Revolutionary Thoughts

[Erik Rutins]

ORIGINAL: NeverMan
The only problem I see here is that the AI play is never going to be good, so at what point do you guys stop working on the AI and start working on making the PBEM (or heaven forbid, internet play) better?

I think "never" is a pretty strong word. I think a lot of people who don't have time for a PBEM just want a competent AI opponent and I'm sure we can get there. No AI has ever been as good as a good human opponent, but I think for many having it play a decent game without any bad mistakes will give them the enjoyment they're looking for.

Also, improving PBEM play is a parallel priority for us and Marshall has been spending a fair amount of time looking into PBEM issues and fixing them. Finding a way to streamline PBEM (ideally as an optional "fast PBEM" system that players can choose) is high on our priority list of PBEM improvements, but there are definitely some limitations on what we can and can't do in a reasonable amount of time.

Internet play would be great, but let's take this one step at a time. First, bug fixing and solidifying the existing AI and PBEM play, then we can go further afield and try to satisfy both those who want "classic EIA" and those who want further evolutions and more streamlining.

Regards,

- Erik

[j-s]

ORIGINAL: bresh
Most of my friends are still waiting to buy it, since they dont think its enough EIA yet.
So if you start making devarations, i doubt those ever come and buy.
There are other war games you know, who satisfy other wishes.

I agree with you. Two days ago 4-5 of my friends asked, that "is it still something else than Classic EiA?". They won't buy game before they can get a classic scenario. Before that they will play something else.

If classic scenario is made some day, then we can start PBEM with our old group, many are waiting for this. Before that, I have to play against AI and wait for classic scenario (with original map, dominant powers, corps strengths and naval rules ect.). Until that I will play by PBEM games via Cyberboard.

[MarshallEllis]

ORIGINAL: NeverMan

ORIGINAL: Erik Rutins

I don't think there's any harm in discussing options, as long as you all keep in mind that our top priorities remain:

1. Bugs
2. Things that hinder vs. AI play
3. Things that hinder PBEM play

After that, there are things like the game itself, for example the scenario editor, making any "deviations" we can either optional or creating a EIA Classic scenario that avoids them. Obviously, some of these things happen in parallel but discussing something doesn't mean that it's suddenly become the top priority.

Regards,

- Erik

The only problem I see here is that the AI play is never going to be good, so at what point do you guys stop working on the AI and start working on making the PBEM (or heaven forbid, internet play) better?

The AI will be good (already better in this current release) but it will "never" be like your buddy across the table. To coin a phrase just mentioned yesterday "Humans make better mistakes". Emulating human flaws in an AI system is tricky. AI's are much better at "black and white" issues than they are the grey issues.

I hear you on the PBEM / IP side and you are not alone BUT honestly these communities (pbem vs solo vs IP) are about the same sizes (Based on my communications) but that is how it is today. These two genres are not always different nor are they always the same all of the time ... if that makes sense???

In summary, nobody ever said that you can't ask for both

[NeverMan]

ORIGINAL: Marshall Ellis

ORIGINAL: NeverMan

ORIGINAL: Erik Rutins

I don't think there's any harm in discussing options, as long as you all keep in mind that our top priorities remain:

1. Bugs
2. Things that hinder vs. AI play
3. Things that hinder PBEM play

After that, there are things like the game itself, for example the scenario editor, making any "deviations" we can either optional or creating a EIA Classic scenario that avoids them. Obviously, some of these things happen in parallel but discussing something doesn't mean that it's suddenly become the top priority.

Regards,

- Erik

The only problem I see here is that the AI play is never going to be good, so at what point do you guys stop working on the AI and start working on making the PBEM (or heaven forbid, internet play) better?

The AI will be good (already better in this current release) but it will "never" be like your buddy across the table. To coin a phrase just mentioned yesterday "Humans make better mistakes". Emulating human flaws in an AI system is tricky. AI's are much better at "black and white" issues than they are the grey issues.

I hear you on the PBEM / IP side and you are not alone BUT honestly these communities (pbem vs solo vs IP) are about the same sizes (Based on my communications) but that is how it is today. These two genres are not always different nor are they always the same all of the time ... if that makes sense???

In summary, nobody ever said that you can't ask for both

Marshall, I understand you must work on the AI for market reasons, that's great, good for you and Matrix.

That said, the AI will never be good, period. It just won't. One developer coding/designing 1/2 of his time to an AI of this magnitude (the learning curve on this game is VERY HIGH complexity) will not yield a solid AI.

Some would argue that it is good now simply because you have pumped PP and VP into it, which I whole heartedly disagree. I don't think by simply giving the AI more PP and VP that it makes it good, call me crazy. It won't be "good" until it can mimic and "average" human player, one that understands the game, can make good moves but also makes some bad moves.

If you really want to see how good the AI truly is, strip away the PP and VP pumping and then play. This will reveal the AI for what it is. I don't think most, or anyone for that matter, would consider it good.

Unfortunately, it also seems that, for some reason, everytime a new patch comes out it fixes some old game crash bugs but also it introducing a few new ones, almost like a 1:1 ratio.

Maybe it would be best for Matrix to call this one Empires in Harms and start a new development that will be Empires in Arms? I don't really know.

[Jimmer]

ORIGINAL: Marshall Ellis
The AI will be good (already better in this current release) but it will "never" be like your buddy across the table.

Actually, I believe gwheelock has played in a couple games where the real, live humans were no match for your AI. (And, no, G, we don't need their names posted again.)

I've also seen some aspects of the AI that have the seeds of greatness built into them. The problem is that to REALLY have a kicking-AI, you have to incorporate some fuzzy logic and other game theory aspects. Unless you have training in mathematics at the 3rd and 4th year college level, that would be hard to pull off. Look at how long IBM had to work before they could finally defeat all human Chess players. They had a staff of thousands working on that project for over 30 years before it happened (probably longer -- I only know for sure about 30). And, Chess is an extremely simple game compared to EiA.

[pzgndr]

That said, the AI will never be good, period. It just won't. ... I don't really know.

You're right about the second part. You don't really know. [8|]

[Mardonius]

Jimmer's comments illuminate what I meant by diminishing marginal returns. At some point the AI is going to be about as good as reasonably expected. Each marginal return would demand a heavier investment of time (=money) that could be perhaps better spent on some variant development.

Don't neglect reasonable AI improvements/bug reduction, but keep other options open when that brick wall has been reached.

Some of these variants could be from the "good book", i.e. omitted variants from the original EiA game and others could be more HERETICAL options like those I previously noted, such as:

Minor Country Improved Morale with training. (Yours truly wrote his MA thesis on this little matter for the Portuguese. Moreover, I have had some first hand exposure to training/modernizing certain armies myself.) The old AH rules allowed for an optional rule the Brits to improve the Portuguese and the Hanoverians to a 4.0 morale with 24 months of continuous training. I would love to see other nations have the option of training them as well to say 0.5 below their own morale. So the French to 3.5 and the Spanish et alias to 2.5. I would expand this to other nations as well up to a certain limit. I would allow the gaming group to select what countries (say Holland, Denmark for example) could be trained and to what level and by whom. AND FOR THOSE WHO SAY THIS NEVER HAPPENED, REMEMBER (1) BARON STEUBEN IN THE US (2) FRANCE AND THE PONDICHERRY SEPOYS (3) BRITAIN AND THEIR INDIAN ARMY (4) EUROPEAN (MOSTLY FRENCH)OFFICERS AND THE TURKISH NEZAMI CEDID (5) FRENCH TRAINING OF MANY OF THEIR MINORS AND THE TRANSITION TO SKIRMISHER AND COLUMN TACTICS (6) THE LATE WAR SPANISH ARMY... Same holds true for other eras, including the current one.

Increased Corps Size. Allow nations to spend money and time to increase certain of their Corps. Must be in a step function and within reason.

Increase number of corps a nation can produce. Start with their smallest and perhaps could be improved as noted above.
Could apply to some minors as well.

Increase number of depots with expenditures of time and money.

Increase number of fleets with expenditures of time and money.

Allow certain minors (e.g. North Africa) to build light fleets.

Improve fortifications or harbor guns with expenditure of time/money.

I would also consider having a portion of the Privateer monies go to the privateering state. This was the common practice of the day.

added option not from original post: limit use of North African Manpower to the Turks as no other power was able to use their manpower despite repeated occupations

added option: not from original post: Increase guerrilla geographies to reflect historical realities (Portugal, Naples, Sicily, much of North Africa)

A scenari editor could allow us to change this one camapign to a series of campaigns covering from around 1700 to around 1860. Think of the possibilities (and customer base, which equals $) that could be gained by having a Seven Years War Scenario or a War of Italian Unification campaign.

Yes, we do need the Classic EiA scenario. Just don't forget that perhaps this same engine can pull a whole lot of other weight if marketed correctly.

Just some more heretical thoughts.

Semper Fidelis,


Mardonius the EiA Apostate

[NeverMan]

ORIGINAL: Jimmer

And, Chess is an extremely simple game compared to EiA.

Are you being sarcastic?

There is nothing "simple" about Chess other than learning how the pieces move.

pzn: I guess I really don't know, but I'm sure we can all wait and find out.

[gwheelock]

ORIGINAL: NeverMan

ORIGINAL: Jimmer

And, Chess is an extremely simple game compared to EiA.

Are you being sarcastic?

There is nothing "simple" about Chess other than learning how the pieces move.

pzn: I guess I really don't know, but I'm sure we can all wait and find out.

Actually; Jimmer is right. I can write the rules of chess on a single piece of paper, the EIA rulebook is 48 pages.
Chess never had a rule 8.2.1.2.2.2.2 (I think thats the deepest drill-down in EIA).
Chess has 6 types of unit (King, Queen, Bishop, Knight, Rook, Pawn); I'm not even sure
just how many EIA has if you count all the various morale variations; 0, 1 or 2 step guard comit options, etc.
Chess is completely deterministic game (ie. there are no dice rolls to screw up "the best laid plans...")

And yet; after 30 + years (probably closer to 40), millions of $ and tens of thousands of manhours of effort by some of
the brightest minds in academia & industry - we still do not have a complete "map" of the
chess universe & a chess program that can beat a human grand-master is very recent
phenomana (& THAT was only managed by applying enough brute computer horsepower
to be able to look at thousands of result sets that would appear after just a couple of <potential> moves)

[SamuraiProgrmmr]

Perhaps the more complicated the rules, the easier it is for a computer to beat a human.
&nbsp;
For example, try beating a computer program that playes tic tac toe.
&nbsp;
There was one computer game published that had a very strong AI.&nbsp; It was a port of Steve Jackson's OGRE.
&nbsp;
The reason its AI was strong was that the situation was simple enough to use Brute Force to calculate the possibilities far enough ahead to find near perfect or perfect play.
&nbsp;
I would also point out that there have been chess computers for many many years that could handle 90% (conservatively) of the players out there.
&nbsp;
So, again, the pattern holds.
&nbsp;
Just my opinon on AIs in general.&nbsp; Not making any comments about the AI for this game.
&nbsp;
&nbsp;

[NeverMan]

ORIGINAL: gwheelock

ORIGINAL: NeverMan

ORIGINAL: Jimmer

And, Chess is an extremely simple game compared to EiA.

Are you being sarcastic?

There is nothing "simple" about Chess other than learning how the pieces move.

pzn: I guess I really don't know, but I'm sure we can all wait and find out.

Actually; Jimmer is right. I can write the rules of chess on a single piece of paper, the EIA rulebook is 48 pages.
Chess never had a rule 8.2.1.2.2.2.2 (I think thats the deepest drill-down in EIA).
Chess has 6 types of unit (King, Queen, Bishop, Knight, Rook, Pawn); I'm not even sure
just how many EIA has if you count all the various morale variations; 0, 1 or 2 step guard comit options, etc.
Chess is completely deterministic game (ie. there are no dice rolls to screw up "the best laid plans...")

And yet; after 30 + years (probably closer to 40), millions of $ and tens of thousands of manhours of effort by some of
the brightest minds in academia & industry - we still do not have a complete "map" of the
chess universe & a chess program that can beat a human grand-master is very recent
phenomana (& THAT was only managed by applying enough brute computer horsepower
to be able to look at thousands of result sets that would appear after just a couple of <potential> moves)

You're both wrong because you are equating the difficulty of the game with the amount of rules. Simply because a game has many many rules does not make it difficult and vice versa.

Jimmer was saying that EiA is a much more complex and harder game to master than Chess, and I simply disagree. GO is another game with very simple rules yet is extremely difficult to master.

I honestly can't believe that anyone thinks otherwise.

[gwheelock]

No; what Jimmer & I are saying is that the programming difficulty of a game
is proportional to the number of options in the game (rules, counter types, etc).
&nbsp;
Each one of those rules has to be incorporated into the code.&nbsp; (Rules ARE
code - its just code for humans instead of machines) & to create
an AI that can handle all of those options is also proportional.
&nbsp;
And I DO believe that to MASTER (at the level equivalent to a chess
grand-master) EIA IS more difficult than Chess or GO - its just that
we don't HAVE an "EIA Grand-Master" level of player

[NeverMan]

ORIGINAL: gwheelock

No; what Jimmer & I are saying is that the programming difficulty of a game
is proportional to the number of options in the game (rules, counter types, etc).

Each one of those rules has to be incorporated into the code.  (Rules ARE
code - its just code for humans instead of machines) & to create
an AI that can handle all of those options is also proportional.

And I DO believe that to MASTER (at the level equivalent to a chess
grand-master) EIA IS more difficult than Chess or GO - its just that
we don't HAVE an "EIA Grand-Master" level of player

Actually I believe that Jimmer was saying to program the AI in EiA was more difficult than to program the AI in a Chess game. I agree that the more rules in a game the more prog., just not about the AI.

Well, I guess we will have to disagree that EiA is more difficult than Chess or GO because I don't believe that. Is RISK (which I haven't played since I was 12) more difficult since it has more rules than Chess?

[MarshallEllis]

Well for my 2 cents, from a programming standpoint EiANW was VERY complicated! Don't get me wrong, I think the EiA design is awesome and very well balanced but balance comes in different ways.

For example:

Turkey is poor BUT she has the feudals.
Great Britain has few troops BUT she has a huge navy.
Spain is lower middle class but has a decent navy, Spanish gold and guerillas.
bla bla bla

My point is that this game is not a linear balanced system like Risk where everybody starts with the same amount of armies and nations. This creates all sorts of special code for a specific nation i.e. Turkey can take a ONLY a feudal unit on the board and move it to its home province and raise it to full strength BUT only Turkey can do this and ONLY in December LOL! I'm surprised I didn't get a drug problem from those type of requirements!

As far as complexity? You bet! No one can tell me that doing a supply step or eco phase is easier than Risk. This is simply a complex game but these are 2 different things IMO. The game flow is complex because of all the rules but this is not what makes it difficult to master that is where the design comes in. It's simply a great game design.

Anyway, my unsolicited two cents.

[RayKinStL]

ORIGINAL: gwheelock

No; what Jimmer & I are saying is that the programming difficulty of a game
is proportional to the number of options in the game (rules, counter types, etc).

Each one of those rules has to be incorporated into the code.  (Rules ARE
code - its just code for humans instead of machines) & to create
an AI that can handle all of those options is also proportional.

And I DO believe that to MASTER (at the level equivalent to a chess
grand-master) EIA IS more difficult than Chess or GO - its just that
we don't HAVE an "EIA Grand-Master" level of player

While I agree the AI is harder to program in EiA, this is stictly because of the brute force approach computers can take to chess. Even then however, you clearly have no understanding of the YEARS it took programmers from huge companies like IBM to get computers to understand simple things like piece value, positional weaknesses, and how to attack with a plan. On a technical level, most AI for chess prgrams are dumb. They just have a simple calculation of piece value and utulize the massive processing power available nowadays to play a competent game until a human player makes a mistake.

That said, if you HONESTLY believe that mastering EiA would be harder than mastering chess, you are dillusional and have obviously never tried to play chess at a high level. While the "simple rules" make it easier for an AI to be coded and take advantage of brute force, it is also these simple rules and the number of possible moves that make chess FAR FAR FAR more complex than EiA. As both a chess player and an EiA player, I can assure you that while it took a game for me to get to a respectable level of play as far as EiA is concerned, it took YEARS before I really saw vast improvements in my chess game. Further, I sit here now, having played chess for 14 years almost, and I am still a simple class B player. While I really do respect everyone's opinions on these forums, the notion that EiA is harder to master or more complex than chess is simply ludicrous.

[NeverMan]

ORIGINAL: Marshall Ellis

Well for my 2 cents, from a programming standpoint EiANW was VERY complicated! Don't get me wrong, I think the EiA design is awesome and very well balanced but balance comes in different ways.

For example:

Turkey is poor BUT she has the feudals.
Great Britain has few troops BUT she has a huge navy.
Spain is lower middle class but has a decent navy, Spanish gold and guerillas.
bla bla bla

My point is that this game is not a linear balanced system like Risk where everybody starts with the same amount of armies and nations. This creates all sorts of special code for a specific nation i.e. Turkey can take a ONLY a feudal unit on the board and move it to its home province and raise it to full strength BUT only Turkey can do this and ONLY in December LOL! I'm surprised I didn't get a drug problem from those type of requirements!

As far as complexity? You bet! No one can tell me that doing a supply step or eco phase is easier than Risk. This is simply a complex game but these are 2 different things IMO. The game flow is complex because of all the rules but this is not what makes it difficult to master that is where the design comes in. It's simply a great game design.

Anyway, my unsolicited two cents.

Marshall, the whole Risk analogy was purposefully silly (since Risk is such a silly game) for guys like Guy and Jimmer who believe that simply because a game has more rules that it is harder to master, which is just absurd.

Please don't compare Risk to EiA, as I have not. Risk is a VERY simple game that I have had no desire to even look at since I was 12 (started playing EiA at 15).

It is also obviuos that niether Guy nor Jimmer have played Chess even at the most intermediate level.

[RayKinStL]

One other thing I will add on the chess vs EiA debate.&nbsp; Like it or not, the game still comes down to die rolls.&nbsp; I can make the best decisions all game, but if I roll nothing higher than 3s, I wont get very far.&nbsp; Like it or not, EiA is still a game of chance/luck.&nbsp; Yes there are rules that seek to reduce this variable when compared to a game like Risk, but EiA will never possess the intricacy and strategy involved in a game of chess.

[gwheelock]

ORIGINAL: RayKinStL

ORIGINAL: gwheelock

No; what Jimmer & I are saying is that the programming difficulty of a game
is proportional to the number of options in the game (rules, counter types, etc).

Each one of those rules has to be incorporated into the code.  (Rules ARE
code - its just code for humans instead of machines) & to create
an AI that can handle all of those options is also proportional.

And I DO believe that to MASTER (at the level equivalent to a chess
grand-master) EIA IS more difficult than Chess or GO - its just that
we don't HAVE an "EIA Grand-Master" level of player

While I agree the AI is harder to program in EiA, this is stictly because of the brute force approach computers can take to chess. Even then however, you clearly have no understanding of the YEARS it took programmers from huge companies like IBM to get computers to understand simple things like piece value, positional weaknesses, and how to attack with a plan. On a technical level, most AI for chess prgrams are dumb. They just have a simple calculation of piece value and utulize the massive processing power available nowadays to play a competent game until a human player makes a mistake.

That said, if you HONESTLY believe that mastering EiA would be harder than mastering chess, you are dillusional and have obviously never tried to play chess at a high level. While the "simple rules" make it easier for an AI to be coded and take advantage of brute force, it is also these simple rules and the number of possible moves that make chess FAR FAR FAR more complex than EiA. As both a chess player and an EiA player, I can assure you that while it took a game for me to get to a respectable level of play as far as EiA is concerned, it took YEARS before I really saw vast improvements in my chess game. Further, I sit here now, having played chess for 14 years almost, and I am still a simple class B player. While I really do respect everyone's opinions on these forums, the notion that EiA is harder to master or more complex than chess is simply ludicrous.

You say that you learned EIA in 1 game; fine. You can also "learn" chess in 1 or 2 games. What I am saying is that
to TOTALLY MASTER EIA to a level equivalent to a chess master or grandmaster would also take many years. By defintion;
a grandmaster can ALWAYS beat EVERYONE of a lesser rank - ALWAYS. We don't have anyone of that rank in EIA - AT ALL
(not one will ALWAYS be able to beat their opponants). You would be better off for comparison purposes to compare how
long it takes for someone to learn CASUAL chess - not expert ... because that is what most current EIAers are.

[RayKinStL]

ORIGINAL: gwheelock

ORIGINAL: RayKinStL

ORIGINAL: gwheelock

No; what Jimmer & I are saying is that the programming difficulty of a game
is proportional to the number of options in the game (rules, counter types, etc).

Each one of those rules has to be incorporated into the code.  (Rules ARE
code - its just code for humans instead of machines) & to create
an AI that can handle all of those options is also proportional.

And I DO believe that to MASTER (at the level equivalent to a chess
grand-master) EIA IS more difficult than Chess or GO - its just that
we don't HAVE an "EIA Grand-Master" level of player

While I agree the AI is harder to program in EiA, this is stictly because of the brute force approach computers can take to chess. Even then however, you clearly have no understanding of the YEARS it took programmers from huge companies like IBM to get computers to understand simple things like piece value, positional weaknesses, and how to attack with a plan. On a technical level, most AI for chess prgrams are dumb. They just have a simple calculation of piece value and utulize the massive processing power available nowadays to play a competent game until a human player makes a mistake.

That said, if you HONESTLY believe that mastering EiA would be harder than mastering chess, you are dillusional and have obviously never tried to play chess at a high level. While the "simple rules" make it easier for an AI to be coded and take advantage of brute force, it is also these simple rules and the number of possible moves that make chess FAR FAR FAR more complex than EiA. As both a chess player and an EiA player, I can assure you that while it took a game for me to get to a respectable level of play as far as EiA is concerned, it took YEARS before I really saw vast improvements in my chess game. Further, I sit here now, having played chess for 14 years almost, and I am still a simple class B player. While I really do respect everyone's opinions on these forums, the notion that EiA is harder to master or more complex than chess is simply ludicrous.

You say that you learned EIA in 1 game; fine. You can also "learn" chess in 1 or 2 games. What I am saying is that
to TOTALLY MASTER EIA to a level equivalent to a chess master or grandmaster would also take many years. By defintion;
a grandmaster can ALWAYS beat EVERYONE of a lesser rank - ALWAYS. We don't have anyone of that rank in EIA - AT ALL
(not one will ALWAYS be able to beat their opponants). You would be better off for comparison purposes to compare how
long it takes for someone to learn CASUAL chess - not expert ... because that is what most current EIAers are.

First of all, no most people don't learn how to play chess in 1 game. They may learn HOW the pieces move, but to even play the game at a beginner level takes numerous games. Second of all, that is not how ratings work at all. Often times, grandmasters lose to International Masters, Masters, or even Experts. There are no absolutes in chess. The reason we have no one like that in EiA is very simple to understand and I can explain it with one word...DICE! The fact that this game involves a die, and thus luck and chance, means that it will never be even close to on the same level as the game of chess. In chess, when you drop a piece againt a competen opponent, that is almost guaranteed to be the end. But in EiA, you can select Cordon as a defense against an Outflank, and the die rolls can still see you lose, even though you are in a favorable position. I love EiA, but the fact is that if someone called it a glorified version of Risk, I could not argue against that. Yes the rules are more intricate and there are different gameplay elements implemented to reduce chance and reward strategic thinking, but at the end of the turn, it still comes down to who rolls the better numbers on the die, and such is the same case in a game like Risk. It is absurd to even compare EiA to chess, and the fake that you even attempt to shows me how you know little to probably nothing about the game of chess. Please just stop arguing such a stupid point because it really is not even close to valid.

[Jimmer]

ORIGINAL: NeverMan
It is also obviuos that niether Guy nor Jimmer have played Chess even at the most intermediate level.

Yes, I've played chess competitively. 1600s somewhere when I retired (class C, as I recall).