|
| Thread: New idea: Captives. |  This thread is pages long: 1 2 3 · «PREV / NEXT» |
|
niteshade
Known Hero
|
posted November 01, 2001 10:06 PM |
|
|
"I disagreed about this. I believe it will work just fine and making a good AI is no problem. I don't think we will come any longer about it. We just have to disagree. "
Well considering that making a good AI is traditionaly the single hardest thing about making a strategy game, and considering that H2 and H3 had pretty bad AI (well H3 was alot better then H2), I'd say you are obviously wrong about the "making a good AI is no problem" statement. In fact I would be hard pressed to think of a statement that was more incorrect then that.
And the fact that there is no diplomacy or attitude system in heroes would make it impossible for the AI to judge if he was currently still at war with you, something which would be absolutely essential to making a sound decision on prisoner trading/ransoming.
"=> You'll lose that high level magic possibility because of game changes. "
Um....what exactly are you basing this statement on?
"Anyway, it's not uncommon that you take them out but usually it only happens for low-level heroes with small armies. If the fight is even it's a lot more difficult. "
Very true, but it still happens especialy against the computer (another poor AI problem which further proves my point above). But it's supposed to be difficult.
"=> The enemy would not know in which castle the Hero is. You'd have to protect every castle you own. "
Oh no.....your enemy is now going to try to take over your cities. Sheesh, that's the exact same thing he was trying to do before you captured his hero. Nothing has changed but he is down a hero. Even you should be able to figure out this is a good thing. Are you saying that if you didn't have a hero captive you wouldn't make any effort to guard your cities and you would just let anybody waltz in and take over? I wish my opponents were like you......
"=> Better to kill the Hero. In that case you really DO know where the enemy is going to show up. "
And your enemy knows exactly where he has to go, and has a much easier time getting there. Even you should be able to figure out this is a worse thing.
"=> I don't see the benefit. Imprison scouts or low-level heroes? Whats the point? Defeated their main hero(es)? It's usually game over for that player. Pretty much a moot point to imprison the hero(es), if you have beaten the bulk of their army. "
Your the one who was saying in a different post you would spend the time and resources developing a scout to be important. If I caught some hero with expert scouting and expert stealth, or a backup hero with expert nobility, estates, and mining you can be pretty damn sure I'd want to imprison him instead of leave him on the map to be ressurected when I move away.
"=> The point was taken up that people often accept win-win proposals. This is common in chess. Say it's the last round of a Tournament. Player A needs 0,5 points to win. A offers a draw. Player B knows that he has little chance to beat the top player normally accepts. It doesn't always go that way, but it's not uncommon. "
Well sounds like a great deal for player A, and a horrible deal for player B. If the two of them are matched in the final round doesn't that mean they are both the top players of the tournament? I could see it if by drawing they could both get to the finals or something like that. But it still has no relation at all to what we are discussing.
" The inital point was, people generally don't reject win-win proposals, and if you do you may very well find that the opponents have an edge. I find your viewpoint in auto-refusing win-win proposals a bit strange. "
That's because the proposals you mention would only be win/ win in very rare circumstances in large free for all games. In 2 player games, or 2v2 games they would be win/lose.
"=> It's like you're offered something for 10 Gold, which normally costs 15 but you say no because you know the seller only paid 5 Gold.
"
Ok that analogy makes no sense and has no relation
"=> Or to take a possible Heroes example. Let's say a new level 1 Hero costs 2500 Gold to recruit. You've taken a level 1 Hero prisoner from another player.
=> The set ransom for a level 1 Hero is 1250 Gold. Would you offer that hero for ransom? "
Seems like that would benefit you and your enemy equally, which once again would be a pretty stupid thing to do in a most heroes games since you don't benefit from it. As I said it might be fine with somebody you are not at war with in a multiplayer free for all game, both things which are unlikely to occur, and together extremely unlikely.
"=> If the roles are opposite, would you pay the ransom of 1250 Gold? "
In a multiplayer free for all game? I might not. It might just encourage my enemy to keep capturing my heroes and trying to sell them back to me for a profit. And if he was a neighbor I was currently fighting I wouldn't even consider it. It doesn't benefit me that much, and for all I know I may have given my enemy just enough gold to get his lvl 4s by the end of the week to come kill me with.
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
|
posted November 01, 2001 11:26 PM |
|
|
Well considering that making a good AI is traditionaly the single hardest thing about making a strategy game, and considering that H2 and H3 had pretty bad AI (well H3 was alot better then H2), I'd say you are obviously wrong about the "making a good AI is no problem" statement. In fact I would be hard pressed to think of a statement that was more incorrect then that.
=> Look at today's chess computers. It's also a stratgeic problem. Today the best computers are all but impossible to beat for normal humans.
=> And wasn't there talk about them toning down the AI because it was too difficult to beat in Heroes 3? I'd say the real problem is that you couldn't make the AI too difficult to beat because then people would tire of the game because they always lost.
And the fact that there is no diplomacy or attitude system in heroes would make it impossible for the AI to judge if he was currently still at war with you, something which would be absolutely essential to making a sound decision on prisoner trading/ransoming.
=> But we agreed that there was neither peace nor war in Heroes. So why would judging war/peace aspects be an issue at all????
"=> You'll lose that high level magic possibility because of game changes. "
Um....what exactly are you basing this statement on?
=> The removal of spellpower and what Maranthea said about very-high-damage spells in one of the interviews.
Very true, but it still happens especialy against the computer (another poor AI problem which further proves my point above). But it's supposed to be difficult.
=> I've noticed that this usually occurs when it's the last Hero of that Player, and I'll vanquish the player if I win, or when the Hero is low level with no artifacts. When playing on Impossible, it's dismal to note how very difficult it is to actually accomplish this. I'd say the AI is fairly well programmed when it comes to retreating.
"Oh no.....your enemy is now going to try to take over your cities. Sheesh, that's the exact same thing he was trying to do before you captured his hero."
This is not always the case.
"Nothing has changed but he is down a hero. Even you should be able to figure out this is a good thing. Are you saying that if you didn't have a hero captive you wouldn't make any effort to guard your cities and you would just let anybody waltz in and take over? I wish my opponents were like you......"
He's down one Hero regardless if you Tombstone the Hero or if you place the Hero in the Prison.
It's also a reality in Heroes 3 that you can leave much of your towns unguarded... And is there any player who splits their troops to guards all their towns? If you do i wish you were my opponent!
Once you learn Town Portal, Heroes 3 pretty much allows you to leave your town with little guard. Usually, you have a large area around your Towns scouted so this is not a problem. The problem you may face is DD:ing heroes, but then you're in serious problems whatever your strategy is.
"=> Better to kill the Hero. In that case you really DO know where the enemy is going to show up. "
And your enemy knows exactly where he has to go, and has a much easier time getting there. Even you should be able to figure out this is a worse thing.
=> You can't be sure of that. This is very situational.
"=> I don't see the benefit. Imprison scouts or low-level heroes? Whats the point? Defeated their main hero(es)? It's usually game over for that player. Pretty much a moot point to imprison the hero(es), if you have beaten the bulk of their army. "
"You're the one who was saying in a different post you would spend the time and resources developing a scout to be important. If I caught some hero with expert scouting and expert stealth, or a backup hero with expert nobility, estates, and mining you can be pretty damn sure I'd want to imprison him instead of leave him on the map to be ressurected when I move away."
=> Sure. But you will not catch that Scout, because you wouldn't develope a Scout yourself, according to your own tactics  And also beware that you'll probably not be aware of which skills a Hero has.
=> You'll NOT catch the Nobility hero either. If you do then this is the final battle of the map and you've smashed my full army.
"=> The point was taken up that people often accept win-win proposals. This is common in chess. Say it's the last round of a Tournament. Player A needs 0,5 points to win. A offers a draw. Player B knows that he has little chance to beat the top player normally accepts. It doesn't always go that way, but it's not uncommon. "
Well sounds like a great deal for player A, and a horrible deal for player B. If the two of them are matched in the final round doesn't that mean they are both the top players of the tournament? I could see it if by drawing they could both get to the finals or something like that. But it still has no relation at all to what we are discussing.
=> No, it usually doesn't. It depends a bit. If it is all-meets-all then it could be anyone. If it's a monrad system you play, the best players would usually already have met, so you would meet someone with a slightly-better than average power. (It depends also on how many rounds you play and how many players are in, but overall it's just chance.)
" The inital point was, people generally don't reject win-win proposals, and if you do you may very well find that the opponents have an edge. I find your viewpoint in auto-refusing win-win proposals a bit strange. "
That's because the proposals you mention would only be win/ win in very rare circumstances in large free for all games. In 2 player games, or 2v2 games they would be win/lose.
"=> It's like you're offered something for 10 Gold, which normally costs 15 but you say no because you know the seller only paid 5 Gold."
Ok that analogy makes no sense and has no relation
=> The analogy is exactly the same as the second example below. Paying a ransom of 1250 for a hero, when you know if you purchase it in your own town you have to pay 2500. The only reason you reject is because the seller also makes money on the sale. (First subtract 5 Gold from the numbers, and then multiply by 1250, and it's the same numbers!)
"=> Or to take a possible Heroes example. Let's say a new level 1 Hero costs 2500 Gold to recruit. You've taken a level 1 Hero prisoner from another player.
=> The set ransom for a level 1 Hero is 1250 Gold. Would you offer that hero for ransom? "
"Seems like that would benefit you and your enemy equally, which once again would be a pretty stupid thing to do in a most heroes games since you don't benefit from it. As I said it might be fine with somebody you are not at war with in a multiplayer free for all game, both things which are unlikely to occur, and together extremely unlikely."
=> I really think that once you have 2-3 heroes from different colours in your prisons that you will have all other players uniting against you to ransack your towns.
=> And if I had made the AI I would have penalized your behaviour by making the opponents (with captured heroes)agressive specifically towards you, leaving the other colours alone. And isn't this how you would have behaved if you were playing those colours?
=> An example:
Red: Gain 1250 Gold
Blue: Gain 1 low-level Hero
Tan: Nothing
Pink: Nothing
Purple: Nothing
Green: Nothing
Even if Blue perhaps gains more than you... You gain more than Tan, Pink, Purple, and Green. That's why this is win/win for Red/Blue and Lose for all the others.
"=> If the roles are opposite, would you pay the ransom of 1250 Gold? "
In a multiplayer free for all game? I might not. It might just encourage my enemy to keep capturing my heroes and trying to sell them back to me for a profit. And if he was a neighbor I was currently fighting I wouldn't even consider it. It doesn't benefit me that much, and for all I know I may have given my enemy just enough gold to get his lvl 4s by the end of the week to come kill me with.
=> Ah well... Your style of playing Heroes is a very greedy one. Not like mine.
=> I sure wouldn't want to be a hero employed by you.... My paymaster doesn't even have the decensy to pay my ransom, so I have to suffer in prison?
=> I say it would serve you right if the Hero pledged loyalty to the enemy!
____________
"A brilliant light can either illuminate or blind. How will you know which until you open your eyes?"
|
|
niteshade
Known Hero
|
|
posted November 01, 2001 11:57 PM |
|
|
"=> Look at today's chess computers. It's also a stratgeic problem. Today the best computers are all but impossible to beat for normal humans. "
Chess is a million times simpler then heroes. All actions has quantifiable results and their are are only a tiny fraction of the variables to consider that their are in heroes.
"=> And wasn't there talk about them toning down the AI because it was too difficult to beat in Heroes 3? I'd say the real problem is that you couldn't make the AI too difficult to beat because then people would tire of the game because they always lost. "
Quite frankly, no there wasn't. The AI is still considered bad in H3, and all talk has been about improving it. They are not trying to make it unbeatable, they are just trying to make it as challenging as a good human player would be. I don't know where you got all that from. And it's rather scary if you are saying the computer AI is too good for you. It makes me seriously question anything tactical coming from you.
"=> But we agreed that there was neither peace nor war in Heroes. So why would judging war/peace aspects be an issue at all???? "
You don't seem to understand how things work. The only time it would be worth trading a hero is if you are at peace with an enemy. Since the computer has no way to figure this out without war/peace status he will never know when it's good to do. A human on the other hand can make deals in the game.
"=> The removal of spellpower and what Maranthea said about very-high-damage spells in one of the interviews. "
That will have no effect on hero killing strategies which mostly involved using such spells as blind and slow rather then high power damage spells.
"=> I've noticed that this usually occurs when it's the last Hero of that Player, and I'll vanquish the player if I win, or when the Hero is low level with no artifacts. When playing on Impossible, it's dismal to note how very difficult it is to actually accomplish this. I'd say the AI is fairly well programmed when it comes to retreating. "
I have killed high level AI heroes quite a few times in the mid to early games. It's not hard to do, if you worked on your strategy a bit I'm sure you could manage it as well.
"Oh no.....your enemy is now going to try to take over your cities. Sheesh, that's the exact same thing he was trying to do before you captured his hero."
This is not always the case. "
Ok maybe you don't try to take over your enemies towns, but an inteligent player does.
"He's down one Hero regardless if you Tombstone the Hero or if you place the Hero in the Prison. "
The tombstone is much harder to guard then the town. The town you have to guard anyway because your enemy is always trying to take it. The tombstone adds a new location you have to guard as well as your town.
"It's also a reality in Heroes 3 that you can leave much of your towns unguarded... And is there any player who splits their troops to guards all their towns? If you do i wish you were my opponent! "
Very true, and all the more reason not to worry about the prisoner in your town. Usualy you guard towns more by intercepting them on the way into your territory with your own hero.
"Once you learn Town Portal, Heroes 3 pretty much allows you to leave your town with little guard. Usually, you have a large area around your Towns scouted so this is not a problem. The problem you may face is DD:ing heroes, but then you're in serious problems whatever your strategy is."
Once again this is quite true, and once again it only goes to prove my point.
"=> Sure. But you will not catch that Scout, because you wouldn't develope a Scout yourself, according to your own tactics And also beware that you'll probably not be aware of which skills a Hero has. "
Actualy somebody could very easily identify the scout with their own scout, and use their main force to nab it. And the fact that you are unaware of the skills of a hero you nab is all the more reason to imprison them because you never know which are important. However the fact that the hero only showed up when he got close to your own scout or something else which might reveal hiding heroes is a good sign about his abilities.
"=> You'll NOT catch the Nobility hero either. If you do then this is the final battle of the map and you've smashed my full army. "
That's probably true on a small-medium sized map, but not on a large map where you could easily catch him in his castle and still not be ahead in the game.
"=> The analogy is exactly the same as the second example below. Paying a ransom of 1250 for a hero, when you know if you purchase it in your own town you have to pay 2500. The only reason you reject is because the seller also makes money on the sale. (First subtract 5 Gold from the numbers, and then multiply by 1250, and it's the same numbers!) "
Except that your enemy paid no gold to get things in the second example and in fact he commited a hostile act against you to get it, and unlike the first example he is your enemy and you are giving him a profit.
A better example would be if your enemy stole a 10 gold item from you and offered to sell it back for 5 gold. Most people would not accept a deal like this cause it just encourages him to steal more from you.
"=> I really think that once you have 2-3 heroes from different colours in your prisons that you will have all other players uniting against you to ransack your towns. "
Only if they are really important heroes, and you are playing a free for all multiplayer map which as we said almost never happens. The computer is certainly not that smart even in games with the best AI.
"=> An example:
Red: Gain 1250 Gold
Blue: Gain 1 low-level Hero
Tan: Nothing
Pink: Nothing
Purple: Nothing
Green: Nothing "
That sounds like a good deal for red and blue unless they are at war. If they are, it will benefit one more then the other, though it's hard to judge how much. Maybe red really needs that gold for his army that is about to come and smash blue. Maybe blue is the same way and needs to save all the gold he gets. The problem is that the AI never knows if he is at war or not (he's always at war).
And your idea about the computer hating you more if you refuse to trade does not work in a game with no relation system or diplomacy system.
"Even if Blue perhaps gains more than you... You gain more than Tan, Pink, Purple, and Green. That's why this is win/win for Red/Blue and Lose for all the others. "
Unless red is fighting blue. Then it's only a win for red, and blue gets the shaft as he aids his enemy who will probably proceed to use that aid to kill him.
"=> Ah well... Your style of playing Heroes is a very greedy one. Not like mine.
=> I sure wouldn't want to be a hero employed by you.... My paymaster doesn't even have the decensy to pay my ransom, so I have to suffer in prison? "
lol, I'd rather win the game and free the hero after my enemy was dead then lose the war because of bad strategy and have all my heroes executed.
|
|
Darion
Promising
Famous Hero
|
|
posted November 02, 2001 01:01 AM |
|
|
Umm... just so that you know, there was a game called Lord of Magic that tried the thing with the exchanging prisoner thingie... and even though the game was pretty bad, the computer would never make good decisions. It is a heckuva a lot harder to program
For instance, say that you want to trade a Lvl 1 Warlock back for a Lvl 1 Druid. In order to make a smart decision, the computer would have to make some choices difficult for a human to make! It would have to weigh the attack, defense, power, knowledge as well as any skills the hero would have (ex. If it has pathfinding on an all-grass map rather than on swampy ground or order magic when you have no order spells at all and no way to get any, or eagle eye) and that would make it difficult for the computer to make any kind of decision at all (not to mention adding in artifacts, expert skills, creatures and proximity) and that would just mess the computer ai up so much.
____________
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
|
posted November 02, 2001 01:34 AM |
|
|
=> It seems we're getting nowhere. We'll just have to disagree about these things. Will make a short reply on some things.
Chess is a million times simpler then heroes. All actions has quantifiable results and their are are only a tiny fraction of the variables to consider that their are in heroes.
=> Sure, but in Heroes you can cheat by letting the AI look at things the AI player wouldn't know. Can't cheat in chess.
=> The rules are simpler in chess.. But I'd say that complexity favours the AI if the programming is properly done.
"And it's rather scary if you are saying the computer AI is too good for you. It makes me seriously question anything tactical coming from you."
=> No, normally I don't have any problems with the AI. Though impossible setting can be difficult on some maps.
You don't seem to understand how things work. The only time it would be worth trading a hero is if you are at peace with an enemy. Since the computer has no way to figure this out without war/peace status he will never know when it's good to do. A human on the other hand can make deals in the game.
=> There's no visible war/peace but the AI should handle it internally, just like a human would. And again you're underestimating a good AI. (No need to reply to this... You've already said you believe the AI will be poor.)
=> About killing heroes. We won't know how easy it will be in heroes 4 until we see the game. Anything we say would be pure speculation.
"I have killed high level AI heroes quite a few times in the mid to early games. It's not hard to do, if you worked on your strategy a bit I'm sure you could manage it as well."
It does happen, but all too often they retreat like cowards with _huge_ amount of troops left  Of course, usually it means I loose less troops as a result. My tactics would be similar to yours. Without Blind and tricking the AI with Haste and/or Slow, you very rarely defeat them.
"Ok maybe you don't try to take over your enemies towns, but an inteligent player does."
=> Usually, I take the towns but care a lot less about hunting down their heroes. (unless they're becoming so weak that they'll soon be out of the game.)
"The tombstone adds a new location you have to guard as well as your town."
=> There is a strategy involved in whether or not you guard the Tombstone.
Once again this is quite true, and once again it only goes to prove my point.
=> I'm obviously disagreeing. It proves my point.
A better example would be if your enemy stole a 10 gold item from you and offered to sell it back for 5 gold. Most people would not accept a deal like this cause it just encourages him to steal more from you.
=> It depends. If your Hero is taking over mines close to their towns, then they've caught a Bandit. The loss of a Hero is not necessarily the result of a bad action from an enemy. There's also the possibility that the town with the Prison has been taking over by a third player, and they did not defeat the Hero at all to begin with.
The problem is that the AI never knows if he is at war or not (he's always at war).
=> We disagree about this. I believe this should be evident for the AI. (And especially if the AI plays players of your type. The acts you're doing are hostile to the AI players that it affects.)
And your idea about the computer hating you more if you refuse to trade does not work in a game with no relation system or diplomacy system.
=> We disagree about this.
lol, I'd rather win the game and free the hero after my enemy was dead then lose the war because of bad strategy and have all my heroes executed.
=> Well, you still can't blame a hero who doesn't want to rot in jail, and decides to join the enemy instead.
____________
"A brilliant light can either illuminate or blind. How will you know which until you open your eyes?"
|
|
niteshade
Known Hero
|
|
posted November 02, 2001 05:19 AM |
|
|
I'm going to cut short the argument on strategy and weither or not prisons are useful because it's primarily based on speculation, and because you've pretty much stopped backing up your arguments at all making the discussion pointless. As for the Ai issues.......
"=> Sure, but in Heroes you can cheat by letting the AI look at things the AI player wouldn't know. Can't cheat in chess.
"
Which is true, but it's often not enough and it will not help with the matter we are discussing.
"=> The rules are simpler in chess.. But I'd say that complexity favours the AI if the programming is properly done. "
And you would be wrong. Very wrong. In any given turn of chess there are rarely more then 100 possible moves or variations you can make to your turn, and usualy quite a few less. That makes it easy for the computer to plan ahead in advance. If we assume 50 possible moves per turn, then for the computer to plan 5 turns ahead requires it to consider a huge number of scenarios, but nothing that it's processor cannot handle. This combined with some basic chess strategy, opening moves, and plays makes it possible to make a quite smart computer. With more powerful processors you can plan further and further in advance making the computer very hard to defeat indeed.
In Heroes every turn has an almost infinite number of possible variations. How many troops you recruit, how many heroes you hire, what structures you build, exactly where you move on the map, how you split up your troops, etc. etc.. Calculation of all possibilities is impossible. It would require more processor power then plotting everything that could happen in the next 10 turns of a chess game. Alot more. As a result this approach can not be taken to the heroes AI. This is what makes chess AI easy to program, and other computer games much harder.
Yes in theory in the year 3000 complexity might be an advantage, but you obviously aren't well well versed with programming in the year 2001. Hopefully this will clear some things up for you. A human can handle all these variables because we have judgement, common sense, and intuition. A computer has none of these things.
"=> There's no visible war/peace but the AI should handle it internally, just like a human would. And again you're underestimating a good AI. (No need to reply to this... You've already said you believe the AI will be poor.) "
Ah so your saying the AI should think and use intiution just like a human does. I guess your right there is no need to reply to something silly like this. BTW I should say I don't necesarily think AI in H4 will be poor. Much of the game is streamlined to a level that the AI might even have an easier time dealing with things. So I have hopes that it will be better then it was in H3. It's definetely getting there as long as nobody adds something stupid that the computer can't handle to the game.
"=> We disagree about this. I believe this should be evident for the AI. (And especially if the AI plays players of your type. The acts you're doing are hostile to the AI players that it affects.) "
It's easy to say this but without somekind of hostility system where the AI either likes you or more or less based on specific actions, it can't be done. Usualy this is done by having a numeric hostility value which is effected by specific actions.And if they make a system like this it certainly will be displayed for other players to know. The problem is they are not doing this, and they certainly will not waste time doing it just so they can add your system in. The other problem with systems like this is they can be abused easily and it's hard to judge exactly what the right number to add to the hostility value of any action should be.
"And your idea about the computer hating you more if you refuse to trade does not work in a game with no relation system or diplomacy system.
=> We disagree about this. "
Once again you give no reason why you disagree. Do you really think computers have emotions and can be made to hate people without a complex relationship system being programmed into them?
"=> Well, you still can't blame a hero who doesn't want to rot in jail, and decides to join the enemy instead. "
Well considering they have specificaly said this will never happen, I'm not too worried about it.
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
|
posted November 02, 2001 09:01 AM |
|
|
I'm going to cut short the argument on strategy and weither or not prisons are useful because it's primarily based on speculation, and because you've pretty much stopped backing up your arguments at all making the discussion pointless. As for the Ai issues.
=> Actually, you're saying things that just goes outside the scope of Strangers and which I don't agree to at all.
Which is true, but it's often not enough and it will not help with the matter we are discussing.
=> Cheating won't help? You can't be serious!!! Do you actually know anything about programming???
=> How would you feel if the AI peeked into you spellbooks and knew every spell you had at every time.
=> How would you feel if the AI automatically knew the position of each of your heroes and which skills they have.
=> How would you feel if the AI automatically placed their heroes just out of range from your heroes each time.
=> How would you fell if the AI automatically had all the information in the "Thief Guild"
=> How would you feel if it knew how much of the Shroud you had removed?
And this is just the start...
The game would be unbeatable if it cheated like this and then optimized the move according to movement.
And if you take Combat. It's like chess. there's only a finite amount of possibilities of moves and a finite amount of pieces. And if it looks at your spell-book it could see if you can counter its strategy.
"And you would be wrong. Very wrong. In any given turn of chess there are rarely more then 100 possible moves or variations you can make to your turn, and usualy quite a few less. That makes it easy for the computer to plan ahead in advance. If we assume 50 possible moves per turn, then for the computer to plan 5 turns ahead requires it to consider a huge number of scenarios, but nothing that it's processor cannot handle. This combined with some basic chess strategy, opening moves, and plays makes it possible to make a quite smart computer. With more powerful processors you can plan further and further in advance making the computer very hard to defeat indeed."
You're very wrong. The number crunching strategy works only so far. And the experience with computers shows that it's not the number crunching which is important. It's the evaluation of the positions, and to select which moves to analyze further. Also like a human, a computer would have to develope some form of game-plan to do well. The analysis can never be done to the end unless you're close to check-mate. If you start with move 1 and want to analyse ahead, for each move you have to multiply with 100. Only 10 moves would require 10 exp 20 analysis!!!
"In Heroes every turn has an almost infinite number of possible variations. How many troops you recruit, how many heroes you hire, what structures you build, exactly where you move on the map, how you split up your troops, etc. etc.."
=> The computer doesn't really need to know all of these things. Number crunching only won't help. Some of the things you mentioned would only be valid in certain cases.
=> Calculation of all possibilities is impossible.
=> It's impossible and a bad strategy already in chess.
It would require more processor power then plotting everything that could happen in the next 10 turns of a chess game.
=> Even in Chess this is impossible today, if you want to make the move within a reasonable time.
Alot more. As a result this approach can not be taken to the heroes AI. This is what makes chess AI easy to program, and other computer games much harder.
=> Chess AI programmers would disagree with you. Count out possibilities is dead easy and usually totally hopeless.
=> I've never said that number crunching was the way to go. It isn't and I wouldn't even contemplate building an AI that way.
Yes in theory in the year 3000 complexity might be an advantage, but you obviously aren't well well versed with programming in the year 2001.
=> Well enough to know that the best computers doesn't really use number crunching to beat grand masters.
Hopefully this will clear some things up for you. A human can handle all these variables because we have judgement, common sense, and intuition. A computer has none of these things.
=> Number crunching only doesn't make an good AI. If it is going to be effective it has to resort to other things. And strictly speaking i don't call "Number crunching" only for an AI.
BTW I should say I don't necesarily think AI in H4 will be poor.
=> Oh yes, you are. At least in my opinion. The gist of what you're saying is that the AI can't possibly be able to handle even a simple barter system. And it can't remember this or act on that. That's not only poor, that is a shamefully bad AI.
It's easy to say this but without somekind of hostility system where the AI either likes you or more or less based on specific actions, it can't be done.
=> Of course I assume a hostility system!! It's just that the player doesn't see the ratings, the very idea. Then it could be used against the AI! Only the AI knows. And you don't need Strangers to add it in. This is a basic feature of an AI. Something it must have anyway.
"Once again you give no reason why you disagree. Do you really think computers have emotions and can be made to hate people without a complex relationship system being programmed into them?"
=> It doesn't have to be complex. Just measure a few key factors and you have a good system.
Can we finally return to discussing Captives?
____________
"A brilliant light can either illuminate or blind. How will you know which until you open your eyes?"
|
|
thunderknight
Promising
Famous Hero
|
posted November 02, 2001 10:17 AM |
|
|
Hmmm, it sounds there are some interesting discussions here.
I just want to share something from another strategic game: Sango series, a strategic game based on history.
In even the early installment of Sango (i.e. Sango II which was at least 10 yrs ago), the prisoner/captive thing already existed. In Sango II & III, after you defeat your enemy armies, you have several choices on the captured generals:
- behead them on the spot,
- release them (I seldom do it,  )
- persuade them to join your side,
- take them captives,
Will the defeated generals join you or not depends on:-
- Charisma of the lord under the player's control
- Historical facts i.e. some ppl may be your enemy of destiny and so they will NEVER join you
- The ˇ§warˇ¨ attribute of the defeated generals i.e. something like overall battle abilities of the generals, the higher the attribute, the more difficult for them to join
- The leading general of your own army when you defeat enemy generals. If you use some legendary generals as your leader, it will greatly increase the chance for the enemy generals to join
- You can use multiple persuasion, offer gold, offer war horse or legendary weapons available as an incentive for the defeated generals to join
If you cannot persuade the defeated generals to join you at the spot, you can resort to the remaining 3 alternatives. Behead or release will end the story. If the one you kill is the lord (i.e. the leader) of the enemy, it has to choose another lord and it will declare you as its forever foe. (I like that.  )
You can choose to keep them captives and bring them back to your castles to keep them in the prison there. However, they can escape on the way. The chance of escape depends on the war attribute of them and the overall strength of your army keeping them captives. Some of the legendary generals will nearly ˇ§alwaysˇ¨ escape on the way.
If you bring the captured generals to your lord, you have more options:-
- ask for ransom from your opponent
- persuade them to join you by offering some title like ˇ§Da Jian Junˇ¨ and make them your own generals
- you can even offer a princess in exchange for his loyality (ya, 99% of the generals in the game are male)
Since the early installments of Sango series are for single player only, most of the time you are dealing with AI. But it seems AI is quite good in the game. They ALWAYS ask for a very high ransom and will ALWAYS refuse to pay the ransom if it exceeds certain % of its total $$$.
There is no limit for the number of generals that a prison can hold.
In a game of > 3 players, if you capture a city where generals of other AI players are held, the situation is just like as if you have captured them on the battle field. You can persuade them, kill them, release them or keep them in prison.
I know nothing about programming. However, if it is possible in a game 10 years ago, I think it wouldnˇ¦t be a problem now. The only problem I can see is the same old issue:
Simplicity or Reality ??? What do you want ???
____________
Choose what you love
and love what you choose.
|
|
niteshade
Known Hero
|
|
posted November 02, 2001 05:46 PM |
|
|
"=> Actually, you're saying things that just goes outside the scope of Strangers and which I don't agree to at all. "
I assume you mean captives not strangers. And this all relates to the original argument.
"=> Cheating won't help? You can't be serious!!! Do you actually know anything about programming??? "
Sheesh talk about taking my statement out of context. I said being able to know everything will not help the AI figure out if he should trade heroes or not. It will help the AI in other ways. Why don't you try reading what I write before I respond to it.
"=> How would you feel if the AI peeked into you spellbooks and knew every spell you had at every time.
=> How would you feel if the AI automatically knew the position of each of your heroes and which skills they have.
=> How would you feel if the AI automatically placed their heroes just out of range from your heroes each time.
=> How would you fell if the AI automatically had all the information in the "Thief Guild"
=> How would you feel if it knew how much of the Shroud you had removed?
"
I would feel that none of this would help it know weither it should trade a hero or not. Many of the other things would help it in other aspects of AI, and in fact some versions of heroes have employed these "cheats" for the AI. The other problem of course is knowing the info and knowing what to do with it are two different things.
"The game would be unbeatable if it cheated like this and then optimized the move according to movement. "
It would hardly be unbeatable. H1 cheated in extactly these ways and more but you could still beat it. The problem is that poor AI in other ways makes up for this information.
"And if you take Combat. It's like chess. there's only a finite amount of possibilities of moves and a finite amount of pieces. And if it looks at your spell-book it could see if you can counter its strategy. "
Well in theory it's like chess however.......
In turn 1 of chess there are only 20 moves you can make.
In turn of 1 of your average combat each troop has about 20 different moves it can make, plus it can wait, defend, plus
there are about 20 different spells that could be cast.
So assuming 7 troops on each side.....
Number of possible moves for each troop (counting spells) 400. So the total number of possible moves for a single round of battle in heroes would be 400 to the 17th power. I'm not going to calculate that out, because I don't feel like typing in that many 0s, but let's just say it's probably much more then a million times more possible moves per turn then chess. So in conclusion your battles being as simple as chess anology is just wrong.
"You're very wrong. The number crunching strategy works only so far. And the experience with computers shows that it's not the number crunching which is important. "
Well your right, number crunching only works so far, but it's still a vital part of all strategy considerations in chess programs. and it's something you can't do at all in heroes.
"It's the evaluation of the positions, and to select which moves to analyze further"
True, but in heroes with a near infinite number of possible moves per turn it can't even do this.
"The analysis can never be done to the end unless you're close to check-mate. If you start with move 1 and want to analyse ahead, for each move you have to multiply with 100. Only 10 moves would require 10 exp 20 analysis!!! "
Well most turns you have alot less then 100 moves, but yes your right you are limited in how many turns ahead you can plan. But this is still a huge advantage over heroes where you cannot plan at all.
"=> The computer doesn't really need to know all of these things. Number crunching only won't help. Some of the things you mentioned would only be valid in certain cases."
The computer still needs to decide all of these things. It has to decide how many troops to switch to, which buildings to make, what to recruit, how to spend his money. So he damn well needs to know what he is doing for his turn. And like I said he can do an almost infinite number of things. So there is no way for him to use number crunching to figure out what to do. Somebody has to program essentialy a very complicated if x then y program to code the AI. Often assigning numeric values to various actions depending on the circumstance. This is pretty much how things work currently in Heroes if you've read the developers updates and things like that (they can be very interesting)
" Calculation of all possibilities is impossible.
=> It's impossible and a bad strategy already in chess. "
Chess computers do still plot out all possible outcomes when making a move. That is the key part of how they are able to still function when a situation they have not been specificaly programmed for arises.
"Alot more. As a result this approach can not be taken to the heroes AI. This is what makes chess AI easy to program, and other computer games much harder.
=> Chess AI programmers would disagree with you. Count out possibilities is dead easy and usually totally hopeless."
You say chess AI programmers would disagree that other computer games are harder to program AI for then chess. Then why is it that chess has great AI and most other games have bad ones even after years of development. That alone proves your point wrong.
"=> Well enough to know that the best computers doesn't really use number crunching to beat grand masters. "
But obviously not well enough to know how important number crunching still is to the equation, and the limited number of possible moves per turn.
"=> Oh yes, you are. At least in my opinion. The gist of what you're saying is that the AI can't possibly be able to handle even a simple barter system. And it can't remember this or act on that. That's not only poor, that is a shamefully bad AI. "
There isn't going to be a simple barter system in H4. That's why the AI is not going to be bad. Thankfully 3D0 is smart about these things even if you are not. But I should say it probably could handle one if an entire diplomacy and hostility system was built into the game. But heroes has never gone that route, and they are not going to start now.
"=> Of course I assume a hostility system!! It's just that the player doesn't see the ratings, the very idea. Then it could be used against the AI! Only the AI knows. And you don't need Strangers to add it in. This is a basic feature of an AI. Something it must have anyway. "
Well it's something that heroes has very traditionaly not had. Deal with it. Maybe some day they will add it, but not in this version. And if they add a relationship system they will probably add aliances, and things like that too. Then prisoner trading might be feasible.
"Once again you give no reason why you disagree. Do you really think computers have emotions and can be made to hate people without a complex relationship system being programmed into them?"
=> It doesn't have to be complex. Just measure a few key factors and you have a good system. "
You have a system, but not a good one. Good AI systems are always complex. It's the nature of the game. I'd rather have no system then a bad one.
"Can we finally return to discussing Captives? "
Yeah sure I'm getting pretty tired of this argument anyway. I will rest happily knowing that the actual game will not be screwed up by this idea. And who knows if heroes 5 has diplomacy captives might actually work there.
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
|
posted November 03, 2001 12:08 PM |
|
|
I assume you mean captives not strangers. And this all relates to the original argument.
=> Yeah. I meant Captives. The discussion of AI programming was related to Prisons, and how to strengthen the Prison structure. That's fairly unrelated. The Prison structure itself is related since Captives interacts with it.
=> If you mean the tactical choice involved when having defeated an Enemy and choose between killing or taking captive, then our discussion very quickly dodged this question.
The other problem of course is knowing the info and knowing what to do with it are two different things.
=> An programmer would have no difficulty at all in turning the knowledge against you. Some of the information is more useful than other.
In turn 1 of chess there are only 20 moves you can make.
In turn of 1 of your average combat each troop has about 20 different moves it can make, plus it can wait, defend, plus
there are about 20 different spells that could be cast.
=> This may appear true, but the mechanism is different in chess. For one thing you don't have a King to protect in Heroes, which reduces the complexity of the analysis a lot.
True, but in heroes with a near infinite number of possible moves per turn it can't even do this.
=> It can count out how much damage each of the creatures could do in a combat for the round to come. This will makes it possible to surrender / retreat in time. So at least for retreating it should be fairly easy to count out if and when there's a risk.
Well most turns you have alot less then 100 moves, but yes your right you are limited in how many turns ahead you can plan. But this is still a huge advantage over heroes where you cannot plan at all.
=> You don't need to plan in the same way. As a human player, I think a LOT longer on my next chess move than on my next move in Heroes. The same goes for the AI.
"Chess computers do still plot out all possible outcomes when making a move. That is the key part of how they are able to still function when a situation they have not been specificaly programmed for arises."
=> No, they don't. If they counted out all outcomes you wouldn't be able to beat them. And the computer would never be able to do its next move, because it would never finish the analysis. (You probably meant that the AI counts X moves ahead, and then selects some variantions analyse them for Y moves, and so on.)
You say chess AI programmers would disagree that other computer games are harder to program AI for then chess. Then why is it that chess has great AI and most other games have bad ones even after years of development. That alone proves your point wrong.
=> Because there's a lot more people trying to program chess AIs than the AI for any one single computer game. And because programming chess AIs have been going on for a much longer time.
Well it's something that heroes has very traditionaly not had.
And if they add a relationship system they will probably add aliances, and things like that too. Then prisoner trading might be feasible.
=> Why would allied parties need to trade prisoners? They would just release prisoner's belonging to the ally.
=> And for trading heroes of different sides, you're in the same situation as before. There's not that much new because you add alliances. The options you have for Prisons (Set Ransom, Pay Ransom, and Exchange Prisoner) is the same as before. And for taking Captives you still have the options of "Kill Hero" or "Take Captive".
=> I've not suggested to add alliances. That's something which players will be more likely to "misuse" than captives. (And it would be more difficult to stop the "misuse", because the single "Barter" system model I proposed no longer applies. Basically, the AI would need to have "human intuition" and things like that or it would be too easy to swindle and backstab it.)
You have a system, but not a good one. Good AI systems are always complex. It's the nature of the game. I'd rather have no system then a bad one.
=> The complexity of the system I had in mind is not much more difficult than the Marketplace, and the options the AI has to face there. (Of course other parts of Heroes are much more complex, but that's beside the point.)
|
|
niteshade
Known Hero
|
|
posted November 03, 2001 08:31 PM |
|
|
Sigh....I think I'm going to drop this argument now. Your last post like many others has tons of errors in it, and alot of places where you say things that show you completely do not understand what I said before (ie when you say our argument only relates to the tactical use of prisons when it was directly related to the AI's use of prisoner exchange). It's getting rather frustrating correcting you each time, and having you say something back which shows a complete lack of understanding of what we were discussing. Perhaps it is a language barrier on your part, I assume English must not be your native language (or at least I hope not).
So rather then having a futile argument with somebody with no knowledge of programming, AI, or the computer game industry I will just drop things now. I'm sure it seems to you like I'm just giving up, but perhaps when you learn english a little bit more you will understand my frustration better.
Anyway your original idea is not completely without merit, and perhaps in a later heroes with a real diplomacy system and states of war and peace between players it would work quite well. I doubt we will see a heroes like that but who knows.
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
|
posted November 03, 2001 09:59 PM |
|
|
Niteshade, I never wanted to discuss AI issues at all. The Altar if for the wishing so why would I worry about the implementation of what i wish for.
As for programming you're wrong. I've programmed games, not as complex as heroes but games anyway. I've also made a game of my own a few years ago. I don't have any deeper insight in the Heroes game engine.
The real challenge about the AI probably lies in making it both FAST and GOOD.
You've made some valid points, but basically we're probably discussing different things in some of the passages.
I said straight out that prisoner exchange was not to be possible for Captives. That was an advantage proposed for the Prison structure to make it more advantegeous compared to having the prisoner as a Captive. Thus prisoner exchange and captives were not directly related, although they were indirectly realted through the prison structure. (See my second post in the Thread.)
I don't think it's the English. I believe we're not always talking about the same things.
For one thing I somewhat misunderstood what you meant by diplomacy in some of the first posts.
Xenophanes: I don't think you can get imprisoned heroes to join you in Heroes 4.
But if we want to expand the captives idea, if the owner of the Hero doesn't pay within 1 week, then you could be offered to hire the Hero by paying the price for the ransom yourself.
Wyvern: It could perhaps be possible to convert heroes to your cause, but the artifacts, spells etc. required would have to be very difficult to obtain. Still it's a valid wish. The idea that if the original owner don't pay for the ransom you'll get a chance to do so.
Darion: Interrogation of prisoners is a good idea. The problem is: what information should you get ut from them. Feel free yo suggest it was your idea.
Thunderknight good post of yours. Basically I feel the same way about the programming issue. It has been done before, so why should there be a problem doing it.
On the other hand, diplomacy (even if it's one-sided, so only AI keep track for it internally) is too late for Heroes 4 if they haven't already included it. The fastest they could do it is for an expansion, or in Heroes 5. After all, programming the AI takes some time.
____________
"A brilliant light can either illuminate or blind. How will you know which until you open your eyes?"
|
|
niteshade
Known Hero
|
|
posted November 04, 2001 05:33 AM |
|
|
|
Well you do have a point that the altar is for wishing, not implementation discussiosn. I don't disagree that if it could be implemented properly it is a reasonabel idea as long as they were changing the direction that heroes is going in and having more diplomatic boards where you made deals and agrements with your enemies. I suppose it's not entirely possible it could find it's way into a system that did not have these things, but the AI programming that would be required to lay the groundwork for it would be the same groundwork required for all the other stuff. So I can't see them putting one in and not the other.
|
|
Darion
Promising
Famous Hero
|
|
posted November 04, 2001 09:52 AM |
|
|
Ooooh!!! Fight fight fight!!! LOL.... let's get back to the discussion and play nice here boys, instead of arguing about vague philosophies that take up lots of space and are confusing to read. Ahem... now about the Perfect Heroes IV with perfect units all balanced out (*cough* DRAGON GOLEM!*cough* )...
Gentlemen and women, please resume your noh-mal dis-kau-see-yons!
____________
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
posted November 04, 2001 11:43 AM |
|
|
Boys???? Who's Darion calling Boy!!!!
(Trying to stop Djive.... Tooo late!!!!!)
*** Zap ***
Sorry Darion, you're now a slimy Toad for a few days...
Also see it this way... The discussion was mostly on-topic (or at least related to either Captives or Prisons). If you don't want to read a discussion then, don't. Nobody is forcing you. No need to send in a post to complain.
Back to the discussion of Captives.
Pardon? Turning you back to an Elemental??? I'm afraid not, you'll just have to wait it out.
____________
"A brilliant light can either illuminate or blind. How will you know which until you open your eyes?"
|
|
Darion
Promising
Famous Hero
|
|
posted November 04, 2001 06:12 PM |
|
|
LOL... oops.
Hehee... anyways, it was a good conversation. You both have good arguments. Course, that won't help me being a toad now...*ribbit ribbit!*
____________
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
|
posted November 07, 2001 05:28 PM |
|
|
You're lucky the *zap* spell only lasts three days. See you're an elemental again.
Be careful about what you call Djive... She's a Witch and doesn't like to be mistaken for anyone of the other sex.
Hrmm... I just thought of something for Prisons... If the Hero is already dead on the battle-field at the end of combat...
What do you need a Prison with guards and all for?
Wouldn't you want a great oven for cremation, or a guarded mausoleum instead? (For the oven: Let's see if you can raise the Hero from the ashes...)
Is it perhaps so that Heroes will be unconscious until next turn instead?
Hehe... I just realized that for the "Captives" idea I've assumed that the Heroes are unconscious rather than dead and Tombstoning (alt. Captivity//Imprisonment) only happens if your army loses.
And another thought... If Heroes aren't resurrected within a time-frame, will they then perhaps be raised as undead?
And *lol* what would happen if the enemy decides to put a captured hero through the Skeleton Transformer?
And for that matter decides to use Necromancy to raise the Hero as a Skeleton?
|
|
Dingo
Responsible
Legendary Hero
God of Dark SPAM
|
|
posted September 25, 2003 01:31 AM |
|
|
A good idea! But it might get confusing when playing SP. AI better be smart enough to handle this
____________
The Above Post/Thread/Idea Is CopyRighted by, The Dingo Corp.
|
|
Gerdash
Responsible
Famous Hero
from the Animated Peace
|
|
posted September 27, 2003 04:45 PM |
|
|
(ok, i've been busy-busy-... and i would like to comment on some old topics that i have thought about in the meantime like leaders and vampires, but i dont't have the time for it all.)
1) as far as i have read, it's not honorable for a noble to kill an enemy noble that has surrendered. nevertheless, sometimes prisoners were killed. maybe killing an enemy hero would result in a morale penalty.
2) all this tombstone and hero resurrection thing is something that i dislike.
2a) it feels artificial and cheesy. and the hero is already comfortably in valhalla or somewhere, would be unfair call it back from the dead.
2b) if you want to kill the hero, why don't you do it so that it cannot be resurrected?
2c) what do you need to build a prison for? you could just kill the hero and have it's tombsone in any town of yours. and if the enemy captures the town, he can resurrect the hero.
3) in real life, i guess it was often the relatives of the captured noble that payed the ransom. this might be sort of difficult to implement without possibility to cheat.
4) of the ideas that i have read (not all of them) i liked the one where hero would be taken captive and would have a chance to escape every day. or you could send a stealth hero to free the captive. when the army visits a town with a castle it can throw the hero into a prison in the town castle (i wouldn't recommend a special prison structure). and you could take the hero out of the castle and transport it as a captive again. and when a certain amount of time has passed without ransom payed, the hero might be killed (sounds extremely reasonable to me), i.e. removed from the game for ever or reset to a fresh level 1 hero.
5) if a certain amount of time has passed without ransom, maybe the enemy could get some weekly chance to recruit an imprisoned (but not captived) hero, based on the alignments. maybe an imprisoned hero would loose levels over time, depending on the torture levels. and maybe the more you torture, the higher the chance of recruiting the hero. so if you need the enemy hero faster, it's level will probably be lower.
========
ok, i guess i stop here. the main idea was that it's not honorable to kill a surrendering noble.
|
|
Djive
Honorable
Supreme Hero
Zapper of Toads
|
|
posted November 20, 2003 10:04 PM |
|
|
Normally, the expected consequence of killing an enemy leader or hero would be a morale boost, not the other way around. Enemy heroes are often seens as ruffians, villains or fools by the other side. The normal case if a holy priest kills an enemy vampire leader would be rejoicing and thankfullness. If the vampire kills the priest, then it'd be good rythms, and tasty snack! I don't think either side would ever contemplate remorse.
A morale penalty could be included in the paladin kills paladin case, but then one paladin is not really supposed to kill the other. The penalty would only apply for good killing good, since evil tend to like killing all enemies, whether they are hated evil rivals, or good enemies.
2. It's still in the best interest of the other side to kill the hero. I've seen some cases where they just killed the leaders, and let the troops go. this because the leader is the person causing the long term threat. Both Good and Evil would require a Ransom for the Heroes. This was a common practise in some cultures, but not used that often in others.
The prison is there for holding the hero until the ransom can be paid, and the hero freed.
The problem with killing heroes when winning is that sometimes neutral stacks can win over hero stacks, and it's bad for the game if the neutrals will auto-kill the hero. Perhaps, neutral can also offer ransom for Hero.
"Pay within a week, or send Hero back in pieces!"
The player can get back hero in several ways:
- A stealth hero might be able to rescue the prisoner.
- An army can wipe out the neutrals.
- The player can pay the ransom.
- The Hero escapes of own accord (should be unlikely, and Hero should probably have some specialized skill to pull it off.)
Neutrals could be of different types, and perhaps be marked a bit differently on the map. Like:
- Always kill Hero on win.
- Always kill Hero of opposing alignment on win.
- Always hold hero for ransom.
3. On ransoms. if a holy man was held ransom I think it was often the church that paid. For the best military leaders. I'd say the army. If a civilian was captured and held for ransom then it was relatives. I don't see it as a problem that the employer (=the player) pays.
4. I don't think you need to build a prison to keep prisoners. There will be cells to hold prisoners anyway. However, if you want "high security" cells then it could be an add-on build.
5. Losing levels is tricky. Are you going to remove learnt secondary skills? Are going to keep track of other possible random bonuses the hero gained at the level-up?
If you torture a Hero, then the likely event is that the hero escapes at first oppurtunity. You might get obedience by lips, but wouldn't get obedience by heart and will.
____________
"A brilliant light can either illuminate or blind. How will you know which until you open your eyes?"
|
| |
|
|
