Is there any way to gauge the statistical impact of advantage vs. abilities like reliable talent (trance of order, etc.) which convert a 9 or lower to a 10 on d20? In particular, I am trying to assess the value of one or the other for attacks.
For comparison your roll average with advantage is 13.825 and with reliable talent is 12.75 (lower than I though it'd be), compared to a normal roll average of 10.5.
Of course reliable talent can also get dis/advantage.
As DxJxC points out, reliable talent doesn't really look like much when you take the average. I found most of this online, and I probably still put more work into it than I should have. If you're trying to beat a low number, reliable talent is a guarantee, but doesn't help you at all if you're trying to beat an 11 or higher. Advantage helps across the board, with the amount of help it provides diminishing as you try to hit a higher target. If the DC you are trying to beat is 10 or lower, reliable talent is better. If you're trying to beat an 11 or higher, advantage is better. If you're trying to hit a natural 20, advantage doubles your chances.
ROLL TO MEET/EXCEED
DISADVANTAGE
NORMAL
ADVANTAGE
RELIABLE TALENT
20
0.002
0.05
0.098
0.05
19
0.01
0.1
0.191
0.1
18
0.022
0.15
0.278
0.15
17
0.039
0.2
0.359
0.2
16
0.062
0.25
0.437
0.25
15
0.089
0.3
0.51
0.3
14
0.123
0.35
0.576
0.35
13
0.16
0.4
0.639
0.4
12
0.202
0.45
0.698
0.45
11
0.249
0.5
0.751
0.5
10
0.303
0.55
0.798
1
9
0.361
0.6
0.84
1
8
0.424
0.65
0.877
1
7
0.492
0.7
0.91
1
6
0.564
0.75
0.938
1
5
0.64
0.8
0.96
1
4
0.723
0.85
0.978
1
3
0.811
0.9
0.99
1
2
0.903
0.95
0.998
1
1
1
1
1
1
Reliable talent with advantage is the same as regular advantage, but you will meet/beat a 10 100% of the time.
Because I enjoy any excuse to use Google Sheets recreationally I made a quick sheet that computes the probability to equal or beat a given number via a normal roll, reliable talent with normal roll, and roll with advantage. I think plotting these values out on a graph really helps to understand just what the advantages of Advantage are and you can take a look here: https://docs.google.com/spreadsheets/d/1TkKoQZS-eSlp6yp1SfNVaiSFodYc3TC1v5XzfkDMXeA/edit?usp=sharing
For Reliable Talent though I think an intuitive understanding is sufficient to understand how it is helpful. If you need to roll an 11 or higher it doesn't help at all but if you only need to roll a 10 or less it guarantees success.
If you really want to know which is more helpful though you need more information than just this though. You also need to know the distribution of what number you need to roll to succeed. What this looks like will vary from DM to DM and character to character as both the DC set by the DM and the modifiers you character has that apply to the check will impact what number you actually need to meet or beat on the d20 roll.
Its two separate rolls, so they must be treated as two separate rolls.
You have a 5% chance to roll any number the first roll. And its the same for the second roll. You could roll two 1's Rolling twice just gives you two chances at rolling a good enough number. It doesn't increase your chance of getting any single number.
Its two separate rolls, so they must be treated as two separate rolls.
You have a 5% chance to roll any number the first roll. And its the same for the second roll. You could roll two 1's Rolling twice just gives you two chances at rolling a good enough number. It doesn't increase your chance of getting any single number.
Yeah but probability is stupid so it doesn't actually work that way.
Rollback Post to RevisionRollBack
Paladin main who spends most of his D&D time worldbuilding or DMing, not Paladin-ing.
Its two separate rolls, so they must be treated as two separate rolls.
You have a 5% chance to roll any number the first roll. And its the same for the second roll. You could roll two 1's Rolling twice just gives you two chances at rolling a good enough number. It doesn't increase your chance of getting any single number.
I meant to link to this in my previous post but the math YouTuber Matt Parker did a video about calculating the probability of rolling a given number when rolling two or more dice and taking the highest result: https://youtu.be/X_DdGRjtwAo
Its two separate rolls, so they must be treated as two separate rolls.
You have a 5% chance to roll any number the first roll. And its the same for the second roll. You could roll two 1's Rolling twice just gives you two chances at rolling a good enough number. It doesn't increase your chance of getting any single number.
You really arguing that advantage and disadvantage don't increase or decrease the likelihood of getting a good roll compared to a normal roll based on the 5% chance that they roll the same number?
Here is the way you can visualize probability, make a chart/table of possible dice rolls. 20×20 has 400 possible combinations of rolls so lets use d4s, but the results will be similar.
11 12 13 14
21 22 23 24
31 32 33 34
41 42 43 44
I've highlighted in green all the combinations that result in a 4 with advantage. Without advantage the odd of rolling a 4 is 1/4 (25%), with advantage as we can see it is 7/16 (43.75%). Similarly the odds of rolling a 3 or higher jumped from 50% to 75%, on a d20 this is exactly the same odds as rolling an 11 or higher with advantage.
The second roll has the same 5% chance to roll ANY number.
Which if your talking about a rogues Reliable talent you also add in any other bonuses he gets which should push things well above what is normally needed.
Reliable talent works on a rather narrow scope of rolls whereas Advantage works on a far larger number of rolls and is pretty easy to get.
The second roll has the same 5% chance to roll ANY number.
Yeah, and since the thing you're taking is the higher of the two, they're not independent events. Advantage absolutely gives you a higher chance of rolling any specific number above 10. DxJxC demonstrated that.
The second roll has the same 5% chance to roll ANY number.
Sure, go ahead and ignore the entire math lesson that visually demonstrates how advantage improves high roll odds and double down on an irrelevant factoid that you are misunderstanding.
Yes, 1 d20 has a 5% chance of rolling any 1 number (including the 1 number the other die landed on, which is what I was referring to). If you are only rolling 1 die and at a time, your odds of succeeding at any 1 roll are not higher than any other 1 roll, but if you roll multiple dice and only need 1 to succeed, your odds of success are higher than with 1 roll.
I think 20 and even 4 are just too big of numbers to get this point across. Let's use a coin or d2 if you prefer. Let's say you win if a coin lands on heads. Flip 1 coin, 50% chance of winning. Why are you arguing that if you flip 100 coins, there is still only a 50% chance of getting a single heads result? Because that is what you are doing.
Keep in mind that reliable talent doesn't work for attack rolls. It's just for ability checks that the character is proficient in. Rogues get this at level 11 so they will have +4 to proficiencies. This is doubled by anything with expertice (commonly things like steath, thieves tools, perception, investigation). This means that an 11th level rogue will succeed on dex ability checks with a DC 23 or less assuming they maxed it earlier. By level 17, it jumps to 27. Assuming you hide in combat, this would also give you advantage, and the enemy would have to roll very high to notice you. For perception/investigation, it makes your active rolls the same as your passive at a minimum. So if you have a DM that hates using passive for checks, this literally makes it the same, but you could still roll higher. If you have expertice in perception, and even a +2 in wisdom, you're looking at a 20 minimum which will typically let you notice most anything.
So if you're really just looking at attacks, then advantage is where you want to go, but if you're looking for skill boosts, reliable talent is awesome. It's really focused as others mentioned but those are typically the things you want to do well anyway.
I was referring to the class of abilities like reliable talent that convert a number <10 to 10. I am not aware this roll modification type has a given name (like "advantage") and thus cited reliable talent as it is the most well known example. Other examples (e.g., Trance of Order) do affect attack rolls.
Ok I see. So what are you trying to do exactly? Trance of Order, for example, is a level 14 Sorcerer ability. That would put the attack roll at a minimum of 20 which will hit most things. I'd much rather have that than adavantage but that's a very specific high level ability that does other things as well. Are you trying to homebrew something or multiclassing? What's the source of the advantage (recklass attack, steady aim, etc.). Level of play also matters. Some additional context would be helpful for evaluation.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Is there any way to gauge the statistical impact of advantage vs. abilities like reliable talent (trance of order, etc.) which convert a 9 or lower to a 10 on d20? In particular, I am trying to assess the value of one or the other for attacks.
Advantage is generally thought as being the same as +4 or +5, depending on the exact target number.
Any ability that converts a 9 or lower to a 10 is guaranteeing that you get 10+modifier, so works best if your modifier is high.
For comparison your roll average with advantage is 13.825 and with reliable talent is 12.75 (lower than I though it'd be), compared to a normal roll average of 10.5.
Of course reliable talent can also get dis/advantage.
As DxJxC points out, reliable talent doesn't really look like much when you take the average. I found most of this online, and I probably still put more work into it than I should have. If you're trying to beat a low number, reliable talent is a guarantee, but doesn't help you at all if you're trying to beat an 11 or higher. Advantage helps across the board, with the amount of help it provides diminishing as you try to hit a higher target. If the DC you are trying to beat is 10 or lower, reliable talent is better. If you're trying to beat an 11 or higher, advantage is better. If you're trying to hit a natural 20, advantage doubles your chances.
Reliable talent with advantage is the same as regular advantage, but you will meet/beat a 10 100% of the time.
"Not all those who wander are lost"
Because I enjoy any excuse to use Google Sheets recreationally I made a quick sheet that computes the probability to equal or beat a given number via a normal roll, reliable talent with normal roll, and roll with advantage. I think plotting these values out on a graph really helps to understand just what the advantages of Advantage are and you can take a look here: https://docs.google.com/spreadsheets/d/1TkKoQZS-eSlp6yp1SfNVaiSFodYc3TC1v5XzfkDMXeA/edit?usp=sharing
For Reliable Talent though I think an intuitive understanding is sufficient to understand how it is helpful. If you need to roll an 11 or higher it doesn't help at all but if you only need to roll a 10 or less it guarantees success.
If you really want to know which is more helpful though you need more information than just this though. You also need to know the distribution of what number you need to roll to succeed. What this looks like will vary from DM to DM and character to character as both the DC set by the DM and the modifiers you character has that apply to the check will impact what number you actually need to meet or beat on the d20 roll.
I think you guys are calculating advantage wrong.
Its two separate rolls, so they must be treated as two separate rolls.
You have a 5% chance to roll any number the first roll. And its the same for the second roll.
You could roll two 1's
Rolling twice just gives you two chances at rolling a good enough number. It doesn't increase your chance of getting any single number.
Yeah but probability is stupid so it doesn't actually work that way.
Paladin main who spends most of his D&D time worldbuilding or DMing, not Paladin-ing.
I meant to link to this in my previous post but the math YouTuber Matt Parker did a video about calculating the probability of rolling a given number when rolling two or more dice and taking the highest result: https://youtu.be/X_DdGRjtwAo
You really arguing that advantage and disadvantage don't increase or decrease the likelihood of getting a good roll compared to a normal roll based on the 5% chance that they roll the same number?
Here is the way you can visualize probability, make a chart/table of possible dice rolls. 20×20 has 400 possible combinations of rolls so lets use d4s, but the results will be similar.
I've highlighted in green all the combinations that result in a 4 with advantage. Without advantage the odd of rolling a 4 is 1/4 (25%), with advantage as we can see it is 7/16 (43.75%). Similarly the odds of rolling a 3 or higher jumped from 50% to 75%, on a d20 this is exactly the same odds as rolling an 11 or higher with advantage.
Not a 5% chance to roll the same number.
Its a 5% chance to roll any number on a die 20.
The second roll has the same 5% chance to roll ANY number.
Which if your talking about a rogues Reliable talent you also add in any other bonuses he gets which should push things well above what is normally needed.
Reliable talent works on a rather narrow scope of rolls whereas Advantage works on a far larger number of rolls and is pretty easy to get.
Yeah, and since the thing you're taking is the higher of the two, they're not independent events. Advantage absolutely gives you a higher chance of rolling any specific number above 10. DxJxC demonstrated that.
Sure, go ahead and ignore the entire math lesson that visually demonstrates how advantage improves high roll odds and double down on an irrelevant factoid that you are misunderstanding.
Yes, 1 d20 has a 5% chance of rolling any 1 number (including the 1 number the other die landed on, which is what I was referring to). If you are only rolling 1 die and at a time, your odds of succeeding at any 1 roll are not higher than any other 1 roll, but if you roll multiple dice and only need 1 to succeed, your odds of success are higher than with 1 roll.
I think 20 and even 4 are just too big of numbers to get this point across. Let's use a coin or d2 if you prefer. Let's say you win if a coin lands on heads. Flip 1 coin, 50% chance of winning. Why are you arguing that if you flip 100 coins, there is still only a 50% chance of getting a single heads result? Because that is what you are doing.
Keep in mind that reliable talent doesn't work for attack rolls. It's just for ability checks that the character is proficient in. Rogues get this at level 11 so they will have +4 to proficiencies. This is doubled by anything with expertice (commonly things like steath, thieves tools, perception, investigation). This means that an 11th level rogue will succeed on dex ability checks with a DC 23 or less assuming they maxed it earlier. By level 17, it jumps to 27. Assuming you hide in combat, this would also give you advantage, and the enemy would have to roll very high to notice you. For perception/investigation, it makes your active rolls the same as your passive at a minimum. So if you have a DM that hates using passive for checks, this literally makes it the same, but you could still roll higher. If you have expertice in perception, and even a +2 in wisdom, you're looking at a 20 minimum which will typically let you notice most anything.
So if you're really just looking at attacks, then advantage is where you want to go, but if you're looking for skill boosts, reliable talent is awesome. It's really focused as others mentioned but those are typically the things you want to do well anyway.
I was referring to the class of abilities like reliable talent that convert a number <10 to 10. I am not aware this roll modification type has a given name (like "advantage") and thus cited reliable talent as it is the most well known example. Other examples (e.g., Trance of Order) do affect attack rolls.
Ok I see. So what are you trying to do exactly? Trance of Order, for example, is a level 14 Sorcerer ability. That would put the attack roll at a minimum of 20 which will hit most things. I'd much rather have that than adavantage but that's a very specific high level ability that does other things as well. Are you trying to homebrew something or multiclassing? What's the source of the advantage (recklass attack, steady aim, etc.). Level of play also matters. Some additional context would be helpful for evaluation.