raffle/random choice software

[Joe Hone]

I thought I read about something like this here on DC before but the search engine is failing me. I'm working with a non-profit foundation that wants to award nominal amounts of cash to charitable groups in the USA on a random selection basis. These are small groups, often church ministries, providing localized services like special needs tutoring, food banks, cholesterol screening, etc. The non-profit doesn't want to use the words "chance" or "lottery" in the application notices, and it most definitely wants to use a software program to ensure the selection process is fair. It expects a hundred or more applications for about 20 awards. I've done web searches, but as always, DC will give me a better start than guessing what software works and what doesn't. Any input or advice?

[Ath]

Maybe mousers World Cup Predictor could be a start for you? It's not exactly what you've asked, but it does give some kind of a prediction

[TaoPhoenix]

Forgive me if I'm getting out of my depth. I have a way of doing that.

But given this statement,
"It expects a hundred or more applications for about 20 awards"

I don't get why something like thirty lines of 1980's grade Basic code doesn't work?

Aka, putting a few frills on the front end, this should be a snap as a Coding Snack here!

[ayryq]

A site I frequent (dcrainmaker.com) sometimes has giveaways; the site owner uses random.org to choose a commenter for the prize. They have many functions, including one that picks multiple, unique, random numbers from a set.

Here's a link to choose one set of 20 random numbers, chosen from 1-100. (You could number your groups)
https://www.random.o...mat=html&rnd=new

They are super geeky about making sure their numbers are random; here for example are their statistics: https://www.random.org/analysis/

[ayryq]

I should add that it is key that you take the (first) output of whatever software you go with. Trust the randomness and don't "try again" if the results you get don't seem "random enough."

[mouser]

Aryq's suggestion of random.org is a good one if you are simply trying to choose m winners from a group of n entrants.

There are a few specialized cases where you may need to look deeper:

1. If you want to select winners not purely randomly, but with some weighted probabilities -- for example if you wanted the chance that an entrant to win to be based on how many years they have been in operation, etc. [doesn't sound like you need this].
2. If you needed to be able to *prove* to some third party rigorously that you only ran the random selection once and these are the independently verified results. [doesn't sound like you need this].
3. If you were going to be running multiple giveaways over time and wanted to make it less likely that people would win twice in a row, etc. [doesn't sound like you need this].

[ayryq]

1. If you want to select winners not purely randomly, but with some weighted probabilities -- for example if you wanted the chance that an entrant to win to be based on how many years they have been in operation, etc. [doesn't sound like you need this].

-mouser (June 09, 2015, 08:32 AM)

You could simulate this by just assigning (some) entrants multiple entry numbers.

[Joe Hone]

Thanks for the suggestions, if random.org doesn't seem right I'll ask for a Coding Snack.

The awards are going to be given for a few years, and winning disqualifies the recipient from winning again for 36 months. I guess the foundation will track entrants to prevent multiple wins by an applicant during the disqualification period. 

[ayryq]

One more thing about random.org:

They have a paid winner-picking service: "The Draw Service offers a high level of transparency to your entrants and also protects you against accusations of tampering by storing a record of the results on RANDOM.ORG."

For fewer than 500 entries (and as many winners as you like) the cost is $4.95. Looks like you can upload a list of entrants (rather than just numbers) as well.

https://www.random.org/faq/#S5

[mouser]

nice -- so that takes care of my point #2 above.

[kunkel321]

I think this would be fairly easy to setup in a spreadsheet like Excel.

EDIT:  Okay, here's a starting point, from Google Sheets
https://docs.google....7Q0/edit?usp=sharing

It's a "dumb" random, with no seed.  And unfortunately it is possible to get duplicates.  Just keep refreshing if that happens.
Paste your list of entrants in one Tab, then enter the number of awards to be given in the other Tab, and copy the list. 

There are no protections, so don't delete/cut the forumulas!   

[app103]

Aryq's suggestion of random.org is a good one if you are simply trying to choose m winners from a group of n entrants.

There are a few specialized cases where you may need to look deeper:

1. If you want to select winners not purely randomly, but with some weighted probabilities -- for example if you wanted the chance that an entrant to win to be based on how many years they have been in operation, etc. [doesn't sound like you need this].
2. If you needed to be able to *prove* to some third party rigorously that you only ran the random selection once and these are the independently verified results. [doesn't sound like you need this].
3. If you were going to be running multiple giveaways over time and wanted to make it less likely that people would win twice in a row, etc. [doesn't sound like you need this].

-mouser (June 09, 2015, 08:32 AM)

Mouser wrote an app that I know takes care of 1 and 3, which he used to select the winners in DC software giveaways of the past, but he has never made this app publicly available.

Perhaps with some urging, we can convince him to release it?