Immaculate-002: Late Age for Noobs (GAME OVER: ATLANTIS/SHATNER WINS)

[triqui]

Yep, that's exactly what I meant with an inform
Very helpful, thank you shaetner. I would had made one if I weren't so damn busy lately, but it's a good thing to do in newbie games, so everybody learns how to extract data from the graph charts.
Nice job!

[arnob]

Impressive !

Now, I know why I'm still a noob

Arnob

Originally Posted by shatner

Here's my take on things...


Provinces
The early game baseline is conquering one province per turn every turn starting on turn 2, so by turn 13 the baseline is 12 provinces. Now that becomes an increasingly meaningless baseline over time as nations get additional expansion squads, bump into nations, engage in early wars, etc, but we just entered our second year so it still has some meaning.

Man, Ulm and Gath are behind the curve in this regard, Patala is exactly on track and T'ien Ch'i, Marignon, Jomon and Atlantis are ahead. No one's line has dropped meaning it is likely that no one is currently warring with anyone else. It bears pointing out that Jomon, Atlantis and, to a much smaller extant Patala, can expand natively under the water. Some of their size might be because they have more provinces to choose from when invading. No one conquered two provinces on turn 3, indicating that no one has an awake SC pretender.


Forts
Only Man, Ulm, Atlantis and Jomon have an additional fort, built in that order. At this stage that means either profound luck (getting a fort creating event, finding a fort creating site, getting a lot of money from events) or having decent scales. Atlantis and Jomon are the two largest empires so they probably have money to burn on forts. Now, LA Man has a lot of synergy with their forts (having strong and high admin forts, good siege-bonus troops and commanders, etc.) and uninspiring sacred troops. As a result, they usually opt for good scales which they can invest in forts to compound that income scales by the high income. This behavior backs that interpretation up. Ulm has many gold-cheap, resource-expensive troops who are effective against indies so they can recruit fully with money left over for forts.


Income
On turn 1, everyone's income is the same +/- their scales. As such we can tell that Man has the best income-enhancing scales and Patala, the worst. The top four incomes belong to the four largest empires, so no real surprises there. The spikes followed by dips this early on usually indicate overtaxing the capital. Note that this graph does not show income from events (whether good or bad), gifts from other players, nor does it reflect upkeep. We are looking at gross income from taxation; nothing more.


Gem Income
Jomon and T'ien Ch'i are both leading the pack by a wide margin, which makes sense as both nations can recruit very diverse mages who make excellent site searchers. Also, Jomon looks to have an awake rainbow pretender and their research curve eased off around the time their gem income started to shoot up so it's possible their pretender is out site searching. That dip in Jomon's gem income is probably from a bad event (barbarians attack, for example) stealing a province from them and then being reconquered the following turn.

Note that this chart gives no indication of income gained from blood hunting; something which both Ulm and Marignon are well equipped to do.


Research
Ulm and Jomon must have awake pretenders with decent research to have such good turn-2 research. Also, Ulm and Jomon's research lines have sharp climbs followed by plateaus which usually mean a pretender who has left the lab to go indie stomping or site searching. Of course, Ulm could be sending out mages to go blood hunting, which would also explain the declines in their research. Everyone else's research lines are more-or-less curving smoothly; this indicates that they have been hiring one mage a turn and sticking them in the lab. The fact that certain lines are curving more sharply upwards indicates that they have better national researchers and/or better magic scales. The rank-and-file mage of Atlantis, for example, pulls in a meager 4 research a turn whereas Marignon's recruit-anywhere Goetic Masters get almost twice that. And if someone has magic-1 or magic-3, the lines get even more divergent.


Dominion
Jomon and Ulm had the top dominions early on, which jives with them having awake pretenders (since a pretender counts as three temples for generating dominion). T'ien Ch'i and Atlantis have since gained in the ranking, hinting at them having either a really high dominion (which would make it likely those nations are relying on sacred units for their expansion, and as such might have a strong bless). Alternatively, they could have built additional temples early on. Man's upswing in dominion coincides with the turn they built their second fort, meaning they likely built a temple there too. Gath's losing dominion might mean they're hemmed in between high dominion neighbors, or they've just had a run of bad luck (there are several events which lower dominion, which make a big difference on the charts early on).


Army Size
First off, note that army size is literally the sum of the actual size of all the units for each force. Therefore, a nation with 35 size-6 elephants will have a larger "army size" than a nation with 100 size-2 humans. Since the absolute values aren't very telling, you instead want to look at the changes. None of the nations present get free-spawn save Ulm, who can hire commanders who spawn wolves. However, the chart doesn't really show that Ulm is building enormous wolf armies, or if they are they are dying almost as fast as they are being raised. T'ien Ch'i, Jomon and Patala all had spikes on their charts. For T'ien Ch'i and Jomon (whose troops are fairly resource intensive), that probably means they hired mercenaries and then threw them at indies until the mercenary force was depleted. That or they had the "good" event where they get a bunch of militia. Since those militia suck but cost upkeep, it is common for the player to use them as cannon fodder or outright suicide them to get them off the payroll. Patala has had the largest spike and they can hire a bunch of troops who are low-resource (monkeys), large (gorillas) or both (elephants) so it is possible that they turned a lucky event into a primate or pachaderm army overnight.

Gath has the smoothest army size line. You get that from recruiting the same thing every turn from your only fort and then never suffering meaningful casualties... which, given that they're giants, could be the case.

Beyond that, Jomon, Patala and Ulm all had moderate losses this most recent turn. Normally that would hint at war but the province chart doesn't back that up, so it could just be a coincidence that they all attacked difficult indies this turn and took losses. A quick glance at the Hall of Fame shows that Marignon's prophet is dead, so that's probably related. T'ien Ch'i and Man both seem to be growing their armies aggressively for the last 4 turns. Either they've got lots of indies to stomp, or they're gearing up for war.


Anyway, that's my take on things. Take all that with a grain of salt because I'm still a noob and Dominions is a complex game where a lot of stuff can happen which can cause the charts to be misleading.

[BewareTheBarnacleGoose]

Sorry about the taking so long to send in my turn, everybody. We lost power here, and I couldnt get online.

[Immaculate]

Stuff happens. Seems like we are all good now.

[triqui]

Originally Posted by arnob

Impressive !

Now, I know why I'm still a noob

Arnob

There are quite several tricks that we all need to learn, it's much better if we share experiences here, that make the learning curve much faster.

For example, several new players ignore the fact that in your first turn, it's often the best play to put taxes at 150% in your capitol and put your commander to patrol to fight unrest, instead of becoming a prophet. That's true often for all non-blessed nations, the impulse you get from that extra half a turn in gold is better than the slight increase in combat efficiency that 1 smite per turn gives against indies. Of course, this might prove false depending on your nations, several nations *need* that divine blessing and/or the sermon of courage+smite spam.

Some other common first turn tricks is to attack blindly a province with a scout scripted to retreat. This way you see a perfectly accurate report from one of the province, which may help to decide what to attack in the first turn (your scout get experience and goes into the hall of fame as a side bonus :P)

Other thing that often is overlooked by newbies, is manual taxation. When you start to play, it's common to leave the taxes in "auto", so the game move the taxation acordingly to your unrest. That's suboptimal. First, you can tax 110% most of your provinces. Unless the unrest is 6+, there's no drawback. So you can set all your provinces with unrest 0-4 to 110% taxes, and leave it to 100% taxes when it goes unrest 4-5, without any drawback. Second, the AI overreact to unrest. You clean 1 unrest for each 2% taxes left. So if you have unrest 5, the AI will go tax 90%, to remove 5 unrest. However, with 6 unrest, the AI will go to 80% tax, to remove 10 unrest, That means the first 10% tax gives you 5 unrest removed, and the second 10% tax cut gives you 1 unrest removed, which is not efficient. Third: if you keep all your provinces with unrest 0, you are losing the benefit of several good events (those that lower unrest)

Another common mistake is to forget to raise Province Defense in every conquered province. An undefended province is vulnerable to even a scout attack. PD 1-2 cost next to nothing and avoid this. The best way to never forget this is combining it with the manual tax: go to Nation Overview (in the statistic section) and check every province has Def >0 and every province that has unrest <5 is set to 110% tax.

Then an often overlooked aspect of the game for new players is long term planning. Often we tend to do what is "needed" for the next turn, without looking what is needed 2-3 turns in advance. One good example of this is castle building. If you recruit every turn everything you can, you often can't build a castle in the first turns. That is a very big hindrance for middlegame, you HAVE to be building a castle before year 1 is over. Same goes with site searching: often we don't put mages on site searching becouse "we really need that extra 6 research *this turn*". This leads to midgames where you have research, but nothing to do with it becouse you have a lack of gems.

[Immaculate]

tell us more about the unrest events.

[triqui]

There are several beneficial events that reduce the unrest in a province. If you have 0 unrest, they do nothing. If you have 5 unrest, they reduce it to 0. So if you have taxed your province at 110% tax rate for a couple of turns, and you hit one of those events, you have got extra free gold. If you don't, just go back to 100% taxes and leave the province at any unrest level from 1-5, which does not hinder you at all (you need unrest 6+ to lose gold). You get a few extra gold coins this way, and in the begining, that might mean the difference quite a lot. Buying one extra mage at turn 3-7 means a LOT of research through the game. Buying a few extra soldiers might mean the difference between taking 2 provinces in turn 4, or taking 1, and so on.

Another popular strat with some nations (Man, Tien Chi, bogarus, Pangea) is to patrol your capitol at 200% tax rate. Those nations have very good patrolers (either free patrol bonus, or very cheap high movement or flying units, such as Pangea). This way you get an extra punch of gold in the begining. Overtaxing your capitol and/or main provinces in the first year might give you an extra 5000 gold, which mean 3-6 castles (specially for nations with good Administration in their castles, such as Bogarus or Man). That's a nice bump. Getting a good headstart might give you a much more confortable midgame: it's way easier to fight midgame if you are leading in provinces, income, castle, or resarch (or any combination of those), than if you are a mediocre nation (or the underdog).

[BewareTheBarnacleGoose]

Thanks, Triqi, this has been extremely enlightening! Since we are on the topic of unrest, I have a question- I've read that friendly dominion decreases unrest. Is it known by how much it does so, and does it scale with dominion strength?

[Calahan]

That’s some pretty good newbie advice you’ve been giving out there shatner and triqui, much credit to you both for that (and indeed some well seasoned players would do well to listen to what is being said regarding the autotax as well, since while it cuts down the MM loads, it really is pretty inefficient at its job as triqui says)

But thought I’d correct a few points in passing (since one of my many faults is hating to see misinformation, even if it’s only slight in this case, but it's best to drill-in correct info from day 1 IMO, rather than aim to correct it later once the pupil becomes more educated)

Originally Posted by shatner

Income
On turn 1, everyone's income is the same +/- their scales.

(what follows are very minor points, but thought I'd mention them for those like myself who like to be anal about every minor detail )

1 – The starting income is highly influenced by scales as you say (and so again as you say, the turn 1 income graph is a great source of intel on the enemy scales in this regard), but it is not the only factor as you suggest. Since it has to be mentioned that the type of fort a nation has at their capital also affects that nations turn 1 income. i.e. if you start a test game with EA Ermor and EA Agartha, and take max positive income scales with both, then you will see a distinct difference in their incomes due to EA Ermor having a Great City of admin 60, and EA Agartha having a Cave City of admin 30. And those 30 admin points can make a noticeable difference in turn 1 incomes. (watch out for differing populations throwing the figures off though. See next)

Also the starting population is not exactly 30k for each nation, as it can be anything between 29500 and 30500 (or maybe 29501 and 30499). And while the 1k difference between the max and min values doesn’t make the biggest difference in the world over the long haul, 1k of population will probably add something like 15-20gp to the income if a nation has taken good scales

Originally Posted by triqui

Unless the unrest is 6+, there's no drawback. So you can set all your provinces with unrest 0-4 to 110% taxes, and leave it to 100% taxes when it goes unrest 4-5, without any drawback.

Second, the AI overreact to unrest. You clean 1 unrest for each 2% taxes left.

2 – Your unrest data is a bit off triqui. Firstly under-taxing reduces unrest by 1 for each 3% below 100%, and not for every 2% as quoted above. (this formula is given in the manual (and the Wiki - http://dom3.servegame.com/wiki/Province), and it is correct on this one). So setting taxes to 90% reduces unrest by 3.3 (rounded down to 3) and not by 5. Although unrest can be reduced by more than 3 with 90% taxes due to the effect of positive dominion. As positive dominion can reduce unrest by between 0-2 each turn (which might be where you are getting your figure of 5 from). But note that there is no guarantee that positive dominion will reduce unrest, only a chance that it will. So 90% taxes is only guaranteed to reduce unrest by 3 and not 5 (excluding all other factors of course), but unrest can be reduced by 5 with 90% taxes if you get a good roll from a positive dominion unrest reduction.

Also any amount of unrest will affect income, as it is certainly not ignored until 6+ as you state. IIRC you lose something like 2% of income per unrest point. This is easy enough to test as well by starting a new game (take neutral growth), making note of the cap income for a few turns (although watch for variations due to temp scales changes), set 110% taxes for a turn or two to generate 1-2 points of unrest, then reset taxes to 100% and take note of the income figure. You should see that income is down by a few percentage points due to the unrest in the province (and by too much for it to be just the loss of a few tens of populiation caused by the overtaxing).

So running provinces with unrest until they hit 6+ before looking to reduce it is not likely to be very efficient IMO. (speaking as the MM loving freak that I am, I often run income provinces in positive dominion at 110% until they hit 2 unrest, and then reduce to 100% and let positive dominion reduce them to 0. Then rinse and repeat. Which it seems is roughly what you are suggesting, albeit a less efficient version IMO due to your figures being off, but the basis of your idea is good none the less IMO).


Although it should be noted that there are several events, both good and bad, that are triggered by having unrest in a province (As you indicated in regard to unrest reducing events). Not sure if this can really be factored into an “allowing unrest” strategy or not though. I wouldn’t completely rule it out, but I can’t imagine it ranking anywhere high on an efficiency scale. Those interested in the mechanics behind the random events should follow Edi’s signature links to the list of random events and their triggers for more info on the subject. (Edit - link to event list - http://forum.shrapnelgames.com/showthread.php?t=45185)



And I hope you will all please excuse this intrusion, especially if it was unwanted (it's one of my many bad habits as I said), and hope you both keep up all the good newbie education that you are doing

[triqui]

Thanks for the corrections, enterteining read. The admin in the capitol is quite good advice, I overlooked that. Certainly, Man would probably lead just becouse of that, even with average scales. Good advice.

It has been a lot since the last time I read through the math of the game, so my memories might be wrong. However, what you says makes me remember it was so.

Just that I thought 5 unrest was the right move regardless, becouse 4 unrest gives you -8% income, while you get +10% from having 110% taxes, and that's why I decided to stop once you get to 5 unrest and then swinging it back. Probably it's not the best way to do, though, but having some extra gold in the begining is better than having some extra gold in the long run. That's why you pay an interest for a credit :P