# Talk:Scroll of cleansing

Jump to: navigation, search
This talk page is for discussing the Scroll of cleansing page.

## Untitled

Anyone know how this scroll effects Overloads? Does it save the Torstol or all the potions apart from 1? 00:29, November 3, 2010 (UTC)

With overloads it does save the torstol's, i tried it with 270 torstol's.. ended up with 310 potions. Amun erebos 22:12, December 9, 2010 (UTC)

On the first couple days it used to save the avantoe and dwarf weed on extreme attack and extreme strength respectively. Shame they changed it. I took full advantage of that! 81.97.184.92 01:34, November 5, 2010 (UTC)

Torstols are saved, extreme potions are not. (I've seen the effect happen personally.) Volons (talk) 06:11, July 15, 2013 (UTC)

## Data Collection

This reward is pretty awesome and the supposed 10% efficiency can have quite a variance,I suppose depending on "luck". On my first test I made combat pots (harralander + goat horn) beginning with 318 harralander, resulting in 416 combat pots (30.8 % efficiency). It would be interesting to see other similar data reported. Sanctumanon 07:47, January 3, 2011 (UTC)

I am having similar luck on my first 500 potions... rate seems to be closer to 25-30% rather than 10%.... --Condor85 20:46, January 3, 2011 (UTC)
You two need to remember that it's 10% for each action, so getting 25% more potions is less surprising when you remember you actually did two actions: vialing and then adding a secondary. I do agree, though, that the rate is slightly higher than 10%: from the 10,000 irits I just vialed I got 11,264 unfs, which is 11.2% efficiency. (wszx) 06:14, March 2, 2011 (UTC)
If you start with 10,000 herbs, saving 10% leaves you with 1,000, which when used leaves 100, and then 10, and then 1. This adds up to getting 11,111 herbs out of the original 10k, for an effective rate of 11.11% more herbs than you really had, even though you're only saving 10% of the herbs you used. Your 11,264 unfinished potions from 10,000 herbs is only 153 (less than 2%) more than the expected value.
Getting 25-30% more potions than herbs used is a bit odd, though, if the effect's only meant to be for the herbs and secondaries. It should be something like 1000 herbs + 1000 secondaries + 1111 vials of water = 1111 potions, with all ingredients used up. --MarkGyver (talk) 06:31, March 3, 2011 (UTC)
^The above is a sound mathmatical argument, though through my herblore experience with the scroll (I'm 99) i've found on multiple occasions of samples of 10k potions that the results are closer to 12.5% bonus (1/8, similar to the speed increase effect). I think 4-5 samples of 10k potions is more then enough to say that 11.11% and 12.5% are significantly different. ~Rory
^^The above is *not* a sound argument. The expected bonus potions is 1111 for 10k herbs used, thus the 153 extra bonus potions are 153/1111 more than expected. That's almost a 14% difference from expectation, and is a result that seems to happen consistently. I would argue that the actual rate is at least 11%, if not closer to 11.5%, rather than the 10% stated in the literature. This would yield an expected 112.5%-113% of potions made per herbs used.Perlmutter 17:56, February 20, 2012 (UTC)

2199 kwuarms mixed into 2510 kwuarm potions (1.14 times the kwuarms, or 14.1% rather than 11.1% expected bonus) - Hell Man 2nd (Perlmutter 09:06, June 23, 2011 (UTC) -> didn't remember to sign before)

2100 kwuarms mixed into 2380 kwuarm potions (1.13 times the kwuarms or 13.3% rather than 11.1% expected bonus) - Hell Man 2nd Perlmutter 09:06, June 23, 2011 (UTC)

5151 toadflax mixed into 5774 toadflax potions Perlmutter 00:41, February 21, 2012 (UTC)

Based on these results, and those of my own experience, I think I could confidently say that the probability of saving an herb is greater than 10%. I would hazard to estimate 11.5%. Beginning with exactly 14k clean herbs, and making 14k unf pots, I wound up with 1626 remaining herbs (herbs saved). This, of course, is about 11.6%. Assumin the probability is 11.5%, the total herbs saved from about 10k unf pots made would be 1150. Continuing to unf these 1150 herbs would save an additional 132.25 on average. From unfing these, one would save about 15.2 herbs on average. Doing this one last time would yield about 1.75 herbs saved on average. Adding up all of the herbs saved we get 1299.2, or about 13% more unf pots than the 10k herbs we started with. I think this is pretty accurate with the above samples. Maybe it's even higher than that, but considering the common problem of people sharing unexpected results more commonly than expected results, I think 13% is fairly accurate. Further testing should be taken into account. I will try to share my results here in the future. 15:24, March 9, 2012 (UTC)

From a test of 37,079 herbs made into unf pots, 4193 herbs were saved. This is an average of 11.31%. Based on this, I conclude that The probability that an ingredient will be saved is about 11.0-11.5%.
For future testers, I leave this advice. To calculate the average herbs saved, take the difference between your starting herbs, and final unf pots, and divide that by your final amount of unf pots. Where A is the number of herbs you begin with, and B is the number of potions you end with:
Herbs saved = (B - A) / B
This will give you a simple, and effective result. 05:20, March 13, 2012 (UTC)

Let's call the percentage of herb saved t. The total number of unfinished potions produced from a starting pile of X herbs (and as many vials of water as you need) will be

${\displaystyle X+Xt+Xt^{2}+Xt^{3}+...}$

which sums to

${\displaystyle X{\frac {1}{1-t}}}$

Any time a measurement is taken using a starting pile of herbs and unf'ing until all the herbs are used, the ratio of final unf pots to original herbs will be ${\displaystyle {\frac {1}{1-t}}}$. This allows us to isolate t, just as Dan did with his own special measurement (thank you for that, by the way), but we can now run a calculation on every data submission listed above (including one of my own, which I just did recently). To find t from the number of starting herbs and ending potions, we take P for total unfinished potions made, H for number of starting herbs, and use

${\displaystyle t=1-{\frac {H}{P}}}$

And we get:

• 3807 potions from 3318 herbs (harallander)[1] gives t=12.84%
• 11264 potions from 10000 herbs (irit) gives t=11.222%
• 2510 potions from 2199 herbs (kwuarm) gives t=12.39%
• 5774 potions from 5151 herbs (toadflax) gives t=10.79%
• Skater Dan Man gave a direct measurement of extra herbs (unknown herb) produced, giving t=11.31%
• 5598 potions from 5000 herbs (toadflax) gives t=10.68% (my measurement)
• 3361 potions from 3000 herbs (dwarf weed) gives t=10.74% (my measurement)
1. ^ This is the sum of two trials, the first harallander trial in this talk page, and the one from the person who posted below me

Conclusion 1: The percentage of herbs saved might depend on the level of the herb, possibly with higher level herbs having t closer to 10%.
Conclusion 2: The pencentage of herbs saved does appear to be higher than 10%.

I'm about to conduct an experiment with 3000 Dwarf Weed. EDIT: Done, added to the above table. Volons (talk) 11:19, July 15, 2013 (UTC)

I also suggest someone conducts a trial with 3K+ harallanders so we can add the two harallander trial results together (to account for possible extremeties in the above harallander trial). --Volons (talk) 05:48, July 18, 2013 (UTC)

Wait, and the Harralander data I posted a few days ago won't do? <.< ... IP83.101.44.209 (talk) 05:55, July 18, 2013 (UTC)
Ooh, an interesting conversation =3 . Anyhow, I dug up my personal notes on some Herblore I did earlier this year. I made Energy potions, starting with 3000 Dirty harralander herbs. Those 3000 turned into 3391 Harralander potion (unf)s. My 3000 Chocolate dust allowed me to make 3386 Energy potions (aka, I was 5 Chocolate dust short). The noted "23.6%" for Harralander above seems very odd. All other percentages (roughly) seem acceptable amounts of variation on the listed 10%. Will take a while before I do any more Herblore, but I'll try and remember this Talk page when I do ^^ . IP83.101.44.209 (talk) 11:26, July 15, 2013 (UTC)
The t-value should be 11.11% (if the 10% is accurate) since you make potions from the herbs you save and therefore end up with 11.11%. (109.58.113.122 10:29, July 19, 2013 (UTC))
Incorrect, the average amount of herbs saved will be 11.11%; the t-value in the formula above is the 10% that the effect saves you per herb. With t = 10%, the modifier ${\displaystyle {\frac {1}{1-t}}}$ for X (the number of herbs) is 11.11% as expected. IP83.101.44.209 (talk) 10:42, July 19, 2013 (UTC)

Oh I'm sorry ! Made a mistake (happens sometimes). I'll add your harallander data to the table. --Volons (talk) 17:52, July 20, 2013 (UTC)

## multiple scrolls of cleansing

Can you buy multiple scrolls like these ? Will the effect add up ?

No, the effect will not stack. I don't know if you can buy more than one of them, but doing so would be a waste of tokens, wouldn't it? --Andorin (Talk) (Contribs) 08:19, January 16, 2011 (UTC)
I wondered this as well. There seems to be nothing to stop you from purchasing multiple single-use-permanent-effect items, unless it blocks you at the point of pressing the purchase button - And I'm loathe to waste tokens on it to see if Jagex left a little bug in there letting you do so. I've taken to writing down everything I buy from Marmaros so I don't accidentally get the same thing twice just in case it is possible. 09:12, January 16, 2011 (UTC)

can some one add the info in the pic

It's not possible to buy more than one Scroll of Cleansing, or any other dungeoneering scroll:

I'll add this to the main page. EDIT: Nevermind, already done. Volons (talk) 11:29, July 15, 2013 (UTC)

## This is called Wasteless Herblore

Both in-game guide and at the shop.

ff cook me plox. check it yourself.

Wasteless Herblore is used in the in-game guide because Jagex did not change it from its original concept name. I'm not sure about the shop though. talk 08:40, April 29, 2011 (UTC)
I just checked, it is still called the scroll of cleansing in the shop. 09:09, April 29, 2011 (UTC)

The description of the scroll in the Daemonheim Rewards shop is Reading this scroll permanently unlocks the wasteless herblore ability, ...' Volons (talk) 11:33, July 15, 2013 (UTC)

## Fastest source of Herbs!

Considering the principal that an herb saved is an herb gained, this scroll makes making unf potions one of the fastest sources of herbs in the game! In my experience, I have been able to make 2800 unf potions/hr. Assuming that you save 10% of all herbs used (even though testing suggests the agerage is significantly greater than that), you could effectively gain 280 herbs per hour. Granted, this is a method that requires constant attention, while farming and miscelania involve massive amounts of afk time, and brief periods of attention for some considerable herb yield. I don't know much about the best methods of gathering herbs through monster killing. I haven't tested cannoning banshees or abberant spectres, or other good gathering methods. However, I can say that this is definitely better herb yield per hour than gathering at the sorceress's garden (even on a one clickable world without bots, as if they exist anymore). Sorceress's garden and monster killing also yield a variety of herbs, while making unf potions you gain any herb you have the level to use. Making snapdragon potion(unf) = about 3M+/hr by current prices. 15:45, March 9, 2012 (UTC)

Must have been a hidden update because the page previously stated it didnt work on guam and guam potions unfinished, but im making attack potions right now and ive saved guams and eyes of newts so UPDATE :)  —The preceding unsigned comment was added by Dark reave (talk) on 05:06, March 15, 2012.

## Terminology

Can we please be more careful about the terminology we use in this article? A lot of the text on the article gets the meaning across, but is technically incorrect. I've corrected some of it myself, but haven't yet corrected all of it. The type of stuff i'm talking about is:

-Not stating that something is an average rather than a set number.

-Using incorrect formulas.

-Stating the average % of herbs saved in a way that implies that it is the average percent that a particular herb is saved.(WoHenRe (talk) 16:46, December 15, 2013 (UTC))

## Portable Well speed

When used on a Portable Well, does the 1/8 chance to make potions faster actually work on a Portable well? I'm not talking about the ingredient saving. Has anyone confirmed that potions will still be made faster on a Portable Well?

Puretppc (talk) 17:04, September 12, 2014 (UTC)

I can confirm this. Alercah (talk) 20:51, 10 May 2020 (UTC)

## Combinations potions

The scroll also work when doing combinations potions. I just saved an overload while doing a Searing overload potion. This is amazing! I know when you make overloads the "secondary" that is usually saved it the herb itself, but what in combinations potions... More testing required Proulxs (talk) 01:36, November 19, 2014 (UTC)

It's completely random. It's also updated so that overloads can save any potion or the herb. 01:37, November 19, 2014 (UTC)

## Herblore level

Sadly I didn't keep precise track of herbs saved, but when I reached level 92 I bought 8K fellstalk and created unf potions. Previously I had done that with torstol, and 8K would save me a little over 1K, but at level 92 I saved a few less than 1K. However, when I reached level 93 I bought another 8K fellstalk and I saved slightly more than 1K, to the extent that at the end of the 16K (8 of which mixed at level 92 and 8 at 93), I had saved a total of 2032.

My hypothesis is that one's herblore level plays some part in how many herbs are saved. Maybe it is gradual, maybe it is with a few big steps.

TheBramNotABram (talk) 19:25, August 1, 2017 (UTC)

## Barbarian Potions

Just had some Caviar saved while making barbarian potions. Perhaps it was changed in an update?

173.64.93.192 03:10, 26 June 2019 (UTC)

## Issues with the calculations

There's a subtle error in the above analyses, which is using the mean of the distribution (which works out to a negative binomial) as the estimator for the probability t. Fixing that error probably isn't going to have a large effect on the results, but it's worth bearing that in mind. With proper Bayesian analysis, it is possible to combine these into a single estimate, but it's needlessly complicated to do that. We can do better by just measuring directly how often the scroll of cleansing triggers, like Sk8r Dan Man did many years ago, rather than trying to measure the outcome of this negative binomial distribution (which is what I'm assuming all the other data are measuring). Then the analysis to get a parameter estimate, with confidence, is dead simple. So I am going to try to get more data on this as I work on herbs over DXPW, in hopes of nailing this down and being able to give a confident estimate. Alercah (talk) 05:15, 30 April 2020 (UTC)

I made some mistakes in data gathering so I had to discard some, but I still got a few thousand potions of good data. The methodology deserves a tad of explanation: for the extremes, I was calculating by picking a target number of potions and going until that target was obtained. This wasn't done by picking a fixed number in advanced, but the stopping point was dependent only on potion duplication and 4-dose creation chances. These should be independent of the chance of scroll of cleansing proccing; assuming that is true, these are equally valid as if we had used a fixed quantity in advance. The number of "potions created" was the number of super potions used, i.e., the number of times the recipe was crafted. The supers were done with a fixed quantity of unfinished potions, so they are the ideal of data gathering.

• 252 extreme attack used 213 avantoe. (39 saved)
• 272 extreme strength used 245 dwarf weed. (27 saved)
• 282 extreme defence used 258 lantadyme. (24 saved)
• 242 extreme magic used 224 ground mud runes. (18 saved)
• 236 extreme ranging used 1150 grenwall spikes. (230 sets of 5; 6 saved)
• 1379 super magic used 1215 potato cactus. (164 saved)
• 1496 super ranging used 1339 Wine of Zamorak. (157 saved)

I also counted procs for overloads, 342 overload crafts saved:

• 4 torstol
• 3 extreme strength
• 6 extreme defence
• 14 extreme magic
• 7 extreme ranging

This is clearly not powerful enough to determine if the savings are unbalanced overall, because of the low rate of saving any specific ingredient, but more data could be gathered in the future to work that out. Overall, however, that amounts to 34 ingredients saved over 342 crafts.

This is a total of 4501 potions crafted, and 469 procs. 469/4501 ~= 0.1042, so the best estimate we can get is 10.42% proc chance. Applying Bayesian analysis gives a posterior distribution of Beta(470, 4033)---Beta(successes+1, failures+1), which we could use to get a rigorous estimate of accuracy, but I've already xpwasted enough. This should be augmented with more data, but with a Bayesian approach it's fine to just throw more data on the end here and this particular analysis won't have any weird effects from that, so we can narrow down our result further and further. Alercah (talk) 20:45, 10 May 2020 (UTC)