Previously, I asked about how Explore teams were generated. This was partially because I wanted to know and partially because I wanted to write a tool to answer a different question: When exploring for event gems, which kingdom has the highest yield?
I wrote a tool that, for each kingdom, does several thousand iterations of this algorithm:
Choose a challenge or quest team from the kingdom’s roster.
Fill in any empty slots with a random troop from the kingdom.
Choose one troop from the event chest lineup and replace a troop on the team with it.
The end result is more or less an average for each color per combat in that kingdom. I’m missing the “substitute an owned mythic” rule, but I’d have to ask you to give me your invite code per-account to make that work and I don’t want to deal with that. It turns out event chests and mythics only have a tiny impact on the results and don’t change the relative order anyway.
I’ve run it for the past few weeks, but was waiting on a week where you could get the gems from Explore matches to unveil it. I want to add some more features, like making the table sortable, but that takes effort and it’s the holidays right now
So, this week gives you event gems for red enemies, and Grosh-Nak is consistently the kingdom with the highest distribution of red enemies. No other kingdom comes close, I don’t think there’s a useful 2nd choice.
That’s not exactly a good guide. While Grosh-Nak is the most productive red kingdom, it’s pretty bad for the other colors. I stopped counting after 10 that were better at green or brown.
Yeah. I just go by the first color. The rest just means they’re more likely than the only 2 other possible Mana types. But the first color is always the best out of 5.
Sorry to revive a dead thread, but I love this tool (and have recommended it to others, even) and l have a question: does the tool account for summons-on-death cards?
My assumption is no, but l’d be thrilled to be told that, on a Brown week, for instance, Drake Rider counted twice anytime he showed up in the tool, thanks to his being replaced by a Drake.
(I know then you might have to account for which Drake Riders have the trait versus not, so it might be more trouble than it’s worth. Just thought l’d ask if you already pursued the angle!)
The tool is “sort of dead”, but I should look into doing an update periodically. Long story at the end of the post.
It does not account for summoning traits. I think, in the grand scheme, it wouldn’t move the numbers enough to change a kingdom’s rank. But I’m curious. Part of the reason I wrote the tool is people telling me things like “the event kingdom doesn’t shift the best kingdoms” and it only took about a month to find out they’re wrong.
So what I’d need to know is how to model the trait randomness. How does the game decide to assign traits to Explore teams?
Long story:
I worked on this (and some other tools) when, for about six months, I had a job but no work to do. This coincided with a LOT of time when my wife would be out of town. I don’t like being idle so I taught myself JavaScript and a handful of tools under the guise of “making some GoW tools people might like”. Now I have a new job, that means I have real work, but also this job completely blocks the GoW servers. It’s a longer commute, so I’ve had less free time. I’ve been playing a lot of Stardew Valley, but that’s starting to lose its charm so I might dive back in to this.
I think it’s good. I regularly took your tool into consideration when I decided which kingdom to farm. Summoning on death is very annoying in Explore and considerably slow down average time needed for a win. Despite Grosh-Nak have the best number of red enemies, according to your tool, I ignore that kingdom because of Drake Riders.
It’s random. The same troop can be without traits, with 1, 2 or 3 traits in different explore matches. However, chances are not equal for all of these cases - for me, 2 or 3 traits are much more common than 0 or 1 trait and I’m not sure if it because I traited these troops in my collection or it’s the common behaviour.
I can’t get into it. Everything about that game feels like I would enjoy it more, but I can’t get past a few hours in. I think it is the energy system. My guy is going to bed at 8pm because he gets exhausted so easily.
I loved Stardew Valley, but I’ve never finished it because each successive year feels too much like the last. I’d love it to be a little more dynamic, with different events and (maybe) character changes (death, growing, etc…).
But if Concerned Ape ever makes another thing, I’m buying it, no question. Especially if he keeps the price point at $15! XD
I’ve been playing Harvest Moon since SNES so I guess something about those games scratches some absurd itch for tedium I have.
Fun fact: for the cost of Stardew Valley, Gems of War wants to offer me enough traitstones to upgrade Starflower. That’s a pretty good deal, I think I could get minutes of play out of that!
Almost all farming games are fun to me because the itch they’re scratching is a similar one to Gems of War: resource management.
I like that l collect things, and that there’s a right and a wrong way to do it from a min/max standpoint. I also really love seeing the (in farm games, almost literal!) fruits of my labor.
But to be on-topic: please let us know if the tool is ever updated. And whether it’s currently kicking out bad info—l don’t think it is anecdotally, but l could obviously be wrong and would hate to be XD
With regards to traits on Explore troops, I know that different levels of Challenges have differences in traiting, which might account for some differences.
Aside from that, I’m not sure to what extent your own collection influences it. I do know that I played through the Sin of Maraj quests on a low-level character, and none of the four Horsemen were fully traited(!), making it much easier, whereas at least two of them were when I played the quests on my high-level main.
At the same time, sometimes the Mythic troops from my collection that appear don’t always have full traits (I think???), so .
Hmm. If I do the traits thing I think what I might do is generate a range. As-is I do about 100k battles to count colors, what I can do is add a “what if” so each battle generates a “with summons” and “without summons” value. Then instead of printing one number I can show the range for each kingdom. Sort of complicated, so if I get that far and it looks like it barely matters I might cut it back out.
As of right now though, it looks like I need to figure out how to convince GoW my tools have upgraded to 4.3.5 Shaking the dust off…
I was reading the data files to find out what troops were in the event chests and using that for my pool. That data’s been completely removed, so I’m going to have to figure it out myself. Problem: I’m hearing that’s not as simple as “the troops from the kingdom”, apparently some event chests can have troops from other kingdoms if there isn’t a wide enough array of troops.
But I can’t get that data, so we’ll just have to settle with “the numbers will be slightly less accurate once I sort this out.”
Starting February 2018 (slightly before the 3.3 patch), event chests have only included troops from the current weekly kingdom. This change was not announced publically at the time, but I confirmed in PM with devs at the time that it was intended. Every event chest now should only include troops from the current event kingdom.
.
Shaking the dust off…