It’s interesting they got lot of heroism but only got 1 magic. So yeah they got more easy wins against the boss but still probably got harder battle on other rooms
15 complete guilds updated
3 Likes
Hi @HugeOgre,
You have some good data compiled here, but your methodology is flawed on scroll percentages is flawed and needs correction.
Currently, you are stating that the Unlock scroll has a 25% probability of appearing in a room. This is incorrect.
Yes, on a given floor, 25% ( 1 in 4 ) of the rooms will be the Unlock scroll. But, that is not the same as saying that there is a 25% for it to appear in a given room. If that was the case, there would be a possibility for multiple scrolls to appear on a floor (or none at all).
The game logic is likely this in regard to scrolls:
– Roll a random room. Assign that room the Unlock Scroll.
– For the other three rooms on the floor, roll on the the Scroll Drop Table( ← What you are looking for).
So, what you need to do is ignore the Unlock scrolls in the data and recalculate the scroll drop probabilities from the other 75 data points for each guild. That should get you the accurate scroll drop rates.
2 Likes
True 
I do think that both analyses would be useful, though.
@HugeOgre, I’m sorry to say, I decided to just beeline the last few floors to 25, since our guild was unlikely to reach the next reward stage; and I also accidentally deleted half of our scroll data on Discord 
. When I decided not to make a spreadsheet, I was much more confident in my ability to not accidentally press ‘Delete’. We now have a spreadsheet (albeit incomplete). Definitely happy to help next time, depending on how we go .
Hi @Lyrian
yes i knew this going in. However, with unlock scrolls being a known variable/quantity per floor, this has no impact on the distribution of other scrolls. i just took a look at the data without unlock and it is simply offset. the ratio of scrolls to one another doesn’t change, the stability of a scrolls drop rate compared to others doesn’t change, and the STD remains constant. basically this is just another very similar way to look at the same data.
| Scroll | Min | Ave | Max | STD | DRC | DRT |
|---|---|---|---|---|---|---|
| Unlock | 0 | 0.00 | 0 | 0.00 | 0.00% | 0.00% |
| Armor | 14 | 19.33 | 25 | 2.79 | 25.78% | 25.60% |
| Life | 12 | 17.93 | 23 | 3.26 | 23.91% | 24.10% |
| Attack | 6 | 11.47 | 20 | 4.26 | 15.29% | 14.98% |
| Haste | 5 | 8.67 | 13 | 2.26 | 11.56% | 12.05% |
| Luck | 5 | 5.33 | 7 | 0.62 | 7.11% | 7.15% |
| Magic | 1 | 5.27 | 10 | 2.55 | 7.02% | 6.70% |
| Power | 0 | 3.93 | 7 | 2.28 | 5.24% | 5.05% |
| Fireball | 0 | 2.00 | 4 | 1.41 | 2.67% | 2.71% |
| Heroism | 0 | 1.07 | 6 | 1.58 | 1.42% | 1.66% |
| Total Rooms | 1,125 | 1,125 | 1,125 | 1,125 | 1,125 | 1,328 |
However, now i am thinking i could have got a lot more data if i took in everyone’s results from over 100 and excluded Unlocks and Forge scrolls. in this way my data collection was flawed…
(edit) well that would still throw off the Average/Min/Max/STD, those needs to be same quantity sets. but this would allow me to take in more data for drop rates
I got the info for room 27 to 51 if you want
what happened to 26? lol
sure, i shouldn’t turn down data. ill find someway to integrate it
Edit: nevermind i got it the infos from 26 to 52
starting from floor 26 the boss dropping random scroll, do you want me to specify when it’s boss loots?
Apart from what you mentioned in your post, I want to add a few things.
- It is not known if type of scroll depends on enemies or room level. It very well might. Or might not.
- After floor 25, the chances of various scroll might be different from those below floor 25.
- It is not known whether the tower is generated as a whole with some seed number or on floor basis for each guild. If there is an initial seed or configuration number, there is nothing random about it starting from floor 2. If each floor is generated separately, there might be some limited randomness. Alternatively, there is a possibility that some script runs on Sunday night and generates every floor for every Doom tower for every guild. That would have some randomness into it as well.
As all these things are unknown and the distribution of chances is also unknown, the only way to approximate the rate is to collect sufficiently large amount of data. However, I see little point in that if the chances are not independent. You cannot use the table to approximate the drop rate of a scroll based on these chances.
However, if developers would be nice and would stop endless trolling for a moment, they would have set a fixed number of scrolls per tower in random locations on the first 25 floors. It actually makes sense. For example, you are guaranteed to get 1 luck, 1 fireball, 1 heroism, etc. Somewhere. It would also be fair to the Ladder activists.
Thank you so much for this post! 