Architect Archive

Using Only Steel with UT/SR/HR over 950 each, and Chem with no UT, you can and usually do get to 97% LOMU and OMU's. I have never been below 93%, and usually get to 97.

I wonder if it might not be your ore that's the headache. If in the final assembly, your crap ore might be holding you back unless you nail a few amazings. Shrug, who knows.

As I said - I know where the problem is. The crates.

Last night had a fella order a couple medium chems and gas. Didn't have any currently in stock so decided to make these quick manually. Was using 950+ resources and all the ore and fluid units were 97%. I tried 4 times and each time I failed out to a ber 9. Though it was kinda odd as normally a ber10 is fairly easy to make.

Took a closer look and saw I was trying to make a medium mineral harvestor instead of a chem. (think I would have figured that out when no wall units were being used). Anyway, switched over to the chem and then gas. Four tries and four ber 10's and I'm done.

For some reason the minerals were extremely hard to make last night to ber 10.

I then went back to the medium minerals and decided to try something different. Instead of using 1 exp point at a time as I have always done, I highlighted the whole row in a one shot effort. I think it took something like 8 of the 10 points. After the 8 were used I took a look and saw that even with the 2 points left, there was no was i was going to get close to 97% on the final assembly. Maybe it figures in a failure if due when using mult points at once...dunno.

Using 950+ resources, 97% mining/fluid units and the one point at a time exp method, if I don't fail I will produce a harvestor of either ber10 or ber13 (depending on medium or large) 100% of the time...no doubt about it.

Maybe we have these people using mult exp points at a time or the one shot deal.

My 2 ore units worth anyway

Actually, this is what I've been doing for quite a while, with very good results.

1. do the initial assembly
2. use 6-8 exp points in one gulp
3. if it was great (which it almost always is) then I'm usually at BER12 (for heavies)
4. spend 1-2 more points to hit BER13
5. spend the last few points (in one gulp) on storage

I don't know what the distribution is from "amazing" down to critical, but since I think that you always have some percent chance (5%?) to critical, and since with good other factors you are heavily weighted up around "great", the *less clicking the better*. Essentially, the more you click the more likely you are to get a CF. And the first CF ruins the whole thing.

I've always had the best luck using 3 skill point boxes at a time. Not sure why that is, but if I get greedy (with my time, that is) and start using 4, crit fails seem to happen more often. If I use 1 at a time, again, it seems like crit fails will happen more frequently (although amazings happen more frequently as well).

An exception to this is when I final assemble a heavy harvestor. Then I only use 2 skill point boxes on the first experimentation. If I crit fail that first one, the heavy is basically worthless, and using only 2 skill point boxes I've had my best luck.

One thing I don't like about experimentation: If you crit fail the first one, you can end up unable to experiment at all. If you crit fail the last one, you just get a less than ideal result. It really shouldn't matter: first or last. The final possible result should be the same, but for whatever reason, the devs decided first is more important than last...which I'm not a huge fan of.

The % of the final outcome, assuming you have great or amazing success on each experiment, and 10 exp points, on the ore ming unit for example, can be calculated from the weighted average of the component quality i.e. % = ((UT + UT + SR + HR)/4)/10. So if you have UT = 900, SR = 860, HR = 860 then final % would be 88%. I have been playing with the break points, i.e. what % do u need to get a particuliar BER on the ore mining unit.

It may be interesting to know about the OMU breakpoints, but it appears not to matter. By that I mean, if you knew that you could get BER6 on the OMU with (for example) 86% and just stopped, you would be screwing yourself. The "rating" on the OMU doesn't tell the whole story - although it *displays* a whole number, it looks like under the covers it keeps the full value. So there is a big difference between a 6.0000001 and a 6.99999999 OMU. You want to experiment that mining component as high as you possbibly can. I make 93-96% ones and label the schematic accordingly.

The reason is that the OMU quality is averaged with the quality of your final assembly to get the final BER. If you "settle" for just enough on the OMU to hit an apparent "breakpoint" in the display, then you will force yourself to hit a higher quality final assembly to get to BER10 or BER13.

So get that mining component as high as possible and make a schematic. You'll be very happy that you did.