Post/Author/DateTime	Post
Longarms 08-06-07, 01:31 AM	I like the 4d6 rolling method, BUT I think it gives generally too high of scores. Can any who is better at statistics than I am tell me what effect the aformentoned system will have? I really can't figure it out easily. I fall back method would be to do every combo, I guess :cool: My players might like it better than 3d6 :eek: EDIT: THANKS TO SILLYROBOT, we have the numbers. As suspected, this method turns out to be, IMO the much needed middle ground between the harsh 3d6 and the too easy 4d6 drop the lowest. For example: Chance of rolling 14 or higher 3d6 = 10% 4d6 drop the lowest = 35% 4d6 drop the lowest cant drop 1's = 25% Chance of rolling 7 or lower 3d6 = 10% 4d6 drop the lowest = 6% 4d6 drop the lowest, cant drop ones = 12%
Dragonsblood 08-06-07, 02:00 AM	It doesn't really matter what the average is. With a little bit of luck, you get a 50 point buy character, or with a little bad luck you get a 10 point buy, often with PCs in the same party. That's why I never use rolling methods for stat generation in my games.
Ack 08-06-07, 02:04 AM	4d6 drop the lowest has a stand-alone average of 12.24. However, when taken in combination with the PHB rules for which stat arrays are 'unacceptable', it gives an average of 12.5 over the whole array. Rerolling 1s will nudge the average up again slightly. Now, I know I'm opening the same old tired debate in this thread as well, but have you tried point buy? You, as the GM, get to set the power level for the game, and players can sort out their own stats without requiring oversight.
Furion Thamior 08-06-07, 02:08 AM	For the most part it's going to be the same, and probably wouldn't prevent high scores so much as randomly make a decent roll into a bad one, since the highest score one could get if a 1 was rolled is a 13.
Insanit 08-06-07, 02:21 AM	Can't drop ones sounds like an awful idea to me, just think of when someone rolls two ones.
Mock26 08-06-07, 02:28 AM	The not dropping the one could work for you or against you. In a "best" case scenario where someone rolls three 6's and a 1, as someone else pointed out you'd wind up with a 13. But, what if they rolled 4, 4, 4, 1? While that could have been a decent score of 12, with your rule it would become an 8. Over all it will work in your favor of keeping the scores down, but it could turn some decent scores into poor ones. You might be better off trying to do the 3d6 with allowing the rerolls of 1's.
unfortunate_crow 08-06-07, 02:54 AM	3d6 re-rolling 1's works well for us and most of my players prefer it.
Nom 08-06-07, 07:33 AM	Following on from Mock26, it will make your distribution somewhat bi-modal. It won't majorly affect the really good rolls, but will pull many average rolls poor.
Paladin_Hiroshi 08-06-07, 07:38 AM	How about you roll 4D6, drop lowest dice. Roll 3 sets of stats with it, then ask your DM which one he prefers. It's really not that hard. ^^ just a soloution if you don't want to use point buy.
SillyRobot 08-06-07, 08:28 AM	3d6 4d6 4d6 keep 1s 3 1 1 21 4 3 4 6 5 6 10 22 6 10 21 43 7 15 38 74 8 21 62 112 9 25 91 97 10 27 122 134 11 27 148 158 12 25 167 163 13 21 172 146 14 15 160 100 15 10 131 91 16 6 94 70 17 3 54 42 18 1 21 17 216 1296 1296 Expect 10.5 12.24 11.25 Mode 10/11 13 12 Outliers chance 3 0.50% 0.08% 1.60% 18 0.50% 1.60% 1.30% Chance of penalty modifier 37.5% 17.5% 28.9% Chance of bonus modifier 37.5% 61.7% 48.5% The big change from a player's perspective is the chance of getting a 3 is greater than the chance of getting an 18.
High Octane 08-06-07, 08:41 AM	I like the 4d6 rolling method, BUT I think it gives generally too high of scores. Can any who is better at statistics than I am tell me what effect the aformentoned system will have? I really can't figure it out easily. I fall back method would be to do every combo, I guess :cool: My players might like it better than 3d6 :eek: If you are concerned over high ability scores, use point buy. You have concrete control over the power characters have, while still leaving characters customizable.
Warforged Goblin 08-06-07, 08:45 AM	What do I drop in this situation? 1, 1, 1, 6
Roog 08-06-07, 08:50 AM	What do I drop in this situation? 1, 1, 1, 6 Thats easy - drop the 6 for a stat of 3. The hard question is: What do I drop in this situation? 1, 1, 1, 1
SillyRobot 08-06-07, 09:21 AM	Thats easy - drop the 6 for a stat of 3. The hard question is: What do I drop in this situation? 1, 1, 1, 1 Simple enough, take any three of the four -- you get to pick! :evillaugh
Roog 08-06-07, 10:06 AM	Simple enough, take any three of the four -- you get to pick! :evillaugh That doesn't work, the thread title says "can't drop ones". Maybe it gives you a stat of 4.
SillyRobot 08-06-07, 10:14 AM	That doesn't work, the thread title says "can't drop ones". Maybe it gives you a stat of 4. I place that stipulation in the same category of shorthand as 4d6, drop lowest doesn't stipulate only drop a single die. A set of 6,6,6,6 doesn't convert into a 0 score even though all four dice are lowest.
Longarms 08-06-07, 12:01 PM	Silly - how did you generate that chart, thats awesome ;) I like what I see with the chart. Under 3d6, on a per roll basis, someone only has about a 16% chance of getting a 14 or higher, which I think is a little too low. Under 4d6 drop the lowest, can't drop ones, that chance increases to 25%, which I think is about right. The big improvement over normal 4d6 in my eye, is that the low rolls don't go away... BTW - I am in the group of "can't stand point buy." I don't like the absence of low scores (less than 8) because I think those scores, when paired with some good rolls, make some unique characters. I also can't stand point buy because I like to see interesting stat combinations, e.g. a wizard with high STR, which would likely never happen under point buy. Since this is a non-competitive game, I, and most of the players I have had, are not overly concerned about PCs sometimes having unequal power - so the biggest advantage of point buy doesn't appeal to me at all.
SillyRobot 08-06-07, 12:53 PM	Silly - how did you generate that chart, thats awesome ;) I wrote a small program to test all permutations. The actual rules I used were the following Roll 4 dice Total three dice using the following priority Keep all ones rolled (to a maximum of 3 dice) If less than 3 ones were rolled, use the highest rolls of the remaining dice
Johnny_Angel 08-06-07, 01:00 PM	4d6 without dropping ones is what I use; it works fine for me. If you're really wanting to try something different that involves dice rolling, try this: Everybody starts with an 8 in each stat. Roll a d10 for each stat and add that result to the 8. I have no scientific data to prove that this is any better or worse, but it's different than the standard method.
Ace_of_Night 08-06-07, 01:28 PM	Silly - how did you generate that chart, thats awesome ;) I like what I see with the chart. Under 3d6, on a per roll basis, someone only has about a 16% chance of getting a 14 or higher, which I think is a little too low. Under 4d6 drop the lowest, can't drop ones, that chance increases to 25%, which I think is about right. The big improvement over normal 4d6 in my eye, is that the low rolls don't go away... BTW - I am in the group of "can't stand point buy." I don't like the absence of low scores (less than 8) because I think those scores, when paired with some good rolls, make some unique characters. I also can't stand point buy because I like to see interesting stat combinations, e.g. a wizard with high STR, which would likely never happen under point buy. Since this is a non-competitive game, I, and most of the players I have had, are not overly concerned about PCs sometimes having unequal power - so the biggest advantage of point buy doesn't appeal to me at all. A wizard with high strength? So either you're also telling players to invest them in the order they roll it in (First roll is strength, second Dex, etc.) Or are you assuming that someone will actually only have scores of 14 or above with "4d6-can't drop 1s", because that's a 1 in 4096 chance :P Also, where do you get the logic that with 4d6 without dropping 1s you lose the low scores? Keep in mind that the average is a 10, and if you use this new system of yours there's about a 40% chance for each stat to be average or below. Also keep in mind, 3-4 is very far below and could well penalize your character deeply. To get back at what I first said, I would -never- even use the normal 4d6 system, let alone this one, with putting stats in the order you roll them, a 3-4 in constitution is a guaranteed death. ( a d4 HD character will have a 1hp per level average, a d6 one 1.5hp, a d8 one 2.25hp, a d10 one 3.1hp and d12 4hp.. Needless to say that's too little)
CeilingNinja 08-06-07, 01:33 PM	Everybody starts with an 8 in each stat. Roll a d10 for each stat and add that result to the 8. I have no scientific data to prove that this is any better or worse, but it's different than the standard method. 1/10 chance of getting every number between 9 and 18.
Longarms 08-06-07, 03:14 PM	@ silly - you wrote a program just to check this? You are very talented, glad to have someone like you on this thread :) Thank you @ Johnny: "4d6 without dropping ones is what I use; it works fine for me." Really? How long have you been playing that way? Where did you hear of the system, or did you come up with it yourself?
Dan Hemmens 08-06-07, 03:53 PM	The thing I dislike about this idea is that it would have no effect on most rolls and a huge effect on some, which is going to increase the differences between characters. Furthermore, you actually have a *higher* probability of getting some very low results than on a standard 3D6: You're much more likely to get a 3, for example.
WiscDM 08-06-07, 04:09 PM	I've been pondering an idea that may or may not go over well with this crowd... When my players are rolling characters, someone invariably gets some lucky rolls and hits one or more 17's or 18's for his/her stats. Good for them for having lucky rolls, but it can cause some balance problems. I was going to suggest for a houserule for my game that no stat can be higher than 16 and if you do roll a 17 or 18, you can move that point or two over to another stat. This might give someone who rolled an 18, but everything else 9-12's, a chance to bump up another stat to fit into whatever class they were trying to get into. Just an idea... I'm not sure how good/bad it is.
Forgarn 08-06-07, 04:30 PM	@Johnny: "4d6 without dropping ones is what I use; it works fine for me." Really? How long have you been playing that way? Where did you hear of the system, or did you come up with it yourself? That is the way we do it as well. I actually take it one step further and have my players come up with a 6 x 6 grid. They can then take the best score they see: column, row or diagonal (1x1 - 6x6 or 1x6 - 6x1). I have never had a problem with over powered characters. They may seem that way to some, but in comparison to the "average commoner" they are the heros that D&D characters are supposed to represent. And I've been playing that way since 1st edition in 1981.
Mock26 08-06-07, 05:49 PM	I've been pondering an idea that may or may not go over well with this crowd... When my players are rolling characters, someone invariably gets some lucky rolls and hits one or more 17's or 18's for his/her stats. Good for them for having lucky rolls, but it can cause some balance problems. I was going to suggest for a houserule for my game that no stat can be higher than 16 and if you do roll a 17 or 18, you can move that point or two over to another stat. This might give someone who rolled an 18, but everything else 9-12's, a chance to bump up another stat to fit into whatever class they were trying to get into. Just an idea... I'm not sure how good/bad it is. So you are in essence punishing some players for having some luck rolling the dice? As others have pointed out, there are already a bunch of unbalanced aspects of the game that unbalanced stats more than likely won't make a difference in the game. If some of your players are so immature as to be jealous over someone else's stats (or rather, jealous over the stats of someone else's fantasy character) then they probably shouldn't be playing the game as they will probably get jealous when someone else's fantasy character rolls better on hit points or some other non-important issue.
High Octane 08-06-07, 05:55 PM	So you are in essence punishing some players for having some luck rolling the dice? As others have pointed out, there are already a bunch of unbalanced aspects of the game that unbalanced stats more than likely won't make a difference in the game. If some of your players are so immature as to be jealous over someone else's stats (or rather, jealous over the stats of someone else's fantasy character) then they probably shouldn't be playing the game as they will probably get jealous when someone else's fantasy character rolls better on hit points or some other non-important issue. Aww maaan why does his starting gold equal 3d6 x 10 and mine is only 2d4 x 10? It's not fair! *runs off and types on his livejournal and cuts himself*
Mock26 08-06-07, 06:20 PM	Aww maaan why does his starting gold equal 3d6 x 10 and mine is only 2d4 x 10? It's not fair! *runs off and types on his livejournal and cuts himself* You better cut even more because my sword does 1d8 and your dagger only does 1d4!
High Octane 08-06-07, 06:21 PM	You better cut even more because my sword does 1d8 and your dagger only does 1d4! MOOOOOOOOMM!! Mock won't play fair!!
Mock26 08-06-07, 06:24 PM	MOOOOOOOOMM!! Mock won't play fair!! My d30 has more sides than your d20!!!!
Johnny_Angel 08-06-07, 07:14 PM	@ silly - you wrote a program just to check this? You are very talented, glad to have someone like you on this thread :) Thank you @ Johnny: "4d6 without dropping ones is what I use; it works fine for me." Really? How long have you been playing that way? Where did you hear of the system, or did you come up with it yourself? I got it from the Player's Handbook. Roll 4d6 and drop the lowest. Example: I roll 4d6; my dice rolls come out to be a 6, 3, 5, and 2. The 2 is the lowest, so I drop the 2 for a score of 14. -------------------------------------------------------------------------- If you want to try something a little more random, but slightly more balancing, have each player roll 4d6 once until you have 6 numbers. Everybody gets those 6 numbers as their stats, but they are free to place those numbers in whichever stat they wish. Example: You have 5 players and you (the DM) Player one rolls 4d6; drops the lowest for a score of 14 Player two does the same and ends up with an 18 Player three does the same and ends up with a 3 Player 4 rolls a 13 Player 5 rolls a 10 You (the dm) rolls a 16 The PCs now all have the following stats 14, 18, 3, 13, 10, 16 Player one decides to play a fighter; as such places the 18 in STR and the 16 in CON. Player two decides to play a sorcerer; as a result, places the 18 in CHA. etc, etc, etc...
Sockslayer 08-06-07, 07:43 PM	I wrote a small program to test all permutations. The actual rules I used were the following Roll 4 dice Total three dice using the following priority Keep all ones rolled (to a maximum of 3 dice) If less than 3 ones were rolled, use the highest rolls of the remaining dice Just curious, could you do the same for 5D6 drop 2 lowest? if it's not too much work of course ;) I tried doing it myself, but I'm not good enough at programming...
navar100 08-06-07, 08:21 PM	I like the 4d6 rolling method, BUT I think it gives generally too high of scores. Can any who is better at statistics than I am tell me what effect the aformentoned system will have? I really can't figure it out easily. I fall back method would be to do every combo, I guess :cool: My players might like it better than 3d6 :eek: Are you the DM or a player? If a DM, what's so terrible about a player character being good at something? If a player, having a low score does not make you a better player. A DM should not be afraid of a 1st level character with an 18 as much as a player should not be afraid of a 1st level character with an 8.
Mock26 08-06-07, 08:36 PM	A DM should not be afraid of a 1st level character with an 18 as much as a player should not be afraid of a 1st level character with an 8. But should an 8th level player should be afraid of an 18th level DM! :D
SillyRobot 08-06-07, 08:55 PM	Just curious, could you do the same for 5D6 drop 2 lowest? if it's not too much work of course ;) I tried doing it myself, but I'm not good enough at programming... 5d6 3 1 4 5 5 15 6 41 7 90 8 170 9 296 10 470 11 665 12 881 13 1055 14 1155 15 1111 16 935 17 610 18 276 7776 Expect 13.4 Mode 14 Outliers chance 3 0.01% 18 3.50% Chance of penalty modifier 7.9% Chance of bonus modifier 77.5%
Nom 08-06-07, 09:52 PM	If a DM, what's so terrible about a player character being good at something?Inherently? Nothing. But it can be like setting up an adventure course for toddlers and then having a bunch of 10 year olds run through it. They will blitz it. If all they are doing is having a laugh, then fine. But if they are expecting to be challenged, they have just asked the "DM" to try to upgrade everything on the fly to account for the fact that these guys are above nominal. You can play D&D at any power level you like. However, the game designers assume a certain power level and try to make the game mechanics work at that power level. The first question asked of any DM complaining "my players find the game too easy" is generally "did you use 25 point buy or calibrated equivalent?", and the answer is almost invariably "no".
navar100 08-06-07, 10:10 PM	Inherently? Nothing. But it can be like setting up an adventure course for toddlers and then having a bunch of 10 year olds run through it. They will blitz it. If all they are doing is having a laugh, then fine. But if they are expecting to be challenged, they have just asked the "DM" to try to upgrade everything on the fly to account for the fact that these guys are above nominal. You can play D&D at any power level you like. However, the game designers assume a certain power level and try to make the game mechanics work at that power level. The first question asked of any DM complaining "my players find the game too easy" is generally "did you use 25 point buy or calibrated equivalent?", and the answer is almost invariably "no". If the party is smashing through equal CR using up 10% or less resources the solution is simple, use CR+1 bad guys, maybe CR+2. The game is "too easy" because the DM is making it easy. Orcs can have levels too. Maybe the players are just playing smart. What a concept. Don't punish them for it.
Mock26 08-06-07, 10:13 PM	You can play D&D at any power level you like. However, the game designers assume a certain power level and try to make the game mechanics work at that power level. The first question asked of any DM complaining "my players find the game too easy" is generally "did you use 25 point buy or calibrated equivalent?", and the answer is almost invariably "no". And to me that assumed power level seems to be based on the 4d6 drop the lowest number system of generating stats. After all, that is the ONLY method listed for generating stats in the Player's Handbook (which is the PRIMARY source for creating characters) and all other stat generating systems are listed as being OPTIONAL systems.
Longarms 08-06-07, 11:30 PM	"If a DM, what's so terrible about a player character being good at something?" Well, first of all, you are setting up a strawman. You are defining my opinion for me, as being "against" the player being good at anything - and then arguing against that strawman. The strawman you set up doesn't represent my opinion at all. What I think is, the game has better gameplay for everyone when players have more mediocre stats, including some really good stats and some really bad stats. For one thing, the DM can let the PCs gain power more liberally, without constantly worrying about them spinning out of control. I'm too tired to list all the other gameplay improvements right now, but they are there. Ohh another thing, if I didn't think there would be gameplay improvements, I would be crazy to even care what the stats are. They are arbitrary numbers, which as someone else pointed out, just end up being compensated for by in-game challenges. Got a +6 STR bonus? Great, now you will just be fighting 3 orcs instead of 3 goblins. Hope you have a +6 CON bonus too so that your defenses correspond with the increased challenge...
Mock26 08-06-07, 11:34 PM	Ohh another thing, if I didn't think there would be gameplay improvements, I would be crazy to even care what the stats are. They are arbitrary numbers, which as someone else pointed out, just end up being compensated for by in-game challenges. Got a +6 STR bonus? Great, now you will just be fighting 3 orcs instead of 3 goblins. Hope you have a +6 CON bonus too so that your defenses correspond with the increased challenge... If that don't have that +6 constitution bonus the DM could simply compensate by using 2 1/2 orcs instead of 3. :D
Longarms 08-06-07, 11:44 PM	"If that don't have that +6 constitution bonus the DM could simply compensate by using 2 1/2 orcs instead of 3." Heh, good one :) I once read a guide to DMing that really left an impression on me. They talked about how, generally speaking, given the option, players will increase their PC's offensive power before increasing their defensive power. The guide went on to say, that such characters end up causing "swingy" battles than other character types. For example, it is harder for a DM to prepare a challenging encounter (e.g. not too deadly, but dangerous) when the game includes such offensively-weighted PCs. In other words, much greater chance of the DM making a TPK in such circumstances. Maybe thats because when an offensively-weighted PC goes down, the ability of the remaining party to handle the encounter drops relatively drasticly, increasing the chance that one or more other PCs will die. I am not sure exactly how this relates to stat rolling, but I thought it was an interesting observation.
WiscDM 08-07-07, 12:14 AM	So you are in essence punishing some players for having some luck rolling the dice? As others have pointed out, there are already a bunch of unbalanced aspects of the game that unbalanced stats more than likely won't make a difference in the game. If some of your players are so immature as to be jealous over someone else's stats (or rather, jealous over the stats of someone else's fantasy character) then they probably shouldn't be playing the game as they will probably get jealous when someone else's fantasy character rolls better on hit points or some other non-important issue. It's not about the players being jealous of another character's roll, i was only thinking about the scenerios from a DM point of view where one character's stats are head and shoulders above the others and would dominate the game (at least in the early levels). As far as the posts that followed Mock's, I thought they were immature, unnecessary, and uncalled for. I'm certainly not trying to start any kind of flame war and any one that knows me or has read my posts knows it's not my style. I was mearly trying to offer a suggestion which is what the OT was asking for. And my suggestion was really nothing different than the point-buy system which many feel is perfectly acceptable. To suggest that moving points from one high stat to a lower one is the same as whining about daggars to longswords and starting gold in a mocking tone is rediculous and inappropriate for what I assumed was a serious thread. Even if you were only joking around, I didnt think my suggestion was any less serious or viable as any other i've heard on here.
Mock26 08-07-07, 12:21 AM	It's not about the players being jealous of another character's roll, i was only thinking about the scenerios from a DM point of view where one character's stats are head and shoulders above the others and would dominate the game (at least in the early levels). As far as the posts that followed Mock's, I thought they were immature, unnecessary, and uncalled for. I'm certainly not trying to start any kind of flame war and any one that knows me or has read my posts knows it's not my style. I was mearly trying to offer a suggestion which is what the OT was asking for. And my suggestion was really nothing different than the point-buy system which many feel is perfectly acceptable. To suggest that moving points from one high stat to a lower one is the same as whining about daggars to longswords and starting gold in a mocking tone is rediculous and inappropriate for what I assumed was a serious thread. Even if you were only joking around, I didnt think my suggestion was any less serious or viable as any other i've heard on here. If you are referring to my exchange with High Octane I can assure you that from my side they were not directed to anyone at all. It was simple brevity at the expense of any overly immature players out there. Again, they were not directed at you in any way.
MageOfDancingDragons 08-07-07, 03:05 AM	Personally I had my players rool 4d6 drop lowest and then go subject to DM discretion. One of them had a 15 and a 3 and some other average scores, to make the game more interesting and a little more balenced and I moved the 15 to an 18. In my games I prefer keeping the bad scores and upgrading to good scores into great ones. this leads to interesting characters with great potential for epic acts.
Nom 08-07-07, 03:05 AM	Maybe the players are just playing smart. What a concept. Don't punish them for it.Who said anything about "punishing"? I merely observed that DMs who are having trouble setting effective challenges almost invariably use "powerful" ability arrays. The game can work just fine with the "elite" ability array. It can also work just fine with other "arrays". I'm simply saying that it's generally easier for inexperienced DMs to run a balanced game with the elite array or equivalent, and that not doing so is one common factor in balance problems. Slightly more contentiously, I believe that viewing the elite array as "weak" or "unheroic" (or that asking players to use it is "punishment") is an issue of expectations rather than game mechanics. And to me that assumed power level seems to be based on the 4d6 drop the lowest number system of generating stats. After all, that is the ONLY method listed for generating stats in the Player's Handbook (which is the PRIMARY source for creating characters) and all other stat generating systems are listed as being OPTIONAL systems.All true. However, the game was (and is) playtested with elite array. The PH standard system results in an ability array whose mean bonus is in the same ballpark as an optimised 25 or 28 point array. The problem is (as the OP points out) that the PH system has a very large variance with an odd skew and long tail. Combined with the observation that a difference of about +4 or +5 net bonus is roughly worth a +1 LA, it doesn't take a particularly lucky (or unlucky) roll to effectively grant the equivalent of an extra level. A really lucky roll might be worth 2 levels (really unlucky rolls qualify for a reroll). Pitting ECL 3 characters against a level 1 adventure is a noticeable difference. It certainly can be accounted for, but an inexperienced DM its yet another pitfall. I once read a guide to DMing that really left an impression on me. They talked about how, generally speaking, given the option, players will increase their PC's offensive power before increasing their defensive power. The guide went on to say, that such characters end up causing "swingy" battles than other character types.Was that olshanski_home on Regdar's? It's an idea that has been mentioned frequently, but Olsh put in the effort to run some numbers.
Eater_Of_Souls 08-07-07, 03:49 AM	For a truly random figures use a dice roller, comes with a feature of sending emails to roller and a GM to ensure there can be no cheating. http://www.irony.com/mailroll.html the one I use for forum games.
Puggins 08-07-07, 10:39 AM	If you want fairly balanced characters that have the possibility of extremely low or high abilities, do this: (1) roll 3d6 three times. These are you first three scores. (2) Subtract each score from 23 and write down the result. These are your next three scores. (3) Add two to any one score. This will give you characters that are balanced somewhat reasonably AND can have some wildly divergent scores. The good news is that every horribly ability score is offset by a killer stat. Note that you can have a 20 in this system- ample compensation, I think, for having to stick a 3 somewhere. Of course, the characters won't be exactly balanced, but you won't see any outliers that are far and away better than other characters, I think. I honestly forgot who proposed the system that I based this on- Justisaur, maybe? Anyway, this isn't my original idea. Whoever posted the original system can take a bow.
Iskitim 08-07-07, 04:00 PM	A suggestion for thinking about this issue: Take the data that Sillyrobot provided and use it to make a graph for each stat generation method. 3d6 will be a classic bell curve. 4d6 drop lowest will be that same bell curve, only skewed. This reflects the goal of the 4d6 method of eliminating extremely low scores. 4d6 drop lowest non-one score doesn't really mimic a bell curve at all. It's sort of there, but it has weird random peaks. This bothers me a lot, and I don't understand why anyone would find it desirable. If you like having low scores in your game, I'd recommend 3d6. If that makes it too low, try 3d6+1(don't add 1 to natural 18s). That way you preserve a more reasonable distribution.
Longarms 08-07-07, 05:18 PM	"4d6 drop lowest will be that same bell curve, only skewed." An important characteristic of bell curves is that they are symmetrical. The curve produced by 4d6 drop the lowest isn't symmetrical at all - you get way more chance of high rolls than you do low rolls. I think its sort of ingenuine (e.g. your are saying it primarily to support your point) to say that this mimics a bell curve. That said, probably the best suggestion from my point of view is to just stick with 3d6, maybe rolling three different sets.
Roog 08-07-07, 11:20 PM	"4d6 drop lowest will be that same bell curve, only skewed." An important characteristic of bell curves is that they are symmetrical. The curve produced by 4d6 drop the lowest isn't symmetrical at all - you get way more chance of high rolls than you do low rolls. I think its sort of ingenuine (e.g. your are saying it primarily to support your point) to say that this mimics a bell curve. It is, however, the easiest way to describe it.
Sockslayer 08-08-07, 11:32 AM	5d6 3 1 4 5 5 15 6 41 7 90 8 170 9 296 10 470 11 665 12 881 13 1055 14 1155 15 1111 16 935 17 610 18 276 7776 Expect 13.4 Mode 14 Outliers chance 3 0.01% 18 3.50% Chance of penalty modifier 7.9% Chance of bonus modifier 77.5% Thank you Silly :)
Aratex 08-08-07, 12:50 PM	Just chiming in with my thought on ability score generation. I personally use Otto's Random Array method (http://forums.gleemax.com/showthread.php?t=622415), with the added warning that any above average creatures they run into (bosses, minibosses, critical NPCs, etc) will be built with the same array the players chose to use. It turned out to be a pretty good threat, I never realized how many Monster Manual races have +10 or more to their key ability score. :)
Iskitim 08-08-07, 01:38 PM	An important characteristic of bell curves is that they are symmetrical. The curve produced by 4d6 drop the lowest isn't symmetrical at all - you get way more chance of high rolls than you do low rolls. I think it's sort of disingenuous to say that this mimics a bell curve. The symmetrical nature of a bell curve is an important characteristic of the normal distribution. It is not the only important characteristic. A skewed distribution(such as the one provided by 4d6, drop lowest) still mimics the classic normal distribution in a number of significant ways. The most important in the case of D&D is that there is a single value with the highest probability of occurrence, and that the probability of occurrence is strictly decreasing as you move away from that value. This is violated by 4d6, drop lowest, keep 1s, which is why I'm leery of the system.
Longarms 08-09-07, 01:37 AM	"The most important in the case of D&D is that there is a single value with the highest probability of occurrence, and that the probability of occurrence is strictly decreasing as you move away from that value. This is violated by 4d6, drop lowest, keep 1s, which is why I'm leery of the system." Color me surprised. I thought that might have been what you were trying to say, until I went back to the chart and noticed that the "4d6, drop lowest, keep 1s" only has two values which do not follow that strict pattern, and then to a degree that seems so slight (4's and 9's should be somewhat more common), that I didn't think that could have really been your issue. So I guess what you are saying is that if "4's" appeared about 1.7% of the time instead of 0.5%, and "9's" appeared about 9.3% of the time instead of 7.5% of time, you would be fine with the system. That's what is keeping you away from it. I don't know, looking back, I think it was maybe reasonable for me to infer that *something else*, such as symmetry, must have been your issue.
Eled_the_Worm_Tamer 08-09-07, 10:10 AM	If you want players to roll, meaning that you want things random don't stop them from discarding 1s, then your saying 'I want things random, but ideally only when it hurts you' as this capriciously diminishes good rolls (I'd be ticked if I roll 6.6.6.1. the only time In my life and I was made to record it as a 13) likewise if you use dice rolls to create characters and there in front of you the starts are right, the deal at the cross roads pays off and the mythical 6 18's character is rolled as you watch, you have no right to demand a re roll, faint as it is, that was a possibility. If you want more control of PCs ability scores, then short term use a point buy system with a value you like, and / or restrict access to spells such as cats grace, items like Gauntlets of Ogre power. Particularly I would find it a little spitefull in a Dice roll character gen game if the Dm increased the difficulty due to my getting good rolls if he would not have correspondingly decreased the difficulty had I rolled poorly. Ultimately a big part of why many of us play these games is to feel extraordinary, and having high ability scores if we can get them goes to that. 'Challenge' is a danerous word in D&D as a yard stick, go with 'fun'. At the end of the day its not how much you were 'challenged' defeating the ork sentry, its how much fun you had doing it. (Though to an extend chalenge is part of it) As A Dm the only time I take into account my players ability's is when planning named foes, as I'm generous with character creation this gives a slightly cinematic feel, the heroes wade though lesser foes like titans, but find them selves pressed hard by higher rankers. tactics never stop being an issue, play dumb and I will kill you, but I dont see why I should punish a strong character. (It may help that in my group Charicter creation and advancement is involved, with personalities and goals discussed before a die is cast, and characters are advanced in my presence with my advise and permission saught on spell, feat and Prc choices, so no suprsies that might destroy a plot)
zorgling25 08-09-07, 12:28 PM	What I think is, the game has better gameplay for everyone when players have more mediocre stats, including some really good stats and some really bad stats. For one thing, the DM can let the PCs gain power more liberally, without constantly worrying about them spinning out of control. I'm too tired to list all the other gameplay improvements right now, but they are there. Except how bad are those really bad stats? Con below 8 - Afflicted with severe illness Wis below 8 - Borderline, or Autistic Int below 8 - Borderline, or Mental illness In terms of humans, any stat below 8 represents a medical disorder, not just "below average." Do you want every character your players generate to have some sort of medical condition?
RealSorceror 08-09-07, 12:41 PM	We always use 4d6, drop the lowest, reroll 1s and 2s. Then we roll 3-4 sets. It usually produces high stats, but we always fight higher than our CR, so it works out.
Onikani 08-09-07, 12:59 PM	(1) roll 3d6 three times. These are you first three scores. (2) Subtract each score from 23 and write down the result. These are your next three scores. (3) Add two to any one score. In straightforward addition - this makes the pc's stat total always equal to 71 (every pair of stats will add to 23 so 3x23+2). Old school dnd (boxed set versions) had a system where DM could give their players a suggested amount of points (72 was the suggested number IIRC) and players were free to assign their stats as necessary as long as the resulting total was the chosen number. Ganted, back then you needed a 13-15 for a +1 and an 18 was only +3, so a 'balanced character (all 12's) wouldn't have any bonuses. Oh well... Back on trhe main topic -i played in a campy recently that was 3d6+1. Yes it gives a range of 4-19, and yes one player was lucky enough to have a 19 as a starting stat, But we didn't see any real balance issues, and no one felt nerfed. Good times.
Man in the Funny Hat 08-10-07, 02:29 AM	I like the 4d6 rolling method, BUT I think it gives generally too high of scores. Can any who is better at statistics than I am tell me what effect the aformentoned system will have? I really can't figure it out easily. I fall back method would be to do every combo, I guess :cool: My players might like it better than 3d6 :eek:Well, it seems to me that using this method you have indeed reduced the high scores somewhat, but you've shifted the probabilities into the LOW scores rather too heavily as a result (as opposed to increasing the probabilities of being average). Really, though, I think you're placing WAY too much concern on the possibility of "high" scores. But, the change in methods does not provide THAT large a reduction in high scores, probably not even where you could notice it unless you start generating dozens of characters. Which just pushes me to again ask the question of why you think 1%, or even 5% fewer "high scores" is THAT important to your game. On top of which you seem to forget that it IS still random. While the probability curves are indeed there, LOOK at the number of rolls you have to make to really start to see the pattern and feel the effects of that subtle change in an ongoing campaign. Seriously, if "high" scores are really THAT much of a concern to you then you REALLY need to drop the idea of RANDOM generation of scores entirely. Move to an array system or point buy or something similar. It's just going to be the nature of random dice methods that they CANNOT be manipulated as subtly you want and to SEE that manipulation reliably reflected in the results The number of player characters being generated is just going to be too low to see your fine manipulation have an effect. You really have to first accept that if you, as DM want to allow random results then you're SACRIFICING fine control - you WANT disparate results because that is what random rolls will give you. I've got a wide variety of random and non-random generation methods listed at http://home.earthlink.net/~duanevp/dnd/stat_generation.htm and I think one of the non-random methods might be better suited to addressing your concerns.
Kaya Kenobi 08-10-07, 04:27 AM	meh... I prefer my ideas, note.. I powerDM, and I encourage powergaming, so take this with a grain of salt: Often I do this, 4d6 reroll 1s and 2s. PCs should be heroes, PCs should not be weak. With the method, the average is 9-18, with 12-15 being the standard with a couple of 18s. In another game I ran, I had players roll 3d12 and reroll 1-3.
Roog 08-10-07, 07:58 AM	In terms of humans, any stat below 8 represents a medical disorder, not just "below average." Do you want every character your players generate to have some sort of medical condition? Why?
McPoyo 08-10-07, 10:44 AM	When I use a rolling method (which is most of the time, since only one player actually prefers using point-buy as it allows him to fine tune and customize his stats, especially because he always tanks his physicals and plays a caster), I do 4d6 drop the lowest, but restrict the total Modifier. For instance, if a player rolled nothing above a 12, or did not have an end modifier of +1 or higher, they reroll, just like normally. If their total mod, though, is over a +8, I have them reroll as well. It helps cut down on the problems caused by people who legitimately roll 18,18,18,16,15,14 (Yes, I actually had that one), and then someone else ends up with 11,11,11,10,10,13. I've yet to have any serious complaints about it, other than the occasional player whining that someone got a 17 and he didn't get anything higher than a 15, even though the guy with the 17 has two negatives and the whiner doesn't have any. I've found it's best to apply a folding chair to players when they do that, though.
Arrowstorm 08-10-07, 02:26 PM	I don't like rolling or point buys. I have done all the roll strategies you all have talked about including 4d4+4, 21d6 put it where you will etc. I hate to rely on luck to allow me to play what I like. Our group also likes decent stats so that we can play many variations. After all, some classes need good multiple stats. The point buy I find too restrictive and we haven't used it in years. The past few campaignd we went back to a technique I first experienced in a campaign 20 years ago. Everyone starts withe the same numbers, 17, 16, 15, 14, 13, 12 and puts the stat where they will. Immaturity over poor dice rolls isn't immaturity, it's just pure frtustration at lady luck. After all, you are stuck with the character for awhile. We've done the 17 down for several campaigns and it appears to work. Arrowstorm
Scipio 08-10-07, 03:16 PM	Except how bad are those really bad stats? Con below 8 - Afflicted with severe illness Wis below 8 - Borderline, or Autistic Int below 8 - Borderline, or Mental illness In terms of humans, any stat below 8 represents a medical disorder, not just "below average." Do you want every character your players generate to have some sort of medical condition? What page number are you getting this off of, I don't recall seeing it anywhere.
McPoyo 08-10-07, 03:17 PM	Trust me, in the case above I listed it was immaturity.
Longarms 08-10-07, 04:38 PM	"Well, it seems to me that using this method you have indeed reduced the high scores somewhat, but you've shifted the probabilities into the LOW scores rather too heavily as a result (as opposed to increasing the probabilities of being average). Really, though, I think you're placing WAY too much concern on the possibility of "high" scores. But, the change in methods does not provide THAT large a reduction in high scores, probably not even where you could notice it unless you start generating dozens of characters. Which just pushes me to again ask the question of why you think 1%, or even 5% fewer "high scores" is THAT important to your game. On top of which you seem to forget that it IS still random. While the probability curves are indeed there, LOOK at the number of rolls you have to make to really start to see the pattern and feel the effects of that subtle change in an ongoing campaign. Seriously, if "high" scores are really THAT much of a concern to you then you REALLY need to drop the idea of RANDOM generation of scores entirely. Move to an array system or point buy or something similar. It's just going to be the nature of random dice methods that they CANNOT be manipulated as subtly you want and to SEE that manipulation reliably reflected in the results The number of player characters being generated is just going to be too low to see your fine manipulation have an effect. You really have to first accept that if you, as DM want to allow random results then you're SACRIFICING fine control - you WANT disparate results because that is what random rolls will give you." Whoa, I think you read too much into my posts. I don't remember indicating that I wanted to "subdtley" control my players' die rolls. What I said that 4d6 generates rolls that are "too high", and if you want to pick nits, it would have been slightly more clear for me to say "too high too often." That's all. "On top of which you seem to forget that it IS still random." In response, I quote the geico caveman commercial... "Whhat???"
Marcotic 08-10-07, 05:22 PM	I use standard, highest to an 18, my roomy uses standard lowest plus 2
Mock26 08-11-07, 02:33 AM	Except how bad are those really bad stats? Con below 8 - Afflicted with severe illness Wis below 8 - Borderline, or Autistic Int below 8 - Borderline, or Mental illness In terms of humans, any stat below 8 represents a medical disorder, not just "below average." Do you want every character your players generate to have some sort of medical condition? Proof, please? If you can't provide proof please refrain from trying to pass off your opinion as fact.
Archus Forix 08-11-07, 09:26 AM	well, when I get my players to make characters, I use the 4d6 and drop the lowest number, regardless. So if the lowest is a 1, drop the 1! But I also make sure that each one has a minimum of 10. If it is lower, then reroll. The idea behind this is that your characters are heroes. They possess unique skills and abilities that set them apart from NPCs. They are going to be above average in a lot of areas. If that makes my PCs too powerful, just remember I can up the ante with the CR too... :D
Malraux 08-11-07, 10:46 AM	Proof, please? If you can't provide proof please refrain from trying to pass off your opinion as fact. Well, you could cross compare the distribution of 3d6 to something else well understood and easily evaluated. For example comparing INT/WIS to an IQ table. Since both have a standard distribution, you can make an evaluation about roughly what it means to have an INT/WIS of 8; roughly in the bottom 75% of the population. True mental impairment only applies to the bottom 1%.
Longarms 08-11-07, 01:43 PM	"Since both have a standard distribution, you can make an evaluation about roughly what it means to have an INT/WIS of 8; roughly in the bottom 75% of the population. True mental impairment only applies to the bottom 1%." Heh. I have had players running PCs with a 7-8 INT before, and they tended to initially play their characters as if they were retarded, e.g. unable to follow a simple plan, talking weird. I had to pull them aside and explain that their characters are only in the bottom 25% of the population, which does not mean they have to screw up simple plans or talk all messed up. Silly players not understanding statistical distributions ;) Hey, I just check some stats on the system that is the subject of the present discussion: 3d6 has only a 10% chance of rolling 14 or higher on any one roll; same thing for rolling 7 or lower. In contrast, the present system has a 25% chance of rolling 14 or higher, and only about 12% to roll 7 or lower. So the cries that this system was cruel or produced lower rolls than 3d6 appear unfounded...
zorgling25 08-11-07, 01:59 PM	What page number are you getting this off of, I don't recall seeing it anywhere. Look at the tables in the PHB giving scores for various other creatures and compare using relevant skill checks, etc.
Neutronium_Dragon 08-11-07, 03:21 PM	> And to me that assumed power level seems to be based on the 4d6 drop the > lowest number system of generating stats. No. All of the iconics were built with and the power level was gauged against the elite array method, since that kept everything at a standard level of power. (This is no big secret - the developers talked about the playtesting back at the time of 3E's release.) This is also why all of the NPC tables and so on use the elite array method for PC-classed characters (and creatures).
navar100 08-11-07, 06:09 PM	> And to me that assumed power level seems to be based on the 4d6 drop the > lowest number system of generating stats. No. All of the iconics were built with and the power level was gauged against the elite array method, since that kept everything at a standard level of power. (This is no big secret - the developers talked about the playtesting back at the time of 3E's release.) This is also why all of the NPC tables and so on use the elite array method for PC-classed characters (and creatures). Right, but player characters are supposed to be more powerful than NPCs. If NPCs had the same magic gear as PCs of equal level, for example, when the party defeats them they'd double their magic item wealth.