# Can someone explain me the math behind polls?
**URL:** https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900
**Category:** Support
**Created:** [Novembre 22, 2015, 9:13 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900 "2015-11-22T09:13:20Z")
**Posts on this page:** 7
**Page:** 1
### Author:  [@terraboss](https://meta.discourse.org/u/terraboss)
#### Post date: [Novembre 22, 2015, 9:13 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900/1 "2015-11-22T09:13:20Z")
It’s a little bit confusing for me and others to understand the mathematical summary behind the percentages.
I’m the opinion that all answers per poll should result to 100% and not 200%. Even for two votes. That’s not logical.
Is this a bug?
---
### Author:  [@jomaxro](https://meta.discourse.org/u/jomaxro)
#### Post date: [Novembre 22, 2015, 9:19 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900/2 "2015-11-22T09:19:25Z")
Wasn’t this just fixed?
> [@Polls percentage accounting is not accurate](https://meta.discourse.org/t/polls-percentage-accounting-is-not-accurate/35772):
>
> It seems when calculating percentage for each option in polls Discourse uses floats truncating instead of rounding. This can result into inaccurate percentage values. For instance, a poll with 2 options, 11 voters and votes spread as 5 for the 1st option and 6 for the 2nd option indicates 45% and 54% which is 99% in total but should be 100%. The math looks to be the following: 100/11\*5 is 45.45, which rounded to 0 decimal places is 45, and also truncated to 0 decimal places is 45. 100/11\*6 is…
---
### Author:  [@terraboss](https://meta.discourse.org/u/terraboss)
#### Post date: [Novembre 22, 2015, 9:20 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900/3 "2015-11-22T09:20:27Z")
I’m running the latest-release +59. ☹
---
### Author:  [@gerhard](https://meta.discourse.org/u/gerhard)
#### Post date: [Novembre 22, 2015, 10:59 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900/4 "2015-11-22T10:59:36Z")
The meaning of that poll result is:
- 49% of all voters selected “Montags”
- 24% of all voters selected “Dienstags”
- …
This seems to be a multiple choice poll, so it is perfectly normal that the sum of the percentages is greater than 100%. But the -1% definitely looks like a bug.
---
### Author:  [@zogstrip](https://meta.discourse.org/u/zogstrip)
#### Post date: [Novembre 22, 2015, 1:23 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900/5 "2015-11-22T13:23:55Z")
Yeah I will deactivate the “rounding up to 100%” on multiple polls. Nice catch 👍
---
### Author:  [@zogstrip](https://meta.discourse.org/u/zogstrip)
#### Post date: [Novembre 23, 2015, 10:35 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900/6 "2015-11-23T10:35:01Z")
@terraboss, I just pushed a fix for multiple choice polls 🍤
[https://github.com/discourse/discourse/commit/9d2128dad30aaa96a0b13be666daabbbba2993f8](https://github.com/discourse/discourse/commit/9d2128dad30aaa96a0b13be666daabbbba2993f8)
---
### Author:  [@zogstrip](https://meta.discourse.org/u/zogstrip)
#### Post date: [Novembre 23, 2015, 12:35 UTC](https://meta.discourse.org/t/can-someone-explain-me-the-math-behind-polls/35900/7 "2015-11-23T12:35:09Z")
This topic was automatically closed after 2 hours. New replies are no longer allowed.