Hormalakh

Why would they? A lot of people have voiced this concern. Apart from a few overly sensitive types I see this thread as nothing but win for Obsidian. They're making the game they wanted to and getting feedback from their players. We make comments on what we've seen in the past and they consider it. It's much better than a producer telling them what to do and there being no reasoned discussion about it. I live in America though and believe that the best ideas come from the reasoned discussion and evaluation of ideas. I don't think Obsidian is the type to be overly sensitive about what their varied gaming demographic says. It's pre-production afterall.

Yes, that is a consequence of normalizing ranges, so again this comes back to asking players the question, "How much chaos do you like?" In many cases, this is a personal preference. I have, for instance, seen people request elements like the fabled Ars Magica/Rolemaster botches and crits of old, which were wild and crazy. I have been thinking about this and I think the answer to the question lies somewhere near "enough chaos to give the player the illusion of choices in combat." If the chaos is sufficiently interesting to make gauging combat scenarios and whether a party should engage or not an interesting choice, then I think the job is done. There should be times where players should not be absolutely sure whether they can take on an enemy or not. Doing so makes players feel that the game is less linear than it would otherwise be.

An 80% chance to hit is a 20% chance to miss; IIRC, three consecutive misses is (.2)^3 = 0.008, or 0.8%. My mistake, yes. It's 0.8%. Edited previous comment. You know this might be a mistake on my part. Looking at the DoTA/LoL games, those games really are dependent on gear (very little character customization) and so I might have jumped to a mistaken conclusion there. Obviously the DnD 4e games wouldn't be affected this way. As long as the absolute difference between a hit and a glancing blow from a weapon is significant enough that character skill matters, this really shouldn't be a problem.

I would actually be very happy with this. I think that it's the beginning of striking a good balance.

I can't count the number of times I've seen players make a complaint in the vein that their XCOM units had 90% chance to hit and missed three times in a row. "Impossible!" An 80% chance to hit means you miss 1 out of 5 times (20%) for the next shot. An 80% chance to hit means that 1 out of 125 times (0.8%) a player will miss every shot for the next three shots. Players don't realize that each shot is an independent event however, and the second line "makes more sense" to them, where in fact, after each shot, they need to realize that their probability has once again "changed back" to 20%. Probability is hard for some. [h/t: rjshae]

Long post. I've bolded the important bits. Sorry! Thanks Josh for taking the time to have this conversation with us. I've been thinking about this since yesterday and I finally got a chance to experience the game mechanic you are proposing in a game or two. A few things struck me and I think I'm better able to articulate my concerns. I firstly realized that, as you had pointed out, the question here can become one of variability over random conflict resolution and how to make this interesting for the player. It seems to me (and I could be mistaken) that you believe limiting this range might be beneficial for the game in some way. For example, instead of so-called "chaotic" ranges in probability, we tone down the chaos. I considered this aspect in certain games and tried to evaluate them in the context of cRPGs and what I find to be interesting about them. I first noticed (in games like LoL, for example) that the HP bloat wasn't addressed, but that turned out to not really be the biggest issue for me. I should firstly restate that I am speaking strictly about melee combat here. I think different systems can utilize slightly different "ranges" in probablity distributions, but what worries me the most is the lack of "chaos" when it comes to melee combat. When I considered melee combat in probability-based conflict resolution, I realized quickly that conflict resolution more immediately became less reliant on the skill of my character and more reliant on the loot/equipment I was carrying. As the variabilities for these weapons started to decrease, it became more important to find a "stronger" weapon to increase the base damage than it did to increase my character's skill with the use of the same equipment. I think this becomes mainly the biggest of the problems for me. I also realized that I was missing the "frustration" of early level combat but at the same time, there was a more linear approach that I should be taking towards combat. If my variances fall within a certain range, my character can only approach a smaller subset of combat situations at any time. I know the average damages that I can produce at any time and the combat situations I put myself in must fall within the appropriate risk/reward scenarios. As these variance ranges of probability decrease, my options of "viable" combat scenarios decrease. Dodging enemies allows you to sometimes risk fighting enemies at a higher-level than you, even though the the risks are high. Yet the rewards for such a fight are also high. I would thus propose that you consider increasing the "chaos" of your probability-conflict resolutions, but either tier them based on skill, or utilize thresholds. One possibility is to utilize poisson distributions and other non-normalized distributions to show character skill. What this does is allows a character to use the same long sword they got in Chapter 1 as a viable weapon, but because the skill in the character has increased, the probability ranges of the weapon has changed. This makes the character the actual weapon and the long sword the tool. I have a previous post here that tries to further explain this. ------------------------------------------------ I also have a few comments about the considerations that you have made here about XCOM and player's reaction to RNG. I will approach the RNG first. I, like you have already explained, have sometimes found it difficult to clarify concepts of probability to people who look at such problems from a different perspective than I. It thus seems to me that the problem isn't probability per se when it comes to conflict resolution, but the perspective in which it's framed for the players. I wonder if changing the terminology might help players better accept this. Changing "Chance to Hit" to "Chance to Miss" for example, would be an interesting experiment to try with your testers. See if changing how the probabilities are described to them changes the way they see the situation. And finally about XCOM. One of the issues I have with you using XCOM as the go-to for RNG failure is because of the way that their RNG works. They used pre-seeded RNG that meant meta-gaming was part of the game with a saved seed at the start of the game. Your RNG would never change every time you reloaded. It’s basically allowing the player to know the dice rolls for the next few rolls. The probability values change completely when you have a priori knowledge of those rolls (see the Monty Hall problem for an example of this). The probability of a 6-side die rolling a 1 is 1/6. But in X-COM, meta-gaming meant that the probability of rolling a 1 is either a 0 or 100%. This was meant to fight against the degenerate game-reload, but the issue is the player mentality when it comes to risk/reward structure. Like I’ve said before, a high risk/resource action is more likely to initiate player reloading if a roll is bad (disintegrate save-or-die), but a low resource action that is being performed several tens of times is less likely to initiate player reloading for a few bad rolls. In my most recent experience with Arcanum which had quite brutal critical fails, the thought of gaming the system didn’t cross my mind because I would have to save/load so frequently as to make it frighteningly boring.

Wow.... I guess I'm in the minority here. I'm not sure how missing (0 damage) suddenly equates to RNG-hate. Can someone explain this to me? The HP bloat problem really hasn't been addressed as well as the issue when considering risk/rewards in this sort of structure. While I don't necessarily have a problem with resource-heavy things like spells having minimum damage, resource-light mechanics like melee battle, seem a little less interesting when not considering the risk/reward of missing/hitting for higher damage. You are looking at averages over time and seeing that they play the same, whereas it's not the averages that matter or make the older approach interesting, it's the spikes and valleys over time that make one different than the other.

lol @Malekith. I was just reading those posts.

There is a question that has been asked on Josh's Formspring and I wanted people to share their opinions on the matter. Right now, Tim, Steve, and I are experimenting with with a set of defenses that cover the basic "did you get hit?" mechanics of all sorts of attacks, from melee swings to arrows to fireballs to mind control spells. Currently, characters have a defense against melee attacks that attackers try to overcome (like AC without the armor component -- but with shields). A "miss" against any defense translates to half minimum damage inflicted or half minimum duration on any sort of status effect. I.e. there aren't "full" misses, but mitigated effects. A hit is the standard damage/duration. A hit that is within the critical hit range does 150% max damage or duration. This system is already implemented and seems to be working pretty well, but we'll continue to experiment with it. Some people asked a few follow-up questions. Josh's answers have been highlighted in green. Q: Wait...so neither the enemy nor your party members can ever miss? As in causing 0 damage/duration? Also what about critical misses? Not currently, no. There is no special effect for a "critical miss". Q: What made you decide that there shouldn't be a 0 damage miss. That is to say, what problem did you see with prior implementations of this that made you decide to try a new approach? All-or-nothing results tend to produce large spikes in conflict resolution. On the extreme end, you have traditional AD&D spells like Disintegrate that either annihilate the target completely or... do nothing. More typically you have the standard to-hit roll that either results in normal damage or absolutely nothing. Because the gulf between success and failure results is so large, random chance has a very large impact how the conflict works out. This system normalizes the results. Our goal is to make your choice of tactic ultimately more important than the results of the die roll (though the die rolls still matter). If we're only implementing mechanics that are proven to be fun in RPGs, I'm not sure why we're talking about D&D's THAC0/BAB system. Players generally dislike the all-or-nothing results of those mechanics, which is why you saw a move away from it in 4E. Q: Do you have any sort of source material on which you're basing this system? I'd assumed you're only implementing mechanics that have been proven to be fun in RPGs, ideally CRPGs. As for source inspiration, 4E's dailies' miss results are a pretty good start. Also a lot of RTSs and MOBAs have moved to much more deterministic systems. Q: That doesn't mean you should preclude 0 damage misses completely, especially in something as resource cheap as melee damage. Disintegrate was a resource heavy spell and I can understand that. Why not weigh your probability distributions and still have a 0 damage for those unable to pass a threshold like you are intending with lockpick and other events. Afterall, even a failed lockpick doesn't allow half of the party members through a locked door. Locked doors are a traditionally problematic conflict resolution in games (as are most all-or-nothing checks) and, I think, highlight the problem rather than absolve it. My question is: how do "full" misses make gameplay better than mitigated results? Q: It becomes a problem of victory through attrition. It can also limit the number of enemies attacking a party at one time. If you have 100 goblins and each always gets 1 point of damage even when they miss, that's a problem. Have you considered how this scales with lower-level and high-level party members? I can't simulate this, but does this adversely affect certain stages of the game more than others? We're not planning on hundred enemy combats, but even at normal IE stages, I don't think it's a large problem. As for how it scales, we already know how standard THAC0/BAB scales (poorly), but it is one area we will continue to test. Q: I'd wager that you're underestimating the fun of dodging and missing. It doesn't need to be as prominent as it was in Baldur's Gate-era missfests, but people like making characters that dodge all incoming damage. Also, the risk of doing no damage is fun. I think you're overestimating the fun of dodging and missing. I don't think most players find it particularly enjoyable, and it's exacerbated/amplified in games like the new XCOM where players are constantly in stunned disbelief at the RNG. What are your thoughts? Post them here or at Formspring if you wish. http://www.formsprin...896189576204826

What's less surprising is the lack of love for Barbarians. Give us more information about these guys, Obsidian!

Hey guys, just checking in on this thread. Some very cool ideas and I'm hoping some future modders can work this out. I'm really having high hopes for this game, and if all else is good, a RL mod would be awesome. As for Diablo, I've played Diablo before and I never thought of Diablo as a rogue-like. So there's that. I still don't think Diablo is a roguelike, but to be honest, I've only started getting into the RL scene, so I'm not sure what defines a true "rogue-like." Anyway, most people have understood what I meant here, so that's cool. A true endless path would be endless It's meant to be a fun mini-game - one where you can never win... (or can you?) No, no you can't.

cant wait for the RL mods.

Stash is available at CAMP, HOME, STORE, and probably a few other places.

Think less. Wait more.

All I could think was YABBA DABBA DOOO

I remember having a link to ideas about uncovering fog of war. I'll repost it here: http://forums.obsidian.net/topic/61425-let-us-see-the-whole-area-please/ Many ideas were talked about here, so in case the devs do look at this thread and do want to change FoW and how it's implemented, this would be a good place to start

The game pausing thing is very much one of those "if you want it, you should restrict yourself" kind of things. If you're interested in the challenge provided with limited-pauses, you can go ahead and place that limit on yourself. RT with unlimited pause has worked effectively in the past without sacrificing gameplay or challenge. While I hate this argument, I feel that it works here: don't railroad other players into your own playstyle. If you want to limit your pauses, go ahead. Don't force others to play that way.

You could always just wait for the game to come out :smug:

So an idea: The endless dungeon, Od Nua, is going to be a pre-designed dungeon featuring 15 levels of dungeon crawling. I was wondering whether (either in this game or in the expansion) it would be possible to consider leaving a "secret entrance" to levels below 15 that are randomly generated with increasing difficult monsters and sort of act like a "rogue-like" mini-game? a few thoughts: 1- this will probably require a lot of design work, so I don't want this being part of this game but maybe for an expansion, it would be nice to continue down Od Nua forever? 2- Obviously the balacing would have to work so that you don't really gain too much treasure and experience to unbalance the game. maybe you can use what you find down there and the experience helps you to gain levels, but whne you leave you get level drained back to what you started with. (just an idea) 3- this also requires randomly generated tile-sets which isn't what this game is. Hence more design work. Obviously a rogue-like only works with infinitely procedurely generated tile-sets. Anyway... think about this for a future expansion or a mini-game type thing. It'd be fun! Now destroy my idea everyone! Pile on!

No, no. No anime size, but it would be interesting to know how a halfling carries a two-handed sword... I would imagine a claymore would be too unwieldy for a halfling. Maybe they two-hand regular longswords?

Exactly my point. Which means the addition of a, imo, pointless "deep stash" is there only to limit players from accessing 100% of the stuff they have picked up. Why? Please dont say beacause of "degenerative inventory". Punative. You make it sound like any sort of strategy or tactic is punitive. A game without rules isn't a game. Just wanted to requote this from the OP for you, emphasis mine:

Weapons aren't race-restricted in game. However I do agree with the OP: it would be interesting to see a halfling wield a larger weapon. Not sure if strength checks will be implemented, but it might be interesting to see such a thing. Of course, then we will also see things like FO:NV with the fire hydrant melee weapon.

The fallout version (implement such a thing for important NPCs) would be more than "fun-sized." THAT WOULD BE AWESOME!!! :dancing: :dancing:

Of course, Oblivion isn't what Project Eternity intends to replicate.

Do not mess with the Jesus...