Something Ain't Right with Installation Upgrades [RESOLVED AND IN PROGRESS]

Spillover does now indeed appear huge but it is only applicable to reservoirs and aaltars. Going back to the original problem of mass L1 harvesters, the best strategy was still upgrading reservoirs anyway to support bulk L1 harvesters.

However, can we tie spillover from reservoirs to the average harvester level on a parcel perhaps? That would definitely help encourage leveling up harvesters.

I’d like to introduce the concept of Harvester Spillover Retention Factor (SRF) which is used to determine final reservoir spillover. For each harvester level, it might look like this:

L1 - 35% SRF
L2 - 40% SRF
L3 - 45% SRF
L4 - 50% SRF
L5 - 55% SRF
L6 - 65% SRF
L7 - 75% SRF
L8 - 85% SRF
L9 - 95% SRF

Then when a “withdraw from reservoir” transaction occurs, the final reservoir spillover is calculated by taking the average of all your parcels harvester SRF’s.

Benefits of incorporating this is we don’t actually need to touch the original tokenomics at all but greatly increase a parcel owners harvester yield and achievable ROI as they level up their harvesters.

The lore perspective is that when a gotchi initiates a reservoir withdrawal the “strength” of a harvester dictates how much of the stored harvester alchemica actually makes it to the reservoir to be passed on to the gotchi.

I really liked @notorious_BTC thoughts on making a game that comes with tough decisions. Maybe we don’t need to change the tokenomics (significantly) but can give plenty of other tough decisions to level up harvesters like:

  • Improved SRF
  • Improved HP
  • Maaker level limiting quantity of harvesters
  • Aaltar level up criteria that requires certain level of harvesters
3 Likes

Like with the traits changes, it seems like a simple remapping of spillage might be the elegant solution.

This opens up a third option, where the harvesters have their own spillage modifier, that goes on top of the reservoirs rate. This would be the most gamer friendly option, as there would be many ways to do this successfully. Such as, pushing up to a three, while you still have some ones left, then selling your 1’s to upgrade the 2’s to threes, because that gets half the spillage inside your estate.

I can see it playing out, where you would start at 128 L1s, and and some point, you would start condensing down into higher level harvesters, but it might take a year, because it’s just a slow optimization process, not something to be rushed.

I would love to see either the swapping of or stacking of spillage, combined with making the Humble and Reasonable estates, use the altar instead of lodge. There is a lot of synergy between these two changes, that could resolve both topics at once.

1 Like

Yeah Dan had mentioned something like this in the DAO meeting but my biggest concern with greatly increasing the harvester’s yield is the undeniable addition of major sell pressure that would follow.

Oh I like it, fren. I found my message in the discord on May 04:

"then I guess it will work like this:

for example, you have a FUD harvester lvl1 and it has a 50% spill, but it extracts all 14 FUD into the lvl2 reservoir and when you come to empty it, you get a spill of not 45%, but 72.5%"

1-0.5*0.45

That is, we can set a spill at the harvesters BUT it will be realized only at the moment of emptying the reservoir. Then there will be motivation to upgrade harvesters BUT again not to L9 unfortunately, so we still need to change recipes. But the addition of spills for harvesters does not prevent us from changing the recipes - we can take complex measures.

2 Likes

we need a weighted average. if I have 2 harvesters L1 and L2 (spill L1 = 50% and L2 = 45%) then general spill will not be 47.5% ( (50+45)/2 ), it must be weighted. we need to use rate/day instead of level . (14/(14+29))*50% + (29/(14+29))*45% = 46.63%

we cannot punish a person who has 2 harvesters L1 (50%) and L9 (10%) by making his total spill equal to 30%, it should be (14/14+575)50% + (575/14+575) 10% = 10.95% and rightly so

5 Likes

Oh cool, I thought I’d seen the “harvester level affects spillover” idea somewhere before :slight_smile:

I agree that the final reservoir spillover calculation should be weighted based on daily yields like you’ve shown not just a simple average, that makes much more sense.

If we developed this idea further, should the reservoir spillover also factor into this equation? Or do we just use harvest spillover factor in the calc? i.e. all a reservoir does when it levels up is increase in storage/collection capacity.

The total harvester yield is not changing in this idea. Only the portion of that yield that the owner is guaranteed to obtain. Sell pressure would likely be the same.

1 Like

Yes we could use a weighted average based on the harvest rate, just like we do with Reservoirs and capacity.

4 Likes

on this mechanism…
It always was in my head that the harvesters would spill over and not the reservoirs. perhaps I am crazy here, but I envisioned the harvesters sort of like oil rigs that when they tap the resource, it goes spraying into the air. I thought it was going to be a %%% spillover that sprayed into the air. Ie: a 50% spillover harvester would spill half of its yield one by one as they occur.

I don’t understand why the reservoir would have a spill over? I see the alchemica as already being neatly stacked into a pallet that I simply pull out and deposit…

6 Likes

Really great discussion rolling here frens.

I see there to be several levers we can play with. I’ve been thinking a bit about Coderdan’s bit on the call yesterday and why they chose to have diminishing returns on the harvesters. It makes sense to me. With DR though, we need to have other mechanisms in place which strongly motivate players to upgrade, despite the lesser rate of yield.

This is what I wrote from the end of the call yesterday:

The primary point of focus ought to be discussing how we want the gameplay for the farming to play out.

The models are helpful to see the numbers, but I think it important to have a more philosophical discussion first. The way I see it is there are two major paths we could take when building the models for harvesters.

The best way forward, in my opinion, is the path which creates the most user choice and non-linear paths to victory (as they see it).

All of this, of course, is under the pretence that the numbers can be tweaked in such a way that inflation is relatively the same, regardless of which path is chosen.

Path One: diminishing returns on harvester levels as they are upgraded. This is the current format and in its state is game-breaking. However, a path with DR can also be fun if we play with some of the levers. We must ask ourselves, what limitations and scarcity can we find that force players to upgrade to continue growing?

For example, if the total qty of harvesters to be allowed to be built is limited (and even bound to the lvl of maker), then the player is forced to make a choice: upgrade the maker to get more building space or upgrade the harvesters themselves.

An advantage I see with DR harvesters is that players will be motivated to buy more land. As long as the punishment isn’t too steep with the DR (as in its current state), the choice to upgrade or buy more land becomes an important decision to make (and drives land prices up in the process)

Path Two: the harvesters become more powerful with each upgrade. This is a solution that has been proposed already. The questions to then ask are how can we make this a compelling path where the player also has to really make tough choices on how they wish to progress.


The levers I see so far from the discussion we can play with for exciting gameplay:

  • The rate of yield on harvesters (diminishing or strengthening)
  • The build time on harvesters
  • The max qty build per parcel
    – this can further be played with the maker
  • Spillovers
  • The cost of harvesters to build

While we’re in discussion over these topics, I think it’s important to try and look at the parcel holistically. Now that we know the spillover is massively huge in radius, we have a better idea of how things will play out.

To only adjust spillover %% would do a disservice to the other mechanisms on your land.

While we’re here, we ought to analyse the ratio of harvester/reservoir per each upgrade and make sure the costs are feasible.

The maaker, too, we should look at. In its current state, a lvl 9 maaker is incredibly expensive (~22k GHST of materials) and underwhelming. What would make a maaker juicy enough that someone would invest to build it to lvl 9? And then we bring the cost to match that premium available to those willing to make that investment.

And of course, how can GLTR really step up and shine for those players who want to really push the edge on their farming?


Apologies for the long post, but here we are.

Finally, I think it’s best to first look at the game flow without the numbers. Create the experience and how the player would respond to the environment and then once we have a compelling build, reverse engineer the numbers to fit that game.

5 Likes

Exactly my thoughts too. If the harvester SRF was implemented, in our case the alchemica would be spilling up into the air and getting lost en-route between the harvester and reservoir when a collection at the reservoir is triggered.

It’s almost like “Reservoir” is not quite the correct word for them. Is “Cache” a better word for them? i.e. A temporary storage unit at which high speed retrieval of harvested alchemica is possible.

Put the spillage on the harvester - the harvester leaks continuously, even if the reservoir is full.

Excellent summary. Loving where this discussion is going.

A few ideas for encouraging maaker level ups:

  • Certain installations that only unlock at higher tier maaker levels (e.g. better towers, turrets, special walls)
  • Maakers contributing to defense at higher levels. Imagine their cute robot eyes turning red at the sight of lickquidators and charging around your parcel defending installations during an attack,
  • Customisable maaker cosmetics at higher levels. e.g. L9 Maaker cruising round with a phat GHST token hanging round its neck.
2 Likes

What if makers act like roombas, when you aren’t using them to build? They clean up any alch that lands on the parcel(they throw it in reservoir), when they are not busy. This would justify the high price of the advanced models.

Makers become Takers, when they sit idle.

If priced properly, the L9 Maker could be your full time maintenance crew. A personal yield management squad.

5 Likes

While the model coderdan suggested does move towards closing the gap. Crafting more L1’s is still the most optimal ROI on any timeframe because of the decreasing returns of upgrades Below is the relative ROI chart for that idea:

In my view, some other balancing mechanic like what has been suggested (harvester spillage, limiting # of total harvesters based on maker level, etc) would still be needed to add additional incentive to the player to level up their harvesters.

Great summary, fren! I think keeping the decreasing returns model, combined with some other counter-balancing mechanic like the ones you’ve listed that help to further incentivize upgrading harvesters could very well do the trick.

2 Likes

While we’re on the topic, let’s remember to also consider that currently, according to coderdan in dao-discussion discord a few days ago, a parcel’s aaltar level also limits the level of your harvesters and reservoirs. I do not believe this has been published in the bible, though it was alluded to in the original litepaper, so many players may not be aware of this mechanic.

This is an additional cost to upgrading that must be considered in our models and balanced out with the appropriate incentives to upgrade.

Yes, we are also taking that number into account in our internal sims, as it’s quite important to deducing the overall ROI of a strategy.

1 Like

On a long timeframe it’s the highest ROI, but the counterbalance is the opportunity cost of not having that Alchemica now.

The other counterbalances you mention could also be implemented fairly easily.

2 Likes

A potential pro of this model is that it incentivizes land ownership. If the best ROI is simply to max out L1s, players have a choice between buying more land and upgrading their current farm. It also serves as a boon to early adopters and people who own lots of land already. Otherwise, it makes the most sense to own the fewest lands possible but maximally upgrade them. We also need to take into account the “work” required to tend to the lowest level harvester/rez farms.

3 Likes