tech limits at WS+2 and beyond (Full Version)

All Forums >> [New Releases from Matrix Games] >> Gary Grigsby's World at War: A World Divided



Message

WanderingHead -> tech limits at WS+2 and beyond (10/1/2007 10:05:59 AM)

The maximum per-turn expenditure for a given unit (except heavy bombers) is:

level<=WS: 3
level=WS+1: 6
level=WS+2: 3
level=WS+3: 4
level=WS+4: 5

In general, for level>WS+1: 1+(level-WS).

In my experience, this is like a full stop at WS+2. Other than heavy bombers, you never get passed this level because it just takes far too long.

I think it would make sense to change this so that for level>WS+1: 5+(level-WS). Then we would have

level<=WS: 3
level=WS+1: 6
level=WS+2: 6
level=WS+3: 7
level=WS+4: 8


In this way, you'd still be slowed down, inhibiting the super units. But in the late game you'd still have a chance to make meaningful tech advances.

Any thoughts on this?




Marshall Art -> RE: tech limits at WS+2 and beyond (10/2/2007 1:49:47 PM)

I am in favour of this change. It makes research worthwhile even in 1944 and after. Of course the WA will profit most since they should have enough production available but also Germany could push a few projects forward. We would see a tech race until the end and not a pure war of attrition until the WA have killed enough Axis units. 



WanderingHead -> RE: tech limits at WS+2 and beyond (10/4/2007 9:34:01 AM)

quote:

ORIGINAL: Marshall Art
I am in favour of this change. It makes research worthwhile even in 1944 and after. Of course the WA will profit most since they should have enough production available but also Germany could push a few projects forward. We would see a tech race until the end and not a pure war of attrition until the WA have killed enough Axis units. 


Unfortunately, this is one of those things that needs a lot of voices to change.

It does bother me that the minimum time to go up a step suddenly jumps from 3 seasons (for level < WS+2) to 6 seasons (for level >= WS+2), so much so that I may just have to make this mod-able if there isn't enough support to actually change it.



ArticFire -> RE: tech limits at WS+2 and beyond (10/4/2007 6:01:07 PM)

Seems like a good change to me personally.



GKar -> RE: tech limits at WS+2 and beyond (10/4/2007 9:30:45 PM)


quote:

ORIGINAL: WanderingHead

level<=WS: 3
level=WS+1: 6
level=WS+2: 6
level=WS+3: 7
level=WS+4: 8

I wouldn't mind even bigger increases, but I'll take anything I get. [:)]



WanderingHead -> RE: tech limits at WS+2 and beyond (10/4/2007 11:58:14 PM)

quote:

ORIGINAL: GKar
quote:

ORIGINAL: WanderingHead

level<=WS: 3
level=WS+1: 6
level=WS+2: 6
level=WS+3: 7
level=WS+4: 8

I wouldn't mind even bigger increases, but I'll take anything I get. [:)]


In AWD it was specificlly reduced from a basic 3*(1+level-WS) limit, to slow you down after WS+2 and prevent superunits. That objective still makes sense, but I think it went just a little too far.

The current implementation, a level takes 3 seasons until WS+2 then 6 seasons. With my proposal, it would be 3 seasons until WS+2 then 4 seasons.



Marshall Art -> RE: tech limits at WS+2 and beyond (10/6/2007 11:42:02 AM)

The change above indeed would be a good compromise between the initial too-easy research and the current too-hard limits IMO. Considering that typically there are several or even a dozen or so units on board (i.e. on the map or in production), research in fact still takes one or two turns longer than posted above due to the formula that factors in WS and units count (which I like BTW).




WanderingHead -> RE: tech limits at WS+2 and beyond (10/9/2007 9:53:23 AM)

I think what I would like to see is increasing TOTAL COST for faster development.

Right now, if you spend anything you might as well spend as fast as possible. I think that if there were a cost associated with going faster it would give you the freedom to pursue a few techs quickly (and at the end game you'd need to just in order to finish something in a useful time), or to trade that off for doing more techs slowly.

I think this is the best way to leave strategic options and yet discourage superunits.

I've been thinking about how to integrate this with the existing UI, and what I've come up with so far is a random research bonus whenever you buy research, which would decrease in probability when you buy more research. In conjunction with increased net costs, of course.

For example, roll die(next_threshold/2) and if the roll is greater than the number of points you've invested THIS TURN, and if you invested a non-zero amount this turn, then give 1 extra research point to this attribute. Change the cost from
(level-WS)*(5+num_units/5)
to
(level-WS)*(8+num_units/5).

What it means is that if you invest only 1 research point, you usually get another one for free. But if you invest the maximum number of research points then you usually don't get another one for free. Research would be as much as half as expensive if you do it slowly, instead of quickly. Which encourages you to distribute it and do many attributes, but without forcing you or limiting choices like the limits used today.

No changes to the UI. And also, I mist admit that I like adding a little more randomness to the research process.

Not likely to happen, I suppose, but I think I like the idea.



Marshall Art -> RE: tech limits at WS+2 and beyond (10/10/2007 10:27:57 PM)

Actually an interesting idea, but I would suggest to turn it otherwise - the more points you purchase the less of them you really get. Research slowly and you can almost count on each point invested (say 90% chance). buy 6 points and the probability to really score 6 hits for RP is low, you rather have 3 or 4 points certain, and maybe one more or less depending on luck. The higher the resources spent, the higher the risk of failures in research, as you have to try several paths at the same time. Just an idea.

Anyway the intention to introduce some randomness is a good idea because as it is now I can precisely plan when to have a new tech level available, e.g. advanced CAG torpedo attack for Pearl Harbour, or Fighter range for Sealion. Typically scientists do not deliver just in time like that...



WanderingHead -> RE: tech limits at WS+2 and beyond (10/11/2007 9:18:06 AM)

quote:

ORIGINAL: Marshall Art
Actually an interesting idea, but I would suggest to turn it otherwise - the more points you purchase the less of them you really get.


I think you are right, that is better. Especially because it does not involve changing the nominal cost, which my proposal required.

Let me think some more on a specific way to assign the probabilities.



Lebatron -> RE: tech limits at WS+2 and beyond (10/13/2007 4:08:04 PM)

I really don't feel more luck should be introduced. Perhaps that's just my Chess mentality speaking. In the A&A circles of vet players, I was one of those that converted over early to the low luck battle system and never looked back. It really helped the superior player win instead of losing because of bad dice rolls. What I liked about WAW from the beginning was that a low luck battle system was standard, thus forcing everyone to play the same system. Unfortunately, in A&A the field of vet players is split. Concerning how this applies to the tech system in this game, getting the free tech jump when your 1 point short is more than enough luck for me. Adding more chaos here does not look appealing to me. However, if a fixed scale of you spend X you get Y was developed I might like that. 



GKar -> RE: tech limits at WS+2 and beyond (10/13/2007 4:28:48 PM)

I agree with Lebatron, the "random tech jump" is enough for me as well.



Marshall Art -> RE: tech limits at WS+2 and beyond (10/13/2007 6:39:56 PM)


quote:

ORIGINAL: Lebatron

What I liked about WAW from the beginning was that a low luck battle system was standard, thus forcing everyone to play the same system. Unfortunately, in A&A the field of vet players is split.


I think the argument is not quite convincing -

Everything, from the battle results, over diplomacy, to spying is purely based on luck, i.e dice rolls and probabilities. The ONLY systems that do not use the element of luck are the production system and the supply system. The research system is split as you can save a point when you get the overflow to complete the tech.

So while virtually hundreds of dice are rolled and probably not less probabilities calculated each and every turn, you still feel there is little luck in the system? I think it would be just fair to introduce SOME level of luck to the systems that are purely chess-like, since it feels not right to be able to plan production and research in full detail while all other elements of the game are always somewhat randomized.



WanderingHead -> RE: tech limits at WS+2 and beyond (10/14/2007 9:56:05 AM)

To be clear, noone should expect a change like this in the stock scenarios any time soon (unless everyone agrees, and I can't see that happening). Mods only.

As for luck, I have several arguments for it in this case:
  • war is not like chess, there are real surprises, and not just surprises that are planned by the enemy.

  • anyone who works in engineering can attest to the difficulty of maintaining a schedule, particularly when faced with solving unknown and unforeseen problems. It simply cannot happen that in Summer 1939 you can predict "by Fall 1940 we will have proximity depth charges developed, tested, bug free, and distributed through 75%+ of the fleet" with any confidence. You can't necessarily predict the coal mine fire that shuts down your factory and delays things 4 weeks.

  • in this case, you have many chances to make research every turn. Over the long term, it will average out. The real impact would be that you would either have some uncertainty of when you get something, not likely to be much more than +/-1 turn, or you pay extra in order to buy certainty. That uncertainty is quite realistic.

  • the randomness I envision would only be incurred when you try to do research very quickly, the goal being higher average expense for faster research, cheaper for slower research (with slower research also likely implying more diversified research).
What I have come up with so far increases both the average cost AND the uncertainty when you try to research more quickly.

Let

m=max(1, 1+level-WS).

This is nothing new, simply the existing basic scale factor, if we remove the artificial low cap at WS+2 and above then the research limit is

hard_limit = 3 * m

E.g. at WS and below hard_limit=3, at WS+1 hard_limit=6, at WS+2 hard_limit=9, etc. Keep this limit the same, but also apply the following random limit, which is unknown to the player:

rand_limit = 2*m + rand(3*m)

The player is allowed to do research up to the hard limit. At the end of the turn, when production is finalized, the implemented limit is

imp_limit = min(hard_limit, rand_limit)

If the player has committed more than rand_limit points for a field, then those excess points are not awarded but the resources are expended.

With this system, the hard limit is 3*m and you can do 2*m worth of research with no randomness. As soon as you exceed 2/3 of your maximum expenditure, however, you have a chance that you are not going to get all that you are paying for. At low levels, with hard_limit=3, if you spend 3 then you only have a 1/3 chance of losing 1 point. At higher levels, say hard_limit=12, then if you exert the maximum expenditure of 12 you still have a 1/3 chance of losing something, but that loss could be from 1 to 4 points. I.e. it can be bigger.

The effect is to increase the cost and the uncertainty only when you try to go fast, not when you go slower.





Lebatron -> RE: tech limits at WS+2 and beyond (10/14/2007 10:43:01 PM)


quote:

ORIGINAL: Marshall Art


quote:

ORIGINAL: Lebatron

What I liked about WAW from the beginning was that a low luck battle system was standard, thus forcing everyone to play the same system. Unfortunately, in A&A the field of vet players is split.


I think the argument is not quite convincing -

Everything, from the battle results, over diplomacy, to spying is purely based on luck, i.e dice rolls and probabilities. The ONLY systems that do not use the element of luck are the production system and the supply system. The research system is split as you can save a point when you get the overflow to complete the tech.

So while virtually hundreds of dice are rolled and probably not less probabilities calculated each and every turn, you still feel there is little luck in the system? I think it would be just fair to introduce SOME level of luck to the systems that are purely chess-like, since it feels not right to be able to plan production and research in full detail while all other elements of the game are always somewhat randomized.


I was not saying there is no luck element in AWD. I know additional luck factors were added to AWD that WAW did not have, like diplomacy, spying, and WR increases. But these are side issues compared to the battle system, and it was the battle system I was discussing when I compared A&A to AWD. What AWD has that A&A does not, is that each combat unit has a bell curve to determine chance to hit. When you roll multiple die for each unit instead of a single die you are reducing the chances of wild outcomes. Yet some uncertainty does remain to satisfy the unpredictable nature of war. Before I started using the Low Luck option in A&A wild die rolls ruled the game. Your skill at playing the game would not amount to anymore than 50% of your chance to win when playing an equal opponent. However, when using Low Luck in A&A skill would amount to 90% of your chance to win. In other words, the randomness of the dice under Low Luck had little weight on the outcome of the game. You did not get excuses from losers when Low Luck was used. In my opinion, this made Low Luck far more appealing for tournament level play. AWD plays like A&A using Low Luck in the sense that skill weighs on the final outcome far more than luck of the die. Adding more luck elements only shifts the skill to luck ratio in the wrong direction for tournament level AWD play.



Lebatron -> RE: tech limits at WS+2 and beyond (10/14/2007 10:49:24 PM)


quote:

ORIGINAL: WanderingHead

If the player has committed more than rand_limit points for a field, then those excess points are not awarded but the resources are expended.

With this system, the hard limit is 3*m and you can do 2*m worth of research with no randomness. As soon as you exceed 2/3 of your maximum expenditure, however, you have a chance that you are not going to get all that you are paying for. At low levels, with hard_limit=3, if you spend 3 then you only have a 1/3 chance of losing 1 point. At higher levels, say hard_limit=12, then if you exert the maximum expenditure of 12 you still have a 1/3 chance of losing something, but that loss could be from 1 to 4 points. I.e. it can be bigger.

The effect is to increase the cost and the uncertainty only when you try to go fast, not when you go slower.



A new tech system that doesn't always give you what you invested back may not be a good idea for this game. In operational level or below games, something like this is usually used. However, AWD is grand strategy with turns representing long periods of time. What we have in place now gives us an uncertainty of 3 months when we are just shy by a point. More than enough I would say in a game with only ~25 turns in it. If it had 100+ turns than more uncertainty should be added to when developments are made.

I also see a major problem with the system you are proposing. It would completely change the choices we make when deciding where to deploy our forces. How could any of us be sure we were going to hit a new tech level before deciding to press an attack. Lets say I wanted to attack and take over Stalingrad. I know that from looking at my tech screen that I'm 6 points from reaching 8 EV for my infantry. Thus I can decide it's worth the risk because a Russian counter against 8 EV infantry would be weak. But under this new system, the 6 points I spend may only give me 4 or 5. Knowing this, I would not chance the attack. Some of you may say fine that's sounds good. The problem is that it removes the initiative from Germany. When it was typical to press the attack knowing that you would have better EV or something else at the end of your turn would no longer be the case. You guys really have to ask yourself, do you want to see what I consider a signature aspect of the WAW design to change?

In order to preserve the Axis initiative using a new system for tech would require a large amount of playtesting before it begins to feel like it does now. Chances are that certain advances like 8/8 infantry or 10/9 tanks that you developed a build plan for would become obsolete. Total cost to develop and when they become available would be very hard to keep within current game specs. What would happen if it cost the German player an additional 6 production over current cost and a delay of one more season to reach 8/8 infantry? Most likely he would lose the initiative in Russia, and any hope of a draw or better, all due to more chaos being added to the game.



WanderingHead -> RE: tech limits at WS+2 and beyond (10/15/2007 5:39:36 AM)

quote:

ORIGINAL: Lebatron
A new tech system that doesn't always give you what you invested back may not be a good idea for this game.


That's definitely the wrong way to look at it. It means an investment is a chance for a return, not a certainty of return. Which is true of almost all real investments, except those with low returns (like, say, certificates of deposit). And that is exactly why those have low returns.

The exact mechanism I mentioned can use some work. The fact that it reduces the max achievable rate early on could be undesirable (but it actually allows you to go faster later). But in general, the issue you highlighted should be solvable by simply saying: if you want certainty, you just have to pay for it. Were I planning a risky offensive that relied on new gyroscopic targeting systems for its ultimate success, I would pay extra to ensure they got there, and other programs would most likely necessarily suffer.

Again, this won't even effect the stock scenarios. Indeed, it may not get done at all. Just a discussion stage here.

My real desire is to somehow make faster investment cost relatively more, and slow investment relatively less. This is itself realistic, and is a better way to encourage diverse investment than the current way, which just limits the player's choices in the late game. The random aspect is really just the mechanism I hit on to accomplish it. Any other ideas?





Marshall Art -> RE: tech limits at WS+2 and beyond (10/15/2007 11:02:30 PM)


quote:

ORIGINAL: Lebatron


What we have in place now gives us an uncertainty of 3 months when we are just shy by a point.

How could any of us be sure we were going to hit a new tech level before deciding to press an attack. The problem is that it removes the initiative from Germany. When it was typical to press the attack knowing that you would have better EV or something else at the end of your turn would no longer be the case.




I am happy with the way it is but I would feel better with the changes proposed ASSUMING we get them agreed upon.

First, I never even hope for the lucky roll to get the last poin for free if I WANT the tech. I rather purchase that on RP to be sure. So there is NO luck in reality within the context of this discussion, which is how to develop a tech quickly AND still not be 100% certain how much it will really cost and how long it may take (exactly).

Also, while Germany might get a tech here and there one turn later (open for debate, see below), most certainly the Allies have a much higher price to pay since they have to invest much more into various tech disciplines just to catch up with Germany. Just the A-bomb research alone would probably cost several RPs more per level and would become available about 2 or 3 turns later compared to now. So in fact the WA might be more hurt by the change than Germany, contradicting the statement that Germany would be put into an disadvantage.

Now as far as the delay of tech availability goes - if I know I need say 15 research points in total, I can see that being not able to go 3 + 6 + 6 (=15) but rather go (2-3) + (4-6) + (4-6) + (4-6) would be a delay of one turn. However, there are many more situations where there would not be any delay necessarily, for instance at 18: 3 + 6 + 6 +3 versus (2-3) + (4-6) + (4-6) + (4-6). Note that (x) means that the number x is the range of RP you actually get. So there are times where research takes longer but there are also times where this does not happen, however the investment is slightly increased since most people would buy at the safer end (i.e.higher) of the possibility curve to make sure they get the tech as needed.

Last, if I play Germany, of course I try to "optimize" research results to match with important battles, whether against Russia or the WA. BUT I would never give up my initiative (if I have any left), just because I am missing one tech level here or there compared to what I wished I have.
If I have the initiative, I will be victorious anyway, otherwise I have lost the initiative already and the attack was a bad one from the start. If I start quibbling over the gain-loss ratio and hold back from attacking, the Allies are on the road to Berlin. The only slight difference may be that I have some higher losses, but this argument can be put in reverse againt the allied counterpush as well.



WanderingHead -> RE: tech limits at WS+2 and beyond (10/16/2007 8:27:23 AM)

quote:

ORIGINAL: Marshall Art
I would feel better with the changes proposed ASSUMING we get them agreed upon.


That agreement would be the tough part.

I've had a hard time coming up with anything better than my previous suggestion, bearing in mind keeping the UI relatively unchanged.




Marshall Art -> RE: tech limits at WS+2 and beyond (10/16/2007 11:46:08 PM)

I think there always will be pros and cons for all changes suggested, so maybe we should stick to fixing any bugs encountered rather than try to come up with "enhancements". If the items discussed were free to be modded it might still be a worthwhile discussion?



Page: [1]

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
0.78125