Difference between revisions of "Guide"

Latest revision as of 02:01, 30 May 2017

Growth

If a kingdom is optimizing for growth, always ask your monarch before you take acres from the explore pool or cast paradice.

If you are dicing for growth, you want about 20-25% guilds and as many towers as necessary to cast the spell every tick. If you are exploring for growth, it depends on how you?re doing it, but you?ll want mills and banks and potentially armories to bring down the draft cost. However, for most provinces that you would want to explore, peasants generate more gc than the tb hit from getting aided soldiers so trading is more optimal where you can afford to buy soldiers for TB and get them aided instead of drafting..

Most provinces will grow largely from attacking. Which leads us to -- how do you grow from attacking?

The answer is partly that it depends. There is one central rule, however: THE BEST WAY TO GROW IS TO NOT GET RETALLED. Your province will be in the best possible shape if you acquire safe acres and don?t get retalled. Even if you only get 50 acres per hit when you could get 150 elsewhere, this is the best way of growing. Why? Because every so often the hit for 150 might get you hit back for 250, then you?re trading hits and dealing with military/peasant loss that is bad for your province.

Generally speaking, when you?re whoring, you want to be focused on offense to the exclusion of defense to bring down your nw relative to your targets and let you hit safely. This is part of why you need to be disciplined when whoring -- two ages ago, I whored up to 6k acres with under 6k defspecs and pure offense elites. Growing this way is more art than science; unfortunately you need to have an understanding of whether you?re likely to get waved and if you are, you need more defense. So it?s not something you can do without some kind of deals.

So how do you identify safe hits? First, you look at the province you?re hitting - what?s their OME? If it?s higher than honor + payrate, they have TGs and are an attacker. This means be careful. Do they have horses? Another sign of an attacker. Generally speaking, attackers can hit you back which is what makes them bad hits -- it costs you or your kingdom acres when they hit you back. You want to hit t/ms that are the biggest province in their kingdom in kingdoms with no attackers the same size as you / in range to retal your kd.

Now, obviously this is more of a general rule than hard and fast; it?s impossible to always hit safe provinces, but always hit the safest in a kingdom if you?re randoming a kingdom, and always hit the safest available to you.

When growing, a sample build might be something like: 8% farms 16% TGs 25% guilds 5% towers 10% stables 20% banks 15% hospitals 1% dungeons

The goal here is to have decent econ (banks/hosps), offense (TGs/stables), and still gain WPA.

I got hit, what do I do? Generally speaking, if you got hit you can't hit back. If you can, you should. Retalling slows growth at the top and generally discourages more hits into your kingdom. There are some issues here though -- eventually, target selection reaches a point where you're the most attractive target even if you do retal. And sending out can get you hit more. So you need to have a sense of what else is out there that can be hit.

In ~95% of situations, retalling is the right move for you to make.

If you can't retal, there are some ops you can do, but generally the best move is to rob resources from the kingdom that hit you.

Generally speaking, training offense will make you less likely to be hit (by bottomfeeding SKs) than training defense. It's more NW efficient and makes hitting you risky.

Waving for Acres There will be times when due to setup, global growth mechanics, or cf?s purely safe whoring is not viable. At times like this it is possible to hide your weak whoring ability by running a more war-growth hybrid build and wave kd?s that you believe are unlikely retal. If you random out on a bad hit, then you are much more likely to be retalled than if you wave a kd. However, this is still a far more risky option than safe whoring. This can be a good strategy for when you have a combination of turtle core and suicide core. Knowing when you can get a war from waving, when you will get a cf and acres from waving, and when you will be trapped getting fb?d for weeks is critical to this more risky strategy.

Pumping When pumping, again, you need to have an understanding of the climate. If you?re not likely to get waved, you can pump more aggressively than if you are likely to get waved. When pumping, you generally want to get to whatever the set ?target? defense is, and then have as many peasants as possible. You will need thieves and/or WTs to protect from being robbed.

Scipump When pumping science as a TOG race, a sample build would look something like: 8% farms 22% guilds 35% banks 30% schools 5% towers

Note -- you only want as many towers as it takes to TOG. You want ~15% guilds minimum; we generally prefer to run more (even more than the 22% here) to pump WPA at the same time. The schools/banks distribution is something that will vary as you gain more alc sci and BE sci, but this is a decent starting point.

You?d generally be hoping to achieve about 8 military units/acre, 2 TPA, 2 WPA, and 15 peasants/acre, give or take.

Elitepump An elitepump would look similar to a scipump, except you wouldn?t have schools and would probably have all banks until you want to start drafting. Then you would rebuild to have 25-30% arms, (more/less depending on cost) and train them all in one go.

Homespump A homespump operates similar to an elitepump. The main difference is that ~48h before you begin drafting, you rebuild to add homes to your build -- the number of homes depends on how much gold you have available. This lets you draft from a higher population number, meaning that when you raze the homes, a higher % of your total population will be in military compared to not using homes. Using homes can be a very effective way of getting drafted quickly or getting more drafted than you could have without homes.

WPApump This is pretty basic -- run as many guilds as you can. At least 50%, preferably more.

Kingdom Tools If you use any of these, design any of these, and want to add a little, please let me know.

There are some general tools available for use. There is utopiapimp,and upoopu that are publicly available, and then stinger makes stingernet available and munk makes munkbot available. Lucidbot is an IRCbot that is also publicly available.

Utopiapimp This is the most basic of the sites. It does not link with an IRCbot, or as far as I know, have a browser extension. Forum agents(such as the one included in Angel) can send directly to it. You can then view data on your kingdom or opposing kingdoms along a number of different metrics. The way intel can be viewed is highly customizable on utopiapimp. It can also track ops, which have to be inserted manually into it. Utopiapimp sometimes suffers from display bugs.

upoopu A more optimized webtool that has similar data to utopiapimp. It has a browser extension which makes it easier to get data into the system. It is supposed to be easier to manipulate data once it is in the system as well.

Stingernet Stinger makes stingernet available. It is an ircbot that is linked to a website. It has similar data available in its website. It also has a custom targetfinder that is linked to a NAP system to prevent anyone you have a CF with from showing up in the targetfinder. In order to have access to stingerbot/stingersite you must obtain permission from Stinger. There are limited spots for his service. Stingernet is largely available on the absnet server, although it has migrated to utonet on some kd?s. Having an ircbot means that you can enter a command string, such as !cb zauper and get back my SoT, rather than having to go to the website to view it. It?s partly a matter of convienence, and partly very helpful because there are custom views that are easier to get on IRC than through the website -- a sci calc, for example.

Stingernet also includes an automated WPA calculator.

• • I?m told that Stinger currently has it available now.

Munkbot This is what CR uses. It has similar features to stingernet, listed above, plus a number of other ones. It includes an activity tracker (internal and external), links to prep and province simulators to help you prepare for conflicts or think through strategy, and has fully customizable views that are designed to be exported into other tools for setting up waves. It is both a website and a bot.

It also tracks op success rates over time and will automatically calculate WPA for provinces you collect all intel on. (You can force it with !wpacalc on IRC).

It also has incorporated tools similar to utools below.

You can also generally request a feature from Munk and he will code it after spamming !donate at you for a few minutes. He?s fussy but if you ask nicely he?ll probably set up a bot/site for you.

IRC Bot: orders aid tracks ops tracks attacks show intel integrated with whatsapp and can call/sms. perform calcs (soldswap / wpacalc / defence home / practical offence / offense keeping set defense and much more)

Lucidbot Lucidbot is only a bot. It collects intel from whatever repository you use and is fully customizable for reporting. It comes with many of the same bot features as stinger/munkbot. Lucidbot must be hosted by someone in private channel. Lucid is very easy to obtain and can be used in conjunction with any of the other public intel sites such as pimp or upoopoo. An intel depository site and intel pull script has been in the works.

Other Tools Angel Angel is the classic tool for intel. You take a SoT, copy the intel, there?s an export line so anyone else with angel can read it in. Angel forum agent can be used in conjunction with all bots and intel repositories that lack intel-pulling scripts.

utools A browser extension that duplicates much of angel?s functionality. Alters UI in-game, adding features such as expected gains and beheading advisors.

Targetfinder Munk?s targetfinder (tf.umunk.net) is probably the best publicly available targetfinder. To use it, search for provinces within your nw range (+/- 30% nw), you can set kingdom size restrictions as well, and sort/limit by race. The TF also possess war trackers. Site is sometimes private at start of age.

Overpop Note: preventing overpop is a mechanic that really privileges activity and it is not expected in most kd?s that you do these things when being chained. If you do nothing else but mass LL before troops get home so as to be able to send out easily this is still better than nothing. However, lots of wars are won or lost by how successful chains are, and these things make chains fail. One of the most significant differences between top-tier kd?s and standard war kd?s is their commitment to overpop control and maintaining a strong ?small? group of provs.

So, a few new things to work on for preventing overpop. Preventing overpop as you go down in a chain is good for lots of reasons: 1) if you have enough land coming in with armies you end up losing nothing and the chain is completely useless, 2) without losing defense you are able to maintain higher def, often preventing multi-taps which can also decrease chain success. 3) At the bottom of a chain defense is, perhaps counterintuitively, often more valuable than offense.

So the basics are still the same: have lots of acres incoming, army-in army-out, release to dragon if you have to before sending out. Here are a few formulas and tricks though:

Military desertion formula: ((((total pop-(peasants*.07))/maxpop)-1)/5) * (thieves + eliteshome + soldiershome + specshome)

Note: You only desert if maxpop*1.15 is less than totalpop-(peasants*.07). So if you have 10k total pop and 9k maxpop, you?re not deserting.

You lose up to 7% of your peasants/tick when you?re overpop, and they leave prior to desertions which is why the formula includes that.

If this number is equal to the amount of soldiers you have, you only lose soldiers for the tick. This is why pers/race bonuses that return dead troops as soldiers are are so valuable when lots of people use IRC, because you use the soldier pool to prevent casualty losses. A good kd will use the soldiers generated to try to bounce provs and feed others. A great kd will use the soldiers to make it as if the prov was never chained in the first place.

Trick #2: besides microing tickly loss this way(either by releasing thieves into soldiers to leave, being aided, ect) another strategy to alievate the destruction of chains is to setup LL targets, either via AW or out of control growth. These target should not be targeted by attacks and will serve exclusively as LL farms for chained provs. The only spell not affected by -damage mods in race is LL damage, so regardless of what kind of attacker you are, if you are likely to be chained you should try to maintain 60 mana at all times. That's a lot of LL, and the target will always be very large, and have no wpa. When your land is about to come in, LL to where you would need to be once acres come in + are built and then go use your great new mwpa for fb.

People argue over the value of releasing WPA on chained provs. 1) If you keep the WPA, you wind up with high WPA provs capable of FBing potentially otherwise unoppable provs in the other kingdom 2) If you keep the WPA, you wind up with a lot of lost potential military and your chained provs potentially fall behind 3) If you release the WPA, you can LL less (but you would run out your mana on LL before releasing 4) If you release the WPA, you can be intra-kd fireballed to release peasants, but it usually is better to release the WPA (several ticks) prior to army coming home, as you release WPA into peasants that you then need to lose to solve overpop.