Robots as Spaceships: Nitpicking the Gunbot

How does the Gunbot measure up to everything else?

In the last post we created a Gunbot Mark I, a 200 lb autonomous tank built with the spaceships system. We made four tweaks, to get there, extrapolating size and adding an onboard computer to be the AI. So now we have our beautiful little robot, and we're ready to test him against the other robots in gurps, as well as some humans, to see how his stats measure up.

Most of the comparisons will be about other combat robots weighing from 100 to 300 lbs. Robot is a diverse term, and machines engineered for different purposes will have different statistics. By focusing on combat robots, we will hopefully be comparing apples to apples, or at least apples to pears. Size also makes these comparisons more valid: size has a large effect on most of a robots statistics.

I'm going to talk about the flaws in spaceships a lot. Don't interpret this as me disliking spaceships. I love the series. That's why I'm using it as the base for these robots. Spaceships is the starting point for this system, if you don't own the books, you won't get much out of these posts. If I did not believe in its basic structure, I would not try to improve it.



But on to the comparisons!

Strength and HP

Our little Gunbot weighs about 200 lbs, and has a strength of 30. That's fairly impressive. It should probably be stronger than a human of the same weight, but ST 30 may be a little high. Lets look at the other robots for comparison.

The combat android in Ultratech has a range of weights centered on 200lbs, and it has a ST of 20. Reign of Steel - Will to Live (The fourth edition update) gives the myrmidon and tarantulas each a ST of 25. Finding iconic combat bots in Transhuman Space is harder, but the stats for SM+0 models usually hug ST 10, which feels very low. After the End 2 has the Guardian, which has ST 20, but is only 125 lbs, a far cry from our 200. It also has the TASE, which is the closest thing to our gunbot in all of Gurps: a SM+0 autonomous tracked gun. The TASE has ST 32, but if we look at its 500 lb weight, we quickly realize that while its SM+0 by volume, its SM+1 in spaceships. If we compare the TASE with a SM+1 hull, we find spaceships to be ahead once again, with ST 50.

Campaigns gives a method for calculating hit points on page 483. When we applying the formula, we get 23 HP. ST and HP have an interesting relationship, but they normally match each others, and that includes all of the robots we compared the Gunbot to above.

It appears ST30 is on the high end, and we can find just about every score between that and 10 in various supplements. There does seem to be a cluster from 20 and 25.  ST20 to ST25 also lines up with my gut feeling. So 30 is a touch high, and we'll want to adjust it, but its certainly not going to break anything. Not the way that some of these other numbers will.

Mobility

The gun bot has move 3/30.  A human can stay ahead of it for about 3 seconds and it can move at 60 mph. Note that 60 mph requires good terrain, and is at the high end of its  performance. This still feels rather fast.

The TASE has a similar mobility system, but has move 6 and a top speed of 12. that's double the acceleration and under half the top speed. The TL9 version of Ultra-tech's tech-bot just has plain old move 5. Reign of steel's Juggernaut has move 6/36. Their mechanic bot has 4/12. The boss-bot is 5/10. The Eater is 5/30. The inquisitor is 5/10. The Armored Personnel Carrier in Campaigns has move 1 and a top speed of 20. The light battle tank in ultra-tech has move 2/25. High tech's main battle tank has move 2/19, and the earlier tanks are slower. Transhuman space features a single tracked vehicle: a SM-4 "swarm-dozer" with move 5.

It appears the top speed is a little high. We didn't put a ton of effort into making this bugger go fast, and yet he can outpace most tracked vehicles in the all of the books. The spaceship books will let us go MUCH faster: we're only half-way to maximum. We may want to tone the top speed down a bit.

Basic move, on the other hand, is smack in the middle of two rather tight clusters. Most of the stats for historical and exant tracked vehicles give a move of 1 or 2. Sources focused on robots tend to give them human-scale movement, between 4 and 6. So having a basic move of 3 is within accepted range, but we should probably pick a paradigm we'd like to stick with. On the other hand, taking 3 seconds to catch up to a running human isn't a bad paradigm to work with, and if all robots come from the same system, the two "modes" from other sources won't be noticed.

Sensors

The Sensors given the robots by spaceships are a little comical. They are somewhat abstracted, but include infrared, visible spectrum, ultraviolet, and radio telescopes, as well as active scanners. That would give +3 to perception the rest of Gurps. Nifty Toys. But the sensors end up giving -1 to perception checks. I can only assume these sensors are tiny. We have a negative level of telescopic vision (not sure how that works), and an active sensor range of .05 light seconds, or 10,000 miles. That's enough to reach the horizon. This is also our wi-fi range. So maybe that's why the cameras are so small: we sacrificed them for awesome reception.

We have radar with horizon spanning range, we can get a signal from anywhere we have a line of sight to, and we have a suite of multi spectral cameras that still only give us a -1 to perception. All of this weighs less than 10 lbs, as it has to share the control room system with the parts that let us actually control the bot. If we devote another 10 lbs to sensors, we pick up a level of telescopic vision, our space-reaching comm and sensor ranges increase, and our fancy sensor suit actually gives us a bonus.

Most of our reference points have +0 perception, or at the very least, a low modifier. Combat androids, AtE Guardians, and reign of steel myrmidons all have default perception. Reign of steel Tarantulas and AtE TASEs both have perception 12. Every bonus counts in Gurps, but we do seem to be in the general ballpark.

The active scanning and communication ranges are completely out of control, reaching 10,000 miles with gear that weighs 10 lbs including visual sensors. At TL8, High Tech gives 8 lb radios a 35 mile range, and half pound radios a 5-mile range. Ultra-tech only slightly improves that with a 5 lb radio having a 100 mile range. The 10 lb radar in ultra-tech has a range of 30 miles in search mode. Lots of other systems are in the books, but they are all in the same ballpark.

To be fair, a 5 lb radio in space might actually have an effective range of 10,000  miles. I don't know enough to say for sure. At the very least, we'll want to craft a rule to limit the range when on the ground, and possibly rework the communication and active sensor gear altogether.

Cost

$3700 is pretty cheap. Especially when you consider that the combat android we've been using for some of our comparisons has a $200k price tag at TL10. Reign of Steel and After the End don't give prices, but THS: shelltech robots are usually more expensive than even the combat android.

I don't know that we want to embrace the money madness those sources suggest, but we have a free license to make these robots more expensive than spaceships says they should be.

Weapons

The Gunbot Mark I has a choice between two weapons. The first is 4cm explosive shell doing a massive 6dx10 cr (ex) damage. Our weapons and armor are restricted to TL8, so we'll compare it to other weapons with a 4cm diameter. (or 40mm, as we normally call shells that can fit in one's hand). This means a grenade launcher, and a few of the Light Anti-Tank Weapons. The contrast in damage is sharp. The grenade launchers do around 4d crushing explosive, or 1/15th of what the gunbot does. In fact, their HEDP rounds barely exceed the armor penetration of the proximity explosions for the spaceships weapon. The grenade launchers also have vastly inferior range. Their one redeeming quality is their rate of fire, which is about 7 times faster, but that's still less damage per second. GL's also tend to be slightly lighter than the gun on the bot.

If the bot wants to shoot 1 shot per round, it can do so with a 1 cm gun. This is roughly equivalent to a .50 caliber bullet. This .50 cal does either 15d(2) or it explodes for 15d. The exact type of the damage is painfully not explained here. It could be pi++, or just pi+. The weapon weighs 10 lbs,and has an accuracy of +8.

The closest match to these statistics in High Tech is the Barret "Light Fifty", or M82A1. The Barret lightly under-performs in most areas. Its got the same rate of fire. It deals 6dx2 damage, as opposed to the 15d(2). Its accuracy of 6 compares favorably to the 8 on the spaceships gun. But the weights of these two weapons are really not comparable. The Barret weighs 35 lbs, and still under-performs on every metric except ammunition weight. If you're using a 10 or 12 lb gun you're stuck with 7d+1 pi Winchesters.

Ammo weight is where the true insanity appears. We don't have an official ammunition weight for the 1cm gun, but  the 4 cm gun rounds weigh 2.5 lbs. Which is a lot. Especially when you consider that the weapon system weighs 10 lbs and it comes with 10 free shots, weighing 25 lbs in total. Scaling issues are part of this, but I suspect this mismatch persists even at classical spaceship sizes.

Spaceships combat has been accused of being too lethal, of its numbers being too high. This has been explained as a natural consequence of the high speeds that come with orbit, of scaling issues, and of high tech warfare. We've uncovered another issue: the damage on these weapons is just too high. These velocities are terrestrial, these scales human, this weapons tech level merely modern, and the problem persists.

Fortunately, we don't have to use the weapons from spaceships, because we have an entire book of weapons at the tech level and scale that we want. We will need to make a few adjustments to incorporate them into the spaceships system, but the basic statistics will work just fine.

Armor

 The armor on the gunbot feels a little low. We used 60 lbs of high tech armor on a SM+0 object to get DR 10. The number doesn't feel desperately wrong the same way that the weapons did, but it does feel low.

I've already done a post comparing high tech armor with spaceships armor (Actually, the post grew out of this initial comparison).  High tech armor is all over the place in terms of DR value, but it stays better than the robot's armor in almost every case. 

If we compare the armor on this gunbot with other robots, it still looks bad. The ultra-tech combat android has DR 30 at TL 9. If we step the TL on the gunbots armor up, we get DR15, and to equal the UT android's armor we must spend 60% of all weight on armor. Reign of steel has higher numbers, with the Myrmidon having 30 DR and the Tarantula sporting 60 DR. These are TL10 of course so we should consider the gunbot as if it had 20 DR, which is still lower than the other two. The Guardian from After The End has DR 10, which is the same as the gunbot, but the TASE vehicle in the same book has 60 DR. Neither have a TL assigned, but between 8 and 10 is probably a fair assumption.

The accusation that spaceships armor is too weak is found to be valid. We're removed scale and velocity, and it still measures up short, though not quite as bad as weapons. Which is a good thing, because our armor support from the books is weaker here. We're going to have to delve into armor and tweak the numbers. Fortunately, the armor system for spaceships is very tweak friendly. It has a scaling mechanism and its easy to see why and where the price and stat increases happen.

In Conclusion

We have several issues to address, but the biggest ones are weapons and armor. ST could use a little tweaking, and perception could use some additional thought. We also will want to adjust the top speed of the tracked drive train. This is good. We're not at a loss on how to tackle any of these issues, and we have a decent idea of what the correct numbers will look like. So we have some work, and we have some insight, and we have some progress. And we'll fix these problems in future posts. Next time, we'll talk about armor!