http://www.johntillersoftware.com/SquadBattles/FirstWorldWar.html
I haven’t bought the game but the design notes ( at the bottom of the page) are just one of the most interesting article I’ve read about WW1 tactics. Design notes: the moment where a designer explains his game, rather than reply WAD 
An exercise of humility.
here, for 5 $:
http://www.gog.com/gamecard/imperialism
I could make mine this comment, comparing Imperialism to Paradox production, that may be extended to other current games :
Every time I play it, I wish we could pass legislation requiring all game designers to study it.
1. The interface is clean and everything is interconnected.
2. You research something, you get to build it. You build it, it produces tangible product. If you build a lumber mill, it’s so much more satisfying to get wood instead of a 0.000375% increase in some hidden production value. (I wish the Paradox guys had been awake during Imperialism class.)
2. The game elements are simple but challenging: six or so commodities, two of this makes one of that, and Worker A needs this food and Work B that food. Simple. But really hard to fulfill over the course of game – just when you’re ready to take off, some critical shortage always creeps up on you. (Compare to Paradox games. In EUIII, for example, is there a single commodity that you really NEED? Apart from tiny, largely invisible increases in income, why trade at all?)
3. Same with the transport system. Developed a nice gold mine overseas? Great, now all you have to do is transport the gold back home. But, wait, you need those same ships for trade. Or that grand invasion you’ve been planning. You suddenly need more ships, which means you suddenly need more cloth, which means you suddenly need more cotton….which is also overseas, which means you need more ships…. (Again, I’m talking to you, Paradox. A game mechanic, if well conceived, can be very simple and still very challenging.)
4. I even like the combat (though I may be in the minority on this). Each unit plays a specific role. Tactics actually matter. And I get to do it myself. (Instead of just watching sprites stand on the map stabbing each other (for weeks!). Another Paradox staple.)
5. Finally, turned-based beats RTS for strategy games. (RTS, with its emphasis on speed over thought, is too often just camouflage for a weak AI.)
My own view is that Paradox has taken us about as far as we can go in the spreadsheet as game genre. Here’s hoping we can press the reset button a bit and revisit the elements that made games great in the beginning, before mega-computers spawned game bloat, when game designers had to search for elegant abstractions. That may be the best impact of the mobile platform. (I actually think CK2 is an improvement on the typical Paradox model, though they seem intent on adding yet more layers of questionable complexity with each DLC.)
Thanks. I needed to say this. I feel better now. 
If one day I design a game, it will be on the same subject than this:
This one seems to be a real gem.
Interestingly, as for any WW1 game, question about a Russian Civil War surfaced. Someone replied on the Matrix forum:
In the Russian Civil War there were in general two opposite sides – Soviet Russia (the Reds) and all the others. However relations among the others (Whites, Poland, Ukraine, Baltic states, Germany, Finland, Western Allies etc) were complicated. Some of them did not have a solid alliance with others and their actions were not coordinated (or were poorly coordinated) and some even fought each other (for example Poland and Ukraine). I also wonder, if all this complex politics can be modded in this game…
Just the occasion to remind Fatal Years is currently as far I know the only computer game about RCW displaying Freikorps, Ukraine, Poland, France, Galicians and Anarchist as independent factions, fighting from their own, out of players control.
for 19 $ you may access to one hundred back issues of Strategy&Tactics, Moves and World at War mags ( without the games).
http://shop.strategyandtacticspress.com/SearchResults.asp?Cat=29
Contents are uneven of course and don’t allow to bypass real book readings. However, there are a lot of general presentations of a period, a war or a campaign, often obscure, with some maps.
The files are pdf freely downloable during a full year . I’m reading them on a tablet and my phone.
And the role of cavalry….How the current AJE battle system is simulating this?
http://jonshaferondesign.com/2012/09/10/the-recipe-for-good-ai/
3 points I fully agree:
The AI has failed if a player considers it to be either 1) too random, or 2) dumb.
What the AI does is less important than what the AI seems to do
A designer needs to spend significant time and effort establishing what the goals and focus should be for all systems – the AI is no exception. Simply handing off this to a programmer is usually a recipe for disaster. The job of a programmer is to write code that is efficient, robust, and easy to maintain – it’s a designer’s job to ensure the in-game experience is fun. Those goals do not naturally align. When no direction is given, programmers will typically architect and code a system just like they always would.
The third is very relevant for current AGE system. You can’t design a real good game without addressing AI.
When I will have time, I will test these modified values:
// ********************************************************
// ***** COMBAT *****
// ********************************************************
cbtHitCoef = 600 // is the % chance, in 1/100 (400 meaning 4.00% here) to hit the enemy, for each final firepower point.
cbtAsltCoef = 30 // same, but for assault
cbtMinToHit = 15 // is the minimum % to hit someone (rising this value help the side with the lower quality units)
cbtNbRoundsPerDay = 6 // Number of rounds per day
cbtProtCoef = 90 // coefficient (in 1/100) for each protection point (3 pts = 0.9 x 0.9 x 0.9 damages sustained)
cbtLeaderBonus = 30 // % bonus for each point of stat of the stack leader
cbtUnitLeaderBonus = 15 // % bonus for each point of stat of the unit leader
cbtLdrCasuDiceSides = 500 // Nb of sides for the casualty dice for rank 1 leaders, at the end of combat (twice as low for rank 2, rank 3 excluded)
cbtMaxEntrenchLevel = 2 // Max absolute entrenching level
cbtMaxEntrenchNotArt = 2 // Max entrenching level, if the stack has no artillery (high levels don’t benefits non artillery anyway)
bmbMinEntrenchLevel = 2 // Minimum level to reach by an artillery so that it can bombards or interdict ships& water supply
cbtCohBonusrAutoRetCoeff= 50 // bonus to cohesion as per: autoretreat% = BasePercent * (Coh + bonus) / 100
cbtTQModifierForAssault = 100 // default is 80
cbtMenPerConscriptPt = 500 // each conscript point represents 500 men
cbtChanceDumbHit = 10 // 10% chance to make a dumb hit against naval target
cbtChanceCriticalHit = 10
cbtDamageDumbHit = 50 // damage if lighthit (real dumb hit: 1 dmg & 1 coh)
cbtDamageCriticalHit = 200
I don’t believe it will suffice to solve the whole thing and worse, there are great chances these values will give absurd results. However, only those trying nothing get nothing 
a 120,000 men Army? In Ancient times? 
