Topic: Clicker Heroes /
Gilds vs. QAs: An Early Game Head-To-Head Test
With the introduction of rubies with version 0.17d, there has been continual debate about the best way to spend them. If a new player asks in the chat which option is the best, inevitably an argument breaks out with many claiming Gilds are better, while other swear Quick Ascensions are better, continuing for many pages.
I have personally advocated for gilds for new players based on my own experience and understanding of the game mechanics. I felt that the consolidation of gilds should give a large DPS boost that would propel the player to higher levels, offsetting those gained from a QA. However, the rubies system was introduced after I had started playing so I had little direct experience to work from. Advocates of Quick Ascensions have made a compelling argument that the faster start they afforded would continue to carry the player to faster progress sooner, building a lead that would not be overcome.
After puzzling over the issue for a while, I decided that play testing is the only way to settle the issue. No logical argument or calculation can take the place of actually playing the game. I made two runs to zone 1000 that were as identical as possible, with the only difference being that I would spend my rubies buying Quick Ascensions in one run and Gilds in the other.
The most difficult issue to be settled was finding a consistent protocol for making decisions in the game. There would be many opportunities to introduce bias through arbitrary actions, so I tried to make the playing format as rigid as possible. I was not interested in the most efficient or most advantageous method for each, rather a well-defined and reproducible method to play the two games as close to identically as I could, while minimizing the impact of chance.
Tests were done using version 0.18 of the game. Version 0.19 was released after I started, so games were left idling without closing the browser window to avoid getting updated and having the new relics system interfere with the test. The Quick Ascensions run was done on the Clickerheroes.com site, while the Gilds run was done on the Steam version of the game. Both games were started at the same time on the same day of the week (but two weeks apart, so only one at a time) so that my usual weekly schedule would leave them idling at similar points for similar lengths of time.
Heroes were progressed through sequentially, taking each to level 10, buying the upgrade, then moving on to the next. They were leveled to their highest upgrade when the cost did not impact leveling the most recent hero. Shinatobe was taken to 25, then Grant to 50 to bridge the larger gaps at the end. Frostleaf was leveled to 110, then Treebeast, Ivan, Brittany, Masked Samurai and Forest Seer were used for DPS from then on. They were leveled in increments of 25, along with any other heroes whose random gilding made them useful. In the QA run, Natalia picked up three gilds which made her a good contributor to the group. Dread Knight also picked up three gilds and was used to level 100 when feasible in the QA run. For the gilds run, the main five heroes had acquired enough gilds of their own that others were not useful so I stuck with only them.
After Regilding, only the gilded heroes were leveled for DPS.
Probably the most difficult situation to assess objectively was when to ascend. I finally settled on these criteria:
- Skills would be used to extend ascensions as far as they could go. This was to even out the effects of treasure chest drops and have a consistent cutoff point for them. This also allowed me to have a life to some extent while waiting for skills to cool down late in a run, since I was playing fairly intensely. Ascending as soon as I got to a boss I couldn’t idle past was not sustainable. This was still somewhat arbitrary, but the best I could do.
- Until Vaagur was maxed, runs were ended with a single skills use.
- After that, runs would be extended with a 1234 combo until 7 was necessary to beat a boss. Skill 7 (Superclicks) was then reloaded and a final 123487 combo was used on the last boss. If this boss was not primal, it would sometimes be skipped rather than waiting for the skills cool down to finish.
- Once ancients were leveled to 40, skills would be used (123479 and 123487) on only two additional bosses.
- After the second Regilding, no skills would be used on bosses (since even an 87123 was barely better than idle damage at that point).
- Energized Dark Rituals were used every 30 minutes when possible.
Ancients and Regilding
The plan I used for buying and leveling ancients was based on the walkthrough at Wikia, with some modifications to streamline and clarify it. The object was not to have an ideal buying and leveling system to maximize efficiency, but rather to have clear decision points so that each game would be as identical to the other as possible.
In the scheme, HS refers to the number of free hero souls listed on the screen, as opposed to the total number earned. Ancients were generally bought in groups to minimize reroll costs, but were left at level 1 until the entire group was bought to minimize the impact of their order of purchase.
1. 25 HS, buy Vaagur.
2. 50 HS, buy Siyalatas/Libertas (reroll to get one of them)
3. 75 HS, buy the other
4. 100 HS, level all ancients to 10, without dropping below 100. Take Vaagur to 15.
5. 200 HS, buy one of Mimizee/Mammon/Dora/Fortuna/Dogcog
6. 250 HS, buy another one
7. 325 HS, buy another one
8. 425 HS, buy another one
9. 525 HS, buy last one
10. 400 HS is now the minimum. Level all ancients to 20 without dropping below 400.
11. Regilding 1: Use random regilds to move gilds to Treebeast, Ivan, Brittany, Masked Samurai, and Forest Seer
12. 900 becomes minimum. Level all ancients to 30 without dropping below 900.
13. 1600 becomes minimum. Level all ancients to 40 without dropping below 1600.
14. 2500 becomes minimum. Level all ancients to 50 without dropping below 2500.
15. Buy Kuma and Bubos. Level both to their max without dropping below 2500.
16. Buy Morgulis, Atman, Solomon, and Argaiv without dropping below 2500.
17. Level Morgulis to 2500, then all other ancients to 50.
18. Regilding 2: Move all gilds to Masked Samurai with 80 HS direct move.
19. Level Morgulis to 3600, all other ancients to 60.
20. Level Morgulis to 4900, all other ancients to 70.
21. Continue with Morgulis to (10n)2, others to 10n until zone 1000 is reached.
I chose to treat QAs a hero souls bonus rather than a true ascension. This made direct comparison of the two run more straightforward. The numbers below thus reflect actual ascensions for the QA run.
QAs bought during ascension # (value): 1 (7), 2 (9), 4 (11), 7 (12), 11 (14), 17 (16), 26 (19), 35 (29), 41 (33), 51 (52), 56 (116), 66 (303), 74 (559), 89 (8796)
Total QA purchases: 14
Total HS from QA: 9976
Total HS earned: 188442
Hero souls from rubies were about 5% of the total
109 ascensions (95 actual ascensions, 14 QAs)
152,933 hero souls spent on ancient levels (Morgulis to 19,600, all others to 134)
13 days, 3 hours elapsed from first click, for an average of 3.32 hours per ascension.
Gilds bought during ascension #: 1,1,3,4,6,8,10,15,19,25,30,34,38,41,44,46,51,56,62,65,69,72,78,85 = 24 purchases
Total Gilds bought: 72
Total Gilds earned from game: 91
Gilds from rubies were about 44% of the total.
105,918 hero souls spent on ancient levels (Morgulis to 14,400, all others to 113)
13 days, 20 hours elapsed from first click, for an average of 3.82 hours per ascension.
Landmarks on this graph (Ascension # for following events):
Event QA Ascension # Gilds Ascension #
Siyalatas and Libertas to 10, Vaagur to 15 16 19
Mammon/Mimzee/Dora/Dogcog/Fortuna bought 40 40
Ancients leveled to 20, Regilding 1 55 56
All ancients to 30 64 61
All ancients to 40 68 65
Bubos/Kuma bought and maxed 76 70
Argaiv/Atman/Solomon/Morgulis bought 82 77
All ancients to 50, Regilding 2 86 81
Zone 1000 reached 95 87
Buying gilds seemed to produce better results, getting to zone 1000 with fewer ascensions and lower ancient leveling required.
Observations and Discussion
The QA run did indeed jump out to a fast start, as its proponents have suggested. This was helped by the random number generator, which produced a second crop of 50 rubies only 9 hours after the first one. However, that lead had evaporated within five days, with the results of the two test runs almost interchangeable up to the first regilding. At that point, the gilded runs began to pull ahead and stayed ahead to the end,
The Gilds caught up to the QAs due to the sheer number of gilds and their distribution. By Ascension #25, an additional 30 gilds had been purchased and as a result nearly all the heroes had at least one gild. Thus the 50% DPS bonus applied almost constantly, making the progression through the heroes much faster. It can be seen on the graph above that this is where the Gilds run first began to get to higher levels than the QA run. The extra damage made it possible to get an extra boss or two at the end of the each ascension. It did not take long for the extra bosses to produce as many hero souls as had been obtained from the Quick Ascensions, nullifying their advantage.
Buying gilds had a distinct disadvantage in the cost of Regilding. With the QA run, there were 20 gilds to move at the first regilding which cost 240 HS, about one run’s worth. With the Gilds run, there were 72 to be moved, which cost 660 HS or three runs’ worth. The second regilding was similar, with the gilds run taking roughly three runs’ worth of hero souls to move all the gilds to Masked Samurai.
The main observable difference between the two runs while playing was the slowness of the Gild run. Gilded ascensions had very long tails, puttering along with just enough DPS to beat the next boss for a long time. They ended up at higher levels, but took a long time to get there. In contrast, the QA ascensions tended to end rather abruptly, with skills quickly losing their effectiveness against bosses and producing a much clearer end point.
This time difference was probably an artifact of the ancient leveling scheme I followed, which was aimed at simplicity and reproducibility rather than efficiency. I suspect the effect would largely disappear with a superior scheme, and the fewer number of ascensions would result in a faster overall time to zone 1000 for the Gilds. Thus I believe the shorter time for the QAs run was mostly due to not buying gilds, as opposed to significant time savings from the QAs themselves. Of course, the proper way to test this would be with a control run where nothing at all was bought, but I will leave that test to someone else at the moment. However, based on the present data it is true that buying Quick Ascensions progressed to Zone 1000 in less total time.
Based on these data and with the changes added with version 0.19, I will recommend to new players that they invest their rubies first in the permanent 2x DPS bonus, and then in gilds. They show a distinct advantage in the early game. Quick Ascensions in the early part of the game have little value and do not make much practical difference, whereas gilds are quite scarce and buying extras makes an outsized impact.
Beyond zone 1000 remains untested, however, and I would not generalize these results to the rest of the game. There is no further consolidation of gilds to boost their impact and higher leveling of Argaiv can mimic the effect of more gilds, so it would seem likely that their value would drop after the second regilding. I will leave it to others to demonstrate that, however.
I do not consider this the final word on the issue. This is one test under one set of conditions. Another test using ancients buying and leveling schemes optimized for their respective purchases would be informative, as well. I would encourage much more play testing in general. I will do my best to answer any questions and criticisms in the comments.