An interesting statistical anomaly
- Blackferne
- Archbishop
- Posts:9554
- Joined:Thu May 27, 2010 4:04 pm
- Contact:
Last night in the wolf fight in Trinal's random encounter I rolled multiple 1s on a d20. That is not that uncommon I suppose. But what is really uncommon is that in the first 161 d20 rolls there were zero d20 rolls equal to 1. In an equal distribution of rolls there should have been about 8. In an equal distribution the average should be 10.5 with a standard deviation of 5.92. The collective rolls (myself excluded) to that point had an average of 12.25 with a standard deviation of 5.19.
The point I'm trying to make here is that we are playing on the honor system. If I see a trend that bucks statistical norms it indicates that I have three kinds of recourse:
1) I roll for you. I don't want to do this because I find part of the fun of tabletop is you rolling the die (or virtual die) yourself.
2) I try again to get an in forum roller going. I am hesitant to do this because my last attempt really didn't go well, and the mod is in beta anyway. If I screwed up the forum in a way that erased posts I would be pretty pissed.
3) I will have to adjust encounter difficulties and DC checks higher. This has the problem in that while you might be able to hit a lvl 5 monster if you are fudging die, will your team mate be able to do the same? If A level 5 monster hits you can you deal with the increased damage it would give?
Now if you have been fudging, just stop and be honest going forward. I don't want to see a ton of 1s and 2s to "fix the math", just be honest about it. This is the equivalent of your teacher saying "I'm leaving the room and if whoever stole my Cat Fancy magazine puts it back on my desk, we can pretend this never happened."
The point I'm trying to make here is that we are playing on the honor system. If I see a trend that bucks statistical norms it indicates that I have three kinds of recourse:
1) I roll for you. I don't want to do this because I find part of the fun of tabletop is you rolling the die (or virtual die) yourself.
2) I try again to get an in forum roller going. I am hesitant to do this because my last attempt really didn't go well, and the mod is in beta anyway. If I screwed up the forum in a way that erased posts I would be pretty pissed.
3) I will have to adjust encounter difficulties and DC checks higher. This has the problem in that while you might be able to hit a lvl 5 monster if you are fudging die, will your team mate be able to do the same? If A level 5 monster hits you can you deal with the increased damage it would give?
Now if you have been fudging, just stop and be honest going forward. I don't want to see a ton of 1s and 2s to "fix the math", just be honest about it. This is the equivalent of your teacher saying "I'm leaving the room and if whoever stole my Cat Fancy magazine puts it back on my desk, we can pretend this never happened."
Jounville Blackferne
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Re: An interesting statistical anomaly
To be fair, you haven't really given Reese an opportunity to make many d20 rolls.
Spoiler:
An interesting statistical anomaly
I do promise I have been honest but I do tent to have a higher then normal percentile and usually the dm will at points weight my die to normalize it a bit.
Re: An interesting statistical anomaly
I would aslo add I tend to have long streaks so if I start to get some low rolls I tend to be in a slump for quite a while.
- Blackferne
- Archbishop
- Posts:9554
- Joined:Thu May 27, 2010 4:04 pm
- Contact:
Re: An interesting statistical anomaly
This isn't about one person or even the fact that the group average doesn't equal a uniform distribution. This is just a not so subtle reminder, dnd isn't about winning or losing, but to make it fun there needs to be risk of failure. And when I see a data set that has not a single 1 for 161+ tries that something strikes me as "that's not right".
Jounville Blackferne
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Re: An interesting statistical anomaly
What I had several 1's and 2's.
Admittedly they were all in damage rolls which was massively disappointing. Nothing like making a striker built to hit like a truck and when you do you hit it's like a truck with feather pillow fenders.
Admittedly they were all in damage rolls which was massively disappointing. Nothing like making a striker built to hit like a truck and when you do you hit it's like a truck with feather pillow fenders.
They say when you go on a quest for vengence to dig two graves. They underestimate me.
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel
Re: An interesting statistical anomaly
I think Tak's dicemod is hacked...
After my amazingly bad damage rolls I decided to go through my damage rolls since we started using the dicemod and the results were interesting...
I've used it to roll d8s on 39 occasions thus far for a total of 168, which is an average of 4.3 per roll, which is slightly below average. Additionally I've only rolled one 8 that whole time, but six 1s.
I've used it for d6s on 23 occasions thus far for a total of 69, which is an average of 3 per roll, which is also below average. I have rolled three 6s over that time, but six 1s.
So that also means that in 62 damage rolls I've only hit the max die value 4 times, but the minimum 12 times.
Admittedly 62 rolls is not a large enough sample size to be scientifically sure that it is hacked against me...but god damn does it feel like it sometimes. What good is a striker who hits like he's using a pillow?
After my amazingly bad damage rolls I decided to go through my damage rolls since we started using the dicemod and the results were interesting...
I've used it to roll d8s on 39 occasions thus far for a total of 168, which is an average of 4.3 per roll, which is slightly below average. Additionally I've only rolled one 8 that whole time, but six 1s.
I've used it for d6s on 23 occasions thus far for a total of 69, which is an average of 3 per roll, which is also below average. I have rolled three 6s over that time, but six 1s.
So that also means that in 62 damage rolls I've only hit the max die value 4 times, but the minimum 12 times.
Admittedly 62 rolls is not a large enough sample size to be scientifically sure that it is hacked against me...but god damn does it feel like it sometimes. What good is a striker who hits like he's using a pillow?
They say when you go on a quest for vengence to dig two graves. They underestimate me.
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel
- Blackferne
- Archbishop
- Posts:9554
- Joined:Thu May 27, 2010 4:04 pm
- Contact:
Re: An interesting statistical anomaly
That is interesting.
Jounville Blackferne
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Re: An interesting statistical anomaly
I just tested 60 1d10 rolls and ended up slightly higher than average with 1 - 1 and 3 - 10's
looks good to me
looks good to me
Spoiler:
Re: An interesting statistical anomaly
https://docs.google.com/spreadsheets/d/ ... sp=sharing
Here are the stats for the dicemod as of March 30, 2014.
-Takanudo (is no spreadsheet ninja)
Re: An interesting statistical anomaly
Well, hopefully the link works.
-Takanudo (is apparently no embedding wizard, either)
-Takanudo (is apparently no embedding wizard, either)
Re: An interesting statistical anomaly
Out of curiosity, I updated the dicemod stats spreadsheet. The last time I updated it you guys had made about 1500 rolls. You are now passed 5000. More than half have been 1d20.
The stats on 1d20 are a mean of 10.65681984 with a standard deviation of 5.781629611. One would expect a mean of 10.5 with a standard deviation of 5.766281297. So not perfect, but hopefully close enough that not everyone in the party ends up being killed by bunnies as soon as they start.
Full spreadsheet
-Takanudo (will be working on converting the phpBB 3.0 mod into a phpBB 3.2 extension this weekend)
The stats on 1d20 are a mean of 10.65681984 with a standard deviation of 5.781629611. One would expect a mean of 10.5 with a standard deviation of 5.766281297. So not perfect, but hopefully close enough that not everyone in the party ends up being killed by bunnies as soon as they start.
Full spreadsheet
-Takanudo (will be working on converting the phpBB 3.0 mod into a phpBB 3.2 extension this weekend)
Last edited by Takanudo on Wed Feb 01, 2017 1:45 am, edited 1 time in total.
- Blackferne
- Archbishop
- Posts:9554
- Joined:Thu May 27, 2010 4:04 pm
- Contact:
Re: An interesting statistical anomaly
Takanudo wrote: ↑Tue Jan 31, 2017 1:33 amOut of curiosity, I updated the dicemod stats spreadsheet. The last time I updated it you guys had made about 1500 rolls. You are now passed 5000. More than half have been 1d20.
The stats on 1d20 are a mean of 10.65681984 with a standard deviation of 5.781629611. One would expect a mean of 10.5 with a standard deviation of 5.766281297. So not perfect, but hopefully close enough that not everyone in the party ends up being killed by bunnies as soon as they start.
Full spreadsheet
-Takanudo (will be working on converting the phpBB 3.0 mod into a phpBB 3.2 extension this weekend)
Bunnies...I knew there was a monster I was forgetting.
Jounville Blackferne
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Re: An interesting statistical anomaly
crap I foresee the Monty Python and the Holy Grail bunny making an appearance now.Blackferne wrote: ↑Tue Jan 31, 2017 3:22 pmTakanudo wrote: ↑Tue Jan 31, 2017 1:33 amOut of curiosity, I updated the dicemod stats spreadsheet. The last time I updated it you guys had made about 1500 rolls. You are now passed 5000. More than half have been 1d20.
The stats on 1d20 are a mean of 10.65681984 with a standard deviation of 5.781629611. One would expect a mean of 10.5 with a standard deviation of 5.766281297. So not perfect, but hopefully close enough that not everyone in the party ends up being killed by bunnies as soon as they start.
Full spreadsheet
-Takanudo (will be working on converting the phpBB 3.0 mod into a phpBB 3.2 extension this weekend)
Bunnies...I knew there was a monster I was forgetting.
- Blackferne
- Archbishop
- Posts:9554
- Joined:Thu May 27, 2010 4:04 pm
- Contact:
Re: An interesting statistical anomaly
More likely a swarm of bunnies so Tony Plush gets injured by there amount of tail coming his way.
Jounville Blackferne
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
Archbishop of the Church of Alvis
Dungeon Master of the Tabletop Forum
- Jimer Lins
- Deacon
- Posts:2999
- Joined:Fri Jul 30, 2010 10:53 pm
Re: An interesting statistical anomaly
The Killer Rabbit of Caerbannog, to be precise.
A man may fight for many things. His country, his friends, his principles, the glistening tear on the cheek of a golden child. But personally, I'd mud-wrestle my own mother for a ton of cash, an amusing clock and a sack of French porn. - Edmund Blackadder
Re: An interesting statistical anomaly
Never underestimate the amount of tail Tony Plush can chase/handle...Blackferne wrote: ↑Tue Jan 31, 2017 4:33 pmMore likely a swarm of bunnies so Tony Plush gets injured by there amount of tail coming his way.
They say when you go on a quest for vengence to dig two graves. They underestimate me.
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel
Re: An interesting statistical anomaly
I do not have to be the fastest runner in the group. Just faster than Tony Phlush.Blackferne wrote: ↑Tue Jan 31, 2017 4:33 pmMore likely a swarm of bunnies so Tony Plush gets injured by there amount of tail coming his way.
___
Sorry, got nothin.
Sorry, got nothin.
Re: An interesting statistical anomaly
Pretty sure just telling Tony "Look some tails after you" will distract him long enough to get away, which doesn't really require being faster than him.Maizan wrote: ↑Wed Feb 01, 2017 1:45 pmI do not have to be the fastest runner in the group. Just faster than Tony Phlush.Blackferne wrote: ↑Tue Jan 31, 2017 4:33 pmMore likely a swarm of bunnies so Tony Plush gets injured by there amount of tail coming his way.
Re: An interesting statistical anomaly
Where tail is concerned Tony Plush isn't worried about speed, only endurance.Staris wrote: ↑Wed Feb 01, 2017 6:17 pmPretty sure just telling Tony "Look some tails after you" will distract him long enough to get away, which doesn't really require being faster than him.Maizan wrote: ↑Wed Feb 01, 2017 1:45 pmI do not have to be the fastest runner in the group. Just faster than Tony Phlush.Blackferne wrote: ↑Tue Jan 31, 2017 4:33 pmMore likely a swarm of bunnies so Tony Plush gets injured by there amount of tail coming his way.
They say when you go on a quest for vengence to dig two graves. They underestimate me.
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel
The Plush Family:
DPS - Furiel: Commando, T'ony: Slinger, Kessilee: VG, Ziraava: Sent
Tank - Montae: Shadow, Gaen: Guardian
Heals - Kam'ryn: Sage, Sabran: Soundrel