List of possible AI improvments

[Jimmer]

ORIGINAL: Murat
How did a fleet land a corps when the ports with a fleet and a corp were blockaded?

It's not possible to absolutely blockade all of the French ships, at least not in the boardgame. GB can do a pretty good job, but it's not absolute.

Anyhow, though, I misread your original post. But, part of the problem remains because France can put corps on the map every turn in various ports. This forces GB into constantly switching ports. And, since they can't do it absolutely, they leave an opening.

[bresh]

ORIGINAL: Jimmer

ORIGINAL: bresh

Gibraltar has no value beside a safe port in this version of the game.

Oh, yes it does. It's a safe port from which a sizable British fleet can monitor entrances and exits to and from the Med. You never know when an enemy fleet might ...

Im not sure what you mean ? Its just as i wrote ?
"Gibraltar has no value beside a safe port in this version of the game."
In the board game, GB could use dominat status if lost this and couple more.


Regards
Bresh

[Adraeth]

i notice that the inactive AI is mainly focused when the objective is at a great distance. In a solo play with Prussia (against easy AI) i see smart moves by the France AI against me that forced me to accept a conditional peace.
 
Even the AI vs AI (Austria Vs Spain) in Italy shows really interesting moves, anyway, in the far east the Turkey and Russia seem to be more inactive after a while,and GB sometimes just stands.

[peskpesk]

[font="times new roman"]18) The AI should create one (or if having a strong army two/three) ”killer stack” (stack optimized by good leader, mix of forces etc) more often and try taking out enemy stacks.[/font]
[font="times new roman"][/font] 
[font="times new roman"]19) When advancing/on the offensive, the “killer stack” should be flanked by flanking forces to minimize risk of being cut of from supply, reinforcements etc.[/font]
[font="times new roman"][/font] 
[font="times new roman"]20)  The AI should try to use a expendable, outflanking force more often, to hit isolated garrisons, supply depots, week crops and “annoy” the opponent, forcing him to react on it.[/font]
[font="times new roman"][/font] 
[font="times new roman"]21) The AI should be wearier on facing the double move, (France/ Great Brittan moving last and then first. When calculating the risk of being attacked. Too big risk is taken now.[/font]
[font="times new roman"][/font] 
[font="times new roman"] 22) The AI should not fall for the “pull trick” (A weaker enemy fleet/corps is in range of the AI. The AI moves a force there and attacks it, just to be pulverized the next move by the enemy’s waiting “killer stack” that waits out of range)[/font]
[font="times new roman"][/font] 
[font="times new roman"]23) The locations of the major generals for each nation. Napoleon, Wellington, Blucher etc should war the AI of possible increased risk of war with that nation or increased risk major offensive taking place there.[/font]
[font="times new roman"][/font] 
[font="times new roman"]24) When faced with a two front war (or more), the AI should decide on a strategy. The strategy should be re-evaluated each turn:[/font]
[font="times new roman"][/font] 

• [font="times new roman"]      Try to finish of one opponent and ignoring the other(s) at the risk of facing high cost at the other fronts.[/font][/align]
• [font="times new roman"]      Decide to face them all and divided his forces among the fronts and for each front decide a strategy for the front.[/font][/align]
  • [font="times new roman"]a.       Hold the ground[/font][/align]
  • [font="times new roman"]b.      Retreat[/font][/align]
  • [font="times new roman"]c.       Be on the offensive[/font][/align]
  • [font="times new roman"]d.      Ignore the front due to lack of forces or threat from the opponent.[/font][font="times new roman"] [/font][/align]

[font="times new roman"]25)  When at war try to buy of difficult/strong opponents with if possible “cheap”/quick conational or informal pace. Face the easier opponents in battle.[/font]
[font="times new roman"][/font] 
[font="times new roman"]26) Avoid futile declaration of wars. When declaring war on minors the AI should calculate the risk of the likely controlling major power is at war with another major that in turn moves before the AI nation and has quick access to the minor. (Ex Great Brittan is at war with Russia. The Turkish AI declares war on Georgia. Russian forces moves in and assaults the capital before turkey has the chance. Due the fact that Russia is at war with Great Brittan that controls Georgia and moves before turkey.[/font]
[font="times new roman"][/font] 
[font="times new roman"]27) The Turkish AI should take more use/risks with the feudal forces and “bleed” them in combat/force marsh/invasions etc if December is closing by and the capital of the feudal force is expected to be friendly controlled when renewal phase happens. [/font]
[font="times new roman"][/font] 
[font="times new roman"]28) Avoid declaring war in winter turns. Due to high supply cost.[/font]
[font="times new roman"][/font] 
[font="times new roman"]29) Avoid being in low forage areas in winter turns if no supply depots/chain is present and the AI has enough money for the cost.[/font]
[font="times new roman"][/font] 
[font="times new roman"]30) Avoid being at sea when Fleet maintenance phase is coming up.[/font][font="times new roman"]                      [/font]

[Grognot]

Turkish AI seems relatively passive. Last game I played (as France, heh), Turks never took anything besides Egypt, through 1814. 'War' with Russia with neither side attacking the other.

UK AI, if it's moving fleets into the Channel to stop a French assault, against French ships that are already there, should not suicidally block a Fr/Sw/Du stack with just one 10-ship fleet at a time.

UK should perhaps heavily bribe the Swedes ASAP, to counter a French bribe -- the Swedish fleet is a nice addition to the French one, and Russia is likely to DoW Sweden by March.

AI should be more willing to break a blockade if there's a much more powerful army set to imminently seize the port. Better to lose some than all.

A severely outmatched AI that's at risk of losing a lot of minors and then being invaded should consider surrendering and offerinng to cede those minors, rather than losing those minors pre-surrender and then having additional territorial losses from the surrender itself.

(Edit)
Turkish AI, if allied with Fr or UK, should occasionally request money. They could use it much more than the access that they constantly ask about until granted.

[nappy]

I agree the need for the AI to use stacks is required. I find that the AI like to build big stacks but when it comes to attacking it just sends piddly little waves of understrenght corps. MY best HUH? moment was when Napoleon with 1 corps with 1G, 5I attacked the British in Amsterdam with Wellington and 4 corps.

Yes Turkey almost never builds stacks or really moves more than 1 or 2 corps - ever- that needs fixing.

The AIs need to defend their capitals more pro-actively. It is just too easy to do a cheapie and grab a capital. Half the AI wars seems to consist of stacks sitting off doing nothing while smal; one corps armies keep taking runs at each other's capitals. For example almost anyone can bag constaniopole without effort.

the AI should also not ignore minor neutrals. I often find the AI's are usually so gung-ho to take capitals that they often will ignore undefended enemy minors nearby. In one war AS, PR, GB, SP, RS  all were at war with france, but no-one bothered to take teh 4 or so undefended French Italin minors - giving france a continued economic boost.

The IA has to place and defend its fleets more agressively. France almost always sets up its whole fleet in Brest with a  whopping 1I and no corps within a light year. Its a free lunch for GB.  The idea is that you have to force GB to spread out thin and not make these fleets attractive targets - which also prevents GB from grabbing minors like crazy early on.

Naps

[nappy]

A couple of more Suggested specific AI tweaks
 
*Russia* It seems to have a penchant for making its entire army stack in St. Petersburg. Usually by late 1807 - even if its at war with 3 powers it still keeps its entire army there. I am yet to see a a large Russian presence anywhere else. ALso it seem sto DoW on sweden grab Finland and then leave Sweden to the owner via LoW. [&:]
 
*Turkey* It sets up corps especially feudals in Bessarabia and Moldavia and NEVER moves them. Besides taking Egypt - doesnt do much else.
 
Neither RS or TU seem to even bother with the Caucasus area minors.
 
*Spain* I've seen the AI spain try to take territory in Italy - good! But it has far too much of penchant for just taking portugal and then keeping its entire army in madrid- even when at war. 
 
Hence it seems to me that Spain and Russia are far to paranoid about their capitals (the two that should be the least worried) while France, Turkey and Austria often just let it fall to piddly raids.
 
I have to re-iterate the point I made before again!! The AI far too often takes single coprs (sometimes with leaders) and takes suicidal runs at enemy capitals usually way out of suplly and force maching .. constantly. Its offensives seem to be one-two corps armies. It does seem to form defensive stacks but usually way out of way somewhere or as an over-reaction to something small.
 
Naps    
 

[easterner]

AI Analysis

GB- Navy   generally good. Guards Dover crossing arrow as #1. Foe somereason lets up on Fr. blocade, on 1st game as Fr it let up blocade, moved all to Dover/  Fr invades Portsmouth. GB surrendered.

Other navies: generally good but when a war ends tends to abandon fleets off old enemy coast....for years. Pretty expensive.

AI: Prussia gets it done, other countries need Pr. AI Lessons.
AI tends to not grab minors. Pr. does. Examples. Naples entered Papacy, stopped in Rome, moved out of Rome and never took it, I did 6 mos later while it sat there. In last game Alex entered Circassia early 1805, game ended in 1813 he was swtill there, unmoved one area from capital, never took it.

AI Oddities: Russian Corps sit around Lenningrad like it's 1942. Declaring war on Turkey with entire army at St. Pete's leads to surrender. As others mentioned in never enters Sweden only Fin. and doesn't rebuild cossacks.

GB lands in Oporto, never moves on Lisbon. Throws single Cps away over crossing arrow, never having used this board game option I had failed to register its presence for awhile in my 1st game as Fr. So co-ordinated would have been nasty while I was napping.

SP invades Algeria, builds army by Madrid, never moves.

Au: moves but peacemeal, took Constantinople from me and forced my surrender in 1806. Piecemeal methods though would have failed against stronger power. Needs concentrate better and use Charles who sits near Italy while Au loses in Germ. needs more aggression to grab minors.

Tu. has 5 Cps in Palestine, fails to attack 1 Egyptn Cps. too passive.

In general: leaders with one Cps put in risky places. As GB I captured Nappy, when peace came it cut him off in N. Central Ger with 1 Cps. I entered that minor and overwhelmed his 1Cps of Gds.

When AI detects cps moving in great numbers they should retreat or bring in reinforcements, not leave 1 Cps to die alone.

In 1st game took N. Spain; no sign of Guerrillas, only assigned 1 Cps to this holding action later re-inforced to 2. Since Sp had declared on Fr. They should have been better prepared rather than let Pr go down tube, which they did. Pr put 2 hard hits on my Fr wiping out 2 armies before Nappy got there and set things aright.

[Froonp]

Well, I don't know if it was posted, but here are two improvements that the game could have, from my friend Jérôme, who does not write English, and who bought and is playing the game. He played the French.

- The British should not send small troops in France peacemeal, because he could defeat them is small groups and now Britain is in dire straits in terms of Army.

- The Russians when they go for Sweded should go by sea, and not struggle to go by land and reach it in a very long time.

[nappy]

ORIGINAL: Froonp

Well, I don't know if it was posted, but here are two improvements that the game could have, from my friend Jérôme, who does not write English, and who bought and is playing the game. He played the French.

- The British should not send small troops in France peacemeal, because he could defeat them is small groups and now Britain is in dire straits in terms of Army.

- The Russians when they go for Sweded should go by sea, and not struggle to go by land and reach it in a very long time.

Yes the piecemeal nature of many AI offensives needs to be looked at. It needs to learn how to offensively use a stack.

I also agree about Sweden; Russia has to be more agressive with Sweden. Its essentially always ends up a FS for France; which as human a player I would just throw into the sea war in the Channel. Russia also needs to stop covering in St Petersburg with 80% of his army.

Naps

[carnifex]

Here are things the AI should NEVER do:

Have the lone Egypt corps attack the Turkish force in Palestine. This happens if Turkey attacks Egypt and Russia gets control. Since Russia goes first, it attacks the Turkish stack.




Blockade 12 Heavys and 22 Lights with 7 Lights. This happens every game. First turn Russia declares on Turkey and then sends their piece of shit light fleet to hand Turkey a PP on a silver platter.

[Mynok]

I played Turkey and France last night against the AI and saw the fleet thing for sure.

[Murat]

I have been toying with the concept of lifting the FoW for the AI. In other words, the AI would know which stacks had the highest troops, etc. To make an AI based on basic principles, such as Grant's strategy for defeating the Confederacy - attack the armies by applying the most troops against a smaller force to eliminate it and ignore cities except for the 2 forts anchoring the South, Atlanta and Richmond, may actually be achievable but would require the AI 'knowing' that the 6 full corps with Nappy are more important than the 12  1 factor corps with Bernadotte. Any thoughts?

[AresMars]

With all due respect to the comments made so far, this is a game to be played between 7 PEOPLE and the AI is not really (and rarely is) an effective opponent.
 
If develloper time is going to be spent, I would prefer attention be spent on improving the play experience for 7 human players and removing game bugs then making SOLO play better.....there are other games for that experience.
 
I played EiA f-t-f for years and always enjoyed the experience...time flew past, availibility became near impossible, our group got older, got married, had kids (sigh!gasp!) and time became a premium - this computer game is a chance for our group to play again and I would like to see the experience improve further.
 
Players' make mistakes and I can live with that, as long as the computer game makes it easier for 7 PBEM players to play.....
 
AresMars
 
 

[WYBaugh]

ORIGINAL: AresMars

With all due respect to the comments made so far, this is a game to be played between 7 PEOPLE and the AI is not really (and rarely is) an effective opponent.

If develloper time is going to be spent, I would prefer attention be spent on improving the play experience for 7 human players and removing game bugs then making SOLO play better.....there are other games for that experience.

I played EiA f-t-f for years and always enjoyed the experience...time flew past, availibility became near impossible, our group got older, got married, had kids (sigh!gasp!) and time became a premium - this computer game is a chance for our group to play again and I would like to see the experience improve further.

Players' make mistakes and I can live with that, as long as the computer game makes it easier for 7 PBEM players to play.....

AresMars

With all due respects I think you'll find the majority of players are looking to play this computer game solo. I know there are the sides who want nothing more than a human opponent but I feel from my experiences and my continuous lurking here at Matrix that folks want to be able to play the game when they want too and not rely on 6 other humans.

Marshall has stated that the majority of development went into human vs human playing and not the AI. If they want to make it more accessible to us solo players then it's time for AI tweaking and not just PBEM focus.

Bill

[DodgyDave]

better ai will come, i am not worried about that part, i just want the first patch, so to remove the serious bugs and then hope they will eventually make the game more like the boardgame, like the boardgame turn sheet for each nation, the status sheet is so nice and so on.

but fix bugs, then solve other issues like adding up more options, so each group can decide on what to play with

[AresMars]

WBAUGH, your point is very valid.  Don't get me wrong.....

This game is a re-creation of a boardgame that requires 7 people to maximize the experience -- the game was not developed for SOLO play and thus has a poor AI for that.  You (and anyone else who buys the game) are a better player then ANY AI we could ever hope for.....

I would hazard a guess that once the issues of bug fixes and more options are resolved, SOLO players would find a much better game value and play experience out of PBEM games....IMHO, NOTHING beats playing _with_ and crushing human opponents......

I expect the number of PBEM games to expand dramatically once the first couple patches come out.

You cannot please everyone - I was just mentioning MY particular hope of where the focus of Matrix should (in my opnion) really be....

Like you, I spent the money to get the game -  I worry that the expectations of SOLO players may not be easy to answer in the short term and also feel that Matrix needs to remind potential buyers that this not really a SOLO computer game.

Is that fair?

AresMars

NOTE: I notice several spelling errors and I appologize for my poor english...and I am too lazy to correct them... [:D]

[WYBaugh]

ORIGINAL: AresMars

WBAUGH, your point is very valid.  Don't get me wrong.....

This game is a re-creation of a boardgame that requires 7 people to maximize the experience -- the game was not developed for SOLO play and thus has a poor AI for that.  You (and anyone else who buys the game) are a better player then ANY AI we could ever hope for.....

I would hazard a guess that once the issues of bug fixes and more options are resolved, SOLO players would find a much better game value and play experience out of PBEM games....IMHO, NOTHING beats playing _with_ and crushing human opponents......

I expect the number of PBEM games to expand dramatically once the first couple patches come out.

You cannot please everyone - I was just mentioning MY particular hope of where the focus of Matrix should (in my opnion) really be....

Like you, I spent the money to get the game -  I worry that the expectations of SOLO players may not be easy to answer in the short term and also feel that Matrix needs to remind potential buyers that this not really a SOLO computer game.

Is that fair?

AresMars

Absolutely fair!

[pzgndr]

This game is a re-creation of a boardgame that requires 7 people to maximize the experience -- the game was not developed for SOLO play and thus has a poor AI for that.

This is incorrect. The original board game may have provided the maximum experience for 7 people if available, but the UMP rules were specifically designed to allow less than 7 people to also have an enjoyable game experience. Yes?

The computer game replaces UMPs with AI, therefore anyone who played with UMPs before will be playing with AI now. Was the game developed "for" solo play? Not necessarily, but with functional AI and adjustable difficulty setting for each major power, solo against 6 AI opponents becomes a viable option now whereas it was not before. So it's unfair to dismiss this new option.

The quality of the AI is of course subject to debate. But read the forum comments; Marshall Ellis and Matrix Games acknowledge weaknesses with the initial AI and plan to make improvements. 'delatbable' provided a nice solo vs AI AAR which demonstrates that even with its weaknesses upon release, the AI did "reasonably well." Once the game is more or less finalized with bug fixes and other issues resolved, then the AI can be enhanced for better performance... for 1-6 AI players for any of the major powers.

If develloper time is going to be spent, I would prefer attention be spent on improving the play experience for 7 human players and removing game bugs then making SOLO play better.....

Well, fortunately that is not the plan Matrix has in mind. [:)]

[bresh]

Is it just me or does the movement algorithm not always use the "cheapest-forage" clear-route ?
 
Example walking from Paris to Strasburg, program chooses to walk through the 3 forage forest s of strasburg, instead of the town hex 4-forage west of.
 
Regards
Bresh