Crusaders Of The Lost Idols Calculator

See more of Crusaders of the Lost Idols on Facebook. Forgot account? Create New Account. Related Pages. Real Madrid C.F. Yeah, the calculator optimizes idols per time, not idols per run. If you can get two thirds the idols in half the time, that's four thirds the idol farming rate and hence time efficiency. And yeah, you're expected to use on average as many spawn speed buff cards per run as you get per run through the silver chests.

I've finished a new google spreadsheet talent calculator that includes all new & old talents. Usage is explained over there.If you have any questions (some settings being unclear) just ask. Fast Farm Score: You rush through the Free Play for 31 minutes (until Reset Fatigue times out) and immediately reset. The scores maximize your idols per hour when doing this. Cap Farm Score: You rush through the Free Play until you cap or almost cap all your crusaders, then immediately reset. Again, the scores judge idols per hour.

Pushing Score: You're trying to complete tough objective or push as high as possible. This score maximizes your effective DPS at your level cap. Hey MarioVX,Giving you new notes on your new calculator. By splitting the optimizer into three different types, you have side-stepped many of the issues I have been running into updating some of the old calculators that try to merge idol gain with dps gain.

When I have a moment i will try to do more checking. However, I have a few initial comments.1. Could you include a total idols spent calculation?2.

DPS affects idol gain and idol gain affects dps. I really like how you format your sheets in that you break down the steps into smaller calculations.4. Denote 'in seconds' B2-B55.

B34-36 in minutes?6. Could you put notes on the 4 sub-steps of SuperT & ExtraT to describe what it is doing. So I don't have to reverse engineer it.:-P. Someone else requested this already, guess I'll add it in tomorrow then.

Not anymore, no they don't. At least in the damage - idol direction. When farming, you reset at 31 minutes or when you cap your crusaders, either way you're instakilling all the way up to reset, hence damage increases makes no difference. The instakilling assumption makes this calculator inappropriate for really new players, but since the legendary update the point where this is valid can be reached fairly quickly, such that I didn't see it justifying a 4th column, 'pre-cap, pre-sprint farming score'.

As for idol gain - damage, you mean in the long term feedback, differential equation sort of way, right? Technically correct, and it's the reason why you should basically farm all the time until an objective becomes easy, but the connection is still false for a single run by itself. Since you could potentially respec for a very difficult objective, and then you don't want this indirect feedback to be counted, I think it's reasonable to keep it run-contained and hence separate. Uh, thanks! I wasn't expecting to get a compliment for the format of this sheet, I think actually it hits users in the face quite nastily, and it looks even more chaotic when un-hide the columns beyond I. As for the breaking down, yeah, Divide & Conquer makes things easier.:). Fair point, I'll add that.

You're peaking behind the curtain here!:P Yes, those as well as all time calculations throughout the sheet are in minutes. I'm just requesting the user input in seconds because for the requested time lengths that's a more feasible unit.

I don't think I'll annotate the back-end calculations to keep it small, but I can explain that part since you're curious. For the DPS increase a.k.a. Pushing score, each increment in those is simply a x4 DPS multiplier at cap, as it's 25 more levels. For the cap farming score, it affects it in 4 ways: A) more crusader levels = more base idols.

B) higher area needed to reach cap = more bonus idols from DIS. C1) more time needed to reach cap = more bonus idols from IOT.

C2) longer total run time = less idols per hour due to a bigger divisor. Each of these four pathways is translated to a multiplicative factor to your idols per hour and then they are multiplied together to give the total benefit.Thanks for the feedback!. A2-A5 Is the time in seconds it takes to do that what is specified in the name. Pickup + Transition is the 'fixed time' from one area to the next (transition), and you can put a little more in that if you often miss picking up the quest items, since that then is also fixed time in the sense that it won't be affected by spawn speed buffs. Reset Time is only making a tiny difference, that's the time it takes you to reset and start the next run. It gets longer if you want to reset with mobile since you need to close the game on the computer, start it on mobile, watch the 30 second ad, go through the animation, reselect your campaign, close the game on mobile, load it on your computer.

Normal area are non-boss areas, you need to guess (or measure) your average time it takes you to complete a normal area (without the area transitions, i.e. Only time from the first monster spawning to the last monster you need to kill dying).A6 is the geometric mean multiplier to monster gold from one area to the next.

The actual multipliers are periodic over the area number modulo 5 with the first couple of areas being special, so a good way to test this is in sprint mode, take new gold / old gold from one boss to the next without spending gold in between and then take the fifth root of that ratio (because that was the accumulated multiplicative increase over five areas). The default value (1.745) is roughly what I got for RP2 from some quick testing. Obviously it's difficult to pinpoint due to the low amount of decimal digits the game displays gold numbers with.A20, yep, that definitely needs some explanation, sorry it's not in the sheet. C(ommon) Spawn Buff Equivalent is a metric I created to unify Spawn Buffs of different rarities. It makes rarer spawn buffs count as much more as they are longer compared to the common one, i.e.

Common = 1, uncommon = 2, rare = 3 (rest is irrelevant since you can't get epic buff from silver chest). The value supposed to go in B20 is the average C spawn buff equivalent per silver chest. So the way you can estimate this is by opening up a tremendous amount of silver chests and writing down A=#commons B=#uncommons C=#rares D=#total chests opened, and then calculate (A.

1 + B. 2 + C. 3) / D.

I did it with a sample of 1000 silver chests and rounded the result a little (think I got 0.197, so I just wrote in 0.2). Do it with a bigger sample to get a more accurate value if you want, but I think it's pretty good.I think my rate of sc earn and length of buffs already allows me to permanently single-stack. Is that what it's basing it off of? Or triple stack?

Insight into this is appreciated. Thanks!It assumes you're roughly equilibrium stock holding your silver chests and spawn buffs, including common, uncommon and rare.

They're all the same intensity and only differ in length, so they're all interchangeable. That means your actual uptime (or rather, uprate) can be anywhere from 0 to 3, including fractional values, with the model and reality still fitting. You'd only get a divergence once you're past 3, i.e.

Here's what I've noticed in my (MILLIONS!) of attempts:.Seems to trigger most often at Gym, though I have started the route in Class as well. Seem to toss me into Prom Queen route instead.Tricking Principal Spider actually requires 14+ Smarts for the first option (the one where you set a web trap for him). I haven't had too much trouble with some of the secret endings, but Miranda's Hidden Agenda has been a HUGE pain for me to achieve! Monster prom miranda. Auditorium, Bathroom, Library etc.

If you could maintain a common, uncommon and rare spawn buff nonstop all the time without net-losing some on average. I was indecisive whether I should catch this case with a hard cap, as it would reduce the benefit of increasing the talent further to hard 0, while actually you're still benefitting a little for longer epic spawn buffs during events or when doing objectives, which is impossible to quantify. But as long as you're not over that point the calculations work fine.

I collected data on chest openings from 6 sets of 1000. The averages were: 51.83, 47.17, 28.00 (with variance from 43-59, 36-52, 22-32). That works out to 0.23 across 6000 chests.I'm earning roughly 25 chests per Sprint run, which means an average of 1.3, 1.18, and 0.70 buff cards per run, a weighted average of 5.76 common cards. So each card, assuming no lost time, would have to be only 5.21 minutes in duration to maintain perfect single stack on Sprint runs.

My current cards are 9 minutes each. And I'd want them at 10 minutes (level 7 of the talent) in order to run just 1 common + 1 uncommon. You're welcome!Good question, I'm glad someone is asking this, and thanks for pointing out the not-so-nice messing behavior for values of 0 here. Alright, let me explain.The calculator assumes you always use Thalia to cast Storm Rider during farming (he doesn't need to be in your formation permanently, just swap him in briefly to cast Storm Rider when it comes off cooldown), because he is the only one that can give a Gold Bonus with Storm Rider. Since DPS bonuses are useless during farming based on the calculator's assumption that you re-invest your gold regularly as to not get unnecessarily stuck, it cares only about stuff that affects gold and speed, roughly said. So there's no reason to use any other Storm Rider during farming except Thalia's with his legendary amulet that gives it a gold bonus.

And the only influence of that input setting is on Storm's Building and Instant Satisfaction, because a gold bonus per Storm Rider use results in higher final gold bonus if you can cast Storm Rider more often. Pushing is not affected by it. If you don't have the legendary amulet (you should get it soon then!:P ), just input 0. This means that casting Storm Rider more often doesn't provide a gold bonus anymore, hence the farming score of these talents should be 0. So interpret that #DIV/0!

Error as 0, in these cases. But I'll try to catch them and update the sheet.For Storm Rider during pushes, I have made a trade-off where I gave up a miniscule amount of precision for a gigantic amount of generality by dropping the constant '1+' the Storm Rider damage multiplier starts with for push cases. The longer the push, the more uses of Storm Rider and the higher the bonus per use, the more accurate this generalization gets. And the great thing is, it generalizes over all Storm Rider casters and all total numbers of uses at the same time. So, for this calculator, it doesn't matter which crusader you use to cast Storm Rider, because the damage is comprised of independent multipliers, the relative damage increase due to an increment of an involved talent is also independent of all those factors.The exception to this is Littlefoot, but not in regards to the Storm Rider-affecting talents, but for the pushing score of Fast Learners. Since, if you use her, have her legendary spear, mimic her with Mindy, that all increases the value of an increased XP gain.So, what about Kaine?

Again, his setting only matters in that specific sense that it gives Fast Learners increments a gold multiplier. If you don't use him with a legendary magnifying glass, your gold bonus becomes independent of XP, and hence the farming scores of Fast Learners are supposed to be 0. Again, I should catch that exception so it properly displays it as such, too. I stumbled on to your sheet a few days ago, after using this one and trying to compare between different calcs. I was looking for a calc to decide between IoT vs ET/ST vs Sprint, as I'm not quite intuitive enough with the game yet to know what I should be focusing on. While it doesn't include sprint, your sheet was very easy to just plug a few numbers in and get a rough idea what was best, so thanks for taking the time to make it!

And as is always nice, it seems to match up generally with sheet for me (thanks to you too, MarioVX). Where of course I'm stuck is adding in sprint as a thing. It's what I'm going with now, but as you've worked on your own calc too, do you feel the sprint calc in MarioVX sheet seems right in how it prioritises sprint for the 'idle' build?

For what it is worth, I'm at approx 300k idols, and MarioVX sheet suggests my next purchase be #1 and then #2 in Sprint, and then it essentially cycles between getting extra levels, iot, and sprint. I currently have 23 ET, 1 ST, 7 IoT. It feels like it is a lot of idols to spend on it out of what I have, but I'm assuming it is the right thing to do. I've checked your sheet and found where the difference comes from.Your calculation of the benefit of each of the two talents is correct. You compute it as an additive benefit (absolute increase) while I compute it as a multiplicative benefit (relative increase) but the methods are equivalent.We differ in how we attempt to calculate the efficiency, i.e. Putting the benefit in perspective relative to the cost of the talent level. Translating your approach to its multiplicative equivalent, and illustrating the difference with an example.Say there's a talent A that increases your idol gain by +300% (multiplies it with 4) at the cost of 1100 idols.

By the relative variant of your definition of efficiency, that would calculate as 3 / 1100 or 0.0027. (You'd have the current idol gain as a factor in there since you're calculating in absolute terms, but since that factor is in each talent's efficiency score it would cancel out).

Now consider two other talents B and C, every one independent of (and hence multiplicative with) the other, and they each provide +110% idol gain at the cost of 550 idols. Their efficiency, according to your definition, would be 1.1/550 = 0.002.

So according to your definition, one should buy the one, big upgrade now with their next 1100 idols. However, if you calculate where you end up with, you multiply your idols gain with 4 with the big talent A, whereas you'd end up with 2.1.2.1 = 4.41 times your idols gain if you chose B and C instead, with the very same total idols cost.So what has gone wrong? Your definition maximizes the sum of the benefits of the talents. But when they interact mostly multiplicatively, as most talents in CotLI do, we actually want to maximize their product. To do this, one could either take the 'cost'th root of a talent's effect's multiplier, or - which is equivalent - divide any logarithm of said multiplier by the cost. The ordering resulting from these two definitions is identical, just the numbers are different.Using this on the same example, ln(4)/1100=0.00126 and ln(2.1)/550=0.00135, so B and C win as they should in this example.

My sheet is using this product-maximization variant.Is that the end of the story? I'm afraid not entirely. While most talents are multiplicative with each other, some (and especially the really interesting ones) are not. They're rather multiplicative than additive, so I'm quite sure the product-maximization fits better than the sum-maximization, but it isn't a perfect fit. Basically, any calculator of the sort we are using (selecting single discrete increments based on a value function heuristic) seems to be prone to some error when selecting a step changes the benefit values of any talents. But it seems unavoidable in many instances.

If neither their additive benefits nor their multiplicative benefits are independent, it gets super hard to find some benefit definition that wouldn't display value changes and would still be applicable to all talents. It may in fact be impossible (some talents are synergistic while others are not), and that's the reason why these single-step calculators without lookahead aren't guaranteed to give the globally optimal solution to this optimization problem.To illustrate how exactly your calculator leads you off the optimal path in the previous example without calculating anything wrong: You choose the big talent A because it's the best according to your score at that given instance. If you chose talent B or C regardless, you'd see that this selection has suddenly increased the promised benefit of the other two talents.

If you select the updated other one (from B and C) again, the sum of those two talents' benefits according to your calculator is indeed greater than the benefit that talent A was assigned at the start, at identical cost! So this way isn't calculating anything wrong, but following its advice still doesn't produce the optimal solution to the problem.The product-maximization score works in this case because its scores aren't updated through the intermediate steps. On other strongly intertwined talents it still falls for the same issue. Successive levels of the same talent have the same problem.So there you have it, it's complicated!:DBut I think it's fair to say that product-maximization (log-benefit divided by cost) is a better heuristic, despite still not being always perfect.Cheers!. First of all, thanks a lot for this great calculator.

However, I do have a question.As you can see, next up is Sprint mode, but it costs 800K. I assume the calculator only takes into account the next level of each talent, and there my question arises:Taking Fast Learners to lvl 18 would mean Kaine gets an XP point if i reset at 31 minutes.

This takes around 32K Idols to achieve and is thus a lot cheaper than Sprint Mode lvl 6.Does the calculator take this into account, or does it only look at the improvement by putting in one level?.