That's not how probability works.

It's 6% every quarter. Across a year, that's a 21% chance of a given ship having an event, or >13% chance of a bad event. Across a 5 ship fleet, there's a 50% chance of at least one ship getting taken out of commission. Across a 10 ship fleet, it goes up to 75%.

Work:1-(1-(1-.94^4)*.6)^5

I could preform a more detailed analysis, but I'm not willing to put in the effort. I'm tired of combinatorics right now: I'm taking a combinatorics course, and it's not your high school "What's the chance of a X" course.
I'll just use some high school stuff.

What are the odds of the ship NOT having a major failure in a given quarter year? 94%. What are the odds of NONE of five ships having a failure in a given quarter, given that? (.94^5), or 73.3%. There is a roughly 3/4 chance of ALL the ships surviving a single quarter year with no breakdowns.

The flip side of that is, there is a roughly 1/4 chance of ONE of the ships breaking down, and a (somewhat smaller) chance of two or more breakdowns in the same quarter year. We'll ignore that grim possibility.

Now roll that four-sided die, every quarter year. How long before you roll a one?

Common sense suggests we will expect to see at least one ship experiencing a failure about every three or four quarter years. Every year we'd be having a breakdown.

Okay, if breakdown chance is truly only rolled once a year, things are less bad- but still problematic.

Remember what happened when the Soyuz-class started experiencing mechanical failures once every few quarters? NPCs started calling for the class to be scrapped.

Now imagine what would happen if a brand-spanking-new ship class we'd just spent political will getting built were doing the same thing?

Let's not build anything with a reliability below 95%. Preferably not below 98%.

I mean heck, automobiles get recalled for mechanical defects that have only a 0.001% chance of actually happening... because if they happen, someone dies, and a one in a hundred thousand chance of a lethal accident multiplied by all the cars of that make and model on the road can add up to a lot of dead people.

What are you talking about, we could design an escort with the lower end of the mentioned stat line right now. It would have reliability issues and I'd want to get general tech up a bit first to get reliability over 90%, but if it came push to shove I'd pick 83% reliability over trying to keep the Kepler around until 2370.
Nix, we're going to be designing at least two classes of cruisers, two classes of explorers, and probably two classes of explorers, between now and 2370. Who says we have to design only one class of science vessels?

The basic point of the Keplers would be to get those poor plucky little Oberths off the street and have a somewhat more survivable, flexible ship to do things like gather signals intelligence deep in Cardassian space. That's a science vessel we can feel good about constructing, one that won't be a complete waste of a berth in the 2310s or 2320s.

Will we be able to do better by, say, 2340 or '50? I certainly hope so. And between now and then we should be able to find time to design a better one.

But...

Basically, we have four roles in our fleet: explorer, cruiser, science vessel, and escort. We've already decided our next new design (the Renaissance) is going to be a cruiser. After that, the only question is, do we design an explorer, a science vessel, or an escort? Whichever answer we pick, we're getting a ship designed with 2310s technology.

So if you don't favor bothering to design any new science vessels in the next ten years, on the grounds that anything we do now will be too weak compared to the ideal image of the Intrepids we can build sixty years from now... What do you think we should do next?

No, we don't want ANYTHING sub 95% reliability. I don't care a bout "funny events" because that happens only 40% of the time, while 60% of the time we are missing a ship for at least 3 months(if not more) are out some resources, and maybe even crew depending on sverity. And even with warp core 100% reliability, if we don't also have high hull reliability we can still accidentally the ship.

Basically, I'm getting really pissed at everyone ok with low reliability due to "Funny events! That's so Trek!". No, it is not. Funny events would be the minority, statistically speaking. You have your math ass backwards
Honestly, even the Galaxies, who are kind of the poster children for low reliability, only failed rarely, like less than once per twenty episodes. You only get high failure rates if you count active sabotage. Stuff like alien commandos busting out of the brig, Iconian polymorphic computer viruses attacking the operating system, or Wesley's science fair project eating key control systems.

95% reliability is creeping up into "dumb things will happen so often that if they based a Star Trek series on this class, it would become a meme that the ship is a piece of junk about to fall apart."

Come to think of it, those shiny shiny high stats for the Galaxy and Intrepid should be interpreted with that in mind... In some important ways they were low-reliability ships.

I juggled the comp teams around since there is one tech left so it will get the inspiration bonus. This way we get to apply the bigger bonus from Daystrom to shields which should complete 1 or 2 techs and have the rest within one turn of completion. Otherwise mainly following up on last years research. Kept the Tiger team on Cardassian research to guarantee completion within two years (next year we can switch to Games & Theory Division to finish the research and move tiger to Way of the Elephant.

Edit: Was reminded Renissance needs to be researched once we chose it.
You have my vote. Remind me of that if I vote for anyone else.
 
Last edited:
@Simon_Jester

Break down is worked out once a year not per quarter, on top of that 40% of all break downs are essentially irrelevant 50% are minor to significant and the last 10% is severe. So your numbers are horribly out.

More than that as i keep having to point out failure chance aggregated across the whole fleet tells us nothing about operational sustainability. What's important is how it would impact science missions in a sector garrison where we have two or three of them.
 
Last edited:
You have my vote. Remind me of that if I vote for anyone else.
List is more out there for ideas as we may pick up a tech team this coming snakepit which could change things and we also need to select which offensive doctrine we want to pursue, plus there is a good chance I will be at work or class when the next research turn is posted. Just in dead periods like this I want to get discussion going on what we want to do for the next snakepit and research to make it easier to decide at that time.
 
Military aircraft need something like 20 man-hours of maintenance for every hour of flight. Commercial aircraft aren't much better.

Commercial aircraft are much much better.

That said, the other big thing with the press-your luck design is that it takes much much more research time. That means having to pass on other research, and having more delays till we can build one.
 
If I were part of the crew, I wouldn't want to serve on a ship that has a good chance of breaking down, which presents risks to the crew.

This use of statistics to justify a lower reliability reminds me of the Ford Pinto.

Like Simon_Jester said, each time a ship breaks down is significant, similar to the Soyuz case and real life automotive cases. People will not want a ship that is perceived to be unsafe.

Neglecting the safety of one's product, in this case ships, I find unethical.
 
Last edited:
What are the odds of the ship NOT having a major failure in a given quarter? 94%
False argument, the chance of a not-major breakdown is 99.4%. The other 5.6% are, at worse, not major breakdowns.

Neglecting the safety of one's product, in this case ships, I find unethical.
No one is arguing for lower Hull or Warp. The other systems, however? Yeah, I'd accept a lower reliability for a higher stat. Across the board, the stat will probably come up more often than the ships break down.
 
That's the normal events.

We already destroyed a moon, and got bonus rp.
Note how you had to go six years back, to when the system was much less mature than now?
These breakdowns have a higher chance of bad things happening. If good events were in the majority, I'd agree.
We don't know that, it's quite possible for good events ending up being the majority, depending on how the system works exactly and what the stats and failure distributions are. Even if that was the case, you'd still have to look at expected loss due to failure vs expected gain due to higher stats. Say 1.5% chance to miss the event due to malfunction vs whatever the increased share of succeeded events is. Given that it's a roll of 2d6 then as long as it's not an auto-success with the lower stats increasing a stat by 1 point means an additional success in somewhere between 2.78% and 16.67% of cases. The relative increase would be even higher, particularly if multiple rolls are involved.
FYI, with all 5s, the Kepler can go up to 11 at the cost of 80% reliability.
Leaving aside in how far that can still reasonably be called "the" Kepler you seem to be demonstrating that great stats are possible in a science ship, if you are willing to compromise reliability. And I perfectly agree!
 
Last edited:
We just need to design a 50% reliability "Explorer" called Goody Hut. We'll put our best team at such designs on it. I think they were called "Vault Tec?"

The design principle would be sound. Every malfunction is a chance for research! Every part that flips it shit is a good chance for part innovation!


...I need to write a USS Vaulttec omake.
 
If I were part of the crew, I wouldn't want to serve on a ship that has a good chance of breaking down, which presents risks to the crew.

This use of statistics to justify a lower reliability reminds me of the Ford Pinto.

Like Simon_Jester said, each time a ship breaks down is significant, similar to the Soyuz case and real life automotive cases. People will not want a ship that is perceived to be unsafe.

Neglecting the safety of one's product, in this case ships, I find unethical.

This would only be relevant for hull and warp core which every one agrees should be kept around 100% or as close as possible, no one is throwing crew safety under the bus.
 
No one is arguing for lower Hull or Warp. The other systems, however? Yeah, I'd accept a lower reliability for a higher stat. Across the board, the stat will probably come up more often than the ships break down.
Hull, shields, combat, presence, it doesn't matter what system. All must be reliable.

No one will trust a faulty ship.

The possible impact of ship failures on the society seems to be forgotten, buried in the statistics.

Engineers swear an oath to safeguard public safety. In this case, that would include the safety of the crew on the ships they produce. They are held responsible and accountable if anything happens to the ships they make due to a faulty or unsafe design.

I cannot in good conscience support a design that is unreliable, no matter what system.
 
I agree with Akuz, if we want to continue this discussion we should probably move to the ship design thread, or drop it entirely since it seems we're not going to change each others' minds
 
I agree with Akuz, if we want to continue this discussion we should probably move to the ship design thread, or drop it entirely since it seems we're not going to change each others' minds
I disagree. A lot of people who don't neccesarily want to get into the weeds of the spreadsheet still have opinions about the philosophy of ship design. It's not about trying to optimize C v SR, but rather what we should optimize for.
 
For the shipyards operations turn, if we use 3 Utopia Planitia berths and do a Centaur refit it should be with one of the small berths because the Constituition-B could be scheduled in that berth the year after (the berth freeing up just in time is much of the point of doing a refit now), and locking up a big berth like that for 3 years would be bad. The Oberth (if we build one) could use a big berth and clear it up after 2 years, when the first wave of Constituition-B's is in construction.
 
I agree with Akuz, if we want to continue this discussion we should probably move to the ship design thread, or drop it entirely since it seems we're not going to change each others' minds
There is the issue of transparency though. Yes, the argument clutters the thread, but it does have some serious implications over here. Take it over there, and you may not get the perspective of active players who can't, for various reasons, participate in ship construction. You might turn up your nose at people who dictating your job to you, but HEY!

That's government oversight! :p

But seriously, your kind of on a self-reinforcing loop of "We can only take the most unfailing of ships!" Seriously, your sacrificing stat points for better reliability. This is probably greivous offender:

His is 5sr cheaper but .28% less reliable.
(sorry aledeth) For a quarter of a percent of reliability, this user is advocating making a design 5 SR more expensive, per ship. And the primary thread would never know, because most of us will probably never hop over to check.
 
For the shipyards operations turn, if we use 3 Utopia Planitia berths and do a Centaur refit it should be with one of the small berths because the Constituition-B could be scheduled in that berth the year after (the berth freeing up just in time is much of the point of doing a refit now), and locking up a big berth like that for 3 years would be bad. The Oberth (if we build one) could use a big berth and clear it up after 2 years, when the first wave of Constituition-B's is in construction.

Well you'll probably be the one to post the plan because the call for plans usually happens when I'm asleep, so do that and I'll vote for it.

And I do think we should build another Oberth. The explanation of how ships gets chosen for missions makes me feel a lot better about assigning them to sectors, as they are far more likely to get sent on missions where the Reaction roll required science. A cheap, high science stat can help pick up events that would otherwise have been missed completely.
 
On the other side of this, we can't forget that unreliability comes from all the fudge factors used in a given design. So if you have ~83% reliability, you've probably used level 5 fudge factors in just about every category, depending on design. That's 5+ years of research just for the design, + you still need to build the prototype
 
Every percentage point matters.
Not nearly as much as you believe, when we can get better stats for a slightly less reliable ship design.

And like people have all ready said, myself included in the design thread, there's a whole second roll if the Hilarious Breakdown occurs to see what kind. The chances of something bad happening are pretty low.
 
Last edited:
On the other side of this, we can't forget that unreliability comes from all the fudge factors used in a given design. So if you have ~83% reliability, you've probably used level 5 fudge factors in just about every category, depending on design. That's 5+ years of research just for the design, + you still need to build the prototype


Sounds like the Galaxy-class. :V
 
On the other side of this, we can't forget that unreliability comes from all the fudge factors used in a given design. So if you have ~83% reliability, you've probably used level 5 fudge factors in just about every category, depending on design. That's 5+ years of research just for the design, + you still need to build the prototype

I'm agreed on your point here but no one has earnestlt suggested we make a ship like that I personally dont want to go lower than 95% with out a damn good reason to do so.
 
This is ship design stuff? Shouldn't it be in the ship design thread? It's eating the main one alive.
The problem is, this is turning into a pretty basic debate about design philosophy. You may not care about the spreadsheets in the ship design thread. Heck, I don't have that much interest in them, and I'm a very, very number-oriented person.

But we have a right to know if our spreadsheet people start focusing on ships that are going to break down often enough that it may result in politicians trying to get them recalled.


@Simon_Jester

Break down is worked out once a year not per quarter, on top of that 40% of all break downs are essentially irrelevant 50% are minor to significant and the last 10% is severe. So your numbers are horribly out.

More than that as i keep having to point out failure chance aggregated across the whole fleet tells us nothing about operational sustainability. What's important is how it would impact science missions in a sector garrison where we have two or three of them.
A chance aggregated across the fleet impacts politics, among other things. I'm talking in no small part about the political consequences of low reliability on ships.

Remember what happened when we started seeing regular breakdowns on the Soyuz-class.

Yes they are. Aircraft Carriers spend more than half of every year in dock, undergoing maintenance. Military aircraft need something like 20 man-hours of maintenance for every hour of flight. Commercial aircraft aren't much better.

Nothing in real life is designed for 99.5% chance nothing happens in a year. Not even a Toyota Camry.
99.5% chance of no system going badly wrong enough to kill a redshirt. Because I'll note that all sorts of minor mishaps that don't wreck the ship can still be bad enough to kill Ensign Ricky.

The big problem here is that if we commission a brand new class where accidents occur on a semi-regular basis, and where severe ship-incapacitating accidents happen often, we are vulnerable to something like, oh, this mock exposé.

Aircraft carriers spend that much time in maintenance precisely because that way there isn't a risk of something going massively wrong and incapacitating the ship while it's on deployment. And they're often designed with robust, redundant systems so that a single failure is not incapacitating- in other words, trading off capability for reliability.

False argument, the chance of a not-major breakdown is 99.4%. The other 5.6% are, at worse, not major breakdowns.

No one is arguing for lower Hull or Warp. The other systems, however? Yeah, I'd accept a lower reliability for a higher stat. Across the board, the stat will probably come up more often than the ships break down.
I suspect that if we start blatantly gaming the reliability, we may find it becoming more of an issue. Reliability is supposed to be there to discourage us from pushing too far out onto the bleeding edge of what the technology is capable of.

Remember what happened when someone recommended seven diplomatic pushes in one turn?
 
Last edited:
But we have a right to know if our spreadsheet people start focusing on ships that are going to break down often enough that it may result in politicians trying to get them recalled.
Or, on the reverse, them refusing to make them as cutting edge as possible (therefore expanding there effective life cycle, I believe) because they are unwilling to risk setting the toast on fire.

Opposite opinion, but same consensus: The Federation primary thread has a right to know what Starfleet the secondary thread is doing, and control the direction it's heading.
 
Well, the Soyuz call for scrapping was because the warp core failed the second time, what was the component that failed the first time?
 
Back
Top