The Wolfhound and the Skiff's stats seem to be switched

[sevenperforce]

4 hours ago, Aeroboi said:

I'm also a little lost in what this engine is reminiscent of. The Apollo CSM engine is what it looks like. But it has a ISP of 412.....what?
What fuel mixture ends up with a ISP of 412 that is reminiscent of a CSM engine with a vaccuum optimized nozzle?
I'm not someone who argues that KSP need engine replicas out of real life, although MH is clearly based of such a idea. But if you go there by using models reminiscent of real life engines then why get fictional by adding non corresponding numbers to their statistics? As for the Wolfhound, why then the controversial ISP numbers thrown in?
For balancing purposes in the scale of distances that Kerbins solar system represents I'd expect lower ISP engines, not ones with greater numbers if anything, unless Squad intends to replicate Hydrolox resembling engines.

These by the way aren't Vacuum optimized engines in real life either. They have sufficient but poorer TWR at sea level and become more efficient as they climb.
In space Hydrogen leaks away so it can't really be a vacuum optimized engine if it is to resemble a Hydrogen engine.
Yet the Wolfhound hangs in between RP-1/LOX and Hydrolox fuel mixture ISP ranges. So what is it that it burns? Kerboloxillium? 

Squad accidentally swapped the stats for the Skiff and the Wolfhound. The Skiff is the J-2 analogue, which is supposed to be heavy, with a low TWR but a ridiculously good isp. The Wolfhound is the Apollo CSM SPS analogue, which is supposed to be lighter, with a poor vacuum ISP but a ridiculously good TWR. They simply swapped the stats. It's a coding error.

[Aegolius13]

Re: Rapiers,  there are still some good reasons to use them instead of (or in addition to) a Wolfhound.  One is that if they're already part of your plane as jets, then they're essentially free from a mass perspective as a rocket solution.   Whereas adding a Wolfhound to existing jets will add net mass.  Whether or not this improves your plane depends on a lot of factors,  including your total delta-v needs.   Also keep in mind that the Wolfhound produces a LOT of thrust,  and is likely overkill (and thus maybe not worth the mass increase) for small to medium SSTOs.  

The second reason is that a Wolfhound will add drag,  both for itself and for the attack it will have to sit behind.   The stack might also add useless mass,  like additional nose cones.  

In vanilla KSP, there's a similar debate about whether it's worthwhile to add nukes to an SSTO.  Same trade-off of weight and drag for better ISP.   I find the answer is no if you're just going to orbit, yes for very long interplanetary voyages,  and a vague middle ground in between.   This may be similar, but I don't have a clear idea yet on where the breakpoints lie. 

[RoboRay]

I'm tempted to just mod the Wolfhound to burn Monoprop and cut the Isp to 300'ish.

The game really needs a monoprop engine larger than the dinky little Puff, and the part that is supposed to represent the hypergolic Apollo SPS seems like the perfect candidate.

[Aeroboi]

7 hours ago, sevenperforce said:

Squad accidentally swapped the stats for the Skiff and the Wolfhound. The Skiff is the J-2 analogue, which is supposed to be heavy, with a low TWR but a ridiculously good isp. The Wolfhound is the Apollo CSM SPS analogue, which is supposed to be lighter, with a poor vacuum ISP but a ridiculously good TWR. They simply swapped the stats. It's a coding error.

Ah, makes sense   And I also see that other topic now discussing this, I missed out on it so far.
I thought the J-2 had ISP of near 450 though, so still not a replica but probably a healthy ISP nerf for a KSP edition if that is their intention with it. Errors do happen and I hope it gets fixed soon. As long as it's ultimate TWR is not to great to make concepts of spaceplanes to easy as I described then I'm fine with it.

 

5 hours ago, Aegolius13 said:

Re: Rapiers,  there are still some good reasons to use them instead of (or in addition to) a Wolfhound.  One is that if they're already part of your plane as jets, then they're essentially free from a mass perspective as a rocket solution.   Whereas adding a Wolfhound to existing jets will add net mass.  Whether or not this improves your plane depends on a lot of factors,  including your total delta-v needs.   Also keep in mind that the Wolfhound produces a LOT of thrust,  and is likely overkill (and thus maybe not worth the mass increase) for small to medium SSTOs.  

The second reason is that a Wolfhound will add drag,  both for itself and for the attack it will have to sit behind.   The stack might also add useless mass,  like additional nose cones.  

In vanilla KSP, there's a similar debate about whether it's worthwhile to add nukes to an SSTO.  Same trade-off of weight and drag for better ISP.   I find the answer is no if you're just going to orbit, yes for very long interplanetary voyages,  and a vague middle ground in between.   This may be similar, but I don't have a clear idea yet on where the breakpoints lie. 

Very true statements. It depends per design certainly and I would never make a SSTO using LV-N just to get into orbit either. 
While it has been said that numbers are switched between the Skiff the Wolfhound also produces more thrust per unit of mass compared to let us say, the Swivel engine, coupled with that higher ISP and I'm certain many spaceplane designs alike will benefit from using the Wolfhound, very effectively to be exact.

4 hours ago, RoboRay said:

I'm tempted to just mod the Wolfhound to burn Monoprop and cut the Isp to 300'ish.

The game really needs a monoprop engine larger than the dinky little Puff, and the part that is supposed to represent the hypergolic Apollo SPS seems like the perfect candidate.

Not against the idea for a heavier more powerful monoprop engine. Especially if it can be used as ports to work under RCS to dock certain monstrosities together. 

Edited March 23, 2018 by Aeroboi

[Aeroboi]

Made a topic of my own that aimlessly orientated to conclusions in similar respect considering the Wolfhound. Having read this now it all makes sense.

Not to ashamed for having it posted as it's clearly a very over the top error in coding and it's only healthy to clarify it outside of a single thread.

My findings is that it would make early spaceplanes way to easy to build since you have a engine of ~400 ISP in the Jet flame out territory already acquired in the 90 science node. If the stats were to be remedied according to real world application things will really make sense.

[_stilgar_]

10 hours ago, sevenperforce said:

The Skiff is the J-2 analogue, which is supposed to be heavy, with a low TWR but a ridiculously good isp.

Heavy engine with low TWR and good ISP? Sounds like NERV. 

 

10 hours ago, sevenperforce said:

Squad accidentally swapped the stats for the Skiff and the Wolfhound.

Did anyone from Squad actually confirmed that they were swapped? I'm not so sure it's an error. 

 

10 hours ago, sevenperforce said:

The Wolfhound is the Apollo CSM SPS analogue, which is supposed to be lighter, with a poor vacuum ISP but a ridiculously good TWR.

Which makes no sense in KSP (which is a little bit different than real life). Who would use an engine with poor vacuum ISP for a vacuum stage? TWR is irrelevant in orbit, but ISP is quite important. That's why NERV is so popular. If Wolfhound is light, high TWR, low ISP, no one would use it. For an orbital stage, you either use Terrier (for light ships) or Nerv (for heavier ships). 

If you swap Wolfhound and Skiff stats, people would use Skiff for orbital stages and Wolfhound for second stage... which is exactly as we have now, but with model replaced... but it would make no sense from "building replicas" point of view. 

[sevenperforce]

11 hours ago, Aegolius13 said:

Re: Rapiers,  there are still some good reasons to use them instead of (or in addition to) a Wolfhound.  One is that if they're already part of your plane as jets, then they're essentially free from a mass perspective as a rocket solution.   Whereas adding a Wolfhound to existing jets will add net mass.  Whether or not this improves your plane depends on a lot of factors,  including your total delta-v needs.   Also keep in mind that the Wolfhound produces a LOT of thrust,  and is likely overkill (and thus maybe not worth the mass increase) for small to medium SSTOs.  

The second reason is that a Wolfhound will add drag,  both for itself and for the attack it will have to sit behind.   The stack might also add useless mass,  like additional nose cones.  

In vanilla KSP, there's a similar debate about whether it's worthwhile to add nukes to an SSTO.  Same trade-off of weight and drag for better ISP.   I find the answer is no if you're just going to orbit, yes for very long interplanetary voyages,  and a vague middle ground in between.   This may be similar, but I don't have a clear idea yet on where the breakpoints lie. 

Rapiers are good enough on their own if you are only going to LKO. If you are going very, very far with few or no refueling stops, pick a nuke. The middle ground where a Wolfhound might be good is an SSTO with ISRU capability. That high Isp will get you out to Minmus to refuel but you still have better takeoff thrust on places like Duna and Tylo than you'd get with a nuke.

10 hours ago, RoboRay said:

I'm tempted to just mod the Wolfhound to burn Monoprop and cut the Isp to 300'ish.

The game really needs a monoprop engine larger than the dinky little Puff, and the part that is supposed to represent the hypergolic Apollo SPS seems like the perfect candidate.

The SPS was a biprop hypergolic, not a monoprop. 

Incidentally, the Puff was the highest-TWR vacuum engine smaller than the Rhino, until the Skiff came along.

5 hours ago, Aeroboi said:

I thought the J-2 had ISP of near 450 though, so still not a replica but probably a healthy ISP nerf for a KSP edition if that is their intention with it. Errors do happen and I hope it gets fixed soon. As long as it's ultimate TWR is not to great to make concepts of spaceplanes to easy as I described then I'm fine with it.

The J-2 was a gas-generator-cycle hydrolox engine, so it had a median isp of 421s, which is quite close to the 412s which I insist should have been on the Skiff rather than the Wolfhound.

5 hours ago, Aeroboi said:

Not against the idea for a heavier more powerful monoprop engine. Especially if it can be used as ports to work under RCS to dock certain monstrosities together. 

That should be the case for every engine.

2 hours ago, _stilgar_ said:

Heavy engine with low TWR and good ISP? Sounds like NERV. 

Well, heavy, low TWR, and good Isp as far as bipropellant engines go.

2 hours ago, _stilgar_ said:

Did anyone from Squad actually confirmed that they were swapped? I'm not so sure it's an error. 

Squad hasn't confirmed it, but there are code fragments which suggest a mistake and prerelease descriptions of the Wolfhound talk about its high TWR, I believe.

2 hours ago, _stilgar_ said:

Which makes no sense in KSP (which is a little bit different than real life). Who would use an engine with poor vacuum ISP for a vacuum stage? TWR is irrelevant in orbit, but ISP is quite important. That's why NERV is so popular. If Wolfhound is light, high TWR, low ISP, no one would use it. For an orbital stage, you either use Terrier (for light ships) or Nerv (for heavier ships). 

If you swap Wolfhound and Skiff stats, people would use Skiff for orbital stages and Wolfhound for second stage... which is exactly as we have now, but with model replaced... but it would make no sense from "building replicas" point of view. 

The Apollo CSM SPS was originally intended to be used as the LM ascent engine, so its vacuum TWR needed to be high. I can think of plenty of situations where I needed high vacuum thrust and didn't care as much about ISP.

The J-2 was used as a second and third stage in Apollo. Stage 3 of the Saturn V, powered by the J-2, provided the TLI burn; those BLEO burns are where you really want the highest isp. If you're actually going to go somewhere and use an engine to stop, high TWR and low dry mass is the way to go.

[klesh]

4 minutes ago, sevenperforce said:

 Squad hasn't confirmed it, but there are code fragments which suggest a mistake and prerelease descriptions of the Wolfhound talk about its high TWR, I believe.

 

The only mention of it in the KSP Weekly that I can find says that they are intending it to be “in league with the Poodle”.

 

The current stats are way closer to the Poodle than the Skiff is.   I dont think they are switched at all.  

[RoboRay]

11 minutes ago, sevenperforce said:

The SPS was a biprop hypergolic, not a monoprop. 

I'm aware... but when we only have two flavors of fuel, LF/O and Monoprop, and LF/O is used by every other engine, using Monoprop seems to be a way to distinguish this engine from the rest. It would also offer a Poodle-alternative that doesn't simply remove any reason for using the Poodle.

Edited March 23, 2018 by RoboRay

[Val]

Merged similar threads.

[Aeroboi]

1 hour ago, sevenperforce said:

Rapiers are good enough on their own if you are only going to LKO. If you are going very, very far with few or no refueling stops, pick a nuke. The middle ground where a Wolfhound might be good is an SSTO with ISRU capability. That high Isp will get you out to Minmus to refuel but you still have better takeoff thrust on places like Duna and Tylo than you'd get with a nuke.

Oh absolutely.

I'm talking about spaceplanes to LKO only. I see little reason why to send winged aircraft beyond LKO besides a few other destinations that is.

What I was getting at is that I expect many rapier designs will have their rocket mode ignored since a Wolfhound is clearly better especially considering it's low relative weight. It's still 800-900 M/s after Rapier flameout to LKO. My argument was especially targeted for being able to build much more efficient SSTO's with less capable jet engines from earlier tech nodes.

However, this calamity with the Wolfhound engine is not something I'd expect we have to discuss much longer as it clearly seems the Wolfhound and Skiff statistics issue is going to be fixed soon and we won't have to argue about this any longer  I'm very happy to have learned this is just a silly coding mistake.
Let's hope the first update fix comes along soon.

1 hour ago, sevenperforce said:

That should be the case for every engine.

I like this one very much. I think it must happen. But I don't expect it's gonna happen 

1 hour ago, RoboRay said:

I'm aware... but when we only have two flavors of fuel, LF/O and Monoprop, and LF/O is used by every other engine, using Monoprop seems to be a way to distinguish this engine from the rest. It would also offer a Poodle-alternative that doesn't simply remove any reason for using the Poodle.

You are forgetting Xenon which makes it 3 flavors. Not rendered as a fuel in game even fissile material is a fuel. So somewhere in the LV-N and RTG is something that is considered fuel also, just not visible or numerically used in the UI of the game.

To my knowledge all monopropellant engines have terrible ISP, no mind ranges that go as higher as our mystical Wolfhound engine or anything that burns LF/O like the poodle.

[sevenperforce]

14 minutes ago, Aeroboi said:

You are forgetting Xenon which makes it 3 flavors. Not rendered as a fuel in game even fissile material is a fuel. So somewhere in the LV-N and RTG is something that is considered fuel also, just not visible or numerically used in the UI of the game.

To my knowledge all monopropellant engines have terrible ISP, no mind ranges that go as higher as our mystical Wolfhound engine or anything that burns LF/O like the poodle.

If I had my druthers, I'd say to make high-Isp LFO engines run on an alternate fuel mixture. All engines currently run on the exact same mixture of fuel and oxidizer, but high-Isp engines like the RL-10 or J-2 use hydrogen, which is a richer fuel-oxidizer mixture. So the Wolfhound (well, really the Skiff) should do the same. It wouldn't make it impossible for newbies to use these engines, but they'd end up with leftover oxidizer if they weren't careful. It would give another way to use wing tanks in space without throwing an LV-N on.

Doubt they'd ever do this, of course, but it would get at the complexities of hydrogen storage and excessive dry mass without having to add a whole new fuel type.

I'd also love to see tripropellant engines that can burn LFO+monoprop at high thrust, lower Isp or switch to LFO only for lower thrust, higher Isp.

On the subject of fuel and other consumable management, I would REALLY like to see a way to dump fuel. The best way I can think of to do this would be to add an "active cooling" part, which would consume a small amount of propellant (selectable in much the same way as you can select LF, LO, or mono on the convert-o-trons) in exchange for cooling. It could be set to simply dump propellant outright or to dump it in proportion to heat exposure. Same shape as some of the new MH structural pieces; layer them on the surface of your ship and you can use your whole ship as a heat shield.

[RoboRay]

1 hour ago, Aeroboi said:

You are forgetting Xenon which makes it 3 flavors.

If we want to get pedantic, Xenon is just a reaction mass, not a fuel.  They would all be considered propellants, of course.

1 hour ago, Aeroboi said:

To my knowledge all monopropellant engines have terrible ISP, no mind ranges that go as higher as our mystical Wolfhound engine or anything that burns LF/O like the poodle.

I did not propose Wolfhound-style "mystical" Isp ranges for a Monoprop engine.  I suggested 300'ish.

[tater]

MH was the perfect opportunity to make new propellant types, since by adding failures, etc, they actually have a reason to use lower Isp, but higher reliability hypergolics.

A lost opportunity.

[Canopus]

15 hours ago, tater said:

MH was the perfect opportunity to make new propellant types, since by adding failures, etc, they actually have a reason to use lower Isp, but higher reliability hypergolics.

A lost opportunity.

I'd love if KSP would distinguish Cryogenic, Hypergolic and Monopropellant fuels each with their own advantages and disadvantages, but adding them so late and only in the DLC wouldn't make much sense.

[Tyko]

Lots of ways to balance these. From my PoV the "Making History" mod was about providing parts that allow me to reproduce historical craft. They refer to a lot of components as "analogs" and often provided the names of the real world components they were using for inspiration.

With this in mind, the test for these two engines is to build something that looks like an Apollo-Saturn V stack and see how it performs. When I did this and used the analog engines I found that Stage 2 and Stage 3 were under-powered while the command module would scoot along at close to 3Gs of acceleration.

This to me indicates a miss from the goal of enabling me to Make History 

Swapping the 2 engines (wolfhounds for S2 and S3, Skiff for command module) isn't perfect, but is a lot closer to the TWRs for the Saturn V.

[KerikBalm]

If the Skiff is a J-2 analogue, it is a vacuum engine analogue. The J-2 was used to go from Earth orbit to lunar Intercept. It was also used in the 2nd stage, which was operating in air so thing that it might as well be a vacuum engine. The SPS was clearly a vacuum engine.

Thus we are comparing 2 vacuum engines. 330 Isp for KSP is actually pretty good. Excluding the nukes and ions, the only engines that beat this were: Rhino, Poodle, Aerospike, Terrier. 330 Isp beats the skipper, the LV-T45, the Vector and mammoth, all 0.625m engines, etc. 330 Isp is on par with the poodle

412 Isp is a league of its own...

The skiff seems to be intended to be used with engine plates, and as its "Making History", we can summarize that the ideal situation here is to put 5 of them under a 5 meter tank in a saturn V replica 2nd stage. 5 of those things are pathetic as lifting engines. I need to pack 9 of them in there to have acceptable lifting power (granted I'm playing on 3x, but I've found all other stock parts suitable for 3x gameplay, so I'd think these would be balanced similarly)

The wolfhound is OP. Its got a better TWR than the poodle, its got a ridiculously high Isp, and it can be packed into a 1.875m form instead of the poodle's 2.5m form.

Swapping the Skiff and the Wolfhound stats is a start, but I think one still needs to get hit with a nerfbat.

I did some comparisons based on what the intended use seems to be:

Spoiler

	Vacuum Engine comparison
Diameter		TWR	Isp	Thrust/area
1.25	Terrier	12	345	38.4
1.875	Cheetah	12.5	345	35.5555555556
2.5	Poodle	14.2857142857	350	40
2.5	Wolfhound	15	412	60
3.75	Rhino	22.2222222222	340	142.2222222222
1.25	Aerospike	18	340	115.2

 

	Midrange Engine comparison
Diameter		TWR	Isp	Thrust/area
1.25	LV-T45	14.3333333333	320	137.6
1.25	LV-T30	19.2	310	153.6
1.25	Aerospike	18	340	115.2
1.875	bobcat	20	310	113.7777777778
2.5	Skipper	21.6666666667	320	104
2.5	Skiff	30	330	48
3.75	Rhino	22.2222222222	340	142.2222222222

 

	Atmo Booster Engines		SL values
Diameter		TWR	Isp	Thrust/area
1.25	LV-T30	16.4	265	131.3032258065
1.25	Aerospike	15.35	290	98.2588235294
1.875	bobcat	18.7	290	106.4372759857
2.5	Mainsail	22.9833333333	285	220.6451612903
2.5	Mastadon	26.06	280	208.5517241379
3.75	Mammoth	24.9733333333	295	266.3844797178

My conclusion: Wolfhound's TWR, Isp, and thrust for its cross sectional area are all too high (some worse than others)

The Cheetah's thrust to cross section area is a bit low, but its other values seem fine.

The Skiff has a TWR that is perhaps too high, but a terrible thrust/cross section ratio

The Skipper's Thrust/ cross section ratio is a bit low

The aerospike is terrible as a lifting engine because of its low thrust/cross section (basically determines how tall the stack it can lift is, assuming a stack of the same diameter as the engine), but its a great vacuum engine

The bobcat is a mediocre sustainer/midrange engine, and a terrible lifter engine.

The mastadon has a great TWR, but its total thrust is a bit low for the size it takes up, its also got pretty bad Isp values.

With that in mind, here are my proposals to rebalance the engines (nothing proposed for the bobcat yet:

Spoiler

@PART[LiquidEngineKE-1] // proportionately increase mass and thrust, slight buff to the Isp
    { 
        @mass = 6.5
        @MODULE[ModuleEngines]
        { 
            @maxThrust = 1460
            !atmosphereCurve
            atmosphereCurve
            {
                key = 0 295
                key = 1 285
                key = 20 0.001
            }
        }
    }
@PART[LiquidEngineLV-T91] // proportionately increase mass and thrust
    { 
        @mass = 1.15
        @MODULE[ModuleEngines]
        { 
            @maxThrust = 145
        }
    }
@PART[LiquidEngineRE-I2] // increased mass and thrust,  mass was increased proportionaly more, so TWR is lowered to be similar to the old a skippers, keeping in mind its Isp is better than a skipper's
    { 
        @mass = 3.5
        @MODULE[ModuleEngines]
        { 
            @maxThrust = 760
        }
    }
@PART[LiquidEngineRE-J10] // its now a bit heavier than the poodle, but the TWR is a little worse than the poodle, and the Isp is better but not ridiculously so
    { 
        @mass = 2
        @MODULE[ModuleEngines]
        { 
        @maxThrust = 260
            !atmosphereCurve
            atmosphereCurve
            {
                key = 0 360
                key = 1 70
                key = 6 0.001
            }
        }
    }
@PART[engineLargeSkipper] // increased mass and max thrust. Thrust was increased proportionately more so that it now has a higher TWR than before, and a higher TWR than the new skiff, keeping in mind that the skiff has better Isp
    { 
        @mass = 3.75
        @MODULE[ModuleEngines]
        { 
            @maxThrust = 865
        }
    }

 

[Not Sure]

Can we get a simple MM patch? I might even make it, and I’ve only touched MM once. I’m really for the “they are swapped” position but knowing squad it probably won’t change for a long time. 

The post above looks appealing, but I like where the current engines are at. 

Edited April 9, 2018 by Not Sure

[klesh]

There is nothing in the bugtracker about this supposed stats switch.  If you really think its a problem, one of you should post it there. 

[KerikBalm]

In their current state, I don't think the stats are switched.

I just think they are very poorly balanced.

That said, I still think switching their stats could lead to a better situation, but any engine with the wolfhound's stats is OP'd for the cartoonish size of the Kerbal's home system

[Bill Phil]

18 hours ago, KerikBalm said:

In their current state, I don't think the stats are switched.

I just think they are very poorly balanced.

That said, I still think switching their stats could lead to a better situation, but any engine with the wolfhound's stats is OP'd for the cartoonish size of the Kerbal's home system

 I've found that stages using the Wolfhound can have less delta-V than the same design with a Poodle. It's situational, of course, but the Wolfhound's stats seem to be more launch optimized than not, as a sustainer engine. 

[KerikBalm]

yes, they can have less delta-V than the woflhound, because the wolfhound has a higher mass. So its going to be very situational, and once any design/size gets over a certain size, the wolfhound will be the way to go.

Even if its got a higher mass, it still has a higher TWR and simultaneously a higher Isp than the poodle.

For those cases where the poodle did better... I'll bet a cheetah would have done even better.

[OHara]

I tried to follow the arguments that the stats are switched, but I do not see it.  Generally, KSP engines are lower-TWR versions of the real engines that they look like:

 TWR, sealevel; Isp, sealevel
  86,    60;   316s,   220s  real RS-56-OSA 1.0tonne
  14,    11;   320s,   250s  KSP Swivel 1.5tonne

  66,    54;   452s,   366s  real RS-25 3.5tonne
  25,    24;   315s,   295s  KSP Vector 4.0tonne

 2.6,   1.2;   850s,   380s  real NERVA-2 2.5tonne
 2.0,   0.5;   800s,   185s  KSP Nerv 3.0tonne

  33,   n/a;   312s,    n/a  real AJ10-137 (Apollo SPS) 0.3tonne 
  15,   2.5;   412s,    70s  KSP 1.4.2 Wolfhound 2.5tonne

  70,    30;   420s,   200s  real J-2  1.5tonne
  31,    25;   330s,   265s  KSP 1.4.2 Skiff 1.0tonne

The Wolfhound has strangely high Isp; the Skiff has strangely high TWR (for a KSP engine).   KerikBalm's suggestion above makes them fit the KSP pattern better, while keeping the Wolfhound strongly vacuum-optimized as its big bell suggests.

Edited April 12, 2018 by OHara
Skiff and Wolfhound stats are strange, but it is not simply a swap.

[Tyko]

On 4/10/2018 at 11:09 PM, OHara said:

I tried to follow the arguments that the stats are switched, but I do not see it.  Generally, KSP engines are lower-TWR versions of the real engines that they look like:

 TWR, sealevel; Isp, sealevel
  86,    60;   316s,   220s  real RS-56-OSA 1.0tonne
  14,    11;   320s,   250s  KSP Swivel 1.5tonne

  66,    54;   452s,   366s  real RS-25 3.5tonne
  25,    24;   315s,   295s  KSP Vector 4.0tonne

 2.6,   1.2;   850s,   380s  real NERVA-2 2.5tonne
 2.0,   0.5;   800s,   185s  KSP Nerv 3.0tonne

  33,   n/a;   312s,    n/a  real AJ10-137 (Apollo SPS) 0.3tonne 
  15,   2.5;   412s,    70s  KSP 1.4.2 Wolfhound 2.5tonne

  70,    30;   420s,   200s  real J-2  1.5tonne
  31,    25;   330s,   265s  KSP 1.4.2 Skiff 1.0tonne

The Wolfhound has strangely high Isp; the Skiff has strangely high TWR (for a KSP engine).   KerikBalm's suggestion above makes them fit the KSP pattern better, while keeping the Wolfhound strongly vacuum-optimized as its big bell suggests.

I look at the ISP numbers and believe they're reversed. It's right there on your own chart...The J-2 has 420, the Wolfhound has 412. For 4th stage, AJ-10 has 312 and Skiff has 330...reversing them not only brings the ISPs more in line, but most importantly the relative ISP between the IRL very efficient sustainer engine and the less efficient Command Module engine

[OHara]

On 4/12/2018 at 5:24 PM, Tyko said:

I look at the [vacuum] ISP numbers and believe they're reversed.

Yes, I saw this and started writing the table intending to make a simple bug-report saying "please un-reverse them". 
Then I realized why KerikBalm, and others, say the swapped stats are not much better, and that comparison to real engines is complicated..

The AJ10-137 certainly looks like a vacuum-optimized engine, as if trying to optimize Isp in vacuum.  This real engine does the best it can with the long-storable, self-igniting,  low-energy-density, hypergolic fuel it had to use.  The J-2 gets its better stats mostly by burning cryogenically-stored liquid hydrogen.  KSP treats all fuels the same,  and give engines performance closer to that with hypergolic fuels. 

The engine masses do seem to be reversed, roughly.  Also J-2 analog needs more thrust, as you mentioned earlier regarding how they are used in an Apollo Saturn-V replica.

We still need to put something on the bug-tracker.  Maybe a link to this thread is the best.

https://bugs.kerbalspaceprogram.com/issues/19076 "some new-engines specs are out-liers in KSP"

Edited May 26, 2018 by OHara