xXCrypticNightXx
Terrarian
Thx for sharing .
So I took my measurements and used them to judge my design. Base I call the time every heart needs to travel to the player, this includes the somewhat speculative 0.2 s exspected time to enter the hoik track. 2nd column is the additional time it would take a heart if it is not at the center, but at the edge, cause there is more of a horizontal travel and each level is what is added one story below for the additional vertical travel. Than I wrote in the results.
View attachment 378933
Unsuprisingly the bottom right corner faired the worst. But I quickly realised, that there are a bunch of options (16 in fact) for heart chambers below or to the side that have around equal or even better timer to it. I wrote them in to the side or below, according to where their heart chamber would go.
Given that the design had a efficiency of a little below 50%, I knew most of these options (maybe all) would have to be eliminated - I started this process knowing I had to eliminate existing chambers, not to add more - but given the numbers I thought I might be able to have an additional chamber through an option I had overlooked. Since I had added 16 chambers I then eliminated the worst 16 options. This still left me with 4 chambers that would need more than 3.5 s to arrive at the player, so I eliminated them as well. This is what my result looked like:
View attachment 378935
The next question ahead was, in what order to trigger them? These delivery times can be thought of as a queue I figured. So the problem at hand was to optimize a cue. A problem that can only be solved with state of the art technology. Basic understanding of logistics, a functioning heuristic and a text editor:
View attachment 378966
Now basically the intructions were. No more the 3 o in any row. S, A may share a row. All queue must arrive within 40 steps. Use Tab/Backspace to move them to the left/right. Note: There may not be a viable solution. In that case remove the bottom queue and start over. I arrived at this:
View attachment 378984
Now all I have to do is build a 40 gate long shift register connected to a 1/4 s timer and trigger the corrosponding heart chambers at the exact time the heuristic suggested. Econ as a second mayor is clearly payin' off .
Since I'm quite confident in my measurements and methods, my execution will decide how efficient this version is going to get. There is some randomness with hearts and the 0.2 expected value might be over-/underestimated - but aside from that I'm quite confident. And I have to be. You may have grasped from the heuristic sheet, that I had to eliminate all but 12 heart chambers. That leaves a maximum potential of 3,6 hearts/s. You clocked the last design between 170 and 180 hearts per minute - it had an efficiency of about 50%. To get an equal result. This new design will need to be at an efficiency of 79% on the low end and 84% on the high end. Well that's some order ain't it .
I'll post it when it's done - so far nothing has been constructed .
Edit: It's done. In fact I did it twice. They are basically the same, just that one of them uses a more cautious heuristic approach - but as a trade of has less total hearts to go around. While there will always be outliers, both of these machines should blow the efficiency level through the roof. And I hope they do cause they only have a total of 38 hearts in the frugal one and 36 for the cautious model. I really don't have the faintest clue witch one will fair better or if they can even outperform the previous model. The numbers are limited in there preciseness (and I have to use them in smaller decimals than they were measured at, witch technically is a big NO NO!) as well as their validity, cause hearts can behave in such outlandish ways that you'd never see comimg.
Then maybe my approach was wrong. Nearing the end of the constructions I had some thoughts in the vein of "This won't work, you just added up the number you measured for different ways of movement, but a system is more than it's parts you would have to measure the numbers for each chamber individually in game..." - Well I hope that's not the fact .
Anyway the heuristic gave me some very interesting orders of activation that were a pain to wire .
I'm curious how it will do
View attachment 379083
Cool statistics. I tested both these designs and unfortunately, the times have dropped. Both designs are coming in around ~150 hearts / minute.
I did a deep analyses on it and besides your math, you are forgetting some variables in your formulas. You can only have 10 hearts on the map at once. And the hearts that are moving will come within 23 blocks of another statue cutting of their spawns. These are the reasons why I suggested my design in my last comment.