Well, I hate to double post but this reply will be extra long so I'd rather not just edit. The results to my editor/in-game tests are in and I took extra precautionary steps to ensure thier accuracy.
Since this going to be long, I'll begin with an introduction . The first reply I uploaded in this thread attempted to answer all the questions rasied in here about certain technologies. What better source, I assumed, than techtreex (the game's data file that controls all technology information)to settle the question over gathering upgrades, noteably Winter Harvest. However, DeathAndPain disagreed and stated the game's data files could easily be errored and have a tendency to round values, and thus cannot be totally trusted. As an example, he stated an alleged flaw in Odin's hunting bonus.In my experience, the games data files are completely trustworthy, so my next several posts I expressed my sincere conviction that the data files (techtreex in particular) are accurate, and gave supporting evidence. But later Etendorf came forth and stated the same alleged flaw believed to be present with Odin, is also supposed to be present in Thor's "Pig Sticker" hunting-enhancing technology. But I did not believe this right away, see my ealier posts for my reasons why.
So now I tested it and I took extra care to be sure my results would be valid. The test here is simply pitting DeathAndPain's and Etendorf's what I will call "theory of rounding," which states that the games data files will deviate somewhat from the acual effect in the game, against my what I call my "theory of absolutism" which states that the game's data files are reliable, trustworthy indicators of the effects that occur in game.
At first, I was surprised to find that in the case of "Pig Sticker" and Odin's hunting bonus, techtreex seems to support the theory of rounding, that is, I had always heard that the bonuses here were 10% and so has just about everyone else, but my theory would predict that if the bonuses were 10% then techtreex should say that, instead for the both of them, we find:
<effect type="Data" action="Gather" amount="1.20" subtype="WorkRate" unittype="Huntable" relativity="BasePercent">
<target type="ProtoUnit">AbstractVillager</target>
</effect>Accordingly, the bonus should be 20% and because Winter Harvest has the same stats except that it applies to farms, we would expect the bonuses to be the same. But I was still not convinced. Thus, I conducted numerous tests involving Winter Harvest and Pig Sticker. (And by logical extension, the results of the Pig Sticker test would also apply to Odin's bonus, becuase bonus because it has exactly the same techtree code.)Firstly, to eliminated any inaccuracies that could creep into an in-game test I modified the proto code a little. For one, i added <flag>GatherDirectToPlayer</flag> into the dwarf's stats, that would eliminate dropsites and also any wasted walking distance. Also, I increased the dwarf gather rates for hunting and farming to exactly 1.00, or 1 resources per second, just for the sake of simplicity.Lastly, so that ALL walking distance could be eliminated I added <flag>UnlimitedSupply</flag> to the deer unit, this way it won't deplete (and the dwarf won't have to move on and hunt another). Also, notice that my player is Loki, this is to avoid modifing the outcome due to Odin's extra bonus, or Thor's dwarf enhancements. Even though the screenies say "Heroic age" there are no techs researched, so just ignore that.I thought it highly probable that some would object to my editing the gather rates, beleiving that by itself would tamper with the conclusion, so my first experiment was to show the proto code could be trusted. Now to start the experimenting. You can click on any screenshot to enlarge it.
This first screenie is to show that I begin the game with no food, the deer has infinite food, and that the timer is at 3 but the gathering has not started yet. (Btw, the reason that other objects appear in this scenario is because it was a sloppy test scenario from a long time ago, I assure you that no other technologies or units are affecting the outcome you'll witness.)
http://xs.to/xs.php?f=1.DeerInf.0foodstockpile.jpg&h=xs202&d=06261Okay, now that you've seen that, you need to see when exactly the villager starts gathering. Notice in this next one that the timer (which will be denoted as "t=x" for short) is at 7, and the deer carcass is ready to provide venison but the dwarf has not yet started gathering.
http://xs.to/xs.php?f=1B.t_7s.dwarfnotgatheryet.jpg&h=xs202&d=06261The villie starts exactly one second later, at t=8, here's the proof:
http://xs.to/xs.php?f=1C.t_8s.dwarfgatherNOW.jpg&h=xs202&d=06261Now, remember with out any ugprades or any other workers, this dwarf should gather at 1 resource per second and started gathering at t=8, let's let the timer run and see how that holds out... 10 seconds later...
http://xs.to/xs.php?f=2.gatheringT_10s.10foodpile.jpg&h=xs302&d=06261Viola! 10 seconds later and we have 10 food. But we all know that resource tend to be deposited in "spurts" even w/o dropsites, so let's continue our timing for a decent amount of time to get a much better result.
Waiting...
The dwarf has now been gathering for 30 seconds and what do ya know! 30 Food! One resource per second is looking good!
http://xs.to/xs.php?f=3.gatheringT_30s.30foodpile.jpg&h=xs302&d=06261Now I don't have a screenie for every 20 seconds so skipping ahead a little, t=128, we see our faithful hunter has been working non-stop for 2 minutes, which is 120 seconds. Take a look at the food.
http://xs.to/xs.php?f=4.gatheringT_2min.120foodpile.jpg&h=xs302&d=06261It's at 120! Protox, thou doth not foresaketh me! If anyone is too hard-headed by now to think that protox did foresaketh me, I have last screenie, which skips ahead by half a minute. Unfortunately I didn't hit pause quick enough so we got another +1 to the t. Here ya go:
http://xs.to/xs.php?f=5.gatheringT_2~.5min--1s.151food.jpg&h=xs302&d=062612 1/2 minutes plus 1 second translates into 151 seconds. And Lo and Behold! 151 Food! But, more time can't hurt, I give it another 30 seconds (29 acually). 30 (acually 29) seconds later...http://xs.to/xs.php?f=6.gatheringT_3min.180foodpile.SpotOn!.jpg&h=xs302&d=06261Not bad, at t=3:08, or 3 minutes of gathering we find 3 minutes worth of food at 1 food per second! Moving on.
Now that we can definitely see that these dwarves will gather huntables a 1 resource per second, let's get down to the core! If the theory of rounding as currently stated is correct, then Pig Sticker should add 10% (or something close to it) to that 1.00, but if my theory of absolutism is correct, then 20% should be added to the rate. (See the techtreex stats quoted earlier.) Let's be sure that this time, Pig Sticker is reasearched from the get-go.
http://xs.to/xs.php?f=7.Editor.ActivatePigSticker!.jpg&h=xs202&d=06261Begin the experiment! Firstly, we need to test the scenario as the correct player.
http://xs.to/xs.php?f=8.BeginPigStickTest!.jpg&h=xs202&d=06261Secondly, we must then kill the deer, get the dwarf into position, and then record exactly the second that he starts gathering. See this:
http://xs.to/xs.php?f=9.t_9s.killDEER.dwarfnotgather.jpg&h=xs202&d=06261Observe that t=9, and everything is ready to begin. The next screen...
http://xs.to/xs.php?f=10.t_10s.dwarfgather~1food.jpg&h=xs202&d=06261...shows that our dwarf friend begins gathering soon after t=9 and almost instantly he gets 1 meat -- looks hopeful. But is that due to a 10% or a 20% increase? We must continue to find out.
http://xs.to/xs.php?f=11.gatheringT_20s.24food--20p.jpg&h=xs202&d=06261So it is 20 seconds of gathering later, are you looking at the food? It's 24! Using a TI-83 or your brain, one can realize that the benefit here is 20%! Not 10%. The theory of abolutism is strengthening! Let's add a smidgen more time and see what happens.
http://xs.to/xs.php?f=12.t_39s.30gath.T.36food--20p.jpg&h=xs202&d=06261Another point for absolutism, after 30 seconds of gathering, the food output is, again +20% which gives us 36 food total. But 30 seconds is probably not quite enough to settle a vast matter such as this, so let's kill some more time lazily while Mr. Dwarf continues his work...
http://xs.to/xs.php?f=13.gatheringT_1min.72food--20p.jpg&h=xs202&d=06261Magnifico! At t=1:09, 1 minute into the working shift, we notice a 20% food increase again! Plus one for absolutism. But a minute is still short, so let's continue...Oops, here's an unintetional shot taken 1 second later, interesting to see that the food went up only 1 in this particular second though, isn't it?
http://xs.to/xs.php?f=14.gatherinT_1min~1s.food~by1.Gathers.uneven..jpg&h=xs202&d=06261Why don't we move ahead a tad to a longer time? Don't you agree? By stopping at t=2:09, that would mean that Gimli would have been working with his pig sticker for 2 minutes, or 120 seconds. According to the theory of rounding, as currently applied, this should give us 132 food (120 + [.1 x 120]). According to the theory of absolutism, we should get 144 food (120 + [.2 x 120]). Let's find out!http://xs.to/xs.php?f=15.gatheringT_2min.143food-19.1p.jpg&h=xs202&d=06261Odd, it's 143. Which comes to about a 19.1% increase, but we saw that gathering can be un even before, and 19.1% indisputably points toward one theory over the other. You choose which.
The next picture was also accidental, but intriguing nonetheless, recall earlier that adding 1 second to the time count allowed dwarfy to deposite 1 more unit of food. This time, one second gave him the chance to deposite 2 more, just further proof that resources really are deposited in "spurts."
http://xs.to/xs.php?f=16.gatheringT_2min.1s.145food-19.8p.jpg&h=xs202&d=06261The next screenie takes place 59 seconds later at t=3:09, or 180 seconds of gathering time. The theory of rounding is suffering a little and this shot does no help. Infact it raises the 19.1% from the previous shot up to 19.8%. See for yourself:
http://xs.to/xs.php?f=17.gatheringT_3min.215food-19.4p.jpg&h=xs202&d=06261Now I was going to post another screen at t=3:39, but decieded to skip that as it is near identical to the t=3:09 shot. The final shot for hunting occurs at t=3:41. Yes, it's a strange time, but it only deals another blow to the theory of rounding by further suggesting 20% (This shot is about 19.9% or 265 food/221 seconds of gathering.) Seeing is believing:
http://xs.to/xs.php?f=19.gatheringT_221s.~3min.41s.~265food-19.9p.jpg&h=xs202&d=06261The theory of absolutism is so far triumphing over the theory of rounding. Techtreex tells us that Odin's hunting bonus and Thor's Pig Sticker should improve workers by 20%. And clearly, that's exactly what it does! Just as techtreex reliably tells us what amount of attack/hitpoint/LOS/etc. increases we should expect for any given military technology, it likewise reliably tells us what sort of benefits we should and shouldn't expect from economic upgrades!
But since this thread is mainly about Skadi's Winter Harvest tech I decided to do anoth test involving that. I should point out that neither DeathAndPain nor Etendorf (the theory of rounding's main advocates, in this thread atleast) directly claimed that Winter Harvest does not give the bonus it's alleged to, and thus, the main purpose of this test is to show that it does give a 20% bonus as the threads creater suspects. Comparing "theory of rounding" with "theory of absolutism" is downgraded to a minor purpose here.
First things first, Newby was surprised to hear from somewhere that Skadi's Winter Harvest could increase farming rates by a substantial 20%. I would beleive so, because in techtreex for Winter Harvest we see:
<effect type="Data" action="Gather" amount="1.20" subtype="WorkRate" unittype="Farm" relativity="BasePercent">
<target type="ProtoUnit">AbstractVillager</target>
</effect>So according to techtreex, the improvement should be 20%. Well same rules as before, the gathering rate was raised to 1.00 and two tests were conducted, one w/o winter harvest, one with it.
The first screenshot shows that Winter Harvest is disabled, just to be sure it will not be activated while doing this gathering test.
http://xs.to/xs.php?f=20.Editor.FARM.Wint.Harv.DISABLE.jpg&h=xs202&d=06262Moving on, here's a shot at t=4 and the deer is ready to be gathered from, but the dwarf has not yet started.
http://xs.to/xs.php?f=21.t_4s.DwarfNotGatherYet.0food.jpg&h=xs202&d=06262Next screenshot is simply to show that the drawrf starts gathering at almost exactly, if not exactly, t=5.
http://xs.to/xs.php?f=21.t_5s.DwarfStartGather.0food.jpg&h=xs202&d=06262Okay, to spped things up, I'll simply post the next several screenshots showing several different times and numerous food quantities, just to show that protox is not letting us down and the dwarf is in fact farming at 1 resource per second. Here's a shot at 30 seconds of gathering:
http://xs.to/xs.php?f=22.gatheringT_30s.30foodpile-1.00gathRate.jpg&h=xs202&d=0626230 Food, what can I say. Here's one after 92 seconds of gathering:
http://xs.to/xs.php?f=23.t_1.37.gatheringT_92s.92food.jpg&h=xs202&d=06262Notice I have 92 food, that 1 resource every 1 second. Good. By now, you should have realized that the gather rate listed in protox is totally accurate but in case not here's a shot taken at t=2:36, or 151 seconds of gathering and there's 151 food.
http://xs.to/xs.php?f=24.gatheringT_2min.31s.151food.Gath.Rate--1!.jpg&h=xs202&d=06263That's good enough proof I think. Now to test what Newby believes regarding the 20% increase. Better activate Winter Harvest before starting this.
http://xs.to/xs.php?f=25.Editor.ACTIVATE.WINTER.HARVEST.jpg&h=xs202&d=06263As with all my tests, it will require 2 screenshots to pinpoint exactly the second that the worker started gathering, which must be taken into account when veiwing all later screens. These two, in order, show that the dwarf began gathering at t=3.
http://xs.to/xs.php?f=26.t_2s.DwarfNotGatherYet.0Food.jpg&h=xs202&d=06263
http://xs.to/xs.php?f=27.t_3.DwarfGatherNow.0Food.jpg&h=xs202&d=06263Next, take a look at this shoot which takes place 1 minute after gathering started. We have 72 food in 60 seconds, again use your TI-83 or your brain and the improvement is, in this case exactly 20%. Nice! But we should go longer than a minute. http://xs.to/xs.php?f=29.gatheringT_91s.109Food-19.7p.jpg&h=xs202&d=06263There's one where the gathering t=1:31 (paused a second too late) and the food is at 109 which amounts to about a 19.7% improvement, still pretty good. But how about another one where the dwarf would be gathering for 2 1/2 minutes.http://xs.to/xs.php?f=30.gatheringT_150s.179Food-19.3p.jpg&h=xs202&d=06263179 food, 1 more and the improvement would be exactly 20%. This next shot was acually a mistake, but notice that the second didn't change at all but the food went up by 2, raising the percent to 20.6, weird, but still clearly shows that the improvement here is the promised 20%.
http://xs.to/xs.php?f=31.gatheringT_150again.181Food--20.6p.jpg&h=xs202&d=06263This is the last screenshot for this test, it takes place after 212 seconds (3 minutes, 32 seconds) of uninterrupted gathering. And again we see that Winter Harvest does cause workers to farm 20% faster.
http://xs.to/xs.php?f=33.gatheringT_212s.255Food--20.2p.jpg&h=xs202&d=06263In conclusion then, to answer your question Newby: Yes, Winter Harvest does give Norse farmers the 20% boost it is alleged to, and likewise it coordinates with what the data files regarding it say.
And, dispite the widely circulated belief to the contrary, Odin and Thor's bonuses acually do provide the 20% bonus that the data files claim they do. The data files are not misleading us by rounding values behind our back, they are known to work properly for technologies such as the medium, heavy, and champion military upgrades an likewise they work properly for economy techs with Odin and Thor being no exception.
Even though I disproved the popular inaccuracy that Odin and Thor's bonuses were rounded down to 10% and were in fact spot on with the data files, I still should point out to you that tests, such as the ones conducted here, are highly reccomended if the attribute in question has an external factor that can effect the output or outcome. Then for practical knowledge, a test will help.
But for other attributes (such as attacks) tests will probably not help you as those allow chance to play a large role. Atleast now the people here can be confident that the data files here do not round the values put into them, thats simply a widely-circulated, but false, urban myth. Though before you rely on them in-total, it is recommended you have quite a good understanding of them to be able to predict such things as DaP's Husbandry example.
[This message has been edited by Magnum Pi (edited 09-02-2006 @ 00:32 AM).]