Thanks for your comments and taking the time to elaborate on them.
Very interestingly, some of things you suggest are actually implemented and are the cause of some of the oddities about Research. Though there is certainly a lot of debate about exactly how some of the effects should be applied, or how complex the whole thing has become (ie 'black box').
The 'fixed' max for the research spending (the 'overall cap') is a function of a few elements, though primarily the total size of your economy, and the number of research centres. The latter item is self-evident; the former, the economy size, is part of what makes it harder to work with, since that factor changes so much depending upon whether you play a large region or small one.
Simply put, SR2010 makes it so that you can't spend more than x percent (the number itself is affected by a lot of factors, and I don't have the info handy, for for arguement sake let's say it's 25%) of your total region's GDP on research. The foundation for this rule is the assumption that there has to be a ceiling after which the trained people, support staff, available research infrastructure, run out. There has to be a point at which spending more money simply won't get more results, and that is what the cap is for.
Your example of the Manhattan project is a perfect one for why some of the research restrictions are in place in SR2010. Why didn't Iraq succeed with their own 'Manhattan Project' in the 1980's? Even with the billions spent, they could only go so far so quickly.
This is also the reason for our 'max you can spend on a single project per day' cap - Just because you write out a check for $5 Billion doesn't mean you'll get your nuclear tech facility on your doorstep tomorrow. The Manhattan project was given a huge budget but still required time to get its results - spending more money would not get faster results. If you do try to spend more than your projects can 'absorb', at least we send it to general tech level ('theory research') and efficiency improvements. (In real life it might just end up in a politician's pockets, but we pretend everyone is honest in the world of 2010...)
And when you go back to WWII, history shows us that both the US and Germany had one big 'Project' - the US chose Nuclear Weapons, the Germans chose rockets and missiles. Each poured huge resources into their Projects, and in fact each succeeded at mastering their chosen research areas and eclipsing the other. However, neither one could afford the resources (money, scientists, support structures) to do TWO big projects. The US chose nukes and their rocketry program was a joke; the Germans chose rockets and their nuclear program was a shambles. Yes, they both also researched other goodies - jet planes, tank guns, radar, blah blah, but when it comes to the big ticket stuff, there were limits.
We do, in fact, use the bell-curve approach you suggest, though again this is something that gets us criticism because it is hard to 'see' and understand. You can research a single item faster by making it your only project, exactly as you suggest, but you do encounter some of the 'down the drain' spending you ask for. How much? Where's the sweet spot? Well, we just don't have the ability to report that information in the research screen, even if we could find a way to say it clearly. Among other things, the 'sweet spot' of how many research slots to fill is affected by a lot of other factors, including your economy size, efficiencies, etc.
Sometimes you have to be careful what you wish for - we implemented a lot of what you're asking for, but the end result is generally considered a bit confusing and difficult to use.
One oddity about how the current research screens work is that estimates (ie number of days, or cost, etc) are only accurate when the situation is stable - because of the above factors (bell curve, benefits of less projects, spreading out spending, etc etc). So when a new project is added or removed, the estimates are not accurate until a day or two has passed to allow everything to "settle in". Again, this is a cause of complaint, because we can't show an accurate assessment of how the new project affects your whole house of cards until the project is actually begun. So as you say, the days/costs shown before adding a tech often change drastically when the tech is added to your existing mix, since you've unbalanced your status quo - the bell curves, the spending splits, etc all need to readjust.
You mention variability in research tech times, and except for the use of random events (the "oops" and "eureka" factors), we already do that - research will slightly be affected third party elements such as civilian approval, government type, labour force use, etc. Again, a good suggestion of yours, but it means that research of the same 'project x' will take slightly different times and costs in different regions, causing confusion.
Should we make research more 'random event' based? "The joys of research breakthroughs"? If we do, you can bet as many people will complain about that as will like it...
Here at BattleGoat we discuss Research often, and one bit of consensus is that it needs to be clearer and more understandable. However, you also asked for it to be "more realistic". I would suggest that maybe we are too realistic, and we need to make it more "artificial" so that it is easier to understand and predict.
Last edited by George Geczy on Feb 09 2006, edited 1 time in total.