Research Results

[Dan Fenton]

I would like to get opinions on a theory of mine that certain reseach gains occurs at a higher than expected rate and that other research gains occur at a lower than expected rate. This is after compensating for Tech Level differences.

UK: Above average - Gun Laying Radar and Anti-Tank Weapons.

Below average - Long-Ranged Aircraft and Industrial Technology.

Germany: Above average - Anti-Aircraft Radar, Jet Aircraft, and Gun Laying Radar.

Below average - Anti-Tank Weapons, Heavy Tanks, and Long Ranged Aircraft.

Italy: Above average - Anti-Tank Weapons.

Below average - none that I am aware of.

Soviet Union: Above average - Jet Aircraft.

Below average - Heavy Tanks, Industrial Resources and Rockets.

USA: Above average - Anti-Tank Weapons and Jet Aircraft.

Below average - Heavy Tanks and Industrial Resources.

What are your opinions?

[JerseyJohn]

Pretty much agreed. Italy also seems to consistantly develop Jet Aircraft pretty quickly.

Part of these jumps might be because countries tend to catch up quickly where their enemy holds a technological edge. Hubert put this in writing a few months back; in effect examining captured weapons, reverse engineering and spies.

Since Italy, for example, is likely to have Jets L=0 when Germany and Britain have Jets=2, she'll probably have L=3 when they have L=4, closing the gap. The same would hold true in all the other categories since Italy is lucky enough to start out at L=0 in everything!

Conversely, countries that start attain a quick jump to say L=2 within a year or so, might have a long time to wait till they reach L=3. This might be especially true if nobody else has much in that technology. Say the Brits with sonar and the Germans with U-boats.

[ May 24, 2003, 12:53 AM: Message edited by: JerseyJohn ]

[Dan Fenton]

The reason why I started this post was that I was not getting any Tech advances in certain areas for campaign game after game. I make the majority of my Research Point purchases in 1940. Most of my games last until 1944 or 1945. At a statistical chance of one advance per game year, it did not seem possible to go several entire games without getting a Tech advance in a particular area. I wondered if other people were having this problem.

The Tech that triggered this topic was Long-Ranged Aircraft (LRA). I tried several patches ago to research it as the United Kingdom (UK). It didn't work, unlike other Techs so I quit trying after one 1939 game.

Then I read Terif's post of the virtues of LRA for the UK. I tried for two more games without any success to get LRA for the UK. My current half finished game is the same story. I have also tried as Germany and only got one LRA advance for the entire game. LRA seemed like a cursed Tech.

There is one odd fact. When I ran my Free French conversion test, for fun I had the UK many turns earlier invest one research point in LRA. To my tremedous surprise, I got an LRA advance 4 times out of 50 times when I was reloading and replaying my saved games. Maybe quiting and restarting fixes something. Maybe the first Tech invested is "blessed".

I have gone for two complete 1939 campaign games as Germany without a Heavy Armor Tech increase. This is even more surprising as Germany gets a +2% research bonus from Russia. The Italian could get Heavy Tanks, but the Germans couldn't.

[JerseyJohn]

I put two chits on a single field if it's a priority, three if I need it badly enough, but more than three seems meaningless.

In most games I build it to the maximum ten with one chit each on a broad field. You might want to try the doubling or tripling idea and see if that helps. Most people agree that more than three is a waste of resources.

Also, if the problem is badly enough I'd recommend saving your game files somewhere and uninstalling SC altogether. Remove every trace of it, then reinstall it from the disc and download the v1.07 patch and it's addition, found several posting below the first update in the FAQ sticky thread -- Hubert's posting. It sounds like something might be missing.

[Terif]

Dan Fenton:

First: Welcome at the SC forum

About research:

My experience from many games show me that there is no above or below average rate for certain tech areas. At least not noticable. I have seen Germany reach Lv3 Heavy tanks with 1 chit, as well as Anti-tank and so on...

If I am playing Allies I usually invest 4-5 chits in Long Range and then I nearly every time got Lv 3-4 in LR. When Axis also invests in LR I need less chits.

But that I dont receive a research success in one tech field also occurs often. At least if I only invest 1-3 chits and my opponent doesnt research it. Since research chances doesnt add up, each turn the chances are the same to get a tech. With 3 chits only 15 %. But the chance that you dont get the tech is each turn 85 %. Usually I get 6-10 advances throughout a game with 10 chits over 4-5 years. They are spread by chance over all fields where I put my chits into (as Axis).

Some of your results can also have to do with the catch up system. If your opponent has a higher tech Lv than you, the chance for each chit is increased by 1 % for each Lv difference.

[ May 24, 2003, 07:20 PM: Message edited by: Terif ]

[Shaka of Carthage]

Welcome to the wonderful world of statistical mathamatics. Leave common sense at the door, because it doesn't apply here.

Here is the basic problem with your theory. You can't prove it. And the reason you can't prove it is because you can't run a large enough sample. And what is a large enough sample? How about a few million attempts. And even that is statistically small.

How do I know this? I used to gamble, then I became a card player. Whats the difference? There is no skill when it comes to gambling, it is all luck. But there is skill when you play a game of chance and you are aware of the mathamatical probablities on winning and lossing. Las Vegas was built on the perception that you could make money if you gamble. And its true, if you are lucky.

Back on topic. Let me give you an example of why your test sample will always be too small. Lets take a roulette wheel. Bunch of numbers, half of them are red and half of them are black. Right? Almost. You have zero, and zero is green. And in some places, there is a double zero, which is also green. Lets assume it is perfectly balanced (cause if the wheel tilted, it would favor one side). All of the wheels in the large Las Vegas casinos are balanced. Notice the display of the numbers that have previously hit? Notice something strange about it? You don't have to look very hard to see a pattern. Same string of numbers appears or the same color appears in a string.

That my friends, is what is known as "statistical deviation". There is no pattern. Just like there is no bias in one nation R&D tech compared to another. It just seems that way.

Flip a coin, only two results. Heads or tails. 50% probablity each and every time. Do it 100 times, and by god, you will see a pattern. Just like the above.

Why isn't it a real pattern? Because that coin, that roulette wheel, and the percentage chance of gaining a tech all have one major thing in common. They have no memory. The past does not influence the future.

If you have read this, forget giving me $5, I want $25. I will send you the address. People pay good money to learn what I have just given you above. But when its sold, they usually sell it to you as a Secrets of Winning at Gambling. And it costs alot more than what you have just paid.

[ May 24, 2003, 06:57 PM: Message edited by: Shaka of Carthage ]

[JerseyJohn]

$5 to $25 Holy Inflation Batman!

For a piddling $3 I'll send you the bio's of my two favorite horses, LoseAlot and DieInTheStretch! R. I. P. in horse heaven (also known as The Glue Factory.

--

Great posts from both of you guys, gues my basic premise is correct, spread them and hope that extra 1% level where you're behind will come in handy.

Interesting that Terif puts that many in one field. Still, can't argue with results, the guy's the undisputed champ.

Shaka Really enjoyed your entry, reminded me of an uncle who always had a system either for the ponys or Vegas. He had a high income anyway so he developed the habit of remembering winnings without counting losses. My last conversation with him was over a jewell encrusted white gold ring he'd bought with some Atlantic City winnings. He told me he'd hit a big jackpot at the craps table and he'd bought the ring with the $5,000 payoff (or whatever it was) and I said, "Ahh, so that means you really spent $23,000 on it." He never spoke to me again. Which is a shame, he was one of my favorites. I was too young to know the value of well timed silence.

[ May 24, 2003, 07:23 PM: Message edited by: JerseyJohn ]

[pzgndr]

Welcome to the wonderful world of statistical mathamatics.

Dan has requested that I add something about research probability results in our update to the Strategy Guide. The thing that's difficult to understand is there is no bell curve, no expected result with a mean time to success, that we can count on in any given game. With 1 chit at 5%, you'd EXPECT success in 20 turns (1/.05) or a little more than a year of game time. But you really have a pretty good chance of seeing success much earlier, and we often do. Conversely, you have a pretty good chance of not seeing success for a very long time, and we see this also.

Over many many trials you should see the expected results of a standard distribution, and as Shaka points out this is an incredibly large number of trials for decent results. Mathematically the system appears to be working as designed. It's just frustratingly random! But that's what makes research such an interesting element in SC. If it was more "normal" and we could count on research advances like clockwork, the game would lose an element of suspense so necessary for replayability.

[JerseyJohn]

Bill, Shaka, Dan, Terif -- Hubert

How's this for a little idea. If a country conquers another country it receives all that country's technological advnaces.

[xwormwood]

Great Idea!!!

Example: Germany got tank-technology from the czechs (Pz 35t/38t) in 1938 which was still in service till May 1945 (Hetzer/Marder).

[JerseyJohn]

Glad you like it.

[pzgndr]

If a country conquers another country it receives all that country's technological advances.

I think of research advances as more than just technology. They also represent production issues, doctrinal issues and all that "stuff" that goes into fielding new equipment. That doesn't happen overnight.

Zappsweden I believe has been critical of the catch-up bonus as being unfair. There may be a way of modifying this to better account for captured equipment. Maybe apply the catch-up bonus only when a nation surrenders? Or at least trigger the bonus only after combat involving new technology.

[JerseyJohn]

Agreed that it isn't an overnight transition, nor one that might even be duplicated in the conquering nation. But there should be some added technological benefit from conquering all of a major country.

In HiCom, conquered nations could only manufacture the former country's units with the conquering country's identity.

Still, taking over an enemies capital also means taking over at least some undamaged industry and a large number of technicians, scientists and engineers.

If Germany had conquered all of European Russia I think T-34s and KVs bearing the Maltese Cross would have begun appearing on other fronts.

[ May 25, 2003, 08:08 AM: Message edited by: JerseyJohn ]

[jon_j_rambo]

"Catch-up factor w/ 1.07" --- I think somebody left this out of 1.07. From playing several games on 1.07, it seems likes the "Catch-Up" doesn't work as well. I don't have any hard evidence or statistics.

Punnyworth --- Dude, take it easy on the Spam.

[pzgndr]

"Catch-up factor w/ 1.07" --- I think somebody left this out of 1.07.

It's still there, working like it's supposed to.

[Dan Fenton]

You guys may get a real laugh out of this. I tried JerseyJohn's suggestion about completely uninstalling Strategic Command and installing the lastest 1.07 patch. On my very next Allied turn, my one research point in Long-Range Aircraft generated my first ever UK advance in that Tech. Again, thanks Jersey John.

Either that or I got some of Shaka of Carthage's statistical luck.

And thanks Terif. I was trying to figure out how your "one" research point could so consistantly generate advances in a field that you deem to be so critical. I never in my wildest imagination dreamed that you would be investing 4-5 research points instead. That was very informative.

[Shaka of Carthage]

Dan Fenton

THATS where it went!

No wonder I've done so bad at the tables this weekend.

Could you like pass it back please?

[JerseyJohn]

Dan

Glad it worked -- I reinstall after each loss.

[Liam]

The Tech advances should be a bit more pre-set agreed. Although in WW2 the opening years the German's had an awesome anti-tank gun that is represented in this game as tech 1 anti-tech which would be the 88...

The Italians had tech advances, just not sure they could ever fully utilize them.

[JerseyJohn]

Oddly enough the 88mm was an anti-aircraft gun!

I think the anti-tank gun they were using early in the war had less than half the 88's penetrating power.

In North Africa Rommel really needed the 88's in the rear areas for protection against air raids but made the hard choice to put them up front, probably the only reason he stopped the British Crusader campaign -- everything else proved ineffective against the American Grant medium tanks being used by the British.

At the time most his Pzr IVs were mounted with short barrelled low velocity weapons, an early war armaments mistake Hitler had insisted upon. Later he uncharacteristically admitted it had been his mistake and upgraded both the Pzr IIIs (long barrelled high velocity gun, but obsolescent) and Pzr IVs with more effective weapons.

[ May 26, 2003, 01:23 PM: Message edited by: JerseyJohn ]

[Shaka of Carthage]

German 88mm would represent level 3 or 4 anti-tank tech in SC.

It depends on how you define the level 0 tech.

[Wachtmeister]

If I reall correctly, Rommel first caught on to the dual purpose FLAK/AT utility of the 88 mm FLAK in the French campagn when an Anglo/French (mostly Anglo the French attacked too late) armored counter-thrust caught Rommel's 7th Panzer by surprise in he Arras sector.

As JJ points out, the 88's were likely the difference in defeating the British Crusader campaign. The Germans were astonished that the British never used artillery to eliminate, or at least supress, the 88's before making those calvery like tank charges against German positions. The FLAK 88's had superb optics that permitted tank kills at very long range. However, they were large high-profile guns that would have been quite vunerable to artillery fire. Also, the FLAK 88's available in North Africa were fairly awkward to move and deploy.

The standard German AT gun until about late-'41 was the 37 mm PAK, aka the "door knocker" when fired at a Russian T-34, or UK Matilda. The PAK 37, was followed by the slightly more effective PAK 50 , the quite effective PAK 75, and the excellent PAK 88. The Germans also used a lot of captured Russian 76mm AT guns on the east front, some of which were converted to accept German 75 mm AT ammo.

[JerseyJohn]

Enjoyed it Mr W..

It's odd that an army so aware of tank tactics paid so little attention to countering them! The Arras action might have succeded if the English and French had been more well coordinated and, as you said, in North Africa it took the Brits a long time and a lot of losses to learn from their tactical errors.

[ May 26, 2003, 07:56 PM: Message edited by: JerseyJohn ]

[Shaka of Carthage]

Want strange? How about this.

British "88" ... they had a 94mm anti-air weapon. But since they were only trained to fire at aircraft, no one considered using them against tanks. While the equipment wasn't designed to be used as a anti-tank, it was more the lack of the ability to be flexible that was the British problem.

United States... We had a 90mm anti-air gun. Wonder what our reasons were for not using it as a anti-tank weapon.

British 25 pounder... This is actually a 88mm howitzer.

Everyone is generally agreed that the "88" was one of the deadliest weapons. Makes you wonder why the British and Americans didn't make more use of thier own versions of it.

Hmmmm... now that I think of it, I believe the most "effective" anti-tank weapon was the Russian 76mm. Not so much in terms of the kills, but rather in size, ammunition it used. Hence, there were towed and self-propelled versions of it.

[ May 26, 2003, 07:50 PM: Message edited by: Shaka of Carthage ]

[JerseyJohn]

Shaka

With the U. S. I was wondering if it might have been overconficence in the Bazooka. Only a shortrange weapon of course and I doubt it could have stopped a Panther or Tiger except for maybe knocking out a caterpillar. I don't know if the Brits had an equivalent, if not I guess they used American issue. But even if that were the case you'd still want something heavier for a back-up.