contribution is unneeded in the fight (and I know we all want to feel like that player!).  It makes one player into a sidekick in the combat realm, which I think most players aren't too keen on as a permanent role.

You could argue that a player who uses his character's action to worsen the chances of an opponent to hit might be worthwhile.  But, once again, the reduction in damage needs to be enough to make up for the increased time the fight goes because of the lost damage from the helping character.  I'm not going to crunch the numbers here, but I would suspect the damage reduced needs to be very large to pay off (or the helping character must be highly outclassed by his party-mate).

So, in the end, it comes down to the fact that the chance to do damage is always going to shorten the fight by an amount that doesn't make the usage of skill tests valid.  If anyone knows of a mechanical approach that defies this analysis, I'd love to hear about it.  Because, otherwise, I feel like skill tests are directed towards players with a dramatically underpowered character (with respect to the rest of their group).  So a mechanic to address a Session Zero problem.  And, outside of a few rare cases (which you don't need to recount, because a few singular examples don't disprove the general trend), skill tests just seem to be meant to make the useless character in combat seem to have a purpose, even when they don't...

Quote from: Eirikrautha on June 13, 2023, 04:27:14 PM
Any kind of non-attacking skill test or challenge will never be more effective than degrading the enemies' hit points via attacks, and will often be much less efficient.  So, how do I come to this conclusion?  Well, let's look at combat tests from probably two of the most popular game systems right now: D&D 5e and SWADE.  In 5e, a player may perform a "help" action, which gives the helped creature advantage on the next attack roll against the chosen creature.  In SWADE, a character may use the support action (resulting in a +1 or +2 bonus to a fighting roll for an ally) or perform a test, which results in either a distracted or vulnerable (-2 or +2 respectively to opponents' rolls or allies' rolls) with the possibility to make the target shaken on a raise.

Quote from: Eirikrautha on June 13, 2023, 04:27:14 PM
I'll note that a similar argument has been proffered previously by many folks when it comes to damage or healing.  Generally speaking, healing is less efficient than doing damage, as it spends resources without reducing the threat or capacities of the enemy.  The best healing is the healing that revives or keeps a character in the fight, as it preserves damage output and shortens the fight.

In my experience of 5E, though, healing is vital. Obviously, the best healing should be tactically applied rather than randomly -- but when it is tactically important, healing is incredibly useful. In my experience, heal is a common intelligent action for healing-capable PCs.

I rarely see the generic "help" action in 5E, but I do often see use of spell or special actions that give advantage or disadvantage. Spells like Bless, or the monk's stunning strike, etc. Sometimes the special attacks do a little damage in addition to setting up the target, but it's often the non-damage effect that is the most important.

Just as in healing, the tactical situation matters. If there is a high-damage but low-hit-chance enemy like a clumsy giant, then giving him disadvantage can make a huge difference. Against a hard-to-hit opponent, a lower-damage character like a bard might be better off giving advantage to the high-damage rogue.

I'm still coming up to speed on SWADE, but I think tests are similar. If they're not important enough, then I might house rule to make them more important.

Quote from: Eirikrautha on June 13, 2023, 04:27:14 PM
So, in the end, it comes down to the fact that the chance to do damage is always going to shorten the fight by an amount that doesn't make the usage of skill tests valid.  If anyone knows of a mechanical approach that defies this analysis, I'd love to hear about it.  Because, otherwise, I feel like skill tests are directed towards players with a dramatically underpowered character (with respect to the rest of their group).

I find that healing-focused clerics in 5E are incredibly useful, and not at all underpowered. Opinions differ on bards, but I've had a lot of players who were happy with them. In my experience with older tactical systems like the HERO System, moves like grab, takedown, entangle, flash, etc. were often popular and interesting options compared to straight damage.

I think tests shouldn't be the default best option, but it's good if a system has them as sometimes the best move for a particular tactical situation.

Quote from: Eirikrautha on June 13, 2023, 04:27:14 PM

• I am focusing on the most efficient resolution of fights

...
So that leaves killing/rendering unconscious due to damage as the primary goal.  And tests suck at this.

Eh, no, tests don't suck at this, skills do, simply because there isn't a skill that kills or renders your opponent unconscious. If there were, a "test" for it would be perfectly good at ending fights. Attribute tests are just informalized "skill" rolls.

Honestly, this whole thing seems like a giant strawman. Attribute tests are designed for things other than combat-ending tactics.

My own system uses opposed rolls for all actions -- so for example, a successful Climbing action can prevent taking damage from the missile attack directed at you while you're doing it. Not exactly what you're asking, but it does make skills more potent in combat.

Well, with actual examples, it is hard to judge, but I agree that in the confines of "assuming enemy combatants are equal or greater in number than the player characters" it is true MOST OF THE TIMES.

There are still cases (in 5e) in which giving adv is more useful than attacking. Say, against a foe that in unharmed by nonmagical weapons if you don't have one.

Also, if you're helping a paladin with a holy avenger to get a crit, or helping a rogue, etc. This would require a high AC and a tough monster in general, which is rare if "assuming enemy combatants are equal or greater in number than the player characters".

It seems you're assuming all PCs to have similar damage output which isn't necessarily true, especially against monsters with some resistance - or worse, immunities!

Also, depending on the game, a skill might let you FIND OUT if the monster has a weakness etc.

If I understand the parameters correctly, the first counter example that springs to mind is a well-timed intimidate. Depending on the system, that could be part of a skill and/or attribute test.

A third of the enemy orcs are dead.  Your buddy with the great axe has just lopped of one's head. The GM narrates that the orcs flinch as the head goes bouncing through the ranks.  You step up next to your buddy with your sword poised and ask, "Who wants to be next?"  If the GM calls for a test, and another third of the orcs run, then that's more fled than your attack was likely going to kill. 

Granted, it depends on how the GM runs it, and whether there is such a skill supported.  In my own system, the orcs have a morale check, with rules such that such a scenario might either result in a small minus to that morale check (so not a skill check) or allow a skill check to put them over the edge into checking morale when they weren't quite there before.  Arguably, killing one of them with your sword might also put them over the edge.  However, that's based on details the players don't fully know.  A more direct case is rallying your own side when they fail a morale check, which in my system is nearly always a good use of leadership skill if you have it, because making half your team fight better or at all the next round is likely to inflict a lot more damage than what you do yourself.

The devil's always in the details.  Fantasy Hero and Flash was mentioned.  I ran a version with skill-based powers.  I had to ban Flash, because it was so overpowered.  That's not about the skill, though.  It's about skill-based access to a magical effect priced for a superhero system, where blinding several people for a few rounds was much more powerful than it was when killing the enemy with swords was the alternative, as opposed to knocking them through a building. It's a quirk of the way Hero System powers function.

Quote from: Steven Mitchell on June 13, 2023, 06:22:34 PM
If I understand the parameters correctly, the first counter example that springs to mind is a well-timed intimidate. Depending on the system, that could be part of a skill and/or attribute test.

A third of the enemy orcs are dead.  Your buddy with the great axe has just lopped of one's head. The GM narrates that the orcs flinch as the head goes bouncing through the ranks.  You step up next to your buddy with your sword poised and ask, "Who wants to be next?"  If the GM calls for a test, and another third of the orcs run, then that's more fled than your attack was likely going to kill. 

Granted, it depends on how the GM runs it, and whether there is such a skill supported.  In my own system, the orcs have a morale check, with rules such that such a scenario might either result in a small minus to that morale check (so not a skill check) or allow a skill check to put them over the edge into checking morale when they weren't quite there before.  Arguably, killing one of them with your sword might also put them over the edge.  However, that's based on details the players don't fully know.  A more direct case is rallying your own side when they fail a morale check, which in my system is nearly always a good use of leadership skill if you have it, because making half your team fight better or at all the next round is likely to inflict a lot more damage than what you do yourself.

The devil's always in the details.  Fantasy Hero and Flash was mentioned.  I ran a version with skill-based powers.  I had to ban Flash, because it was so overpowered.  That's not about the skill, though.  It's about skill-based access to a magical effect priced for a superhero system, where blinding several people for a few rounds was much more powerful than it was when killing the enemy with swords was the alternative, as opposed to knocking them through a building. It's a quirk of the way Hero System powers function.

QFT.

I don't build characters who aren't proficient with at least one social skill that can get them out of a jam. The OP posted that '... none of the skill tests create a chance of flight ...' That's incorrect in terms of how intimidation works in most games that have that skill available for characters. It's literally creating fear in a target, which done successfully means the foe flees, or at least backs off.

And skills are mechanical aspects of the game, not fluffy narrative junk. Not for nothing, but when I see people complaining about skills it usually means they aren't any good at using them 

A squishy character fighting defensively, thus likely foregoing dealing damage, to block an enemy from ganging up on the party's primary damage dealer is meaningfully contributing to the fight. Suppression is also a thing, a shower unlikely to deal damage of projectiles would still prevent a combatant from forgoing defenses to maximize damage output. Optimizing combat also leads to an issue where it would run more optimally if one player was controlling all friendly characters. It's the individual risk reward balance against the needs of other characters or the party that makes combat decisions meaningful and interesting. Moreover, if the combat is reduced to HP attrition then nothing but HP attrition will matter, as opposed to movement, positioning, suppression, fatigue, or morale.

I think the most general answer to this type of question is in how it relates to the action economy.

If your PCs are facing more powerful foes, the opposing side usually has fewer actions than the PCs. In this case, damaging is less important than denying them actions.

If your PCs are facing less powerful foes, the opposing side usually has more actions than the PCs. In this case, damaging is more effective than denying them actions.

Pathfinder 2's system actually heavily encourages you to deny opponents actions, rather than using all of your actions on attacks.

Pure hp attrition is really boring.  If inflicting hp damage is always optimal then the system.needs an overhaul.

The best action in most turn-based games irrespective of the details (and even real life) is to either improve your action economy faster than the opposition can, or hinder the action economy of the biggest threats in the opposing force faster than they can respond. Or both, if you can achieve it. Of course, a dead enemy has the worst imaginable action economy. So unless the details of the game make it worth debilitating the enemy, simply burning the opposition's HP down as fast as possible will also happen to be the best option for destroying the opposition's action economy.

Just allowing characters to give each other a small bonuses by helping with a skill or attribute test by comparison is almost never viable, I'd agree. Some people said it best earlier in the thread though: there are exceptions when it's so risky and pointless for your character to try to do direct damage that you're better off just pumping up an ally who has better odds. That can be really unfun too though, if that's a regular occurrence.

To prevent direct damage from being the best way to approach a combat, I think you need to either:

• Tune the math so it's sometimes more risky to go for direct damage than to help an ally who has better odds.
• There are too many opponents or they may arrive in unpredictable order/waves so you can't just nova the enemy to death.
• The enemy that is least-likely to be killed with direct damage is also the one that represents the most active and dangerous enemy in the opposing force (like a dragon).

It helps if debilitating conditions also enable more effective basic actions on those debilitated targets. It's gamey, but something like a mark that can be exploited by different powers can make certain combinations of powers or certain tactics more beneficial from time to time.

Quote from: Old Aegidius on June 14, 2023, 04:07:27 AMOf course, a dead enemy has the worst imaginable action economy.

Nitpicking:  Charm.  Dying friend that isn't dead yet.  Both can contribute to a negative action economy, not just add nothing. 

When reading the original post, I thought about systems where defensive actions, like dodging or parrying, are distinct actions from attacking. So you can choose a defensive action or an offensive action.

From my experiences with fencing and medieval combat simulations, I found that dichotomy inaccurate. A fighter can smoothly combine defensive and offensive actions. For example, fencers will try to follow a parry with a riposte attack. The actions flow together and take a split second. I wonder if a system could allow the player making a successful defensive action a chance to immediately add an attack.

Quote from: warwell on June 14, 2023, 08:33:49 AM
I wonder if a system could allow the player making a successful defensive action a chance to immediately add an attack.

Harnmaster does this with the "Counterstrike" defense (though on equal levels of success both combatants hit each other). Stormbringer/Elric! allow for ripostes on a critical parry when using weapon + shield or two weapons. There are probably other such systems.

tanking of different kinds comes to mind as a possible net benefit in some situations. if one can utilize the environment and/or use dodge, rage, taunt, patient defense, what-have-you to draw the attention away from and/or absorb some incoming damage for the squishies in the party - who are often high damage dealing glass cannons as well -  that might be better than just doing damage.

Many good points on both sides of the issue have been made here. To wit, no one has addressed the greater issue which is game design based upon the encounter. WOTC began with this model, and despite several iterations of implementation, has never abandoned it. In original D&D, which many dismiss due to combat being deemed too deadly, having to roll the dice for ANY reason, was a sub optimal choice. Remember that combat XP awards were paltry compared to that for discovering and looting treasure. Almost every modern iteration of the game has been all about die rolls. Die rolls to do this, do that, almost anything you can think of besides taking a shit. Anytime you roll for something, you are putting your fate in the hands of random chance. Over time, the game has removed agency from both DM and player, and subordinated it to the whims of fate. Players seem to have a gleeful time constantly shooting craps for their characters fate. The deadliness of OD&D combat was a feature not a bug, and very few appear to realize that. In the same way that thief skill chances were misunderstood and implemented ( because they were poorly communicated in the rules), they were taken as a hard measure of even attempting any of the described activities. The intent was that nearly anyone could attempt to be stealthy and possibly succeed based on how intelligent their attempt was. The thief skills table was designed as a saving throw mechanic for these specialists to succeed even when their careful plan was thwarted because they were that good. After all, it isn't as if stealth operations were absent from the game when there were only three classes. The communication mistake was made because the popularity of D&D beyond the relatively small wargaming community was very unexpected. Things that the designers assumed everybody knew were incorrect once the game left that wargaming community.

So, returning to the original topic, the game has moved further from clever decision making by the players being the largest determinate measure of success to an ever increasing reliance on the whims of fate due to encounter focused design. A great cinematic representation of this can be seen in the movie Heartbreak Ridge. The platoon tells gunny Highway that their current location is THE ambush spot, to which gunny replies "who the fuck says we are going to ambush them here?" This kind of illustrates the difference of clever play vs the adventuring environment and encounter based design.