Optimising your gameplay

Building or stealing sub for pp

The best possible pp/manpower building optimal troop composition is the following:

troopppsummary.jpg (29.5 KiB) Viewed 5863 times

Note: I will explain the numbers for optimal pp for troops later in the guide. For now, I am merely including the numbers here to illustrate an important point:

Looking at the numbers from the 3 tables above (see also previous post), it would immediately appear as if it is clearly superior to build a sub from scratch rather than steal a sub. If you just consider the pp/men for building a full city to resource level 15 for a level 3, it is 2,38 pp/manpower, which clearly beats the best possible troop build of 1,32 pp/manpower. So if you had the option of building a sub from scratch or stealing a fully built lvl 15 resource sub, you would actually gain more pp by building it yourself.

However, to determine how much extra pp you would get, you also have to consider, that once you start building your sub, your main building will not yet be level 20. You should of course immediately chain build it to level 20. However in the mean time, you will not be producing 639 manpower/hour, while the option of stealing a fully built sub means you get 639 manpower/hour immediately. To determine how much manpower you “lose” by building your main building to level 20, you have to take into consideration that each upgrade lowers the build time of future main building upgrades, and also closes the gap to 639 manpower/hour. Below I will show how much manpower you “lose” by building the main building to level 20 from scratch.

mainbuilding20.jpg (121.62 KiB) Viewed 5863 times

The interesting column is the lost manpower column. You should read this column as: in the time it takes you to upgrade e.g. lvl 0 to lvl 1 (2.500 seconds), you will have lost 440 manpower compared to a full lvl 20 main buildings manpower output.

When you finally get lvl 20, you will have “lost” a total of 20.774 manpower (the sum of the “lost man” column) compared to if you stole a sub with lvl 20 main building. So lets compare the pp value of your “lost” manpower to the pp value of building the sub from scratch:

buildvsstealsub.jpg (34.96 KiB) Viewed 5863 times

So regardless of your level, it is always more beneficial to build a sub from scratch as opposed to stealing a fully built sub from a pp perspective as long as you chain build the main building to level 20. I know some also consider possible lvl 16+ cropland when deciding to steal or not, but I am merely looking at the pp. However, also consider that building a sub, you get to choose a name for it, which is kind of neat.

If you combine all of the above, you will recall that I mentioned it is the best pp in the game to upgrade resource fields lvl 10-20. This means the best possible sub you can steal is one which has upgraded main building, barracks, warehouse, cranary to level 20 and resource fields to lvl 9. A sub such as this is far superior than building one from scratch because you bypass all of the building upgrades, which gives you less pp/manpower than building optimal troops and you immediately get 639 manpower/hour. So if you don’t want to bother with the trouble of building a sub from scratch and you cant find any with 16+ cropland, it is actually not such a bad deal to steal a sub which still hasn’t upgraded all resource fields to lvl 15.

“The sub factory”

Here is a radical idea: If you truly want to maximise your pp, there is a possibility to exceed any other player’s possible pp from building: You become a sub factory! From what I have seen on s58, many people look for subs to steal rather than build them. I don’t yet have enough experience from s115 to determine if it’s the same on merged servers. You could therefore offer a sub factory service on your server, which would work like this:

- you “buy” subs which have upgraded main building, barracks, warehouse and cranary to lvl 20 and resource fields to lvl 9
- you “sell” subs which has all buildings fully upgraded and a full lvl 6 turret wall (easily destroyed)

Some might be happy just to make that trade. Others might need a little more “persuasion”. If you are willing to spend coins, you could therefore offer to upgrade 1 (or more) cropland to level 16. That would surely attract buyers, and as I already demonstrated, upgrading cropland 16-20 is some of the best pp/manpower in the game, so you actually gain a lot of pp from it.

The “seller” will already have the sub in some fixed location. You of course need to be able to hit farms for resources in the area or use coins to move the sub to some part of the map where you know, where the farms are. As long as you make sure never to reach max manpower of 6.390 in the sub while you are upgrading, building lvl 6 turret wall and troops from the lvl 20 barracks, you are earning more pp from building than any other player can possible compete with. Once it is time to “sell” the sub, it all depends how fast a new “buyer” can take it over. While you wait, you can just build optimal troop batches, so you don’t lose pp.

If you decide to try this out, I would either upgrade to lvl 10 or 13, so you have 3 or 4 “production lanes” going. Advertise about your service every now and then in wc, and soon you will have a reputation on your server where people come to you for your service and you might have to form a queue as it takes a while to get all resource fields lvl 10-15 (or 16) built. Unfortunately I can’t say how long, because the build times I have seen on forum are definetly wrong with regards to build times for lumber mill.

You of course “sell” back with full lvl 6 turret wall. This is easily destroyed by cats with a decent shooting tech and you only built the wall to get the pp from it, not to defend the city. You might even make an agreement with the “buyer” that he sends cats every day to destroy lvl 6 turret wall and keep loyalty down. Then you get more pp/manpower building wall than building optimal troop batches.

Building troops for pp

If you’ve read this far, you will have noticed I have mentioned optimal troop build batches some times. It is because that becomes the measuring stick, once you have fully built your city and wall, as you then have no other way to gain pp from building.

Building troops for pp is different than building buildings or wall in that you control the batch size. When you upgrade a building/wall, the resource cost and manpower is fixed, however when you build troops, you get the option decide the batch size, and the pp is awarded based on total resources used in the whole batch. As the pp rounds off based on total resources spent, you are gaining 1 bonus pp when you build batch sizes when the first decimal of the pp formula result is 5-9 and losing when its 0-4. You would mostly be interested in less than max batch sizes if you want to completely optimize, however the tradeoff is that it requires a lot of micro managing by you. Below I will first cover max batches and later optimum batches.

Building max batches

Sooner or later you will build max batches of troops. A max batch is one that drains a full 449.300 warehouse to minimum for the particular troop. The max possible batches are 2496 inf ; 599 cav ; 1404 arc ; 249 cat.

You only have 2 / 3 build queues available depending on if you vip or not. Here I will show how to build the best possible troop composition in each situation. When you consider the troop options available, you can immediately spot which troops give the best chance of max pp by calculating the base stats for each troop type.

troopbasestats.jpg (22.22 KiB) Viewed 5860 times

The higher the “total res. per men” is, the higher is also the potential for gaining the best pp per manpower using the pp formula. Below I have calculated the base pp per 1 troop and base pp per men per 1 troop.

troopbasepp.jpg (110.84 KiB) Viewed 5860 times

Based on the base stats above for pp per men per troop, it is clear, that regardless of your level, you get the best pp for your manpower by prioritising:

cav -> arc -> inf -> cat

It is not enough to just look at the best pp/men calculations alone. There is another constraining factor: build time. Your main building is producing 15.336 manpower per day, or 0,1775 manpower per second. Your goal is therefore to select a composition of troops which will consume 15.336 manpower per day, while ensuring you get the best possible pp using the priority list above. The man power / second is as follows:

manpowerconsumption.jpg (28.03 KiB) Viewed 5860 times

Optimum troop composition without vip

If you don’t vip, you can only fill 2 build queues per city. The priority list says to build cav and arc in both queues. However, if you add the man / sec for cav and arc (0,1322 manpower drain), it is just not enough to keep up with your main building (0,1775 manpower production). You can also see that substituting cav or arc for inf is even worse, as inf has worse manpower drain ( 0,0439) than either of them. However, looking at cats manpower drain (0,1429) it becomes clear that it is easy to surpass your main buildings manpower production by substituting some arc for cat. It is also clear, that you can fill one queue with just cav and keep chain building it the whole day. The difficult task is then to decide how many arcs/cats to build in the second queue.

The full cav queue uses 60*60*24 / 52,60 * 4 = 1642,59 (rounded down to 1642) * 4 = 6.568 manpower. That means there is 15.336 – 6.568 = 8.768 manpower to be split between arcs and cats. If you say arcs = y1 and cats = x1 you can create a function for combination of arc/cat to reach 8,768 manpower consumption:
2*y1 + 10*x1 = 8.768
<=> y1 = -5*x1 + 4.384

Next you need to consider the build time. There are 60*60*24 = 86.400 seconds in queue 2 to build arc/cat. If you say arcs = y2 and cats = x2, you can create a function for combination of arc/cat that can be built in one day in queue 2:
35,60*y2 + 69,99*x2 = 86.400
<=> y2 = -1,9662*x2 + 2427,08

The optimum arc/cat will happen where the two functions intersect. As we have isolated the number of arc on one side, we can set the right sides equal to each other and thus calculate the number of cats:
-5*x + 4.384 = -1,9662*x + 2427,08
<=> x = 645,03 (rounded down to 645)

Using that x in any of the two first formulas above, we can calculate the number of arcs:
y1 = -5*645,03 + 4.384 = 1158,85 (rounded down to 1158)

To summarize: Regardless of level, the above calculations prove that the optimum troop composition for a non-vip player to spend all daily manpower for max pp from building troops is:

1642 cav ; 1158 arc ; 645 cat

Optimum troop composition with vip

If you vip, you can fill 3 build queues per city. The priority list says to build cav, arc and inf in the queues. However, if you add the man / sec for cav, arc and inf (0,1761 manpower drain), it is just not enough to keep up with your main building (0,1775 manpower production). However, looking at cats manpower drain (0,1429) it becomes clear that it is easy to surpass your main buildings manpower production by substituting some inf for cat. It is also clear, that you can fill two queues with just cav and arc and keep chain building them the whole day. The difficult task is then to decide how many inf/cats to build in the third queue.

The full cav and archers queues uses 60*60*24 / 52,60 * 4 + 60*60*24 / 35,60 * 2 = 1642,59 (rounded down to 1642) * 4 + 2.427,08 (rounded down to 2.427) * 2 = 11.422 manpower. That means there is 15.336 – 11.422 = 3.914 manpower to be split between inf and cats. If you say inf = y3 and cats = x3 you can create a function for combination of inf/cat to reach 3.914 manpower consumption:
1*y3 + 10*x3 = 3.914
<=> y3 = -10*x1 + 3.914

Next you need to consider the build time. There are 60*60*24 = 86.400 seconds in queue 3 to build inf/cat. If you say inf = y4 and cats = x4, you can create a function for combination of inf/cat that can be built in one day in queue 3:
22,80*y4 + 69,99*x4 = 86.400
<=> y4 = -3,0700*x4 + 3.789,66

The optimum inf/cat will happen where the two functions intersect. As we have isolated the number of inf on one side, we can set the right sides equal to each other and thus calculate the number of cats:
-10*x + 3.914 = -3,0700*x + 3.789,66
<=> x = 17,94 (rounded up to 18)

Using that x in any of the two first formulas above, we can calculate the number of inf:
y3 = -10*17,94 + 3.914 = 3.734,58 (rounded down to 3.734)

To summarize: Regardless of level, the above calculations prove that the optimum troop composition for a vip player to spend all daily manpower for max pp from building troops is:

1642 cav ; 2.427 arc ; 3.734 inf ; 18 cat

Note: I have just rounded all fractions down, except cats which was rounded up to 18.

Building optimum batches

The section above covered building optimum troop composition using max batches. When you build max batches, the potential bonus pp from rounding, which you could hope to get for each batch becomes insignificant, as you can only get it once per batch. As such it is reasonable to completely ignore it, and just focus on the base stats. However, as I mentioned, it is possible to further optimize if you try to get as many bonus pp as possible while you are building the optimum troop composition.

For every level, there is an optimum batch size for each troop type, which allows you to train as many troops as possible while still earning one bonus pp per batch due to rounding. It is a little difficult to create a mathematical function for it, so I will instead show an example for building small batches of troops as a level 3:

lvl3smallbatchesexample.jpg (82.58 KiB) Viewed 5861 times

I have only calculated the pp/man for batch sizes 1-8. There is an interesting pattern to look for when you want to optimize batch sizes. If you take cav, you are gaining 1,5 for size 1 and 2. Then the pp/man drops to 1,42 if you build 3, but then slowly climbs to 1,47 at batch size 8. At batch size 9 it again drops to 1,44 and starts to climb to 1,46 at batch size 13 (i have not shown batch size 9-13 in the example table). What this tells you is that the optimum batch size for cav at level 3 is 2 and that there are semi-optimum batch sizes at 8 and 13. You are free to calculate all possible batch sizes for each troop for each level for yourself. I have done so in excel, but am not showing it here as the tables are huge (I will however show the conclusions below). To identify the optimum batch size for each troop type and level you should apply this method:

1) Start at batch size 1
2) Keep looking at higher batch sizes to see if any have higher pp/man.
3) If you find a higher pp/man, check if any following batch sizes have the same pp/man.
4) When you have found the highest batch size with the same pp/man, you have found the optimum.

Example for a lvl 3 player building cats:
1) Batch size 1 gives 1,1 pp/man
2) The next batch size with a higher pp/man is batch size 2 which gives 1,15 pp/man
3) Batch size 4, 6, 8, 10, 12 also all give 1,15 pp/man
4) No higher batch size gives 1,15 pp/man. That means 12 is the optimum.

When you apply the method above to all troop types at various levels, you get the optimum batch sizes:

optimumbatchsizes.jpg (92.2 KiB) Viewed 5860 times

For comparison, I below show the pp/men for the all troop types at the same levels when building max batches:

maximumbatchsizes.jpg (87.91 KiB) Viewed 5860 times

When you look at the pp/men for max batches, the troop priority is: cav -> arc -> inf -> cat for any level. I also showed that earlier.

However, when you look at the pp/men for optimum batches, the troop priority might appear to change! For lvl 0, 3, 4, 7, 10, 13 the troop priority is still: cav -> arc -> inf -> cat. At lvl 16 it changes to inf -> cav -> arc -> cat and at lvl 19 it changes to inf -> arc -> cav -> cat. So you might think the optimum troop composition which was calculated earlier is wrong? Answer: No. The optimum troop sizes might show that it would theoretically be possible to get extra pp, however, at level 16 and 19 you would have to switch between 6 or 7 cities and frantically chain build 1 inf and 3 arcs (at lvl 19) to try and squeeze out the extra pp all without missing even a few seconds of build time. It is just not possible to do that 24 hours a day.

So why show the optimum troop sizes? The reason I show them is that as a level 3 you only have one barracks to worry about and that means it is reasonably possible to optimize. I have actually built more than 5k inf, 2 inf at a time for 3 pp, just to get those extra bonus pp while building a hunting army in the early game. It was a little crazy, but if you have the time in the game, by all means try to optimize wherever you can. It could even be that at lvl 13 you build max batches in your subs but try and build optimum batches in your main when you have time.

I will also draw attention to the “localised” semi-optimum batch sizes. As I already showed in the cav example earlier for a lvl 3, batch size 2 is optimum, however batch size 8 and 13 are semi-optimum, where you gain a little less, but still better than batch size 3-7 and 9+. If you don’t have time to build the optimum batches, but you do have more time than it takes for max size batches to complete, you can use the localised semi-optimum batches, where you still earn some bonus pp. Below I have calculated the semi-optimum batches in batch size 1-99 for each troop type at various levels. You can use excel to calculate them for batch size 100+ if you want, however I don’t see the point in that – if you are building so large batches you might as well just build max batches.
Note: If you build any other batch sizes 1-99 than the one I list below, you either gain less bonus pp or in fact lose bonus pp due to rounding down instead of rounding up.

Batch size priority for a level 0 (smaller number is better)
inf 14, 29, 44, 59, 74, 89
cav 14, 29, 44, 59, 74, 89,
arc 3, 8, 13, 18, 23, 28, 33, 38, 43, 48, 53, 58, 63, 68, 73, 78, 83, 88, 93, 98, 99
cat 1, 5, 9, 13, 28, 43, 58, 73, 88, 99

Batch size priority for a level 3 (smaller number is better)
inf 2, 5, 8, 11, 36, 61, 86, 99
cav 2, 8, 13, 19, 25, 36, 47, 58, 75, 86, 97
arc 4, 12, 20, 29, 37, 54, 62, 79, 87, 95
cat 12, 25, 36, 49, 62, 75, 86, 99

Batch size priority for a level 4 (smaller number is better)
inf 2, 6, 13, 17, 24, 28, 39, 50, 61, 72, 83, 87, 98
cav 1, 6, 11, 24, 29, 34, 39, 44, 49, 72, 77, 82, 87, 92, 97, 99
arc 2, 6, 11, 15, 24, 33, 37, 46, 55, 59, 68, 77, 86, 99
cat 15, 16, 17, 18, 19, 20, 21, 43, 44, 45, 46, 47, 71, 72, 73, 74, 99

Batch size priority for a level 7 (smaller number is better)
inf 5, 6, 13, 21, 29, 37, 45, 53, 61, 69, 77, 85, 93, 94, 95, 96, 97, 98, 99
cav 11, 35, 59, 82
arc 10, 21, 32, 41, 52, 63, 74, 85, 96, 99
cat 2, 6, 10, 14, 23, 27, 31, 44, 48, 65, 82, 86, 90, 94, 98

Batch size priority for a level 10 (smaller number is better)
inf 22, 67, 99
cav 2, 5, 21, 37, 50, 66, 82, 95, 99
arc 4, 5, 6, 12, 19, 27, 34, 49, 64, 79, 94, 95, 96, 97, 98, 99
cat 44, 89

Batch size priority for a level 13 (smaller number is better)
inf 2, 8, 14, 20, 26, 32, 38, 44, 50, 56, 62, 68, 74, 79, 85, 91, 97
cav 1, 4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 34, 37, 40, 43, 46, 49, 52, 55, 58, 61, 64, 67, 70, 73, 99
arc 2, 8, 13, 24, 29, 40, 45, 56, 61, 72, 83, 99
cat 3, 7, 12, 21, 26, 35, 40, 49, 58, 72, 86

Batch size priority for a level 16 (smaller number is better)
inf 1, 4, 11, 14, 17, 30, 33, 36, 49, 52, 68, 71, 87, 90, 93, 96, 99
cav 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99
arc 74, 99
cat 8, 25, 43, 60, 77, 95

Batch size priority for a level 19 (smaller number is better)
inf 1, 3, 5, 14, 16, 27, 29, 40, 42, 53, 55, 68, 81, 94, 96, 98
cav 2, 7, 15, 23, 31, 34, 42, 50, 58, 61, 69, 77, 85, 93, 96, 99
arc 3, 4, 9, 15, 26, 32, 38, 55, 61, 78, 84, 90, 96, 97
cat 1, 4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 34, 37, 77, 80, 83, 86, 89, 92, 95, 98

Expected daily pp from building troops

Above we have determined the optimum troop composition for both non-vip and vip as well as the possible way to build it using either max batches or optimum batches. Below I list the pp per day you can expect to get building like this at various levels.

Non-vip pp per day building optimal troop composition of 1.642 cav / 1.158 arc / 645:

nonvipppfromtroops.jpg (134.13 KiB) Viewed 5848 times

Vip pp per day building optimal troops composition of 1.642 cav / 2.427 arc / 3.734 inf / 18 cat:

vipppfromtroops.jpg (149.64 KiB) Viewed 5860 times

For all troop types and levels, the optimal batches clearly beats the max batches. If you build using the localised semi-optimal batches, you will end up somewhere in between the pp for max batches and optimal batches.

Note that at level 10, you can get 64k pp/day. At level 13 you can increase it to 68k pp/day, but it requires more farming effort. Also, if you level up past level 13, you actually start to gain less pp/day!


Practical application of building optimal troop composition

How do you put the knowledge from the above sections regarding troop builds to use? None of the max batches fits a perfect 6/8/12/24 hour build cycle, so it is not feasible to just build chain build max batches. Instead you should focus on 2 things:
1) Making sure your build queues (2 or 3 per city) are always producing troops
2) Making sure you never reach 6.390 max manpower while you are building.

For a non-vip build, I would do the following:
Preparation: Make sure you end at 4.200-5.000 manpower right before you sleep. To do so, you might have to produce only cav/cat for a lot of hours.
1) Right before you sleep, build 599 cav and 900 arcs. Then farm so you have full warehouse and set alarm for when cavs finish.
2) Sleep
3) When cavs finish, build another 599 cav and farm full warehouse
4) Arcs will finish shortly after, now build 249 cats. Then farm for full warehouse and set alarm for when cats finish.
5) Build another 249 cats and farm for full warehouse and set alarm for when cavs finish.
6) Determine when you will go to sleep that day and build a small cav batch that will finish shortly before then. Then farm for full warehouse and set alarm for when cats finish.
7) Build around 140 cats and farm for full warehouse and set alarm for when cats finish.
8) Determine when your cavs will finish and build a small arc batch that will finish shortly before or after then. Then farm for full warehouse and set alarm for when cavs/arcs finish.
9) Repeat step 1

In step 8 you might also go for building around 1.160 arcs instead, so they will finish at step 4 the next day.

I hope you get the idea of how to build. In general, when you are building cav/cat, you are lowering your manpower from your 4.200 starting point and when you build cav/arc, you are increasing your manpower towards your 4.200 starting point. You can build smaller/larger than 140 cats in the last batch so you still end up around 4.200 when you are ready to build the cav/arc batches that will produce when you sleep.

The above method makes use of as many max batch builds as possible. However, for myself I only rarely build max batches. Instead I try to find combos, where localised semioptimums of both cav/cat and cav/arc builds finish at approximately the same time. Some of the ones I use are:
8 cav / 12 arc
25 cav / 37 arc
58 cav / 87 arc
47 cav / 36 cat

I then just have to set one alarm, because both batches finish at almost the same time. If the batches get out of sync in terms of completion time, I build either 1-2 cav, 1-4 arcs, 2 cat all of which are optimums and get them back in sync. Doing this yields you extra pp/day with reasonable little effort compared to building max batches.

I will let you work out what combos that work for you. You can simply find your level in the batzh size priority lists above and pick batch sizes that suits you according to how often you are able to log on to haypi during the day.

Edit: sections below are not yet reviewed to create proper tables. Also I will insert an extra paragraph here about wall building and how to optimise both wall and troop building. Some quick conclusions here:
1) Being farmed is the best thing that can happen for you from a pp perspective
2) When your wall is destroyed for optimum pp, build cav/arc all day regardless of vip/non-vip and use leftover manpower to build walls.
3) Asking for other to deliberately destroy your low level walls over and over is equal to feeding (in my book)!


Defending for pp
As I stated in techs, you will want to get lvl 10 turrets relatively fast and completely ignore traps. When you are defending, you will want your turrets to inflict as much damage as possible. To this end, the front row actually blocks the second and third row, while the second also blocks the third row. By blocking I mean it is impossible for any one unit to attack more than one row in the same round with its attack. Beware, that if you had high level traps, they have a range above 100 and are therefore actually able to “attack” from 2 rows and by doing this you lose one round of blocking. E.g. if you have 15 lvl 10 traps and cav marches to the front row, you will “lose” 10 traps in that round from both the front and second row. So just don’t ever build traps.

In the early-mid game your opponents will not yet have researched level 10 shooting, so their cats will not necessarily outrange your turrets even if you don’t have level 9 turrets yet. 15 level 8 turrets is still a strong wall and you can win some tough fights just with that.

You will want to pick your battles when you spot an incoming attack. To this end you should try to determine your opponents’ skills, items levels and tech levels before the attack hits. Scout the attacker. If he has lower scout level than you, you will be able to see his pp, attributes and skills. If he spreads his attributes and skills somewhat evenly or has a low speed, it is usually a casual player who may not know how to command well. If he has low-medium pp compared to the general server level – it is also a sign that he is not very battle hardened. If his items are enhanced, they will have a colour green/blue/red. This also means the items are at least level 5, but not necessarily level 15. If his enhancement levels are just green, it is usually not something to worry about, but if they are blue or red, there is a good chance he has used coins and therefore probably also have level 15 items. On the other hand, if your scout report is unsuccessful, you know he had lots of money to tech scouting, and therefore possible also other techs. You should go to the pp ranking table and keep scrolling down 50 ranks till you locate his name in the table. His pp ranking will tell you how battle hardened he might be.

If you are completely outnumbered, there is no shame in simply hunting your troops and donating resources to your alliance till you are below 50% of your cranny level. That way the attacker doesn’t get anything from you. If you are crop negative, you might even want to leave some crop in your warehouse above cranny level, so he only gets crop and might think, that you are crop positive. Also make sure you still leave 1 inf/cav/arc/cat in your city even when you hunt the rest. That way you can block for your turrets, so they can inflict maximum losses on the attacker. See blocking strategy below.

If you estimate, that you will be able to beat an attacker, you will want to prepare for the attack:
- At all times (i.e. also before the attack hits), you should know your movement speed with each unit type. As you get better items and enhancements, the movement speed will go up. You can determine it by attacking a ha_a with 1 inf/cav/arc/cat and see the war report.
- If your troops are hunting, harvest your hunt right when there are 5 seconds left before the attack hits. That way the attacker has no way to withdraw for just 2 coins but instead have to pay 5 coins in case he wants to withdraw.
- If you run inf/cav army, you don’t necessarily have to hunt both armies. Sometimes the cav alone can be enough, so your inf gets a longer hunting cycle. Also if you hunt both inf/cav he might think he has no chance and will pause his troops 20 rounds, which earns you no pp. By only harvesting cav – he might be lured into forwarding his troops right into your turrets. And if he retreats to the back line after taking losses, your cav might be able to finish him off.
- Make sure you have at least 1 of each troop type in your city before the attack hits.

When the attack is in progress, depending on the attackers army composition, you will want to meat shield your turrets as best as possible. An attackers with inf/arc/cat ; inf/arc ; inf/cat army will try to get your turrets in range of their arcs/cats while meat shielding with inf. It is however impossible to target turrets directly – they are only auto targeted if no other troops are in range. Hence you will be moving your troops of 1 forward so they are just in line with the front row of turrets – that’s why its important to know your troops movement speed at all times. The turrets are placed at 0, 100, and 200 out. Before you get lvl 15 items, you will usually forward arc and cat one round then wait them. A level 5 manual and 7 horse with maxed movement tech (reasonably easy to get by the mid game) should make your 1 arc move 218 and stand close to the front row of turrets after 1 forward move.
Your cats would reach 174 after 2 forward movements. Your inf/cav will however move so long forward that he attacker can hit them without being in range of turrets, so you should use arc/cat first till his troops have gotten closer to you.

Once his arc/cats get in range, they will waste their entire attack to kill your 1 arc or 1 cat and your front row of turrets are safe and can attack his meat shield or arc/cat directly. When your 1 arc dies, forward your 1 cat next round, so it dies next. Then forward your 1 inf next round, so it dies third. Finally you can start attacking with your cav. If the battle didn’t go as expected, you can always also use your cav army to shield for your turrets by forwarding them one round and pausing. Depending on troop size, they might even be able to survive a hit from arc/cat. Now, if attacker instead sent a cav/cat combo, you will also want to prevent the cav from getting to your first row of turrets. This means it can be necessary to either forward cav/inf first before forwarding arc/cats. His cav will cover the battlefield fast, so you want to block them as far out as possible.

What can you expect in terms of pp gains? Early-mid game I got most of my pp from defending. I’d say it is around 250k pp total. Best result was one fight where I won 20k pp defending with just 3/3/3/3. Late game I just don’t get attacked anymore. Maybe because people know they don’t get pp or because the pp ranking scares people off.

If you get caught at home with full army home
You should never hunt both you inf and cav right after each other. The only exception is if you have already scouted a target where you think you will need both inf/cav to win the battle. In that case you should immediately attack as soon as you harvest both armies. Don’t wait around doing other tasks in town, ac or wc.

If you follow that guideline, you should never be caught at home with full inf/cav army by someone who ports on you. However, if you do get caught with full army, here is what you should do to not lose pp: Use your inf and cav as blockers as I have described and you should at all times have 1 arc/cat, so you have 4 blockers to inflict maximum losses on attacker with your turrets, so you don’t lose pp. Immediately when the battle is over, hunt your remaining troops in an oasis. You must do this fast, so attacker wont gain pp from hitting your 8% healed troops. You will pretty soon learn where the closest oasises next to your main are. Just hunt in one of them regardless of its level or if you already hunted there earlier that same day. You are not hunting to get hunt rewards or pp, but to prevent you troops from getting killed and from losing pp, so the important thing is that you are able to send the troops hunting immediately.

Then start building your wall again and build some troops. At one point, when you think the attacker is not watching you anymore, harvest your hunt and immediately attack farms for resources, while focusing on rebuilding your wall. In a situation as this where your army got wiped, you might consider buying iron from the market for rebuilding your turrets. As a contingency, you could also have at all times 2 stacks of 449.300 iron for sale in the market at max price. That way they never get sold, and you can cancel the trade when you need the iron for your turrets.

Here is why you will not lose pp even if full army got destroyed. First, you must know the 3 pp formulas for battles in the world map. Below “loss” means the total manpower destroyed, so after heals each killed inf = 1, each cav = 4, each killed arc = 2 and each killed cat = 10.

(1) If winner is same level as loser:
Winners pp = losers loss – winners loss

(2) If winner is higher level than loser:
Winners pp = losers loss – winners loss * min ( ( ( winners lvl – losers lvl ) / 2 + 1 ) ; 5)

(3) If winner is lower level than loser:
Winners pp = losers loss – winners loss / min ( ( ( winners lvl – losers lvl ) / 2 + 1 ) ; 5)

Note: if winners pp is less than zero, the winner doesn’t lose pp, but instead gains zero pp.

If you got caught with full army home and you lost, it means you got attacked by someone higher level than yourself if you are level 3. I simply cannot imagine a level 0, 2 or 3 winning a battle against your full army behind your full level 10 turret wall. As such I expect you to be using formula (2) above.

Let us say attacker took the minimum losses possible. If attacker had cat movement speed 98-99 he would have normally been able to be able to attack your turrets without losses if you were not commanding, provided he had enough to destroy 5 turrets in one attack. Because you are commanding the way I described, he will instead hit your blockers when he forwards his cats, and your turrets will therefore attack first. So regardless of attackers cat movement speed, he will take losses. On s58 I have never been attacked by someone who tried to use cat movement speed 98-99, so I just think most people focus on getting speed attribute to 100, which puts your cat movement speed above 99. By the way, I have read a lot of speculation about what speed can make you take no losses against lvl 10 turrets, and it is cat movement 98-99. Cat movement speed 97 will make you get attacked by row 3 in round 4, while cat movement speed 96 will make you get attacked by row 2 in round 3 and by row 3 in round 4.

It is difficult to predict precise minimum losses, because there are so many variables, e.g. your items, tech and inf/cav size, attackers attributes, skills, items, tech, army size and army composition and of course the commands by both you and attacker during the battle. But for fun, lets say the following:
- you don’t block but just blindly forward all troops which the attacker kills by waiting
- attacker does not have cat movement speed 98-99 to destroy your turrets without losses
- attacker has max speed, tech, items, enhancements and heal skill

In this scenario, lets say the attacker either uses inf/cat army or cav/cat army.
The inf will suffer ( 15 + 10 ) * 255 = 6.375 manpower loss. With heal skill 10, attacker loses 5.100 manpower.
The cav will suffer (15 + 5 ) * 77 * 4 = 6.080 manpower loss. With heal skill 10, attacker loses 4.928 manpower.

Lets also say you have an army size of 10k inf and 6k cav and heal skill 4. Below are the results of the battle against the cav/cat army, depending on the attackers’ level given the above conditions.
Notes:
The attackers heal skill is based on using all available skill levels to upgrade it.
The “pp mod.” is the factor which is multiplied with the winners’ loss as per pp formula 2 above.

att att att att your your your your pp pp
lvl loss heal loss-heal inf cav loss loss-heal mod. loss
4 6.160 0,10 5.544 10.000 6.000 34.000 31.280 1,5 22.964
5 6.160 0,10 5.544 10.000 6.000 34.000 31.280 2 20.192
6 6.160 0,10 5.544 10.000 6.000 34.000 31.280 2,5 17.420
7 6.160 0,12 5.421 10.000 6.000 34.000 31.280 3 15.018
8 6.160 0,14 5.298 10.000 6.000 34.000 31.280 3,5 12.738
9 6.160 0,14 5.298 10.000 6.000 34.000 31.280 4 10.090
10 6.160 0,16 5.174 10.000 6.000 34.000 31.280 4,5 7.995
11 6.160 0,16 5.174 10.000 6.000 34.000 31.280 5 5.408
12 6.160 0,18 5.051 10.000 6.000 34.000 31.280 5 6.024
13 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460
14 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460
15 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460
16 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460
17 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460
18 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460
19 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460
20 6.160 0,20 4.928 10.000 6.000 34.000 31.280 5 6.460

You might object that the pp loss column is actually not zero for all attacker levels, and that you are therefore not immune to pp loss. However this is also a worst case scenario. First of all, you should never get caught with full army home. Second, if you did get caught you should command. And blocking the way I described above will ensure attacker takes enough losses for you to lose no pp. E.g. if attacker was lvl 11+, you would use pp modifier 5. So you would have to destroy 6460 / 5 = 1.292 manpower or 323 cav. You can easily destroy that just by getting one block in. It is very unlikely you will be attacked by a lvl 4-10 player who has maxed items and enhancements (they would have levelled up), so even though the pp modifier is lower for these attacker levels, your blocks should still make your turrets kill enough troops.

I have only lost pp from one battle defending. I got caught with full army home right when I joined an alliance for the first time and was saying hi in ac. I didn’t lose pp on the first attack, but I lost 915 pp on the second attack on the 8% healed troops. Again I will stress the importance of always sending your troops to re-hunt or attack, when you harvest them, before you do anything else.


Wall building strategy
Here I will cover the optimal wall building strategy. Later I will cover the optimal way to gain pp from building your wall.

I already stated your wall should be only turrets and no traps. Some seem to think it doesn’t matter where your turrets stand, but it does. Once all 15 turrets are at level 10, they are of course the same, but until you get all of them fully built, you will want to build and upgrade them in the optimal way. It is important to build in the optimal way, because you might get attacked before you have fully built your wall. So your goal is with each upgrade to always make sure it puts you at your best possible defence for the time and manpower spent so far.

Some things you should know about turrets:
- An attacker is forced to attack the front row before second row and before back row. So even if you only have 1 turret in your front row and 5 turrets in your second row, the attacker is forced to first attack the 1 turret in the front row.
- If an attacker gets an attack on your turrets with one of his unit types, that unit type will make an attack against all turrets in the same row.
- If the attacker is unable to destroy all turret levels in his attack with his unit type, the attack will first destroy as many levels as possible from the left turret. If the attack points exceed the turrets life, remaining attack points will then be used to destroy levels of turret 2 from the left until it is fully destroyed and so on and so forth. That means your rightmost turret will get destroyed last.
- The battlefield range is the range of the unit/turret with the highest range + 300. So even if you had just one level 10 turret, the field would be range 1350. The other 14 turrets can be level 0-9 and the field range would still be 1350.

Using the above information, you can plan the optimal way to build and upgrade your wall so you always have the best possible defence while you are building the wall up. You should also take into account that you want the best possible pp for your invested manpower and you don’t want to waste manpower, which could have been spent better building troops.

Note: If you see a lot of incoming attacks in your action list, skip to the section further down about optimising pp from building wall while under attack.

Building wall for max defence (and pp) while building up
Chain build the bottom right turret to lvl 6. You will eventually want to get it to lvl 10, however for now stop at lvl 6, as you get the best possible pp this way, and your pp starts to drop compared to building troops when you upgrade lvl 7-10. The bottom right turret should always be the one you are focusing on upgrading when you don’t yet have a fully built wall, as it is the turret which will be attacked last by the attacker, so in effect the 14 other turrets block for bottom right turret. Also, each level you upgrade this turret adds 50 to the battle field range, which means all 15 of your turrets will potentially get extra rounds to attack while the attackers units are trying to cross the battlefield.

Immediately build the rightmost turret in the middle and front row to lvl 1. Then stop building front and second row for now. The two lvl 1 turrets, you just built, are enough to block the whole back row from 2 attacks.

Now start upgrading the 4 remaining turrets in the back row to lvl 6. Always start with the rightmost turret and work your way left when you do the upgrades and build them evenly, so lvl 1 -> lvl 1 -> lvl 1 -> lvl 1 -> lvl 2 -> lvl 2 -> lvl 2 -> lvl 2 -> lvl 3 and so on and so forth. When they are all lvl 6, stop upgrading the back row, and start working on the middle row. Again work your way from right to left. This time however, upgrade all 5 one at a time. As the right most is already lvl 1 that means the build path is lvl 1 -> lvl 1 -> lvl 1 -> lvl 1 -> lvl 2 -> lvl 2 -> lvl 2 -> lvl 2 -> lvl 2 -> lvl 3 and so and and so forth. I already stated why the bottom right turret was special, and there is no need to get the middle right to lvl 6 before you start upgrading the 4 remaining middle turrets. When you are done with the middle row, start working on the front row and upgrade them all to lvl 6. Follow the same method as for the middle row as for the second row.

After the above, you have a full lvl 6 wall and will have earned optimal pp compared to building troops. If you chain built all turret levels, you will have spent 5 hours 23 minutes and have spent 4.080 manpower. In that period, your main building will have produced 3.440 manpower. So either you would need to have had 640 manpower saved up or you could slow down chain building, so it takes you an extra hour to get there. When your city is fully built you will be happy to know you have just spent your 4.080 manpower in the best possible way in the game for gaining pp, regardless of your level. A full lvl 6 turret wall is also capable of killing some attacks when you block with troops, however you can have trouble with attacking cats. Lvl 6 turrets have range 850 and if the attackers shooting tech is lvl 4 or higher, he outranges your turrets. However if you use the blocking method I described earlier with your four 1/1/1/1 units, and the attacker does not pause when he has a turret row in range until it is destroyed, he might overstep your turrets range, and you can inflict some losses on even his cats.

Upgrading your turrets from lvl 6 to 10, you should again start upgrading the bottom right turret to level 7 and work your way left in the back row, then left in the second row and finally left in the front row. A full lvl 7 wall from a lvl 6 wall will take you 3 hours 37 minutes 30 seconds and cost 2.400 manpower. In that period your main building will have produced 2.316 manpower, so again you will be a little short on manpower. Building level 7 turrets is almost as good pp as building troops for pp regardless of your level, but you should prioritize first building the barracks troops which yield better pp if you have limited manpower. However if manpower is sufficient to keep cav/arc queues filled, by all means build lvl 7 turrets before cats/inf even if you vip.

Level 8-10 your pp from building turrets will drop so significantly, that you would lose pp compared to if you instead used the manpower to build troops. However, level 8-10 also increases your chance of winning defensive battles, which could earn you good pp. You will therefore have to decide when you think is a good time to upgrade these last levels. In the long run, it is always better to have a level 10 turret wall up compared to a level 7 wall, so you will want to upgrade eventually.
- If you know, you will be attacked by overwhelming forces in the next couple of hours or days, either don’t upgrade any further or only upgrade the bottom right turret to level 10, so you increase the battlefield range and inflict more damage on the attacker.
- However, if you don’t think you will be attacked in the near future, start upgrading to lvl 10. The best way is to chain build the bottom right turret to level 10. Then proceed to upgrade only the bottom row to lvl 8 from right to left, then bottom row to lvl 9 from right to left and finally the bottom row to level 10 from right to left. You now have very strong firepower from your 5 lvl 10 turrets. Getting the back row to lvl 10 from lvl 7 will take you 11 hours 44 minutes 48 seconds and cost you 4.400 manpower. When you are building levels 8-10 you main building actually starts to produce more manpower than you need for the upgrades, so you can use leftover manpower to train cav/arc.
- Now start working on you second row. Upgrade from right to left to lvl 8, then right to left to lvl 9 and finally right to left to lvl 10. After you’re done, do the same with the front row.

Optimising pp from building wall while under attack
At some point in the game, you might find yourself attacked several times in short intervals by players much stronger than you, even if you have no troops out. This can happen e.g. if you upset someone from a strong alliance, or if someone just wants to try and keep you down. You might think this is a terrible situation, but you have to turn it to your advantage. Once your city is fully built, this situation is actually your very best opportunity in the game to earn the best possible pp from your manpower.

Lets say the first attack has destroyed your wall and you can see a lot of extra attacks lined up at various intervals in your action list.
(This section not yet finished - will edit later)

Reserved for future content

Reserved for future content

Reserved for future content

Reserved for future content

Reserved for future content