Reply to thread

Message

<blockquote data-quote="EzekielRaiden" data-source="post: 6866230" data-attributes="member: 6790260">...but it does remove the key balancing element of random scores.I totally get that most people use random scores because they want to be &quot;surprised&quot; by the character they get. That's fine, perfectly cromulent. The problem comes in when we start rejecting only specific kinds of rolls.The <a href="http://anydice.com/articles/4d6-drop-lowest/" target="_blank">average result</a> for the *whole* probability space of 4d6-L is {16, 14, 13, 12, 10, 9}, approximately. But when you cut out all sets of rolls where the net mod isn't at least a certain number, or whatever other standard, you are shifting this average up. Whenever a DM allows a merely &quot;mediocre&quot; roll (say, {14, 13, 12, 10, 10, 8}) to be re-rolled, it's setting the midpoint of the distribution much higher. To the point that, on average, the *kept* arrays will be several points higher than what you could get from a normal point buy. And, with 5e, this is compounded by the fact that there are several results (16, 17, 18) which are not possible to buy, but are possible to roll, while neither system is particularly capable of generating the nasty low-number possibilities. It is a rare DM who forces a player to keep more than one sub-8 stat, yet legitimately rolling multiple 18s, or even multiple 16s, is met with congratulations--despite being the equivalent of 3s and 5s, respectively.So no, it's not &quot;cheating&quot; in the sense of &quot;secretly violating a rule.&quot; But it is &quot;cheating&quot; in the sense of &quot;milking the system for an unfair advantage.&quot; Plying your DM's heartstrings to remove the balancing effect that's supposed--mathematically--to keep rolling from being completely superior to point-buy.Would you prefer the term &quot;exploitative&quot;?</blockquote>

[QUOTE="EzekielRaiden, post: 6866230, member: 6790260"] ...but it [I]does[/I] remove the key balancing element of random scores. I totally get that most people use random scores because they want to be "surprised" by the character they get. That's fine, perfectly cromulent. The problem comes in when we start rejecting [I]only specific kinds[/I] of rolls. The [URL="http://anydice.com/articles/4d6-drop-lowest/"]average result[/URL] for the *whole* probability space of 4d6-L is {16, 14, 13, 12, 10, 9}, approximately. But when you cut out all sets of rolls where the net mod isn't at least a certain number, or whatever other standard, [I]you are shifting this average up[/I]. Whenever a DM allows a merely "mediocre" roll (say, {14, 13, 12, 10, 10, 8}) to be re-rolled, it's setting the midpoint of the distribution [I]much[/I] higher. To the point that, on average, the *kept* arrays will be several points higher than what you could get from a normal point buy. And, with 5e, this is compounded by the fact that there are several results (16, 17, 18) which are not possible to buy, but are possible to roll, while [I]neither[/I] system is particularly capable of generating the nasty low-number possibilities. It is a rare DM who forces a player to keep more than one sub-8 stat, yet legitimately rolling multiple 18s, or even multiple 16s, is met with congratulations--despite being the equivalent of 3s and 5s, respectively. So no, it's not "cheating" in the sense of "secretly violating a rule." But it [I]is[/I] "cheating" in the sense of "milking the system for an unfair advantage." Plying your DM's heartstrings to remove the balancing effect that's supposed--mathematically--to keep rolling from being completely superior to point-buy. Would you prefer the term "exploitative"? [/QUOTE]

Verification