Bruce70

I do not agree. I recently invested 1 point in Jets as the Brits early in the war. I got 2 advances in successive turns and had level 5 by the end of 41. Sure I got lucky, but should it even have been possible. With the system I suggested above it would not have happened. I tend to agree with Mike that it simply shouldn't be possible to get an advance in 1 turn. IRL even if an advance came quickly there would still be time required to implement that advance. But I don't think we can or should expect more than a quick fix for SC1 and I can live with it as is. For SC2 I would like to see research tied to experience. So that if you are using lots of tanks (for example) then tank research will come more quickly.

There was a recent thread about research being a little quick. I am only partway through my first full game and I am not sure myself whether or not this is so, but I had an idea to slow it down a little anyway. If you have 5 points invested in a particular research area and you get an advance perhaps the number of points you have invested should reduce by 1. So if you only have 1 point invested then after you get the advance, research in that area would stop completely. This would mean that you would have to invest more in research but only when you actually get an advance. This could be thought of as simulating the cost of implementing the new tech. Oviously this would have to be play tested to determine whether or not the cost of research points would need to be reduced or not so maybe its best left for SC2.

Or just use a LAN and make your PBEM directory shared.

I'd actually prefer less action points in winter. As it is winter goes by without me even noticing. Its not a big issue, I'm happy with it the way it is, but increasing the turn duration is not the same as reducing the number of action points, which I think would be a better winter solution.

I'm confused, do you play with a time limit or not? Most of my posts re IP play have stated "short turn duration IP" or something similar and where I have not been specific it was implied. PBEM games do not allow unlimited time to contemplate each turn. e.g. I sleep for 8 hrs/day; go to work for 8 hrs/day; travel, eat and look after my kids for about 4 hrs/day; watch TV for 1 hr/day and play up to 5 PBEM games and some single player CM for about 2 hrs/day. (God knows what happens to the other hour) So my PBEM turns ARE time limited to about 15 minutes per turn. And if you think I sit at work all day thinking of CM and posting comments on the forum... hold on, thats what I'm doing now - damn. [ August 13, 2002, 08:30 PM: Message edited by: Bruce70 ]

Now I know why you are called "freaky".

If you were foolish enough to read the whole thread you would find that yes it has been raised before but some people just don't get it. Anyway because I like a good debate I will continue to try and make the point... In CM you can only issue orders at the end of a turn (up to one minute delay) further there is an order delay depending on command etc. This simulates imperfect communication, you don't need to simulate it again (if you think the order delay needs to be extended that is another issue). Furthermore a RL tank commander is intelligent enough to know where a threat is likely to come from and will act accordingly without the OIC doing ANYTHING. And also if you find while playing your IP game that your infantry have spotted a tank do you ignore that information for units on the other side of the map? Neither IP or PBEM is going to give you a RL experience (On that I am sure we all agree). The best you can hope for is that the outcome will be comparable to the outcome that you would expect IRL. IMO playing without an unreasonable time limit gives you the best chance to produce realistic results. This is a very valid point that has been made before and I tentatively agree with this. Why tentatively? Because it remains to be seen whether people who choose to play IP as opposed to PBEM are better players (experience is no substitute for common sense), and because it remains to be seen whether being a good IP player makes you a better PBEM player or vice versa. P.S. A general comment (not specific to this thread): I get a little sick of people saying things like "Don't raise this again just read the post 2 years ago" and "Don't stir up trouble" or "this thread is getting out of hand" etc. There is nothing wrong with rehashing old posts and since most threads have poor subject headings it is difficult to find anything using a search. Also if you are bored with a thread don't read it. As long as a thread doesn't become offensive or personal (and only 1 or 2 posts in this thread have been) there is no need to stop a discussion that some people are enjoying.

IRL you have hundreds of intelligent soldiers using their eyes to gather information. Similarly it is unlikely that IRL these would be missed as often as in CM (including PBEM). I don't know anyone sad enough to spend an hour looking for them either. Which I think is why playing PBEM is closer to RL. Well I think the result is closer to RL, not the actual gameplay obviously. Real soldier on the other hand do and in fact do so without even thinking about it. To have a realistic game you need time to be the eyes, ears and brain of every soldier on the battlefield since the AI does not currently do this for you or only does so in a very limited way. People under pressure do miss things but real soldiers use their own initiative in the absense of orders or at least try to obtain orders if none are forthcoming. In IP games, it seems from this discussion (I don't play IP) that time limits are chosen to create pressure. Therefore the ability to deal with time pressure is an important factor in IP games and the same is true IRL. But that doesn't mean that short turn duration IP games are more realistic and it also doesn't mean that you will be a better player in PBEM games or a better commander IRL (nor does playing PBEM games). [bTW just to give a ball park figure I would say that if I am attacking in a PBEM game then my turns would take between 2 and 10 minutes depending on the amount of action and I would view the replay 1 to 3 times depending on the map size.]

You could also argue that a PBEM player who plays several games in parallel gets a tougher work out than an IP player who plays only one game at a time. I think most people prefer PBEM because they don't have the same amount of spare time as IP players. Hence they probably spend as little time on their turns as possible. P.S. Still waiting for that game "Legend".

I'm not the greatest PBEM player (never player TCP/IP) but I'll challenge you to a game (QB of your choice) and since the format doesn't matter, lets make it PBEM 1 turn per day. You may impose a time limit on your turns if you wish.[/QB]</font>

I'm not the greatest PBEM player (never player TCP/IP) but I'll challenge you to a game (QB of your choice) and since the format doesn't matter, lets make it PBEM 1 turn per day. You may impose a time limit on your turns if you wish. Now having said that I personally believe that PBEM is not less realistic since having unlimited time and more situational awareness compensates the player for all the advantages that a real life commander has. For example a RL commander has: - A better feel for the terrain since they are there. - intelligent subordinates - feedback from subordniates - no restriction on when orders may be given - no ned to issue orders to every single unit - probably a lot more A TCP/IP game with an unreasonable time limit changes it from a simulation into C&C.

I know this is an old post but I fail to see how BB's statement is an analogy, let alone a false one. In fact I think it is probably spot on (unless the majority opinion has changed since the release of the full game).

An idea...if there are not to be any paratroopers, then maybe BOTH the Axis and Allies (read the player and the AI) could start the game with, for example, three paratroop/fallschirm 'counters'. This 'counter' would be placed on the players main interface (the game screen), starting at three, and then at the beginning of each turn, the player (or the AI when its' the computers' turn) could decide if he/she wanted to use one of the the 'counters'... This is not a bad idea for some other WWII game with a 50+km hex, it certainly seems to be a good compromise between the "paras are not appropriate at this scale" group and the "I want 'em anyway" group (although I expect there would still be disagreement about the type/amount of attack bonuses). However, this is just one extra detail of many that various people would like to see added to the game. The beauty of SC is the simplicity, adding extra "features" would run the risk of making it 'not SC'. Has anyone heard of the saying "if it aint broke, don't fix it"?

Yes for SC2 I agree with you, perhaps. But for a quick patch, changing the percentage will reduce the chance of getting historically inaccurate results. Abstraction is the beauty of SC, you have to imagine that the increased strength of a unit is a result of receiving some of the new type of unit.

Wouldn't it be a lot simpler to just reduce the %chance of an advance from 5% per unit investment to 4% or even 3% IF reaearch is widely considered to be a problem (I cannot say one way or the other since I am still waiting for delivery of my pre-order)

And the US production was shared across two theatres. Also MPPs are an abstraction that are supposed to take into account available manpower as well as production. This is not a perfect system but it seems to work pretty well.

I think you should be able to set certain SOPs before a battle but I think it would detract from the game if you were able to change them during the battle.

Then invariably some folks come forth to claim that they see it regularly in normal game play. I would like to apologise for being one of those "folk". I mistook a dasiy chain mine for a AT mine field. I was under the impression that daisy chain mines were marked as such instead of the generic AT label. My tests confirm what everyone else is seeing, ie no spotting (before detonation) of any mines bar daisy chains. As a side issue, I notice that engineers do not require demo charges to clear daisy chain mines, which makes sense I guess.

Surely you must know someone else local who is buying the game. I live in Tassie and even I managed to find one other person to reduce the shipping cost.

I can not say for sure that I have seen engineers spot AT mines in CMBO. But very recently I did witness armoured units spot AT mines WITHOUT running over them. Two tanks (A Sherman and a Stuart) may have caught the edge of a mine field and detonated a mine causing no damage but not only that minefield but another further away was also spotted. There is no way the tanks could have run over a mine in that mine field. Also I routinely see AP mines spotted by all sorts of units, not just engineers. It is not always possible to be sure none of the mines were actually detonated (not something I have payed particularly close attention to) but it would seem not judging from the lack of casualties. I would tentativel suggest that in CMBO armoured units have some chance of spotting AT mines and infantry have some chance of spotting AP mines.

Good news Hubert. The real problem with having no safeguards against cheating is that secretly everyone suspects everyone else of cheating. It isn't good for morale.

I play all my games against the computer with the computer having +50% bonus or more if the computer is attacking. That way even if the computer attacks poorly its still fun nailing all those tanks and APCs. I don't often lose but get a few draws, I think I'll try a short time limit, that sounds like a good idea for attacking.

Affect is a transitive verb that means: to have an influence on or effect a change in. "Inflation affects the buying power of the dollar." Effect is a noun... Anyone spot the error here?

At the risk of boring everyone to tears, I will answer some of your objections Redwolf. So that makes the terrain the most important aspect I think I would agree with that. How would you encode the terrain as input for a neural network. You would have to train it with a big variety of good/bad examples and with an even much larger variety of terrain examples The choice of encoding would be very important, I am not sure what would be best. Some trial and error would be required if a detailed search of the available literature did not uncover anything. At least that is from a developers POV. For myself spatial reasoning is what my PhD is about so naturally I have my own ideas on this. The simplest way to train a RL NN is to have it play itself in a QB type scenario - but without the graphics of course. The company programmed the hell out of a conventional system, but that could only tell deforming were there, not whether they were allowed or not. My friend's neural network based program could do this. Yes I am not surprised that the NN could do better than a rule-based system. Now, and here we come back to our to our CM application, the key in making this work was not at all int he neural network itself. It was a pretty vanilla one and I think you would even have be able to use a commercial offering. Yes it was more than likely a BP network or some related algorithm. This algorithm was developed in the early 80's (?), NNs have come a long way since then. The key is how do you encode real world data into streams of numbers so that distinctive patterns of numbers reflect differences you care about. This is what he banged his head against for 2 years, and he's a real hotshot programmer. NN programming is easy, that is one of the reasons I find this field rewarding since I myself am an average programmer. As you say the difficulty is in choosing a suitable encoding and there is a certain amount of art (for want of a better word) in that. you would need to have the CM community plot like 20000 attack approches. As I said before, have the computer play QBs against itself. The developer doesn't need to come up with good or bad examples, you know if an example is good or bad by who wins. I think you are focusing on the NN too much, you should do some reading on RL (reinforcement learning) in particular have a look at Sutton's "temporal difference learning". RL doesn't even require a NN but it is common to use a NN for the value function especially if the state-space (situational input) is of a continuous nature or prohibitively large (both apply to CM). If the possible actions are of a continuous nature then you do run into some problems and this is the subject of current research. But it is usually possible to design the system so that actions are discreet as I believe is the case here. First I think that the reduces the CPU usage not enough, but it is hard to argue about that. The point is that since the high level AI gets used rarely you can put more (CPU) effort into it. It is this part of the solution that most people seem to be complaining about with other AIs. Not every level of the AI needs to be the way I have explained and that leads to another interesting option. You write a rule-based AI for every level of the code (this doesn't have to be very good) then starting wherever you like (at the highest level probably) you replace that section with your NN/RL code. Once you have trained that level you replace another level and so on. You may even find that at some levels it is easier to keep the rule-based code. More importantly, this AI would probably do a too even approch, violating the (military) principle of mass. I am not sure what you mean, if the principle is valid the AI will find it. But there is no harm in giving the AI a small knowledge base of military principles. There is no need to reinvent the wheel. If there is a SOP that fits a certain situation then you give this procedure a high value before the AI starts training. The AI will be able to change this value assessment based on its own experience but it does save a significant amount of training time and is a common technique. But you have limited your high-level HQ to have blurred vision, and your low-level commanders are short-sighted, they only live from one terrain cell to the next. It is a feature not a limitation that the high level AI has "blurred" vision. The high level AI does not need to know where every patch of scrub is, as in real life that is a job for subordinates. If every platoon leader started sending radio messages to the battalion commander about the exact placement of trees and shrubs on the battlefield I don't think this would enhance his ability to do his job. It is a valid point about the short-sighted subordinates though, perhaps it would be better to have a LOD system for terrain input in this case. So the AI has detailed information about the immediate surroundings but only general terrain information about more distant features. Gotta go, apparently I am not paying enough attention to my wife. [ July 27, 2002, 03:20 AM: Message edited by: Bruce70 ]

The easy way is to declare war on them. But they will declare war on you if you attack all their neighbours. Particularly the Baltic states, Turkey or Iraq.