Post/Author/DateTime	Post
Zagreen 04-10-07, 02:57 PM	Just got Complete Scoundrel and on the whole I liked it quite a lot. I loves me some scoundrel characters and the book seems pretty decent both on fluff and on crunch. As far as the luck feats go, though... am I missing something or do these pretty much suck? Reroll initiative 1/day, that's not too terrible for a feat, but most of the luck feats require you to burn multiple rerolls to get anything useful. Spending multiple feats to reroll an attack roll 1/day? That's just awful! The "turn natural 1 into natural 20" feats look pretty cool since they only cost one luck reroll to use, but they have a clause that says they can only be activated once a day anyhow which defeats the whole purpose. Has anyone ever tried actually making a Fortune's Friend or other luck-based character? Is it really worth spending a handful of class levels and a bunch of feats just to get a couple rerolls per day?
Pyke_Moonshadow 04-10-07, 03:02 PM	I don't get it... Sluck feats? :confused:
CaoSlayer 04-10-07, 03:08 PM	lol, he means supernatural. No, actually he meants to replace a s in place of the l so you have s-uck feats.
Pyke_Moonshadow 04-10-07, 03:15 PM	:doh: I don't think anyone would take them if they were called that though.
Grod_The_Giant 04-10-07, 03:27 PM	that's his point. Can't comment, though, since I don't have the book.
Rhuxley 04-10-07, 03:35 PM	Yeah, totaly not worth it. Would rather have +4 init al lthe time than the ability to reroll one init reroll. Even the basic luck feats are suck feats
j3st3r 04-10-07, 04:14 PM	I'm considering taking Sly Fortune and Unbelievable Luck for my scout. That way He's got better chance of staying unnoticed while scouting/ambushing and better chance of avoiding attacks of opportunity with tumble. Also luck rolls twice a day and/or +2 to characters lousiest save doesn't sound bad at all. All this with two feats sounds ok to me.
sigma999 04-10-07, 06:52 PM	Yeah they do suck.. big time.. I recommend changing that "1/day" limitation to something more frequent but I can't think of anything other than every time you gain another Luck feat you add your character level to the number of 'rerolls' you get in a day, and add one that gives 3+CHA rerolls as well. The Eberron Action Points were a better idea and you can quite possible combine the two feat types into one (yes and without another dumb feat) by considering "Action Points" the same as "Luck". Or however you might say that. Oh, and permanently losing Action Points was another stupid thing, even tho you get more every time you level up. Should have been a per-encounter basis or at least every day you refresh them.
MCage 04-10-07, 07:39 PM	Yeah i dont think they suck, i just dont think their for everyone. I use them with clerics who have the luck domain, get that feat that lets them channel their luck rerolls through that (dont have the book on hand), i dont remember what exact feats i used but i had a level three cleric that had the healing spell reroll, the one that give you two rerolls and a bonus to your worst save and the channel through domain reroll, which made a really fun character
Steidas 04-11-07, 01:29 AM	THe issue is building up the rolls. What I don't like about this feats and Incarnum is that basically you NEED to take them to get anything out of them... and take a LOT of 'em.
Zagreen 04-11-07, 02:15 AM	Yeah i dont think they suck, i just dont think their for everyone. I use them with clerics who have the luck domain, get that feat that lets them channel their luck rerolls through that (dont have the book on hand), i dont remember what exact feats i used but i had a level three cleric that had the healing spell reroll, the one that give you two rerolls and a bonus to your worst save and the channel through domain reroll, which made a really fun character Third Time's the Charm just lets you re-reroll the reroll you get from the Luck domain, which can still only be used once per day (read the feat description very, very carefully). Which means your remaining 3 rerolls can only be used with Healer's Luck, which just flat out sucks; if you want to boost the amount of HP you heal, Augment Healing is vastly superior. Bottom line: still suck. I could see taking one or two for flavor, but as far as going all the way on them I just don't see it working well.
CaoSlayer 04-11-07, 04:57 AM	they arent that bad, the turn 1 into a 20 is scary if you have a vorpal weapon and is almost the same that having improved critical and weapon focus if you have a 20 x3 weapon. The only once per day isnt that awful because is a very rare chance to roll more than a natural 1 per day. The other pure win feat is the "in place of being outright killed you ends with -9 hit points and stabilised", that is a feat I would take for sure If I was playing with a rogue or traveling with a frenzied berserker. And dont infraestimate the power to roll twice, I dont know the match but being able to roll twice is way better than a +2 since ever if you have the greatest fortitude of the world, if you roll a natural 1-4 by example, you die. Yes, the better luck feats requires two uses or another "suck" feat but most of those are worth it.
IonizedChicken 04-11-07, 12:15 PM	Yes, I agree. Luck feats are terrible. Really, there's no excuse for their suckiness. WotC should have published a Tome of Battle expansion instead of this Complete Scoundrel crap. I'd even say the nat. 1 -> nat. 20 feat sucks, and not only because you can only use it once per day. How many times will you get a natural 1 anyway? once per day? twice per day? three times per day? that's a daily limit by itself. Luck feats should provide Luck bonuses to various rolls. That's probably it. Or how about 'lucky days'? characters could pick a day and get a +2 luck bonus on attack rolls, saves, and checks on that day. Something about the 'alignment of the stars' or rather the 'influence of the outer planes'. It might actually be quite interesting. Sure, a +2 might be a little much, but it's not like all the encounters the character is going to face are going to be on that particular day. Just those he or she can control and has the ability to postpone.
Zagreen 04-11-07, 12:36 PM	they arent that bad, the turn 1 into a 20 is scary if you have a vorpal weapon Once per day. and is almost the same that having improved critical and weapon focus if you have a 20 x3 weapon. Once per day. The only once per day isnt that awful because is a very rare chance to roll more than a natural 1 per day. So you're saying it's OK that it only works once per day because it's not likely to happen anyhow? Yeah, okay. And dont infraestimate the power to roll twice, I dont know the match but being able to roll twice is way better than a +2 since ever if you have the greatest fortitude of the world, if you roll a natural 1-4 by example, you die. On the other hand that +2 is going to count on every single roll. Unless you only make one Fort save that day in the long run a flat bonus is better. Taking a luck reroll ON TOP OF that flat bonus feat isn't too terrible if you want extra security but basically there's no reason to take the luck feat instead. Yes, the better luck feats requires two uses or another "suck" feat but most of those are worth it. You can take luck feats and be awesome once per day, or you can take feats that don't suck and be awesome all the time. Take your pick.
jadehawk 04-11-07, 01:05 PM	:doh: I don't think anyone would take them if they were called that though. Hey Pyke, I'm looking for the sarcasm key and was thinking you might be able to help me. With any suck I should be able to find it by the end of the day.
Drachasor 04-11-07, 02:21 PM	Yeah they do suck.. big time.. I recommend changing that "1/day" limitation to something more frequent but I can't think of anything other than every time you gain another Luck feat you add your character level to the number of 'rerolls' you get in a day, and add one that gives 3+CHA rerolls as well. The Eberron Action Points were a better idea and you can quite possible combine the two feat types into one (yes and without another dumb feat) by considering "Action Points" the same as "Luck". Or however you might say that. Oh, and permanently losing Action Points was another stupid thing, even tho you get more every time you level up. Should have been a per-encounter basis or at least every day you refresh them. Adding your Charisma Mod to rerolls makes a lot of sense. For one, it helps give Charisma more purpose, and it also fits with the character forcing reality to his will. I agree that once/day per feat is a bit lame. That's largely because the Luck abilities are lame. You should reroll and take the better of the two results for one. The Luck feats need that at least to be worth it. Additionally, the number of luck rolls you get needs to be fixed. It either needs to automatically go up a bit as you level (say Character Level / 3), or it needs some more significant alteration. Say Each combat encounter you have gives you a number of temporary rerolls equal to your daily total. If you spend all the temporary ones, then you can spend you per day ones. So in a combat encounter, you have a lot of luck to burn (if you have a lot of luck feats). To go over the last idea a bit, say you have 1 luck feat, so you have 1 luck reroll each day. When you go into an encounter, you automatically have 1 free luck reroll for that encounter. You either spend it in the encounter or it is 'wasted'. If you have 2 luck rerolls, then you'd get 2 such temporary rerolls per encounter. Of course, you'd always have your daily limit to draw upon if you used up your temporary ones. Then, I think, the Luck feats would really be worth considering.... The main thing that luck rolls do is tend to diminish the number of bad rolls you have. That doesn't mean you won't have bad rolls, or that a roll of a 10 on a d20 would be sufficient. Considering most of the d20 rerolls cost 2 luck points though, I think the above change wouldn't be overpowered. Afterall, if you got Feats + Charisma Modifier luck points per day, that would only be maybe 4-6 points for a 6-8th level character at best. So a luck-focused character of this level could only reroll 2-3 attack rolls in a given encounter (given my system), or double that if they use up their daily luck (and then they can't have any rerolls out of combat).* This person has given up a bunch of combat feats for this ability though. They probably don't have Power Attack (or at least not Cleave), and might not have Weapon Focus. So, especially as they gain levels, they are going to miss out on a lot of constant bonuses to their rolls, and instead depend on a few rerolls every encounter. Still, it would require some playtesting, I think. That or a computer program that went over various combats and made a comparison. -Drachasor *If that is too much, then perhaps you should only get a temporary amount of luck equal to your remaining daily uses. I don't think that would be too much, however.
CaoSlayer 04-11-07, 04:48 PM	Once per day. Once per day. So you're saying it's OK that it only works once per day because it's not likely to happen anyhow? Yeah, okay. The chance to hapens is a 5% that is exactly the same chance to miss by just one point or roll a 19 and miss a critical hit. At average just one of twenty rolls is a natural, so unless you are rolling about 40 times a day, you are very unlikely to roll two ones. So you could take weapon focus or improved critical that only works with a certain weapon or you can take this feat that works like both. On the other hand that +2 is going to count on every single roll. Unless you only make one Fort save that day in the long run a flat bonus is better. You can take luck feats and be awesome once per day, or you can take feats that don't suck and be awesome all the time. Take your pick. The +2 isnt important at all if you just roll poor, is very likely that you will miss a save by more than 2 and the most important thing you are ignoring: THAT FEAT WORKS WITH ALL SAVES. Yes, is only once per day but you dont use it unless you fail that save, it gives you the chance to try again a save that would kill you outright... And it also gives you a +2 to the weakest one. Sorry but I prefer a unfailable reroll for any save to just a permanent 10% chance for just one type of save.
Zagreen 04-12-07, 12:02 AM	No it does not also give you +2 to the weakest one. A completely different feat does that. Although rewriting luck feats so they did work that way might be a good way to fix them. And they do need fixing. I don't know how your GM runs things, but if I get by in a day without having to roll a dozen saving throws it's been a pretty light day for me. 10% adds up. Hell, even a third of 10% adds up (assuming that all three saves get hit equally). Sure, you can save your luck reroll for The Big One--but how exactly do you know when The Big One is going to hit? You might go for a whole day saving up against The Big One and end up having to burn through a lot of extra potions, spell slots etc. to recover because you kept flubbing your rolls against lots of little ones, against which your Survivor's Luck feat does nothing. Or you might get hit with two Big Ones in a row, or maybe you already spent your luck rerolls on possible Big Ones and don't have any left for the real Big One. And it's not guaranteed to do anything against The Big One, either, because the reroll is just as likely to fail as the first roll. Worse, it requires an immediate action to use--so if you're caught flat-footed, or you like to throw swift actions around in combat, you might not even get the opportunity to take your reroll at all. I'd prefer a permanent 10% chance for just one type of save rather than just getting the possibility of a chance to reroll a save. Although mostly I wouldn't really bother with either--I only brought up the +2 bonus because someone else was singing the praises of the "+2 bonus to weakest save" side benefit of Unbelievable Luck, a completely seperate feat from the one you're talking about. Hell, Lightning Reflexes and all them aren't even necessarily the best at boosting weak saves. If I had a weak save I was worrying about I'd look more into feats like Insightful Reflexes, Force of Personality, Steadfast Determination, Martial Study (Moment of Perfect Mind)--feats which can end up giving you a hell of a lot more than +2. When it comes down to it I prefer to use my feats a bit more proactively. Supposing you do get an opportunity to use your Survivor's Luck on a do-or-die save and you do manage to make your reroll--congrats, you survived, so now what do you do? You've blown all your feats just for that once-in-a-lifetime save, it's not like you've got room for much in the way of useful combat feats left over. If you had useful combat feats, maybe you'd be able to kill more of the bad guys before they spring their nasty save-or-die attacks on you.
Jaen 04-12-07, 01:41 PM	I think that the feat that makes it so that you dont die when going below 0 is wonderful. Yes you can only do that one time a day but if you die more than one time a day what the heck are you doing wrong? Do you really suck that much or is your DM just that big of an a**? If you take just that feat then you have the one luck reroll to do the I'm not going to die now trick. It's beautiful. And besides you should mostly not consider taking the luck feats unless you plan to take more than one, making them in effect a feat tree. Rerolling is wonderful when you roll that 2 you didn't need or absolutely needed a higher number on. Odds are good (50%) that you will roll better than a 10 next time. What I am trying to say is that by getting a reroll you DOUBLE your chances of success for that roll which is often better than a +2 all the time. You get rolls daily and if the party you play in is anything like mine we like to rest for a day in the middle of dungeons, so we get some of our abilities or spells back, and we do that often. I think that luck feats work well and have already used them, to great effect. I think that if you want better saves you should go and buy an item that gives you better saves, because it is more cost effective than taking ANY feat. You only get about 7 and you're going to spend one on saves? What is wrong with you there are cheap items that give pluses to saves and all saves at that. Or there are items that give bonuses to skills too! I very rarely think skill focus is worth it unless you are a wiz/sorc/cleric and are planning on taking the campaign to epic so you want to max spellcraft all the way. I think you crazy people are forgetting that you get better odds rerolling generally than by just adding a +2. Reroll if you roll nat 1 anyone? Yeah! And the plus 2 doesn't help your ass in that situation. You FAIL. Hey guess what the reroll also doubles your chance of a nat 20 aka auto success!
Piropoint 04-22-07, 11:19 PM	Maybe somebody already mentioned this, but if you do decide to go all out on the luck feats, each new one you get increases in power and versitility. If you have 7 luck feats you have 7-8 re-rolls a day which could all be used (more or less) with the same feat if the need arises. (granted that there are some which require two re-rolls, but those are the 'super' effects) however, there is no synergy if you take 7-8 skill focus/spell focus feats. Each one exists in a vacum. (I know people don't generaly do that, but for the argument, I'm just showing one way that luck feats are at least as good and sometimes better than feats that are generaly balanced.) When you start out, you may need to re-roll every time you get below a ten, but as you increase in level, so do your modifiers and you may only need to re-roll when you get below a 5. A character that can re-roll anything below a 5, has a huge advantage over the character who only gets +3 to a skill from a feat.
sigma999 04-22-07, 11:40 PM	I'd prefer Force of Personality, and add Charisma to Will saves rather than take any Luck feats at all, but it would be amazing to see a player with nothing but Luck actually make it work through a good portion of a campaign. I'd like to see that.
pmurray@bigpond.com 04-22-07, 11:49 PM	And dont infraestimate the power to roll twice, I dont know the match but being able to roll twice is way better than a +2 since ever if you have the greatest fortitude of the world, if you roll a natural 1-4 by example, you die. Is rolling twice beter than a +2? After doing a bit of fiddling in Excel, I determined that rolling twice is better unless you need a 19 or better (or 3 or worse). This does not take into account 1s and 20 being automatic failures/sucesses. Furthermore, rolloing twice is a lot better when you need a 13-15. An unmodified roll has a 35% chance of getting a 14 or better. A +2 gives you a 45% chance. Rolling twice gives you a 58% chance.
Ainamacar 04-23-07, 04:24 PM	Edit: I've found an error in my original analysis that changes the values, though the conclusions I come to are essentially unchanged. I have added or changed a few comments, so it may be worth reading again if you are interested. I also included the procedure I used for arriving at these values in a spoiler block. The numerical results below are current as of 4/24/2007. I've crunched the numbers for save, given some reasonable assumptions about behavior. I am reasonably confident in the math (otherwise I wouldn't post it) but I make no guarantees. In particular, there are 6 classes of reroll (success and then stop, failure and then stop, success but failure on reroll, success and success again on reroll, failure and failure on reroll, failure but success on reroll) that I constructed the probabilities for independently, and everything adds to 1, a very good sign indeed. Under the assumptions I make, all the values in the table are exact probabilities rounded to four significant figures when necessary. The leftmost column is the "natural" roll you need to make a given save, without any modifying feats or abilities. They go up to 22 so you can see how the two methods compare for saves that can only be made with a natural 20. The next column is the probability of success with an unmodified saving throw. The third column is the probability of success for the save given a constant +2 bonus. The fourth column is the probability of success if you reroll and keep the higher roll, and also if you only reroll if you know you have failed. (Neither of these are the official rule for luck feats, but they could be common house rules for convenience, if nothing else. The difference between the two cases is in how much information you possess before you decide to reroll. In the former, you have less information but you aren't penalized for making the wrong choice because the better roll counts. In the latter you have more information, and thus the fact that you must keep the 2nd roll isn't a roadblock. The latter method is still better in the sense that you never have to use your limited resource of rerolls when you don't need to. However, *before* you've made the roll or know the DC, the probability of success for these two methods are identical). The fifth through seventh columns use the official luck rules, but because there is player choice involved I must make some assumptions about the situation. The number labeling these columns, k, represents the fraction of times a player makes the "wrong" choice about using a reroll. For example, if the value is .1 that means that 10% of the time the player chooses to reroll a save on which they have already succeeded (but because the don't know the DC they chose to play it safe) and that 10% of the time the player chooses not to reroll a save which they have already failed (again because they lack knowledge of the actual DC, or because they want to save the reroll). The numbers given assume the player never chooses to reroll a 20, and always chooses to reroll a 1. That makes the math a bit more complicated, but the player has perfect knowledge about success and failure on a natural 1 and 20, and so I take that into account. The underlying assumption here is that the player is about as accurate at gauging the actual success or failure of their original roll no matter what roll they actually need to succeed. It is doubtful a real player is actually this uniform in behavior, and so it should be understood in an average-case context. I have provided three sample values for this, since some players will be better at making this judgment than others (through experience, metaknowledge, etc.) Finally, I point out that when the reroll parameter goes to 0 I actually recover the results of the fourth column, representing a reroll only when you need to do so. In terms of p (the chance of success when rolling a single d20, i.e. column 2), p0 (the chance of success given that you have rolled a 2-19) and k (the reroll parameter) the formula for items in these columns is: 1/20+19*p/20-9*k*p/10+9*p0/10-9*k*p0/10-9*p*p0/10+9*k*p*p0/5 p0 isn't shown below, but can be calculated as (19-n+1)/18 where n is the natural roll required. If this formula gives a value below 0 then p0 is actually 0, and if it is above 1 then p0 is 1. Edit: I originally underestimated the value of the reroll because I incorrectly calculated a conditional probability in a few places. See the spoiler block for an example and complete analysis of the math. To make sure the new reroll data is right, lets look at two extreme cases for k=.1. First the case where the DC is exactly 20 above the save bonus. The following are the only ways to succeed in this case: 1. Roll a natural 20 the first time (probability 1/20) 2. Roll a natural 1 the first time and a 20 on the reroll (probability 1/20*1/20) 3. Roll a 2-19 the first time, choose to reroll, and then roll a 20 (probability 18/20*(1-.1)*1/20) This gives 1/20+1/400+162/4000=.093. Note how k enters into the expression. 10% of the time we roll a 2-19 and figure it is good enough (even though only a 20 will do) which means that only 90% of the time do we correctly choose to reroll and hope for a 20. Next, the case where the DC is exactly 1 above the save bonus. The following are the only ways to fail: 1. Roll a natural 1 and then reroll a 1 (probability 1/20*1/20) 2. Roll a 2-19 the first time, choose to reroll, and then roll a 1 (probability 18/20*.1*1/20) This gives 1/400+18/4000=.007, so the chance of success is 1-.007=.993. Again, k plays an important role, because even though a roll of 2-19 is always a success, we don't know this and thus 10% of the time reroll anyway, occasionally getting an unlucky 1. Here is the complete analysis Assumptions ----------- 1. You always reroll a 1, and never a 20. 2. The parameter k, representing the probability you reroll an initial success or don't reroll an initial failure (except in the case of 1 or 20, as above), is a reasonable approximation of player behavior. In any case, a more rigorous examination using a non-uniform distribution to model this behavior greatly complicates the mathematical construction and is unlikely to change the broad qualitative results. Definitions ----------- p: The probability of success rolling a single d20 p0: The probability of success if you have rolled a 2-19 on a d20. Equal to (19-n+1)/18, where n is the minimum roll needed to succeed (ignoring autosuccess and autofailure). This value is 0 if the equation gives a result less than 0, and 1 if it gives a result greater than 1. k: The probability a player rerolls a successful save or does not reroll a failed save. Does not apply to natural 1 or 20 rolls. When you reroll: A. Rolled 1 B. Passed 2-19 but thought (incorrectly) you need to C. Failed 2-19 and thought (correctly) you need to When you don't reroll: D. Rolled 20 E. Failed 2-19 but thought (incorrectly) you don't need to F. Passed 2-19 and thought (correctly) you don't need to Their probabilities A: 1/20 B: 18/20*p0*k C. 18/20*(1-p0)*(1-k) D. 1/20 E. 18/20*(1-p0)*k F. 18/20*p0*(1-k) The probability you either reroll or do not reroll add to 1, as they should. Results from reroll: A->fail: *(1-p) A->success: *p B->fail: *(1-p) B->success: *p C->fail: *(1-p) C->success: *p Thus, the total probabilities are: Autofailure (no reroll): 0 1->reroll success: 1/20*p 1->reroll failure: 1/20*(1-p) 2-19 success->reroll success: 18/20*p0*k*p 2-19 success->reroll failure: 18/20*p0*k*(1-p) 2-19 failure->reroll success: 18/20*(1-p0)*(1-k)*p 2-19 failure->reroll failure: 18/20*(1-p0)*(1-k)*(1-p) Autosuccess (no reroll): 1/20 2-19 success (no reroll): 18/20*p0*(1-k) 2-19 failure (no reroll): 18/20*(1-p0)*k These probabilites also sum to 1. The values in the table are the sum of the success probabilities. If I've made a mistake in this analysis, it behooves me to fix it, so please let me know. Take Nat Raw +2 Best k=.1 k=.2 k=.25 22 0.05 0.05 0.0975 0.093 0.0885 0.08625 21 0.05 0.1 0.0975 0.093 0.0885 0.08625 20 0.05 0.15 0.0975 0.093 0.0885 0.08625 19 0.1 0.2 0.19 0.177 0.164 0.1575 18 0.15 0.25 0.2775 0.257 0.2365 0.2263 17 0.2 0.3 0.36 0.333 0.306 0.2925 16 0.25 0.35 0.4375 0.405 0.3725 0.3563 15 0.3 0.4 0.51 0.473 0.436 0.4175 14 0.35 0.45 0.5775 0.537 0.4965 0.4763 13 0.4 0.5 0.64 0.597 0.554 0.5325 12 0.45 0.55 0.6975 0.653 0.6085 0.5863 11 0.5 0.6 0.75 0.705 0.66 0.6375 10 0.55 0.65 0.7975 0.753 0.7085 0.6863 9 0.6 0.7 0.84 0.797 0.754 0.7325 8 0.65 0.75 0.8775 0.837 0.7965 0.7763 7 0.7 0.8 0.91 0.873 0.836 0.8175 6 0.75 0.85 0.9375 0.905 0.8725 0.8563 5 0.8 0.9 0.96 0.933 0.906 0.8925 4 0.85 0.95 0.9775 0.957 0.9365 0.9263 3 0.9 0.95 0.99 0.977 0.964 0.9575 2 0.95 0.95 0.9975 0.993 0.9885 0.9863 1 0.95 0.95 0.9975 0.993 0.9885 0.9863 What can we make of this? For the hardest saves that only a natural 20 will ever make (and the +2 doesn't help), a reroll is best, but obviously improbable enough that using it in such a situation is suspect. For saves at or just under the natural 20 mark, the +2 gives the largest benefit. In the higher ranges, around natural 16-18 the rerolls start to outperform the +2 on any given single save. It retains this edge throughout most of the lower necessary natural rolls (or remains competitive if you guess wrong a fair amount of the time). If each natural roll required (from 1 to 22) is equally likely, the benefit from +2 and a reroll with k=.25 is nearly equal, on average, for a single roll. For lower k, the reroll is better for any given single save. It isn't shown above, but for k=.5 (i.e. random guessing on when you should reroll), you do a mere constant .025 better than on unmodified saving throws. So if you don't trust yourself to make an even remotely informed decision, its clearly not worth the feat. Now, there are still plenty of intangibles. The reroll is a limited resource, and its value decays the more often you are forced to make saving throws. If you are forced to make two saves of the same type in a day, the chances that you succeed on both is greater, on average, with the +2 bonus than if you can reroll just one of them if any natural roll on the table above is required for that save is as likely as the others. (This isn't a very good assumption since saves are often targeted to a perceived weakness and very hard or easy saves are less common, but it is probably decent to a first approximation. In this case, the +2 bonus is about 2% more likely to make both saves. If you can reroll both saves, it is about 6% better than the +2 bonus. This can be seen by forming all possible combinations of saves between the relevant columns and finding the average). Of course, the +2 bonus applies to only 1 type of save, while the reroll will generally work for any of them. This flexibility can be key, because everyone can still roll a 1 on their good saves. Looking at those issues in more depth is really outside the goal I have for this post. The short story is still: if you have to make significantly more saves per day than you have rerolls, the benefit of the luck feats deteriorates rapidly. If you have a decent handle on the metagame, the relative value of the reroll increases, since you will probably use your rerolls more efficiently than other players. The flavor text for the luck feats also suggests that the feats themselves represent things that happen outside the characters control or knowledge. This can sometimes justify use of metagame knowledge where it might otherwise be frowned upon. In general, my own conclusion is that, for weak saves (especially if you are likely to be targeted with them, such as will saves for a fighter) you are better off taking the +2 bonus, since you'll often need to roll 15+ to make those saves, and then the +2 is key. If you have 2 or more good saves (and especially if your bad save isn't likely to kill you, like reflex for someone with tons of hit points) then the reroll will help most. The +2 isn't doing anything that special if you can make most any save with an 8 already, compared to being able to reroll a low roll and have pretty decent odds of coming out on top. This is especially true since having all good saves means you have less cause to use multiple rerolls per day, and so that limitation of the luck feats becomes less important. I have a character taking luck feats (a partisan with the luck domain, see my sig) and did a rough version of these calculations a while ago, but I payed much closer attention to the details this time around. This won't end the argument, but hopefully it will allow it to be made with greater clarity. I hope everyone finds it useful.
itwasntme 04-23-07, 04:41 PM	Intrusting ... But sure ok giveing a bonus on Will saves is nice but then again theres a feat better then Force Of Personilty Lets your Con mod be added to will and you can no longer fail on nat 1 for fort saves i vote a good combination of luck feats and non luck feats is GREAT so my vote has to be on a combination of the 2 with luck feats treating nat 1 or nat 20 as the other can be nice nat 1 i cant fail or if i do fail is treated as nat 20 meaning 1 and nat 20 will let you auto save a spell you cant normoly save
tarkin 04-23-07, 04:48 PM	The math also ignores the natural 1 fallacy. 1. Which is better +1 for 40 rolls/day. 2. +40 for one roll a day (your choice) 3. Luck re-roll 1/day. If you have 10 rolls each day, then option 2 is better than option 1. If you have 80 rolls a day then option 1 is better than 2. But only option #3 lets you survive when you rolled a natural 1 on a saving throw vs. Disintegrate. As such, the re-roll is in my opinion, a far better protection.
vonmax 04-23-07, 06:09 PM	Hmm I see lost of nice things here but most of it covers the math of it. Here how a old gamer looks at it rerolls can be fun make for good role playing. That time when you fail the roll at wrong time and little good old reroll luck turns the zero to a hero and get the crowd chearing. Good old fun D&D not math figering it to the last number. Just my way of looking at it if you dont like the way they work dont use them but I think they can be lots of fun and I see some other said that in there post about it. And again this just my option I have been playing D&D from way back when it was small books called chainmail and it always been more about fun more then any thing. :) ohh and yes I am old but I am the youngest playing in my group of old D&D players. ;)
Grand Theft Otto 04-23-07, 06:10 PM	I generally agree. The luck feats kind of stink, particularly when weighed against the Golaith's marking feats.
Thalspar 04-23-07, 07:36 PM	i don't like luck feats that much but i haven't noticed that you all didn't count that in that for each luck feat you have you gain another luck roll per day so a guy with possebly 4 luck feats can reroll on those things 4 times a day. on initiative if you roll low on the first and the second you have like 3 throws left if you want to use that
Deep Blue 9000 04-23-07, 08:32 PM	The math also ignores the natural 1 fallacy. 1. Which is better +1 for 40 rolls/day. 2. +40 for one roll a day (your choice) 3. Luck re-roll 1/day. If you have 10 rolls each day, then option 2 is better than option 1. If you have 80 rolls a day then option 1 is better than 2. But only option #3 lets you survive when you rolled a natural 1 on a saving throw vs. Disintegrate. As such, the re-roll is in my opinion, a far better protection. The math certainly does not ignore that possibility. You'll notice that there is normally a 5% chance of failure because of rolling a one while the 1 free reroll reduces that chance to .25%. Ainamacar's probabilities are exhaustive at summarizing the situation for a single roll. My personal opinion is that while rerolls are better in general, they are also disproportionally more expensive. The +2 save feats, for example, are given as bonus "throw-away" feats by many classes and are pre-reqs for some pretty sweet feat chains. The luck feats in CS on the other hand, require you to take a crappy luck feat before you qualify for the good ones.
Magon 04-24-07, 10:26 AM	ehem..... ni'm just making it so that you can use them when they'r needed, not just once pr day. and every other luck feat and 2 rerolls
Ogg 04-24-07, 11:05 AM	I don't think they're that bad. Trying them out with a character right now. One note is a lot of them stack uses. So while your 1st one only gives you 1 re-roll a day. A second one gives you an additional roll to be used on either. Plus that prestige class lets you get even more. I don't know being able to re-roll a disable device is a really nice thing speaking from experience of being the rogue who just blew it in front of the party. I think they're really made to work with each other, if you look at one feat in & of itself it seems weak but once they stack and you have 3 or more of them they can actually make each feat get stronger.
tarkin 04-24-07, 11:09 AM	The math certainly does not ignore that possibility. You'll notice that there is normally a 5% chance of failure because of rolling a one while the 1 free reroll reduces that chance to .25%. Ainamacar's probabilities are exhaustive at summarizing the situation for a single roll. My personal opinion is that while rerolls are better in general, they are also disproportionally more expensive. The +2 save feats, for example, are given as bonus "throw-away" feats by many classes and are pre-reqs for some pretty sweet feat chains. The luck feats in CS on the other hand, require you to take a crappy luck feat before you qualify for the good ones. His math is wrong for the case of a natural 1. If you check, you see the probability changes for k=.1,.2,.25 When you roll a natural one you always know to re-roll. There is NO possibility of making the wrong decision. They should all be 0.9975, regardless of the k value
Piropoint 04-24-07, 02:48 PM	Ainamacar- Amazing work. I think this is just the sort of thing that should be published in Regdar's Repository. I know that authors hate to "dumb down" their work for the general public, but I believe this could become a truly insightful handbook to luck feats, and probability in general. Once again, good work. I'm sure I will refer to this more than a few times in the future.
Ainamacar 04-24-07, 04:17 PM	Well, Tarkin is right and wrong. I did find an error in my math (now corrected in the original post) whereby I was sometimes undercounting the value of a 1 or 20. However, his reason for believing it was wrong is incorrect. I believe he is misinterpreting what the left hand column designates in my table, and I grant that I could have labeled it less confusingly. It does not represent the actual roll, it represents the roll the player needs to succeed on a given saving throw (with no extra bonuses, etc.), ignoring automatic success or failure. However, the probabilities as calculated do not ignore auto success or failure. I did make a mistake the first time I calculated them, but it wasn't in ignoring natural 1s as such. The probably of success decreases with the different k's in the "natural 1" row because the player still has a chance of rolling a 1-20 on his original save as well as on the reroll, if he chooses to make it. Now, any save in the "natural 1" row can be made successfully on any roll except a 1. Note also that the probabilities are identical for the "natural 1" and "natural 2" rows, since in either case the player succeeds on a 2-20, and fails on a 1. Concrete example: Your save bonus is +10, and the DC is 11. That is the sense in which you only need a 1. However, you only succeed on the save if you roll a 2 through 20, and you still fail if you roll a 1. For a single roll without the option to reroll you make it 95% of the time. If you have the +2 bonus, then your actual save bonus is 12 and so you still succeed just 95% of the time on a DC 11 save. If you reroll and always use the higher result, the chance you fail is 1 minus the chance of rolling a 1 both times. So 1-1/400=.9975. For the various k values, there is a chance you thought you need to roll higher to make the save and thus reroll, but get a 1 on the second try. That is why the probability decreases slightly in these columns. If you rolled a 2, but thought that wasn't good enough you might choose to reroll. One this reroll you have a 1/20 chance of rolling a 1 and thus fail the save. The different k values specify how likely you are, in the model presented, to make that unfortunate choice. For the k=.1 case, let's enumerate specifically the ways you can fail. You can roll a 1, and then reroll a 1 and fail. This occurs with probability 1/20*1/20=1/400=.0025. You can also roll a 2-19, and with a 10% probability think your roll isn't good enough and roll again, and on this reroll you have a 1/20 chance of getting a 1 and failing. The probability of rolling a 2-19 is 18/20, and the ten percent chance of rerolling multiplies this by .1. On your reroll, you have a 1/20 chance of failing. Thus, this line of rolling gives you a 18/20*1/10*1/20=18/4000=.0045 chance of failing in this way. Those are the only two ways you can fail, and so the probability of succeeding is 1-.0025-.0045=.993. Now, I will grant, in actual fact a person is probably not going to reroll a 2 and a 19 with the same frequency. However, this is a problem with the model, not with the math that follows the model. And the fact is, I made this assumption, and its shortcomings, explicit in my original post. In any case, I would like to thank Tarkin. Nothing motivates someone to check their own results like skeptical eyes. Also, thank you Piropoint. I imagine that eventually I will submit something to Regdar's Repository. Informed choices are good, but it's clearly a good thing I didn't jump the gun and submit to Regdar right away.
Piropoint 04-24-07, 11:03 PM	I can see the title now- Probability, Luck, and the Metagame: Oh the Possibilities!
Daeva 04-25-07, 04:41 AM	ok for real, get off your ULTIMATE POWER CHARACTER high horse, and have osme fun, play a cleric with the luck domain and make the horrrendous (you wont even consider this) 3 levle hit on your casting and play fortunes friend, or just play a rouge witha level or two of cleric/wizard and play fortunes friend, ocme one three extra luck rerolls, 2 extra luck feats, add levle to all skill checks once per day, add levle to attacks for a round once per day, add levle to all saves once per day, multipleluck uses per round if needed, have some fun i did and its actually really fun, and i am still capable of helping the group just fine as a human rouge 1 cleric 4 fortunes friend 5 with 5 luck feats (two were class bonus), going into shadowbane stalker
Piropoint 04-25-07, 08:58 PM	I posted these a while back, but i think this is a good place for them. My ideas for new luck feats: Concentrated Luck Some days you aren't lucky as often, but when you are it really makes a difference. Prerequisites: Any three luck feats Benefit: Whenever you use a luck re-roll, you may expend a number of daily re-rolls no greater than 1/4th your character level to add a +1 luck bonus to the final result for each re-roll expended. You may add these points after the final outcome of the re-roll is determined. Example: a 12th level character with 4 re-rolls remaining may add a Luck bonus of up to +3 to the final result of their next luck re-roll by using all 4 re-rolls. Note: This feat alters the final outcome, not the actual roll. Natural 1 still counts as failure, and 19 dose not automatically mean success. You gain one luck re-roll per day Lucky Lullaby Your good fortune sometimes shines through your music. Prerequisites: Bardic Music, any two luck feats Benefit: You may expend daily luck re-rolls to produce a number of effects depending on which song is played: Counter Song: add 3 to the result of the Bard’s perform check for each expended re-roll. Fascinate: A creature remains fascinated for an additional round after the Bard stops playing for each expended re-roll. Inspire Courage: Allies receive a +1 luck bonus to their weakest base save. (if two are tied for weakest, choose one) This increases to +2 at 8th, +3 at 14th, and +4 at 20th. Only on re-roll may be expended each time this song in used. You gain one luck re-roll per day Rogue’s Chance Your good luck is bad luck for anyone you catch off guard Prerequisites: Sneak Attack +2d6, any luck feat Benefit: Whenever you successfully deal sneak attack damage you may use one of your daily re-rolls to re-roll the sneak attack damage. You must take the second result, even if it is lower than the first. You gain one luck re-roll per day Coming soon: A Brand New Familiar!!!!!
CaoSlayer 04-25-07, 09:38 PM	The rogue one would be more interesting if in place of re-rolling damage allowed a rogue to turn a normal attack into a sneak attack, like a "lucky strike".
Piropoint 04-26-07, 05:07 PM	Rabbit Familiar- Tiny Animal Stats- Same as base Animal Special- Master recieves 1 luck reroll per day. If the master does not have access to the luck domain or at least one luck feat, this ability does nothing.
Deep Blue 9000 04-27-07, 04:15 AM	In any case, I would like to thank Tarkin. Nothing motivates someone to check their own results like skeptical eyes. Sorry about that Tarkin. My bad. Anyway Ainamacar, if you feel like doing some more modeling, I think it'd be very interesting to see how big of a bonus to saves a reroll is equivalent to at various DCs. I probably wouldn't spend a feat to get +4 to one save per day. And getting a reroll 1/day seems like a steal for a feat, but that may be because its value is hidden in complex probabilities.
Piropoint 04-27-07, 06:14 PM	Anybody want to start a "pro luck feats" thread? A friendlier title might draw more sypathetic conversation. The reason I ask instead of just doing it is because I know people get tired of repeated threads, and someone would probably just link me back to this one. whatever.
pmurray@bigpond.com 04-29-07, 08:10 PM	I think it'd be very interesting to see how big of a bonus to saves a reroll is equivalent to at various DCs. Done! These are the bonuses in different scenarios. The numbers are fractional because there are 400 ways to roll two d20s. The first column is the DC. The second column is the equivalent bonus in a straight "beat the score" roll (eg, a skill check). The third column is when one is an auto fail and 20 is an auto success. (note that I compare this against the same rule for a single roll - that's why the bonuses aren't uniformly higher) If you roll a 1 and re-roll a 20, then that 20 counts as a natural 20. The fourth column is when one is an auto fail and 20 is an auto success, but they only count on the first of your two rolls. The fifth column is when if you roll a natural 1 on the first roll, then a natural 20 on the second roll does not count as an automatic hit (otherwise it does). DC Roll twice 1 and zero first only rerol of fumble is no crit 0 0 0.95 0 0.95 1 0 0.95 0 0.95 2 0.95 0.95 0 0.95 3 1.8 1.8 0.9 1.80 4 2.55 2.55 1.7 2.55 5 3.2 3.2 2.4 3.20 6 3.75 3.75 3 3.75 7 4.2 4.2 3.5 4.20 8 4.55 4.55 3.9 4.55 9 4.8 4.8 4.2 4.80 10 4.95 4.95 4.4 4.95 11 5 5 4.5 5.00 12 4.95 4.95 4.5 4.95 13 4.8 4.8 4.4 4.80 14 4.55 4.55 4.2 4.55 15 4.2 4.2 3.9 4.20 16 3.75 3.75 3.5 3.75 17 3.2 3.2 3 3.20 18 2.55 2.55 2.4 2.55 19 1.8 1.8 1.7 1.80 20 0.95 0.95 0.9 0.95 21 0 0.95 0 0.90 22 0 0.95 0 0.90