Can you take 20 in 4th edition?

[LostSoul]

if you recall the old 'Take 20' rule meant that you can only use it where there is no downside to failure.

If there is a situation where there is no downside to failure then there is no reason why you can't simply retry.

I think the idea is that, if there is no downside to failure, you shouldn't be rolling. You can just succeed.

This is an idea that deserves more explanation, I think.

 

[James McMurray]

What disturbs me is that lockpicking isn't even a trained only skill. Nothing's stopping the Fighter from taking 20 to pick a lock

If you don't want the fighter getting through the door, you either don't put a door there, use a lock stronger than the fighter's untrained score + 20, or apply some consequence to the roll such as a trap going off every time there's a failure.

It seems people here asuume 'taking 20' means an automatic success. It doesn't and never has. It will always depend on the DC of the task.

It's not an issue of taking 20 being autosuccess, but rather an issue of taking 20 being meaningless in the 4e mindset. If you should only roll when something is important and has a chance of failure, then take 20 is unnecessary. Either the DM says "you do it," "you don't do it," or "roll some dice."

You can't take 20 if there's a chance at failure that has something bad happen, or if it states specifially in the skill. Someone here talked about Perception checks to find something. If you read the description, if you fail, you can't try again.

If you read the DMG, not only can you retry perception checks when searching, the DM is encouraged to let you assume you'll eventually roll a 20.

 

[77IM]

I don't allow Take 20 because it's boring. The DM sets the DCs, so with Take 20 in effect, the DM is basically deciding ahead of time whether or not the lock can be picked or the secret door can be found. Since there's no cost, risk, or downside, players feel free to Take 20 on every search check or pick locks check, removing what might otherwise be an interesting decision.

Worse, obstacles that are vulnerable to Take 20 out-of-combat often have DCs so high that you might as well not even bother during combat. It's way more exciting to pick a lock or climb a wall during a battle so the DCs should be low to accommodate.


I don't usually allow retries for checks with no downside. I do have a house rule for getting a bonus on your roll for spending increasingly long periods of time (this provides the check with enough of a cost to allow retries). But I really like Larry's house rule (roll until you get a 10 or better) and may start using that instead.

-- 77IM

 

[HarbingerX]

If you read the DMG, not only can you retry perception checks when searching, the DM is encouraged to let you assume you'll eventually roll a 20.

Right - but this is only when searching a room to AVOID making the players roll. If the player is doing an active Perception check, retry is NOT allowed, 'unless circumstances change' PHB p.186

 

[Morgan_Scott82]

The people who've stated that taking 20 wasn't an automatic success are correct, taking 20 is never explicitly said to be an automatic success. However, I believe it creates an artificial and significant effect on skill checks.

If a check required a roll of 18-20 to succeed take 20 increases the chance of success from 15% to 100%, however if the check required a roll of 21 well then take 20 increased the chance of success from 0% to, well 0%, not much of an improvement on that count.

If PC can take 20, they will. This puts DMs in a difficult design position where they're forced to make predetermined binary judgements about the success or failure of the PCs during the adventure design process. As a DM I know what my party rogues open locks skill is, I know that he'll take 20 on open locks, so when it comes time to set the DC of the lock what I'm really doing is deciding ahead of time if the rogue succeeds or fails, with zero variability. As a DM who wants to craft a particular story I'm going to predetermine the outcome that better suits my story.

Ah but there's the rub. If I'm already predetermining the outcome when I set the DC for "take 20 situations" then why bother with the setting a DC at all? Why not just narrate it away. In what way is it significant how impossible the impossible to pick lock was? In what way does it improve the story to know how easy the easily picked lock was?

I firmly believe that the role of mechanics is to arbitrate situations with variable outcomes. Whether an attack hits or misses has a variable outcome, how much damage you take when you fall into a pit has a variable outcome, if you're setting a DC for a skill check (thereby involving mechanics) that should mean the skill check has a variable outcome. That just isn't the case when taking 20, since as we've already seen taking 20 implies a predetermined outcome with zero variability.

This means that taking 20 has no significant impact on the rules or the story, the result is predetermined, not at all variable, and therefore requires no mechanics.

 

[Guyanthalas]

...
Ah but there's the rub. If I'm already predetermining the outcome when I set the DC for "take 20 situations" then why bother with the setting a DC at all? Why not just narrate it away. In what way is it significant how impossible the impossible to pick lock was? In what way does it improve the story to know how easy the easily picked lock was?

I firmly believe that the role of mechanics is to arbitrate situations with variable outcomes. Whether an attack hits or misses has a variable outcome, how much damage you take when you fall into a pit has a variable outcome, if you're setting a DC for a skill check (thereby involving mechanics) that should mean the skill check has a variable outcome. That just isn't the case when taking 20, since as we've already seen taking 20 implies a predetermined outcome with zero variability.

This means that taking 20 has no significant impact on the rules or the story, the result is predetermined, not at all variable, and therefore requires no mechanics.

I think that it needs to be looked at with a "what is the plenalty for failure" mentality.

Say Rouge Guyanthalas IV tries to pick a lock in 4E. When determining the difficulty, what is the end result suppose to be? Are you really going for "does he get a cool magic item, or not"? I really hate that design style. I think if you're going to put in a piece of treasure, the characters better damn well get it.
Here is my take. No take 20 on lock picks at all, but Guyanthalas only gets one shot at it. Success, he opens the chest. If he fails on the check, trigger a trap. After the trap has been bypassed (destroyed, dismantled, whatever), then allow an instant success on the lockpick. Heck, give the party XP for the trap either way since it was bypassed with a success.

This should be able to adapt to any skill check with enough creativity.
Thoughts?

 

[UngeheuerLich]

In my 3.5 games taking 20 was the standard to do all things when not under pressure... but it takes time as stated: 20x normal time...

considering that searching a 5x5 square is a standard action (3-6 secs) searching a 5x5 square thoroughly is 2 mins (as stated in the rules) so closely examine a complete dungeon is possible, but you should plan in some days of search...

lifting a gate should be a binary check...

lifting a gate is about DC 23... a task that only a strong adventurer can succed at... the wizard should fail no matter how much he tries... (ever seen the gamers... i don´t want such a situation on my table)

how does that translate to 4th edition?

I honestly don´t know...

I would guess i will do attribute checks without level bonus... with the revision of skill DC´s (the +5 was dropped) attributes are always 5 points to low to get a good result... it is always an untrained skill check...

taking 20 on pick locks when heroic locks are DC 20 also seems quite wrong, so i guess for untrained persons the chance of failure will be breaking the tool they used to open the lock... so no untrained fighter with dex 10 opening locks instead of just smashing it...

taking 20 on hide checks to set an ambush... only if there is much time... but maybe its better to just use the highest roll the monsters rolled... using the lowest roll would kill most ambushes with many people involved...

 

[Biko]

I think lockpicking should be handled as a skill challenge. This way it won't be just one single try, if you fail the lock is either broken or jammed.

 

[James McMurray]

There are skills that allow retries with no negative effects. If you don't allow taking 20, do you make the player roll over and over again until he manages to roll a 20?

 

[77IM]

Yes, but each re-roll takes 10x as long as the previous one. If there's no cost/penalty for failure, or no chance for failure, then rolling is a waste of time. Spending tons of time is a way of artificially creating a cost -- the roll now has meaning and creates tension because if you fail, the next one might take more time than you have available... Of course, I also give a +5 bonus on the re-rolls since you're taking extra time on it. I think this strikes a good balance between the automatic, DM-pre-specified outcome of Take 20, and the harsh failure penalty of disallowing re-rolls altogether.

Take 20 was the Short Rest of 3E. Unless there was substantial time pressure or a dangerous environment, you could expect players to do it constantly.

-- 77IM