jelloflechette

Re: How do you handle campaign limits on defenses?

Re: The cost of killing damage When I started reading your post I suddenly realized it can be interpreted in two ways; one where "game" is the actual game, and another where "game" is a metaphor for online discussions of the game. I don't visit these forums often, so I'm not familiar with the personalities of regular posters. In any case, thanks for the heads up; no more responses to Comic out of me.

Re: The cost of killing damage The thing is, it's only the probability of at least one high-damage roll occuring that approaches unity. The cumulative amount of high-damage rolls remains a relatively small ratio of the total damage rolls. Going from the p->1 statement to saying "always nail the KA guy" is only true if you're assuming that the cost of taking a high-roll KA hit is astronomical compared to the cumulative costs of damage from normal attacks in the time it takes for that high-roll KA hit to occur. Regardless of the raw numerical value of the high-roll KA hit, if the cost assigned to it is not high enough, there is no reason to consistently go after the KA character.

Re: The cost of killing damage Thanks, jtelson. What you stated here is what I had been trying to get at. Comic has been adamantly adhering to the Game Theory "mandate" that high std dev variables must always be considered as the maximum value when trying to win the game. This sounded like over generalization to me. I had been considering the following example. There are 3 units: A, X, and Y. X deals {p(100)=0.01, p(1)=0.99} damage per turn, and will be eliminated after taking 4 damage. Y deals 3 damage per turn, and will be eliminated after taking 2 damage. A deals 2 damage per turn, and will be eliminated once it takes 6 damage. X's damage output has mean 1.99 and std dev 9.8, while Y has mean 3 and std dev 0. Unit A goes first. Both X and Y will attack A on their turns. Based on what Comic has been saying: since X has the higher std dev, according to Comic's statement of Game Theory, A should eliminate X first. Using this approach: [Turn 1] A attacks X, X is down to 2 life points. Both X and Y attack A. A has a 1% chance of being dead, and 99% chance of being down to 2 life points. [Turn 2] A attacks and eliminates X. Y attacks A. A is dead. So, after 2 turns, A will die with 100% chance. Consider the alternate approach, where A eliminates Y first. [Turn 1] A attacks Y, Y is eliminated. X attacks A. A has a 1% chance of being dead, and a 99% chance of being down to 5 life points. [Turn 2] A attacks X, X is down to 2 life points. X attacks A. Cumulatively, A has a 98% chance of surviving the second turn. [Turn 3] A attacks X, X is eliminated. So, by eliminating Y first, X has a 98% chance of victory. Clearly then, "eliminating the damage function with the highest standard deviation" is not optimal in this case, because it gets you killed. And since it's not optimal in this case, it certainly cannot be always optimal. I was asking for more information to determine the various parameters of the theorem/heuristic that Comic had been referencing.

Re: The cost of killing damage Out of curiosity, could you state the theorem that describes this? Is there a threshold that the std dev needs to exceed before one treats an RV as its peak value? It sounds like there's a difference between mathematical expected value vs decision making "expected value". What do you mean by "leaving it to chance"? Are you comparing [rolling nd6 killing damage and then multiplying by some k] with the standard method of [rolling nd6 killing damage and then multiplying by {1,1,2,3,4,5}]? In other words, what's the other mechanic that you're comparing the constant multiplier mechanic to? Also, would you mind at least sketching out the reasoning, or maybe providing some formulas? Could you point out some resources regarding this topic?

Re: Old School vs. New School Phil and Hugh, thanks for the detailed info; that's exactly what I was looking for!

Re: Old School vs. New School I have a similar question, but mine's more properly titled "Old Old Old School vs New School". Does anyone here have experience with the original first edition Champions? If so, what are the major differences between it (a 50-some page book, I believe) and either 4th or 5th edition? I'm sure there's differences in the number and details of powers, adv/lim, and disads. My question is more along the lines of whether there are any fundamental differences in mechanics. For example, if there's no vehicle creation rules in 1st edition, or if it lacks detailed rules for movement or perception, or if it uses different dice-rolling mechanics, etc. Oh, and one very specific question: does 1st edition have the "mod the rules however you like" clause? The reason I ask is because if 1st edition is functionally similar to the hero system as we know it today, I'm thinking about picking up a copy. However, if it is significantly less powerful (as far as being able to universally model anything), I'll pass. Anyway, if anyone wouldn't mind sharing their knowledge and experience it would be greatly appreciated.

Re: 6th Edition Hero System I'm pretty late to this discussion. I've read through most of the posts, but I mainly want to comment on the "centralized universal book vs multiple themed" books debate that has been going on more recently. I want to first point out that I whole-heartedly agree with ghost-angel's stance on this issue. The way I see it is this. The "add-on" or "plug-in" model that shnar is proposing (similar to GURPS, Fuzion, 40k and its army codices, etc) makes it so that the core system is handicapped and essentially pointless. To have the rules for whatever genre or situation you end up HAVING to buy the plug in. This is why I always believed the GURPS title is a lie: it is neither "generic" or "universal". The core book comes with a small list of spells and equipment. What if I want to use a different spell or item? I have to buy the "GURPS blah blah" sourcebook. So, there really is no "GURPS System"; there's only an empty framework where actual game material gets placed, in the form of source books. On the other hand, the hero system is a fully stand-alone, truly universal, simulation system. The core rules already allow you to simulate whatever you want to do. Any extra books are guidelines and tips on how to tune the core system to a particular end. Hero is an FPGA that I can program to do what I want; other systems are ASICs, forcing me to get yet another chip for each new thing I want to do. Beyond that issue, I also wanted to state my concern on a 6th edition. The jump from 4th to 5th was valid, I guess. There was a significant amount of clarification and added examples, which is good. The rules tweaks are arguable, but nothing too drastic. The new art is for the most part very nice too. But to me, the jump from 4th to 5th was more of a "Hero Lives Once More" announcement, and to flush the sour taste of Fuzion from everyone's mouths. That alone makes it have value. But I honestly cannot thing of what would justify a 6th edition. The system was pretty much good to go in 4th. TSR sold out long before they were bought out. GW also sold out a while ago. I have more confidence in Hero, since Steve Long seems to actually care about the game system, but I have to say it any way: please don't sell out. I hate planned obsolescence. I do not want to play a system that requires 3+ handbooks just to have the rules. I do not want to play a system that requires any amount of source books to have the rules. I do not want to play a system that constantly overwrites the rules in monthly magazines. I do not want to play a system that has an online subscription component *shudders*. Sourcebooks that present innovative ideas on how to use the core rules are great. Unique settings would be great too. And like Ghost Angel stated earlier, each setting doesn't have to be compatible with every other one, or fit into some overarching timeline. Just don't tell me I need to buy additional books to get the rules.

Re: Gaming Stores in San Jose/Santa Clara area Hi, One possibility is D&J Hobby, on San Tomas Aquino Road in Campbell (just a little south of Santa Clara). Number: (408) 379-1696. They have games, but as their name implies they also have a ton of other stuff (RC, model, arts and crafts, etc). Might save you a trip or two if you're into any of those other things.

Re: Possible New Player/GM with Questions To the original poster, if you haven't already, you might like to check out the Hero System overview that's available as a free PDF on the homepage: www.herogames.com/get/HRO_int.pdf It's a pretty good summary.

Re: i Can't Die! No matter How Much I want To! Perhaps this "ability" can be represented simply as a special effect of death. After all, if in terms of game mechanics the character can't do anything, he or she is functionally dead. This is just another type of death. Maybe add in the physical limitation: "feels pain even when 'dead' ".

Re: Unofficial Welcome Mat (For New Members) Hi! Thanks for the welcome. Posting fee on its way via paypal. Hey, I just noticed there's a "smiley" face for the blue moon killer. Neat.

Re: Sooo... Immunity to magic Hi, I'm very new to these boards, so please pardon my butting in. I was quite interested in the various cost sequences people have been tossing around. Suppose we have a sequence i, starting at 0 and incrementing by 1. Let cost(0) = 15 and DR(0) = 25. These are agreed starting values. This is true if the cost and Damage Reduction functions are defined as follows: cost(i) = 2*cost(i-1) DR(i) = 25*(i+1) From these functions we get: [i, cost, DR] 0, 15, 25 1, 30, 50 2, 60, 75 3, 120, 100 This is also true for the following functions: Z(0) = 1 Z(i) = Z(i-1)^2 + i + 1 W(i) = 5/Z(i) cost(i) = 2*cost(i-1) DR(i) = 100 - w(i)*cost(i) From these functions we get: (i, W, cost, DR) 0, 5/1, 15, 25 1, 5/3, 30, 50 2, 5/12, 60, 75 3, 5/148, 120, 95.95 4, 5/21909, 240, 99.94 5, 5/480004287, 480, 99.99995 Since the exponential in the denominator of W grows faster than the doubling of the cost, the function is guaranteed to not diverge. It approaches 100 as Sean mentioned. Of course, it's not exactly a "simplistic" approach as can be seen from the functions; quite the opposite. This is the problem of curve fitting when given very few sample points. Of course, one could invoke Occam's razor to claim that the first set of formulations should be used. Then again, one could also claim that the "true" sequence is as follows: 15, 30, 60, f(amount of $ paid to GM) where f is inversely proportional of course.