DND at base has a success at a cost idea in the DMG I believe.
Missed the monsters AC by 1? You can go for the hit, but you get disarmed in the process.
Fail the save by 1? You can pass that fireball save, but you're prone.
Fail the check by 1? You can succeed that, but a consequence for something else. You pick the lock - loudly.

One of my favorite things from Critical Role is the “planning phase” though. Why would I ever want to skip it? “At dawn, we PLAN!”

I think people forget that DM's should be following some rules to help the story move forward. If you need the party to do something to forward the story, don't require a roll for it. If you want to have some skill checks that let them do things that will give more information fine, but if the story can not move forward if they fail a thing that is not good DMing.

I like to borrow stunting from exalted.  When players narrate especially cool character actions, they get a small bonus to the action. In Exalted, the reward was 1-3 extra dice to roll. In DnD, I'm thinking just give +1, +2, or +3 to the roll. This incentivizes roleplay and creativity, and is pretty easy to implement. It also allows characters a tiny bit of narrative control. It can even be used similarly to the flashback mechanic, to justify characters being prepared for the adventure, without needing to disrupt the narrative for a flashback.

One mechanic from not a ttrpg, but a video game that seems heavily inspired by D&D and ttrpgs, that I've been thinking of translating into D&D rules is Wildermyth's death system. Basically, if one of your characters dies, they get a choice: they can deal some damage to the monster that killed them, give a buff of some kind to the rest of the party, or (this is the important one) survive, but leave for the rest of that combat and be maimed, which usually gives them some negative effect in future combats, like if your arm gets chopped off you can't use staves or bows. Sometimes it does give you more interesting options like one time I got to turn my character into basically a magic item, but that's the general gist of the system. The main thing that is difficult to turn into D&D is that you can only survive a death once per chapter in Wildermyth, and if you're not running a module, or even sometimes if you are, it might be difficult to figure out what the end of a chapter is.

Expanding on the “success at a cost” part, maybe we could borrow Call of Cthulhu’s mechanic for forcing a success (called “pushing a roll”) and degrees of success, so depending on how well you roll, the better or worse the resulting situation may be; when you push a roll, you get a free reroll, but you must use the second roll, and if you still fail, the worse scenario happens (the GM may allow the player to choose the punishment for the situation).

For example: you’re talking with a noble, and you roll a persuasion check and fail, you push the roll and still fail, your character has offended the noble and they challenge you to a duel so they may regain their honor, if talking to the captain of the guard, they get arrested or for the next week the guards are more rigorous against the party (perhaps arresting them for lesser transgressions).

I love the idea of bonds and obligations.  All too often, players come up with an interesting backstory, but DMs/GMs ignore it because it doesn't fit into the campaign.

Rolling a d% chance of any player’s backstory becoming relevant to the main plot can force the GM to acknowledge the work that players sometimes put into their characters, but without derailing the campaign.

also from 13th Age i really like the "fail forward" rule, that lets the party retreat from a dire situation, escaping with their lives, but at the cost of things getting worse (as they called it, a "camapign loss") meaning something going on is gonna be more difficult, like another fight will have more opponents, a target will have more security, people in town will be more on the hostile side, or they are forced to flee in a way that maks nearly imposible for them to get a secondary objective they had in mind, and so on

I have been planning a run for my shadowrun game where instead of letting the party plan for 3 sessions I was going to dump them in a run, get where everything starts going to shit, flash back to telling them what they're actually doing here and then say you have X 'luckily i prepared for this' points they can use on this run.

That adversity token idea is kinda like my dm's idea of invoking a flaw. If we rolled a nat 1 while trying to do anything we can tie it to one of our characters flaw. We get a point of inspiration and the dm gets to keep this incident in his backpocket to bring up at a later date

Combat Escalation actually works thematically too. You're getting a better idea of how to hit to hurt.

I would use most of them but with Bonds I have another take:
Whenever characters create a bond, whether it was during character creation (we grew up together; we've been dating for a while; we trained under the same master etc.) or during game (we saved each others asses many times; we hanged around the taverns a lot; we sparred for last few weeks) I would give them a card with said bond that has both characters name on it and the name of the bond. During any time it would apply any of the two can use that card to gain advantage or flat bonus to an action that can be explained by the bond. For example: "Wizard wants to gain a bonus to stealth and invokes his friendship with Rogue, who previously gave him a hint that could come in handy". Whoever uses the card flips it face down and it gets unflipped after short rest.

Success at a cost just sounds like Failing Forward but slower. Instead of "bargaining" with the player or even asking if they want to succeed but with a cost, you as the DM just say "you did it, but this also happened". Let's take the rogue picking a lock scenario. The rogue rolls to pick the lock that has a DC of 15 but they rolled a 12. You as the DM could then say "You picked the lock, but it bent your lockpick. Checks to pick locks while using those will be at disadvantage until you can find other lockpicks." No negotiating or bargaining or even asking them, you just do it.

Soft fail states are always an incredible tool in game design! Great video!

13th Age escalation die applies in various ways to different enemies. Dragons get it the same as players (that's a big oh crap moment when they realize it). Other monsters have abilities that unlock at certain escalation levels. For instance, this ability can be used when escalation die is even.

Primarily its a pacing device to help speed combats that are dragging along and (for me at least) makes it easier to remember "in round 3, x happens" events.

Also, for icons, I run them more as organizations. Sure, the Dragon Emperor rules the land, but Sir Adventurer the Brave probably doesn't know him. He does know how the Empire is run, certain figures at different levels of the organization, etc. Also it really helps players develop their backgrounds.

I remember Jacob from xp to level 3 said their cleric died and the paladin was willing to give up a level of paladin to bring her back. fucking boss bro.

The Shadowrun rule for 'timed' repeats is already partially covered by a little rule in the DMG that most DMs either miss, forget, or just ignore.  That is that if the players have a chance of success and no penalty for failure, don't make them roll.  It just happens.  That rule would not apply in your 'pick the lock guard is coming' scenario, because if you fail the guard may catch you.  But it applies in the library scenario.  If the player has access to the library, and the library has the information they are looking for,  don't make them roll, just give them in info.  On the reverse, if the player only has a day to search and an antagonist is trying to find the same books, make them roll.   And 12 hours isn't going to give you exhaustion, unless your days are less than 20 hours long.

Frankly, I would like to bring back the formalized 'Take 10' and 'Take 20' rules from D&D 3.5E

Failing with a cost also needs to scale with the importance of the scene or encounter.  Failing by 2 against a random kobold should cost less that failing by 2 against the final boss.

I'm not fond of the icon rule.  Not everyone has connection, especially at the start.  And just having deus ex machina to have some random super character come in and try to save the day.  It might be okay to add in as an ongoing thing, you develop these links as the game continues.

I also am afraid both the obligations and icons might interfere with agency as applied in D&D, forcing things on the PC.

Frankly the icons/obligations/bonds makes it sound like you would just rather be playing FATE...  You have your traits that are relationships, benefits, and penalties that shape the campaign.

The alternating speed turns is not a simple system.  The order slots does sound interesting.

Flashbacks are good.  If you combined them with the adversity points it makes an interesting system, for things like Oceans/Leverage, where it seems like things are going bad, but wait, the team had something set up to deal with it, or the failure was actually planned in advance.

Ultimately, listening to this,  I'm thinking of a hybrid rule based mainly on the FATE point system.  Players start each session with one 'inspiration point'.  They can earn more points the normal way, or if you fail by 7 or more, you get another inspiration point.  If you fail by 1 or 2, you can spend an inspiration point to succeed instead.  Similarly, if you succeed by 0 to 2, you can choose to fail instead to gain a point.  You can still spend inspiration to reroll a failure.  You can also spend it on flashbacks.  Finally, you can go into inspiration debt, in which case the DM can force you to reroll, or to fail if on a roll that succeeded by 0 to 2, to pay off a point of that debt.

I’m stealing 9 of these. One I like from Fate is how they have qualities that they can activate or be activated by the DM by offering rewards. This is very similar to some of the things you mentioned.

You won me over with this video. Subscribed and followed. Im also going to check out ER Campaign 2.

The one I heard and find really interesting is to replace the Morality Table(Lawful, Chaotic and so on) with the MTG Color Wheel, because it is much dircter in defining your character