Autofire Question

[Klaus Kipling]

In MT tasks were either average, so easy as to be automatically successful, or so hard as to be almost impossible. Not much granularity there.

Wrong. MT's task system has an autofail on natural 2; and a natural 2 aggravates any failure where a 3 would fail.

So no matter how good the character, under MegaTraveller, 1/36 of tasks fail.

I should have said "as to be almost automatically successful". It's just with 4 increments per difficulty level, a task goes from fairly easy to rather hard very quickly.

 

[Settembrini]

So does shooting.

 

[tbeard1999]

No, I have not once conceded that the mechanic was broken. If you're going to misrepresent what I've said it doesn't help your credibility. :roll:

When you agree that I have a point, then proceed to suggest a fix (that breaks the system in a different way), I assumed that you agreed that the system was broken. Silly moi.

This, in no way, is saying that the base timing/effect mechanic is broken, just misapplied in this instance.

Well, I didn't mean this to suggest that you thought that the whole system was broken. However, since you raise the issue...

Here are the odds of rolling a success and rolling the indicated result on each die (percentages rounded to nearest 1%):

-2 Modifier to the overall roll and to the die
2-: 17%
3-4: 83%
5+: 0%

-1 Modifier
2-: 10%
3-4: 50%
5+: 40%

No Modifier
2-: 7%
3-4: 33%
5+: 60%

+1 Modifier
2-: 5%
3-4: 24%
5+: 71%

+2 Modifier
2-: 0%
3-4: 20%
5+: 81%

+3 Modifier
2-: 0%
3-4: 7%
5-6: 93%

+4 Modifier
5+: 100%

So...if I make an unmodified roll, and it succeeds, each die has a 60% chance of being a 5 or higher. The chance of both dice being 5+ is 36%. The chance of neither dice being 5+ is 16%.

Note the tremendous difference between a -1 and a -2. A roll at -1 has a 40% chance of being an exceptional success. A roll at -2 has NO chance of being an exceptional success.

In fact, the exceptional success is the most likely possibility for any roll with a modifier of 0 or more and a 40% at -1. But at -2, an exceptional success is impossible.

And here are the chances of rolling at least one 5+:

No modifier: 84%
+1: 91%
+2: 96%
+3: >99%
-1: 64%
-2: 0%

Does this look like a good distribution of probabilities for an RPG?

As you are in the vast minority on this playtest forum in decrying the mechanic, then my statement "everyone likes it" can stand. I see no evidence here to the contrary.

Really? On a Mongoose hosted public forum, I am in the minority of critics of a proposed Mongoose product? I am shocked, shocked...

I only see 1 lone voice claiming the mechanic is broken.

You don't think the above analysis indicates that the system yields dubious results as it is written?

 

[Klaus Kipling]

Actually, those numbers are fairly good, IMHO.

It's only the definition that needs a lil' bit o' tweaking.

The way I would do it (and I'll include this in my submission to Mongoose control) is:

1-3 Satisfactory/Average/Mediocre success (C- in an exam)
4-5 Good/Professional/Competent result (B in an exam)
6 'Exceptional' result (A+)

Somebody with a +4 bonus is either highly skilled and able or doing a really simple/easy task (or both). That excellence should be common and reliable is natural in such circumstances.

And in most circumstances how that result is interpreted is up to the ref; in fact, effect might only be important in particularly fraught circumstances.

In most cases it might only provide descriptive colour; but that's no bad thing.

As for the likelihood of one die being high most of the time, that gives the player the option of either a quick but potentially scrappy success or taking their time to ensure a slick result. That gives the player a say in how any roll pans out: that gives a player ownership over what otherwise is just being a victim of random chance.

Also, timing and effect are relevant even in failure.

Now, it's possible to assume that a player will always take the highest die as effect in the event of success, and as timing in event of failure, but that's a very simplistic reading. The particular circumstances of any task are unique and how a player might use the results tactically is not as predictable as the probabililties of particular dice rolls.

Getting all that information from a single dice roll, without requiring a look-up table (like in Spacemaster or Zebulon's Guide Star Frontiers), is pretty sublime in my book.

I'm not sure writing off the posters on these boards as Mongoose "fanboys" is very helpful, as many of the most vocal are folk that frequent all the Traveller boards.

So while I'm grateful for your statistical analysis, which in this specific instance (Autofire), has spotted a real glaring flaw, I don't think your thesis that the core mechanic (timing/effect) is a busted flush has been proven.

(I do think weapon damage does need to be revisited, to increase variability: I'd suggest any damage mod of more than +3 or 4 be expressed as an extra multiplier, unless there's a specific justification for the weapon. So instead of 2x +6, it is 4x)

 

[tbeard1999]

Actually, those numbers are fairly good, IMHO.

It's only the definition that needs a lil' bit o' tweaking.

Ah, the post-hoc rationalization.

<snip of completely new replacement system>

Of course, the irony here is that you seem to be agreeing with me that the system is defective. If not, why propose a significant modification?

1-3 Satisfactory/Average/Mediocre success (C- in an exam)
4-5 Good/Professional/Competent result (B in an exam)
6 'Exceptional' result (A+)

With no modifier, your system produces an equally dubious set of outcomes:

1-3: 14%
4-5: 50%
6: 35%

+1 Modifier
1-3: 10%
4-5: 35%
6: 55%

And of course, with even a modest -1 to the roll, there is no chance for an exceptional outcome.

I'm not sure writing off the posters on these boards as Mongoose "fanboys" is very helpful, as many of the most vocal are folk that frequent all the Traveller boards.

<shrug>

You used the word fanboy, not me. In any case, I see no reason to assume that the posters here are (a) representative of Traveller fans in general; or (b) representative of the target market for this product. And given my own experiences, I find that people tend to be rather reluctant to publicly criticize a game designer, especially on his own forum.

In any case, the statistical qualities of the system speaks for iteself. The fact that most folks have not noticed these qualities is irrelevant.

Nor is it terribly surprising that these flaws did not emerge in playtesting. First, blindtesting has been rather brief; only a few months. Second, it takes hundreds or even thousands of rolls for statistical anomalies to become apparent; in my experience, few gaming groups record every roll made and analyze it later. Third, given that the system skews significantly to exceptional results, it's hardly surprising that players would not initially see this as a long term problem; most players are happy when their characters do well.

And just in case someone is, as you say, a "fanboy", I'll only note that you're not doing Mongoose any favors by uncritically praising a playtest rules set. They're publishing this game to make money; and a critically defective game probably has a low chance of making money.

So while I'm grateful for your statistical analysis, which in this specific instance (Autofire), has spotted a real glaring flaw, I don't think your thesis that the core mechanic (timing/effect) is a busted flush has been proven.

So, you don't have a problem with the fact that an exceptional success is the most likely outcome if the modifiers are 0+? You don't have a problem with the fact that an exceptional result is virtually guaranteed with even modest positive modifers? You don't have a problem with the fact that an exceptional result happens 40%+ of the time with a -1 net modifier but is flatly impossible with a -2 net modifier?

Indeed, doesn't the fact that outcomes are so strongly skewed to exceptional deflate your argument that the timing/effect system gives useful additional information?

(I do think weapon damage does need to be revisited, to increase variability: I'd suggest any damage mod of more than +3 or 4 be expressed as an extra multiplier, unless there's a specific justification for the weapon. So instead of 2x +6, it is 4x)

Unfortunately, you cannot vary damage much as long as you tie it to the timing/effect mechanic. The reason is simple -- that system produces highly predictable outcomes. So whether you use a flat amount of damage or a multiple, your damages will always be extremely predictable and lacking in volatility.

And of course, your suggestions are not currently part of the rules as written.

I should also note that this system that you find so "sublime" produces equally dubious results in the case of failures:

No Modifiers
1-2: 52%
3-4: 33%
5-6: 14%

-1 Modifier
1-2: 65%
3-4: 27%
5-6: 8%

-2 Modifier
1-2: 77%
3-4: 23%
5-6: 0

So, a character with no modifiers has a staggering 52% chance of an "abject failure" if he fails the task.

Still sublime?

 

[Klaus Kipling]

Actually, those numbers are fairly good, IMHO.

It's only the definition that needs a lil' bit o' tweaking.

Ah, the post-hoc rationalization.

<snip of completely new replacement system>

Of course, the irony here is that you seem to be agreeing with me that the system is defective. If not, why propose a significant modification?

Wow, willfully misunderstanding my post once again.

In what possible way could shifting a couple of numbers one place be considered a significant modification?

And there's a huge difference between 'needing minor improvement' and defective. So if I break my little finger I should amputate my arm...?

Basically your numbers just confirmed what I'd assumed anyway, no need for me to rationalise anything.

And yes, it was I who used the word "fanboy", but dismissing the validity of the playtest forum you are using to argue a case is shooting yourself in the foot.

The purpose of this place is constructive criticism, which I see plenty of. You're just offering criticism, of the kind, "it's fecked, rip it up and burn it!"

So, you don't have a problem with the fact that an exceptional success is the most likely outcome if the modifiers are 0+? You don't have a problem with the fact that an exceptional result is virtually guaranteed with even modest positive modifers? You don't have a problem with the fact that an exceptional result happens 40%+ of the time with a -1 net modifier but is flatly impossible with a -2 net modifier?

Indeed, doesn't the fact that outcomes are so strongly skewed to exceptional deflate your argument that the timing/effect system gives useful additional information?

Please read my post before blundering on like this. An exceptional success is only 'the most likely outcome on a +0' if the player chooses the high die as effect. You assume he/she will do this all the time; that is an utterly fallacious assumption.

And no, I certainly do not have a problem with the fact that an exceptional success is less likely or impossible with a negative modifier. You've yet to give me a reason why I should be.

I do think the differentiation of the categories of success do need a slight tweak: that is what moving a number up or down one step constitutes - a tweak. Nothing significant.

And of course, your suggestions are not currently part of the rules as written.

Uh, doh. When did stating the obvious constitute an argument? Kinda the point of a playtest board to offer suggestions for improvement.

You're argument is becoming circular. You offer criticism, of the none-constructive kind, and when people give possible tweaks you call them significant changes and dismiss them because they're not in the current version of the playtest rules.

Meh....

[EDIT] I see you've posted more numbers. And I like them. It means that the higher the bonus the more reliable the excellent success and the greater the penalty the more likely it will be mediocre. That's pretty good.

As for the failure, I think tweaks here are in order too, but I won't bother suggesting them here as you'll only dismiss them for not being already in the playtest rules.

 

[tbeard1999]

Actually, those numbers are fairly good, IMHO.

It's only the definition that needs a lil' bit o' tweaking.

Ah, the post-hoc rationalization.

<snip of completely new replacement system>

Of course, the irony here is that you seem to be agreeing with me that the system is defective. If not, why propose a significant modification?

Wow, willfully misunderstanding my post once again.

In what possible way could shifting a couple of numbers one place be considered a significant modification?

Given that your proposed fix doesn't really solve the problem (see my edited post in which I analyzed your fix), the question is moot.

The purpose of this place is constructive criticism, which I see plenty of. You're just offering criticism, of the kind, "it's fecked, rip it up and burn it!"

I suppose I should have mentioned that I spent the better part of a weekend fiddling with the mechanic before concluding that it could not be fixed.

So, you don't have a problem with the fact that an exceptional success is the most likely outcome if the modifiers are 0+? You don't have a problem with the fact that an exceptional result is virtually guaranteed with even modest positive modifers? You don't have a problem with the fact that an exceptional result happens 40%+ of the time with a -1 net modifier but is flatly impossible with a -2 net modifier?

Indeed, doesn't the fact that outcomes are so strongly skewed to exceptional deflate your argument that the timing/effect system gives useful additional information?

Please read my post before blundering on like this.

Howsabout you simply answer the question? And as noted below, you're in a no position to lecture anyone about reading posts.

An exceptional success is only 'the most likely outcome on a +0' if the player chooses the high die as effect.

Nope. Perhaps you should read my post again. The listed percentages are for each individual die. If the player consistently high die as the effect, the chance of an exceptional result increases.

You assume he/she will do this all the time; that is an utterly fallacious assumption.

Good thing I didn't make it then.

 

[tbeard1999]

As for the failure, I think tweaks here are in order too, but I won't bother suggesting them here as you'll only dismiss them for not being already in the playtest rules.

At this rate, every specific application of the "sublime" timing/effect mechanic will need its own special set of "tweaks".

Hmmmm...is this something we traditionally associate with a good game mechanic?

 

[Klaus Kipling]

Howsabout you simply answer the question? And as noted below, you're in a no position to lecture anyone about reading posts.

An exceptional success is only 'the most likely outcome on a +0' if the player chooses the high die as effect.

Nope. Perhaps you should read my post again. The listed percentages are for each individual die. If the player consistently high die as the effect, the chance of an exceptional result increases.

You assume he/she will do this all the time; that is an utterly fallacious assumption.

Good thing I didn't make it then.

Well if you're not going to actually state that in your argument you can hardly fault me for assuming you were expressing it like you were in the previous post. If you're going to blind people with science it's best to be clear and explicit exactly how you're doing it.

And you are assuming the high die is for effect in your argument - you're completely ignoring timing.

So...if I make an unmodified roll, and it succeeds, each die has a 60% chance of being a 5 or higher. The chance of both dice being 5+ is 36%. The chance of neither dice being 5+ is 16%.

And given the fact that you do add modifiers to effect but don't to timing compromises the way you've analysed this.

This is getting very muddled.

 

[tbeard1999]

And given the fact that you do add modifiers to effect but don't to timing compromises the way you've analysed this.

This is getting very muddled.

Nice try, but it's only muddled if you don't like the implications of the analysis.

 

[AKAramis]

So you don't like it, and you've said why....

the question is, have you typed it up in an MS-Word or RTF, and sent it in per the playtest directions?

 

[AKAramis]

Possible fix for autofire (within the constraints of not fixing the task system):

Let autofire pick the Nth lowest die for timing, with N being skill level+1. So level 0 uses lowest, level one second lowest, level 2 third lowest, level 3 fourth lowest, etc...

 

[SableWyvern]

That's a little finicky, Akaramis, but seems workable.

 

[Klaus Kipling]

And given the fact that you do add modifiers to effect but don't to timing compromises the way you've analysed this.

This is getting very muddled.

Nice try, but it's only muddled if you don't like the implications of the analysis.

Actually, the analysis is totally flawed. You focus on effect die, when in fact the timing die is probably more important.

+3 Modifier
2-: 0%
3-4: 7%
5-6: 93%

This is nonsense when applied to the timing die - the higher the modifier, the more likely the timing die shows a full range on a success. So it shows you're measuring effect alone.

And anyway, it's totally irrelevant to timing whether the task succeeds or not, and has no bearing on an analysis of it.

Let's illustrate it another way.

Jonny has a red die and a yellow die. He decides he will always choose the yellow one to be the timing die. (That fits the rules as written)

As timing is not modified, and is relevant win or lose, what are the odds of any particular number coming up for his timing? Well, it's 16.7% for a 1, 16.7% for a 2, in fact it's the standard spread for a d6.

The fact is the easier a roll is, and on a success, timing will show a more standard range of results on a d6, not higher (given player choice, which complicates matters to the point where statistical analysis becomes unhelpful).

So, what does this mean? In hard tasks the timing die has a tendency to be higher rather than lower, and for easy tasks it has a tendency to be more spread out. That does not become a game breaker, as you insist.

Again, the effect die is being used win or lose, but obviously it is important whether the roll is successful or not. So, granted, the numbers need some tweaking - in a general sense; your accusation that there's loads of different caveats and special cases is just plain wrong: in the rules as written there's the standard way and autofire. That's it. Hardly a myriad of different applications.

And anyway, that is not automatically bad - applying exactly the same mechanic to every situation come what may is putting standardisation above detail and subtlety.

There's an inherent bias in your model that assumes the effect die to be the only one worth measuring. This undermines your case to be authoritive. The numbers are slanted, and you've only looked at the probabilities from one perspective, when the game is using them from 2 or 3 perspectives.

That you dislike this mechanic is your prerogative. You still haven't made a case for it to be abandoned as the assumptions you have are not the same assumptions folk here are using.

That it needs work is a given, but you are not interested in helping, just wrecking. You've shown contempt for the members of this board and arrogantly assume only you have the ability to run probability analysis. When people point out flaws in your model, which you lay out in a confused and contradictory fashion, you twist their words in order to justify your obsession.

Since you've misunderstood the intent of the mechanic from first principles it makes it hard to see the validity of your claims.

Sayonara. I'm off to do something useful.

 

[tbeard1999]

Nice try, but it's only muddled if you don't like the implications of the analysis.

Actually, the analysis is totally flawed. You focus on effect die, when in fact the timing die is probably more important.

More accurately, I haven't gotten yet to the fact that the timing die outcomes are defective as well.

In fact, timing is particularly badly broken because tasks get faster the worse the modifiers:

+1 Modifier
1-2: 14%
3-4: 33%
5-6: 52%

No Modifier:
1-2: 7%
3-4: 33%
5-6: 60%

-1 Modifier
1-2: 0%
3-4: 30%
5-6: 70%

-2 Modifier
1-2: 0%
3-4: 16%
5-6: 84%

So the worse you are at something, the less time it takes! Truly "sublime".

Let me guess--you have fix for this as well!

Of course, all of these "fixes" -- which don't really fix anything -- are different for each specific application of the timing/effect mechanic. That pretty much makes a mockery of the idea that it's a "universal mechanic", doesn't it?

 

[tbeard1999]

So you don't like it, and you've said why....

the question is, have you typed it up in an MS-Word or RTF, and sent it in per the playtest directions?

Yes I have. All of my complaints get sent in per the playtest instructions. I usually give the folks here a shot at them first, but ultimately, they all go in as an emailed Word document.

 

[tbeard1999]

Possible fix for autofire (within the constraints of not fixing the task system):

Let autofire pick the Nth lowest die for timing, with N being skill level+1. So level 0 uses lowest, level one second lowest, level 2 third lowest, level 3 fourth lowest, etc...

Congratulations. You've taken an overly fussy mechanic and made it even more so. Not your fault, however. The mechanic pretty much required it.

You also have introduced potentially worse problems. As the player's choice of die rises (from last to second to last to third to last, etc.) hits become nearly automatic, and the average number of hits goes up dramatically. With multiple dice, the odds of rolling at least one high number become nearly automatic. And if a "5" is the timing die, then 2/3 of the shots will hit on average, assuming no modifiers. If a mere +1 is added, 5/6 shots will hit.

Given a choice, I'd keep the original system -- it's no more broken than this idea, but at least its less fussy.

That said, any fix is gonna have to be more fussy than what the designer orginally started with for the simple reason that the timing/effect system is poorly designed to handle multiple shots. So any fix is gonna have to overcome that fact.

This, of course, raises the question why someone would choose a mechanic that's poorly suited for automatic weapons, since Traveller has a lot of automatic weapons...

 

[TrippyHippy]

In the combat playtests I've run, the players have been pretty happy with the set-up. Of course, they don't spend much time analysing the probabilities of rolls - which helps to move combat along, somewhat. They are just interested in getting good rolls and the excitement of the action.

To a certain extent, I can see the statistical arguments tbeard1999 (and others) are presenting - that difficult task rolls with positive DMs are weighted towards garnering high Effects - but I do think he exaggerates the degree of it's effect on actual gameplay. In most of the firefights I have run, players almost always opt to choose their highest dice for Timing - with the high Effect seen as a luxury in contrast. It's also an accepted point amongst players that the major aim in a firefight is to not get hit(!), which tends to be reinforced by high damage amounts, for successful hits. Armour can protect from a certain amount of damage, but it shouldn't be a substitute for ducking for cover! In the case of autofire, I just tend to think it implies uncontrolled bursts, for the purpose of giving cover fire - but I'm not an expert on military tactics, of course.

There is always a process of rationalisation for dice mechanics - and no system truly presents a perfect simulation of reality. I guess our group has unconsciously accepted the principle that skilled characters can turn what would usually be failures for other characters, into 'partial successes' at least. For less skilled characters, however, it's simply 'all or nothing' - you either get a full success (high effect) or totally miss.

I do welcome comments like these in the remit of an open playtest, by the way, but would hope to encourage less of the use of terms like "broken" (particularly if you haven't actually playtested it). It's quite emotive, for gaming geeks, and not particularly helpful. Try to be more constructive please, and to be clear, demanding that the whole core mechanic of the game be scrapped at this juncture simply isn't being helpful at all.

 

[tbeard1999]

I do welcome comments like these in the remit of an open playtest, by the way, but would hope to encourage less of the use of terms like "broken" (particularly if you haven't actually playtested it).

Ah, but I have playtested it.

Perhaps you can answer the question that started this thread (since none has been forthcoming so far):

How do modifiers (skill, DEX, etc.) affect autofire rolls?

Surely, this question has come up in playtesting?

It's quite emotive, for gaming geeks, and not particularly helpful. Try to be more constructive please, and to be clear, demanding that the whole core mechanic of the game be scrapped at this juncture simply isn't being helpful at all.

I think I'll need a definition of "more constructive".

Seems to me that if the core mechanic is broken, it would be a better idea to replace it now rather than in a supplement (or second edition) later on? (I have never bought into gamer conspiracy theories that companies intentionally publish flawed games so that they can sell Second Editions later).

And to be clear, I am demanding nothing.

I am merely noting that the core mechanic seems seriously defective and have laid out the statistical analyses that support this contention. If the mechanic is as good as you seem to think, then feel free to ignore my deranged rantings. Sales and the profitable life of the product line will ultimately determine who is right.

But if the mechanic is indeed seriously flawed, I can't imagine that Mongoose would allow something as prosaic as a deadline to force them to ship a product with a known, serious flaw.

EDIT: I would add that these criticisms are hardly new; I sent Mongoose an email describing the dubious statistical qualities of this system on 12-4-07. So it's a bit misleading to imply that I'm suddenly coming up with these issues at the last minute.

To me, the most telling aspect of this discussion is how virtually every purported defect is first met with the excuse that there's a "very simple" fix that can be applied to the system. Of course, on further study, such fixes tend to (a) not actually fix the problem; and/or (b) introduce equally bad defects. When this is pointed out, the typical response seems to be to crawfish and allege that there was never a problem in the first place.

I've also noted that applying different fixes to each specific application of the T/E system makes a mockery of the idea that it's a universal mechanic. Now me, personally, I have never found universal mechanics to be very useful. But if the system is sold as "universal", yet requires a host of different fixes, isn't that a bit dishonest?

As for the comment that criticisms need to be constructive, well, unlike my critics, who whine "you just don't like the system" (to which one might reply "no sh*t, Sherlock"), I have supported my criticisms with statistical analyses. I've even done statistical analyses of the ill-considered off the cuff proposed fixes.

This, I think, demonstrates reasonable good faith. The reason that I do not usually suggest fixes myself is that I've already gone through possible fixes and can't find anything that does the job. The T/E mechanic seems remarkably resistant to effective modification. And as an aside, it is not really my job to design the game; though my willingness to statistically analyse various proposed fixes are hopefully of some benefit.

I suppose I could design an effective replacement combat system, but no one's hired me to do so. That isn't really necessary, as there is a perfectly serviceable combat system in T4. With minimal adjustments, it could serve very well for the MGT combat system. Or, the systems in Azhanti High Lightning -- no more fussy and far better designed than the current system IMHO.

 

[TrippyHippy]

Ah, but I have playtested it.

Perhaps you can answer the question that started this thread (since none has been forthcoming so far):

How do modifiers (skill, DEX, etc.) affect autofire rolls?

Surely, this question has come up in playtesting?

I'm sorry, I just seem to have picked up on one of your posts, either here or on another forum, that you've just skimread the playtest files. I apologise if this is not the case.

How do modifiers affect autofire rolls in my game? - the same as they do in the core system - they affect the target number rolled against.

I think I'll need a definition of "more constructive".

Seems to me that if the core mechanic is broken, it would be a better idea to replace it now rather than in a supplement (or second edition) later on? (I have never bought into gamer conspiracy theories that companies intentionally publish flawed games so that they can sell Second Editions later).

And to be clear, I am demanding nothing.

Yes you are. You might want to to dress it up in other language, but at the core what you are saying is that you will not be happy unless the core mechanic is changed. In so many words, it's a demand. And saying a mechanic is 'broken' just because you don't like the statistical outcomes on offer as a simulation exercise, is not constructive, either.

Personally, I don't see the argument for a simple fix - merely accepting the system as it is, and enjoying play. This isn't a system that clunks along - it moves at pace, it captures a drama, it's tactical and crunchy. It's fun. It works for me, and my group.

I suppose I could design an effective replacement combat system, but no one's hired me to do so. That isn't really necessary, as there is a perfectly serviceable combat system in T4. With minimal adjustments, it could serve very well for the MGT combat system. Or, the systems in Azhanti High Lightning -- no more fussy and far better designed than the current system IMHO.

Well, thank God they've not hired you if your basic fix is to use the T4 system! :roll: