Mulling Over the Numbers: A Noob Review

You can share your strategies and experiences about how to battle and defense, how to collect resources…

Moderators: GAW Moderator, Forum Moderator

User avatar
boardwalk
Posts: 20
Joined: Sun Jul 13, 2014 12:20 pm

Mulling Over the Numbers: A Noob Review

Postby boardwalk » Wed Jul 16, 2014 3:08 pm

Since GaW is so much a numbers game, and I am so much a numbers guy, I figured I'd compile some rough analysis of things I understand so far. Hopefully it is helpful to others as it is to me. Do keep in mind that I am ignoring many conditionals and just making simple high level comparisons. Regardless, please correct me as you see fit; I have a lot of initial thoughts on strategy that I'm not certain apply (or matter). I've only just started playing and have no delusion that I've discovered any grand knowledge not already known, nor have any superior advice to give. So, consider my "statements" here more as questions. I hope only to help and be helped with a game I am quickly coming to love.


COMBAT SHIPS

As the limits to fleet size are so large, it's easy to look at base stats and make simple comparisons considering only attack and defense, and get a good idea of what to build or send at any given time. In general, it makes sense to have a little (or a lot as you progress) of everything in every fleet, as every ship will indeed contribute in some way. The next level of thought would be to prepare an attacking fleet according to spy results, sending more of what makes a better matchup, and less of what doesn't. Basic strategy would be: send plenty of big stuff for damage, a swarm of little stuff to take hits, plus whatever is strongest against the bulk of their defense. Easy enough.

However, I am interested in efficiency, and so I ran some simple cost efficiency calculations using base stats, without factoring in things like rapidfire, the shielding mechanic, and that combat is six rounds (though keeping them in mind for discussion). Also, potential variation in upgrade levels adds too much "if...then" that I'm just going to omit for now. So, in order of cost effectiveness of base damage output, from best to worst:

Death Star
Battleship
Destroyer
Heavy Fighter
Cruiser
Light Fighter
Bomber
Dreadnought
Missle Chaser

What does this mean? If you were to build a fleet of only missle chasers with the same base damage output as a death star, it would cost 2.5 times as much resources. Though a very rough comparison, this provides a baseline to start thinking about how to cost effectively arrange fleets. And by cost effective, I mean making sure the expensive stuff survives without needlessly wasting cheaper stuff. This list order would certainly be different per attack by adding RF and the target's loadout, which is where much of the attacking strategy comes in.

I'd like to go over each ship and tell my thoughts on them, in hopes of gaining some confirmation, correction, or expansion from anyone interested. My thoughts on fleet build ratios will come after.

[Light Fighter] Beyond the early game, the cheapest combat ship is aptly the best cannon fodder on both attack and defense. (Spy probes are an option, too, but I'll get to that.) Since it's also cheap on fuel, it's easy to bring at least a small swarm with every fleet to steal hits, even if just in the first round. It is very likely that light fighters will deal almost no damage, but if they keep bigger ships alive for an extra round, they can make a big difference. If anything, they may get lucky and take down some of their opposing counterparts, allowing some of your other ships to land better targets (I would think so, anyway. I've read about the destroyed ship "glitch," but don't know it's current status.) Against cruisers' rapidfire, there is the option of just bringing even more to counteract it, given the low cost. Alternatively, there is the more efficient...

[Heavy Fighter] Very good for defense in the early game, with its RF versus small cargo ships. On attacks, the damage is enough to support heavier fire and help drain shields. At higher level play, I don't think I would bring both light and heavy fighters in significant numbers each unless my target somehow specifically merited it, since use of one or the other is a question of "do I want numbers or support?" Once I'm no longer using heavy fighters as a primary damage output, I would likely prefer 2.5 times the number of ships in the swarm (heavy cost 2.5 times light).

[Cruiser] The quintessential early game powerhouse. RF against light fighters and rockets make the difference in low level attacks feel like night and day. Though not as efficient as the heavy fighter, it is the fastest combat ship, and the second fastest ship in the game. Add a respectable jump in defense over the fighters, and you've got a powerful early game workhorse to bring you through to the next stages of upgrades. Later on, it remains an excellent addition to any fleet for cleaning up fodder swarms in the early rounds of combat. A nice set of cruisers and heavy fighters on defense is a good way to keep day 4 noobs off your back.

[Battleship] The second most cost efficient for damage, yet probably the least effective in terms of utility. It's only RF is against rockets, and only 50%. This likely places it lower on the list in most circumstances, as bombers and dreadnaughts, though significantly more expensive, will have much more damage output with their RF. I'll certainly use battleships before having access to bombers and dreads, but probably not so much afterwards, even less so once I have destroyers.

[Bomber] For only 1.5 times the cost of a battleship, you can have many times the damage versus many defenses, plus better survivability. It is a situational ship, which is only worth the expense against planets full of RF targets, though its value goes up as the number of those defenses goes up in relation to ships on the planet. A planet with more defenses than ships on hand is asking for bombers to stop by.

[Dreadnaught] Like the jump in capability granted by cruisers early game, the dread is the jump into high level play. Though it is among the least cost effective for base damage, its broad range of RF makes it great on either offense or defense. With enough dreads hanging around, forgetting to FS is not as big a problem (it still sucks though lol).

[Destroyer] Pretty much just two battle ships in one. Half the speed, but twice everything else (including the cost), it packs a hefty punch if you're lucky enough for it to hit the bigger targets. Against bigger targets, as its info block suggests, it's pretty much a necessity in order to overcome high shields every round. Unfortunately, its RF loadout is barely better than the battleship, so its overall cost effectiveness is almost the same. Given the likelihood that it will remain into later rounds of combat, I would probably prefer one of these over two battleships after the bulk of swarms are cleared. Once there are fewer targets to shoot at, I'd rather my lucky shots be as big as possible. Plus, I'm much more likely to bring these home, so they make battleships seem almost wasteful by comparison.

[Missle Chaser] With RF against most ships, and defense and shield second only to the death star, this is a scary beast. It is complicated to acquire, however, so I'd guess they are not so common. They are least cost effective for base damage, but since they have so many RF options, it more likely sits just above battleships and destroyers. It is definitely another good defense machine, especially since its RF is only against ships and not defenses. The base damage is on par with destroyers, as well, so it can fill the same role as needed. Really, though, it can fit any role. It's pretty amazing.

[Death Star]
Oh, man. That's just nasty. As if being the most cost effective for base damage wasn't enough, it has 96%+ RF versus almost everything, making it even more so. It's even on par with cargo ships for cost effectiveness of storage. Of course, it is ridiculously slow, takes a long time to build, and, oh yeah, requires 10 MILLION resources each. Working up to that amount on hand, on one planet, must be a logistical nightmare (without spending on armistice, anyway). It is definitely worth every penny, though. It can one-shot anything but another DS, and has an average 97.5% RF versus a target with only RF options (which is likely). This means it will almost certainly take out multiple units and/or defenses on every round. To throw some averages* around (these are approximate): on a single round, it has a 75% chance to hit 11 things, 50% for 27 things, and 25% for 54 things. That is to say, over 6 rounds, it will likely shoot 150 things by itself. It fits in any role (if the speed deficiency is tolerable for it), and works fantastically for FS, holding up to enough resources for another DS. As for attacking, departing from somewhere close to your target is absolutely necessary, unless you like the anticipation of waiting a long time. Other than speed, the only thing that might make it better is if you could dock the rest of your fleet on it for the trip (or can you?)

*The "destroyed ships glitch" notwithstanding. It will get that many shots, anyway, even if some hit destroyed ships (as far as I know).


CIVIL SHIPS
There are only a couple comparisons to make here, as not all the civil ships serve the same purpose. As for cost effectiveness of capacity, all the ships used as such are very close, so I won't bother ranking them.

[Small Cargo, Large Cargo, Super Freighter]
Where space battles are concerned, bigger is always better, and these are no exception. The bigger the price, the bigger the capacity, the faster the speed, and the higher the defense. One point of note: the super freighter is the fastest ship in the game, which makes it the best delivery truck.

[Recovery Vessel, Large Recovery Vessel] Like the other flying dump trucks, bigger is just better. A more important point of note here, though: the normal recovery vessel is the slowest ship in the game besides the DS. If you're sending an attack fleet with hopes of shovelling space dust afterwards (and time is a concern), bring the large vessel if you can. Also, if you're in a busy system, the large vessel will have a better chance of beating your opponents to any lingering shiny money clouds.

[Spy Probe] With the option to use probes for upping a fleet's size instead of LF, it's worth comparing to the LF. They are cheap on gas and cost and build time, and are faster than any ship, so I certainly see the attraction. However, the main concern is rapidfire, which everything has against it. To illustrate, any single unit (other than LF) can destroy an average of 3 probes in a single round, whereas most units would only take out a single LF (or none). Cruisers would have to have better damage upgrades than their opponent's defense upgrades to do as well against LF as everything does against probes. So, as a means of increasing a fleet's numbers, we can consider a single LF worth slightly more than 3 probes. Coincidentally, one LF costs slightly more than 3 probes. At high level play where LF are useless except for numbers, it looks like probes are indeed a viable alternative. The question here is only whether you need to counter an opposing fighter swarm.


FLEET RATIOS
The primary concept to understand about randomized targeting is that whatever you have the most of in your fleet will take the most hits. This may be common sense to some, but I don't remember everyone in statistics class keeping up at the same pace. In general, the idea of sending many more small ships than big ships is twofold: to keep the big ships around for more rounds by drawing fire off of them, and to protect your investment in them (like an insurance policy).

Many of the considerations are important for the sake of relatively even matchups, which the honor system tries to enforce. You can always just balk at the more involved math by simply building oversized fleets to attack underpowered targets. In this case, the mantra comes through: bigger is better. However, since failed attacks (draws) are not unheard of, there is clearly a place for running the numbers, if only to save on resources.

There is no "perfect fleet" since there are so many variables and possibilities in battle matchups. Beyond that, your build also depends on your goals for the attack (or defense). Probably the best way to discuss this is with examples. I'll try to stick to concepts that are exclusive to any particular matchup, such as the need for one stat over others (speed, average damage per shot, capacity, etc.)

To start, I'll expand on the primary concept of random targeting. Let's say you have a fleet of 1000 ships. Whatever ratio of that fleet is a particular ship is the ratio of enemy fire it will sustain. If 500 of those ships are light fighters, then your light fighters will take, on average, 50% of all shots fired by the enemy. Note that this is under the circumstance of the "destroyed ship glitch" where otherwise dead ships still take shots until cleared out at the end of the round.

It is also important to remember that your fleet's ratios generally change after any round where you lost ships, as a result of their defense values. Usually, this change would be in the form of your bigger ships becoming a higher percentage of the total fleet. 500 light fighters with 500 cruisers is a 50% to 50% ratio by number of ships (and, therefore number of shots fired on them); however, the same fleet is a 13% to 87% ratio by total defense of each group of ships. In other words, both halves of the fleet will take the same total amount of damage (on average), but they will not both lose the same number of ships. This is another concept that may obvious to some but not others. If you want to have any of a particular ship left after six rounds, you need to consider defense values as well as unit quantities.

A little more on the general concepts of ratios. Say you have 5 different kinds of ships in your fleet, and the ratio by number of units is 20:20:20:20:20. This means, for each type of ship, 80% of the fleet is something else. If your concern is for one particular type of ship, you can adjust its number to fit your goal, then adjust the others as a group to support it.

Stepping back a bit, there is also the cost effectiveness consideration, which depends on your goal. For instance, if you need a fodder swarm for attacking a DS, 1000 LF is obviously a better option that 1000 HF, since LF costs less and the stats don't matter; you're just buying numbers. If you need at least the power of battleships, but don't expect to incur any major losses, don't send destroyers instead; you would just be wasting gas. Also consider similar cost effectiveness where similar options let you focus on the battle. One destroyer costs about the same as 12 heavy fighters, and has similar total stats. In simple terms, 12 HF are an effective replacement for one destroyer, and vice versa, so you can consider only the matchup and not worry about the cost.


EXAMPLE FLEET
To make this simple, I'll pick a single ship for each of the 3 main roles: fodder, damage, and capacity.

Your goal is to attack, win (of course), and haul off, let's say, 50k of each resource from the opponent's storage. You need to bring enough capacity to carry 150k resources. Note that this is the capacity that needs to survive the attack, not just arrive. You decide 10 large cargo ships should do the job. Ignoring the negligible capacity of combat ships, you could lose 4 of them and still have enough room for your goal. Now you have a baseline to build from: keep at least 60% of your cargo ships alive after six rounds of combat (or fewer), and bring enough firepower to win.

You decide dreadnaughts will make a favorable matchup as your core damage dealers. Based on your opponent's defense, 10 dreads will do the job, and you would like to bring all 10 home. Being diligent in calculating what you're up against, you expect your fleet to take 30,000 damage after shields during the battle. As is, you could expect half of the damage to hit each of the two ship groups you've considered so far. 15,000 damage would take out all of your cargo ships, and two of your dreads. This is certainly unacceptable, so now you consider how to support them and prevent such catastrophe.

You're decision for dreads was to counter the volume of cruisers you'll meet. LF is not the best option here, since in order to counter the cruisers' RF, you would need to expend far more resources in LF than HF. So, HF, but how many? You could always go with the overkill option, but it will needlessly cost you in gas. You need to keep the damage to you cargo ships under 5000 or so in order to have enough capacity. That damage could be spread out and not be a problem, but we'll play it safe. Also, you would prefer to stay under 7000 damage to your dreads so there's no chance of losing any. Since you have the same number of ships in both the cargo and damage groups, they would still expect to take the same damage, so we'll take the smaller damage total to ensure safety to both.

So, 5000 damage to two groups leaves 20,000 for the third. Using simple averages for the randomized fire, if 20 ships (10 cargo, 10 dread) will take 10,000 total damage, then 40 ships would take 20,000. However, you can't forget that the battle will go more than one round, and the randomization of fire could potentially favor the important stuff. You could do some more intense calculation to discover the precise numbers (or build a calculator to do it for you hmm...), but rather than spend that time planning every attack, you can more or less just fudge this last bit. I like to round up to the next ten, or hundred, or thousand, etc. We're working with tens here, so ten more it is.

The fleet now looks like this: 50 heavy fighters, 10 dreads, and 10 large cargo ships. Maybe you had 100 HF on hand and could have sent all of them, spending extra gas to spread out shots even more. Probably better, though, would be sending those other 50 with another fleet for more efficient returns on attacks (and maybe even to the same planet).


DEFENSES
Ah, more numbers. I love it! These are quite easy to figure compared to ships, since RF is only a defensive consideration. Keep in mind that the purpose of defenses is not necessarily to destroy incoming attackers, but to prevent them from plundering. Still, here's the order of cost efficiency for damage:

Light Laser
Rocket Launcher
Heavy Laser
Gauss Cannon
Plasma Turret
Ion Cannon

There are a few things that can be said right off the bat. First, unless you're maxed out on all other defenses and want to build more, don't build rocket launchers. Everything else has better stats and usage, has fewer challenges with RF ships, and light lasers are more cost effective anyway. Second, ion cannons are deceptive in their usage. While they are the least cost effective for damage, damage dealing is not their purpose. These have five times the shielding as heavy lasers, but cost the same and still deal more than half the damage. They are very likely to survive a long time and will certainly help bring draws, at least.

As for damage, there are no ships aside from the DS whose shield is not penetrable by any defense. The low cost defenses are therefore not just fodder, but do add a legitimate damage count to battle. The high cost defenses (gauss and plasma) are monsters, with higher damage than most ships and fewer RF challenges than other defenses. A number of ships are good defenders on their own; with the help of significant defenses, your dreams of gaining satellites move closer to reality. Plus, your nanite citizens would probably appreciate a little protection while you're out.

Aside from the stats, there are also considerations of war strategy. With fleet saving, I imagine many don't worry much about getting attacked while there's little to nothing left to steal or destroy. Attackers swing by unimpeded, grab a little bit of free resources, and move along. That last bit is the problem, though. They got away! That same fleet could make run after run, farming points and resources with no punishment. They stole from you and made your nanite babies cry crystal tears, then stole those too! Don't just let them and then steal it back; punish them in the act!

I feel there is a balance to be had between there being enough resources to tempt attackers, enough defense to either draw the battle or do serious damage, and positioning your fleets to keep many active options open. Unfortunately, the lack of out-of-game notifications makes timely responses impossible while not online. Intel is paramount, so getting friendly with an active alliance and making use of some third party chat app to relay notifications can make it possible. Once I've finished figuring my way through the early game, I hope to do just that.


SIDENOTE ON RAPIDFIRE
The numbers I presented were calculated using the concept of probability of consecutive successes. This works because as soon as a RF check fails, the unit stops firing. A DS versus 1000 units it has 99% RF against has only a .004% chance of getting 1000 shots (.99^1000). Also, since otherwise destroyed units remain targettable until the following round, the chance of all 1000 units taking one of the hits is so low, it will pretty much never happen while the "glitch" remains. However, you would certainly hit a lot of them. ;-)


CONCLUSIONS
Well, I like this game. There's only so much I can really conclude so far, as I've only just started playing. Overall, the numbers seem pretty balanced, though I do feel like battleships leave me wanting, and rockets are a waste. Upgrade and fuel costs, as well as timers, all feel pretty good. I've still got plenty to learn, and maybe this will get a "Part 2" at some point.

If you actually read through this, thanks for your time. I hope it was accurate where relevant.

Thoughts?

Talizorah
OFFICIAL
OFFICIAL
Posts: 589
Joined: Sat Jul 20, 2013 10:16 pm
Location: North Carolina, USA

Re: Mulling Over the Numbers: A Noob Review

Postby Talizorah » Wed Jul 16, 2014 8:54 pm

Well worth the read! I'll add this as a link to the New Players post. This probably has the best description of use for ion cannons I have seen so far.

On a side note, Defenses can hit DS, just not Rocket Launchers, Light Lasers, or Heavy Lasers. Ships that cannot hit DS are all Civil Ships, Light Fighter, Heavy Fighter, and possibly Cruiser, depending on each players techs. Also, maybe make it a little more clear that Recovery Vessels sent with your attack fleet, or on attack in general will not collect debris. Only Recovery Vessels sent via the "Recover" command will collect debris.

Excellent breakdown of each ship and its uses.

User avatar
boardwalk
Posts: 20
Joined: Sun Jul 13, 2014 12:20 pm

Re: Mulling Over the Numbers: A Noob Review

Postby boardwalk » Thu Jul 17, 2014 10:16 am

I didn't realize that about the recovery vessels. I had assumed the "debris field" numbers on attack logs were what you collected by sending recovery with your fleet. Are those numbers only what was left in the field that you can collect later? The wording on the wiki sounds as if they do work as part of an attack.

Talizorah
OFFICIAL
OFFICIAL
Posts: 589
Joined: Sat Jul 20, 2013 10:16 pm
Location: North Carolina, USA

Re: Mulling Over the Numbers: A Noob Review

Postby Talizorah » Thu Jul 17, 2014 8:10 pm

The numbers on the battle report list how much debris was created. After the battle, refresh the system the battle took place in (switch systems, then switch back) and a debris cloud will be present surrounding that planet. The amount of debris is approximately 1/3 (needs verification) the cost of all destroyed ships, excluding gas, which will never be in a debris field.

When a planet has a debris field surrounding it, clicking on the planet will make the "Recover" option appear. Provided you have Recs on which ever planet you have active, you will be able to launch them to recover the debris.

Some players like to send a few ships well ahead of their attack fleet to create a small debris field, then time their Recs to land as soon as just seconds after their fleet hits, giving basically no chance for the defender to recover the debris. The trade-off is, that battle report will show up and more quickly alert them to the incoming attack, so long as they are paying attention to the game. So long as debris was present when you launch your Recs, they will continue to the planet, attempt collection and return, even if someone else has beaten you to it. No warnings or notification is given when a player launches Recs at a debris field around your planet, nor will it alert you when it has been collected. It is a separate system from the battle system.

rhetoric
Posts: 7
Joined: Sun May 25, 2014 9:25 pm

Re: Mulling Over the Numbers: A Noob Review

Postby rhetoric » Tue Jul 29, 2014 8:59 pm

Not a bad post, although I believe you are greatly underestimating the power of battleships, they are the bread and butter ships of many a large fleet.

Notes I want to add:
Light fighters are outstanding fodder, late game ratios of 10-20:1 lf:capitol ships will result in the most cost efficient results against most fleets (excluding cruiser heavy fleets)

Debris fields: 21% of the cost of ships (excluding gas) goes to debris. This total is calculated by taking the 70% of fleet actually destroyed after repairs are account for and taking 30% of that, which is the units destroyed to debris conversion rate.

Death stars: not as expensive as you would think, late game their cost isn't all that much to bear.

Talizorah
OFFICIAL
OFFICIAL
Posts: 589
Joined: Sat Jul 20, 2013 10:16 pm
Location: North Carolina, USA

Re: Mulling Over the Numbers: A Noob Review

Postby Talizorah » Thu Jul 31, 2014 7:04 pm

rhetoric wrote:Debris fields: 21% of the cost of ships (excluding gas) goes to debris. This total is calculated by taking the 70% of fleet actually destroyed after repairs are account for and taking 30% of that, which is the units destroyed to debris conversion rate.


Are you sure about that? I am almost positive 100% of destroyed fleet goes to debris (excluding gas). Repairs are completely separate and you lose score for all ships lost. You regain the score when you put the ships in for repair.

NOTE: I believe there may still be a bug where the attacker does not lose any score when they wipe their attack fleet on the defenders planet. This may have been fixed as the devs were aware of the issue.

rhetoric
Posts: 7
Joined: Sun May 25, 2014 9:25 pm

Re: Mulling Over the Numbers: A Noob Review

Postby rhetoric » Thu Jul 31, 2014 10:34 pm

Yep, go test it, destroy 1000 battleships, resultant debris is 9.450.000 metal and 3.150.000 crystal assuming no attacker losses.

21% of the overall cost of the ships destroyed is the resultant debris

LakeSolon
Posts: 16
Joined: Mon Jul 14, 2014 8:09 am

Re: Mulling Over the Numbers: A Noob Review

Postby LakeSolon » Fri Aug 01, 2014 3:58 pm

rhetoric wrote:...
Debris fields: 21% of the cost of ships (excluding gas) goes to debris. This total is calculated by taking the 70% of fleet actually destroyed after repairs are account for and taking 30% of that, which is the units destroyed to debris conversion rate.
...


Does Fleet Repair Tech level affect Debris ratio at all, or is the 30% applied regardless of the tech level of the participants?

rhetoric
Posts: 7
Joined: Sun May 25, 2014 9:25 pm

Re: Mulling Over the Numbers: A Noob Review

Postby rhetoric » Fri Aug 01, 2014 4:32 pm

Fleet repair does affect debris creation seeing as debris is calculated only from ships not repaired.
Debris ratios for different repair levels:
0- 30%
1- 27%
2- 24%
3- 21%

Talizorah
OFFICIAL
OFFICIAL
Posts: 589
Joined: Sat Jul 20, 2013 10:16 pm
Location: North Carolina, USA

Re: Mulling Over the Numbers: A Noob Review

Postby Talizorah » Sat Aug 02, 2014 11:28 pm

Hmm... I was really curious so I asked via in-game feedback. Their response is that repair tech does not affect debris creation. Do you have any BR's you could share with me showing the debris? I'm very interested to see if this response was only valid for Scorpio server, and if the other servers are calculated differently. If you have LINE, you can send them there. My LINE ID is in my profile.

Thanks!
-Tali


Return to “Strategies Sharing”

Who is online

Users browsing this forum: No registered users and 0 guests