Minecart performance experiments

[Meh]

I had a nagging feeling that the higher level minecarts weren't worth the upgrade, so i did a couple of experiments to test out the performance of the different carts. I tested the basic, large, fast and bulk carts. I did this to test the carts and not the line as a whole.

The first test setup: Horizontal travel, 64m. With a matter mover to beam the delivered items back to start. Conveniently, the matter mover also counts the items. I ran 4 identical lines side by side for 30 minutes. One with 4 basic carts, one with 2 large, one with 2 fast and one with a single bulk cart.
The results: 4 basic handled 33 items/min, 2 large handled 30, 2 fast handled 33 and the bulk handled 25.

Next test was the same, but with 3 turbo tracks on each line. The results were similar, but even more in favor of the basic carts: 87/70/70/50 items/min respectively.

Then i did the same tests, but with vertical movement. Here the performance was much more even. First test without turbo: 20/20/20/18 items/min
Second test with turbo: 37/33/38/32 items/min.


A quick calculation for the first test. Mine carts move 1m/s as far as i know. Loading/unloading happens at 175 items/min.
a single basic mine cart moving cargo 64m would have to spend 17s loading/unloading and 128s travelling (or a bit more including buffers). That gives a total of 145s to move 25 items or about 10 items/m - or 41 items/min for 4 carts.
Similar calculations for large, fast and bulk carts (moving at 2m/s) give 37, 31 and 28 items/m respectively. Pretty close to the 33/32/33/25 i saw in the first test.

TLDR; the basic carts are clear winners and turbo tracks are worth it.

Edit: Added data for fast carts

[steveman0]

I would really like to see the test with the fast carts. Seems odd to leave them out of the equation as they are clearly an important comparison on how speed factors into the equation especially given the significant effect the boosters had. It would also be helpful for my purposes in considerations for cart balancing/new cart design for freight carts.

Also, loading/unloading is theoretically 300/min. I guess you were factoring in stopping/acceleration?

[Meh]

As far as i understand, fast carts are the same speed as bulk carts. I've never used them, so i could easily be wrong.
Their performance should be somewhere between bulk and large.

I would humbly suggest that the large and bulk carts get a boost to their loading/unloading speed and the fast and bulk get a boost to their movement speed. If the boost is big enough, this will make fast carts preferable on long tracks and large preferable on short. With bulk being preferable overall - especially if it gets an additional boost over the others.

[steveman0]

Load/unload speed is locked to the LFU without significant rewrite of the code. I think the solution I will go for in balance will be to remove the duplicate cart cost so that, at least organic rock-wise, the winner is clearly to upgrade.

These numbers will help guide me on determining how to balance the other costs. The fast cart data would be valuable to determine how much a boost the double speed gives relative to double capacity. I suspect to really understand this it would require testing at different track lengths as well since mathematically the difference should disappear as track length goes to infinity.

[Meh]

I'm running the test now with fast carts (vertical with turbo). Will post the results shortly.

[Meh]

The results of the test:
4 basic: 37 items/m
2 fast: 38 items/m
2 large: 33 items/m
1 bulk: 32 items/min

Not quite what i expected :-)

[steveman0]

Nor I! Interesting result. I'm not sure how to explain it...

Still not worth the investment though without the recipe change. I suspect I'll add that to freight carts on the next release.

[epgeek]

Love that data! I would love to see at least the horizontal track re-tested with many more carts and stations to simulate their moving T4 ores to bulk lifts. I'm thinking maybe 3 loading and 3 unloading stations and a dozen bulk carts?

I have noticed that the couple lines I upgraded from large to bulk required adding several more carts to keep the throughput up.

[Meh]

Re-ran the tests to include fast carts. The data is in the top post.

[DustBust]

Yeah, I remember someone doing this a while back as well and since then, have never messed with the other minecarts, seems like this needs to be revisited so each minecart has a better advantage or reason for use

[Demerzel]

Originally posted by Meh:

The results of the test:
4 basic: 37 items/m
2 fast: 38 items/m
2 large: 33 items/m
1 bulk: 32 items/min

Not quite what i expected :-)

This clearly shows how suck the whole minecart concept in the game.

[DjArcas]

Originally posted by Meh:

The results of the test:
4 basic: 37 items/m
2 fast: 38 items/m
2 large: 33 items/m
1 bulk: 32 items/min

Not quite what i expected :-)

Sure, but let's also examine framerate:

4 basic - 25fps
2 fast - 50 fps
1 bulk - 100 fps

Stations needed to service each one:

4
2
1

It's not entirely down to items/minute. And if you are talking about 100m or 10km, then the Fast cart will give you a very different latency time.

[DjArcas]

*much discussion on discord takes place*

[steveman0]

To summarize for anyone who missed out, basically agreed upon that the upgrade scheme now is indeed crap.

Originally posted by DjArcas:

Originally posted by Meh:

Stations needed to service each one:

4
2
1

It's not entirely down to items/minute. And if you are talking about 100m or 10km, then the Fast cart will give you a very different latency time.

And this gem which is incorrect because gates exist and DJ himself wrote them

But anyway... pushing a change that reduces the upgrade cost of carts to only 1 cart per upgrade. This way you're encouraged to upgrade for better framerate (and better network bandwidth). Other costs will remain the same.

[DjArcas]

Well, despite the dozen negative reviews telling me I "can't accept feedback", you're totally correct, and I'll be fiddling with the recipe during Patch 18.

[Ecclesia]

Wait, to be clear: so, both Fast, Large AND Bulk Minecart all require 1 Basic Cart? (instead of 2,2,4). That does make a bit more sense since you upgrade the same cart basically.

Now all of a sudden 8 Alloyed Blocks per (Bulk) Cart doesn't seem that bad at all, the only problem now is that that cost is put on the first tier, not the third :P

Hate to be nitpicky on that item, but if the Basic Cart remains the expensive one to make (cheap upgrade to Large and Bulk), they are not gonna be touched upon until FF expansion, at which point you upgrade them to Bulk anyway, kinda defeats the purpose of 3 tiers, yes?