really like this idea
(11-20-2020, 11:15 AM)Sermokala Wrote: At first when reading it I was fully in support of it but for some reason I have no idea why about halfway through I started to dislike it. hmm i wonder.... (11-20-2020, 11:22 AM)slate Wrote: Instant runoff is great and would solve a lot of issues. 1. Yes, only top 3. Things get a lot more annoying for me programming if I have voters rank all choices. I also don't really have a problem with voters leaving some people off of their ballots so long as they vote on a top 3. 2. This is a two pronged thing I think. First, I'm not planning on putting the vote counts on the stream anymore because it would be really complicated - instead I'd lay it out like I did in the article when publishing the results. As for the lack of trust for awards - I think that's on me making bad mistakes. Awards Committee has gotten a lot less controversial with their picks recently and I think this proposal would help make sure the best player gets the award.
Did you consider changing the point values? Changing to something like 5/3/1 or 8/4/2 might be a simpler way of weighting first place votes more heavily. I don't know if it accomplishes your "make sure we find the best" goal, though.
Draft Steal (retired S35 CB) - Profile/Update | Wiki Troen Egghands (retired S22 DE) - Profile | Update | Wiki
11-21-2020, 02:01 PM
(This post was last modified: 11-21-2020, 02:01 PM by 37thchamber.)
IRV is a decent option. I still think the beatpath method would be better, though, partly because IRV also forces a strict order (edit: and assumes transitive property). When a voter thinks two candidates are equally deserving, they should be allowed to reflect that opinion in their ballot.
I also think limiting the ranking to 3 can have a detrimental effect and could (probably) be mitigated without dramatically increasing the complexity of the computation. I have an exhaustive ballot "simulator" lying around somewhere that I'm fairly sure can be tweaked to use IRV vote reassignment rather than simulating a fresh round of votes. Handles up to ten candidates iirc. Still, this is good stuff. If I can get something working to allow more than 3 selections to be ranked, would you consider removing the limit of three? I impersonate a programmer for a living
Father of the League Wiki • Friendly Neighbourhood Angry Black Guy™ • NOT British
Originator of the Sim League Cinematic Universe (SLCU)
Super capitalists are parasites. Fite me.
Alternatively, if you agree, you can support a grassroots movement dedicated to educating and organising the working class by buying a digital newspaper subscription. Your support would be greatly appreciated.
11-22-2020, 10:46 AM
(This post was last modified: 11-22-2020, 10:47 AM by Memento Mori.)
I like this idea overall but agree with 37th that setting a limit of 3 choices while using an IRV system isn't ideal. In extremely tight races it also opens up the possibility that in the final two rounds there can be ballots that aren't counted at all which may not go down with voters or the users they voted for.
A more complicated and less transparent voting system (particularly if the results aren't shown on stream as you suggested) also requires people to have more trust in the awards committee, which (of course mistakes happen and this isn't a criticism) is probably not in a great place right now with votes being displayed incorrectly on stream and I've heard from DSFL GMs that some of them received incorrect lists of nominees and their statlines. I still prefer it to the current system even with these concerns, though. [OPTION]S24 (PHI): 16 GP, 73 tackles, 1 TFL, 2 FF, 3 sacks, 5 INTs, 10 PDs, 2 TDs
[OPTION]S25 (PHI): 16 GP, 67 tackles, 4 INTs, 13 PDs, 1 TD [OPTION]S26 (OCO): 16 GP, 68 tackles, 1 TFL, 1 sack, 2 INTs, 10 PDs [OPTION]S27 (OCO): 16 GP, 116 tackles, 4 INTs, 23 PDs, 1 TD [OPTION]S28 (OCO): 16 GP, 84 tackles, 1 FF, 1 FR, 3 INTs, 20 PDs, 1 TD [OPTION]S29 (OCO): 16 GP, 99 tackles, 3 FF, 1 FR, 5 INTs, 23 PDs, 1 TD [OPTION]============================================================= [OPTION]ISFL Playoff Stats: [OPTION]S23 (PHI): 1 GP, 2 tackles [OPTION]S26 (OCO): 1 GP, 5 tackles, 2 PDs [OPTION]============================================================= [OPTION]Trophies and Achievements: [OPTION]Drafted 35th Overall by Myrtle Beach in the S21 DSFL Draft [OPTION]S21 Ultimini Champion [OPTION]S21 DSFL Pro Bowl Selection [OPTION]S21 DSFL Defensive Back of the Year Nominee [OPTION]Drafted 4th Overall by Philadelphia in the S22 ISFL Draft [OPTION]S23 ISFL Pro Bowl Selection [OPTION]S23 ISFL Cornerback of the Year Nominee [OPTION]S23 ISFL Defensive Performance of the Year Nominee [OPTION]S24 ISFL Pro Bowl Selection [OPTION]S24 ISFL Cornerback of the Year Nominee [OPTION]S26 ISFL Pro Bowl Selection
[OPTION]S26 ISFL Returner of the Year Nominee [OPTION]S29 ISFL Pro Bowl Selection [OPTION]S29 ISFL Cornerback of the Year Nominee [OPTION]============================================================= Player | Update | Wiki | Twitter (11-21-2020, 02:01 PM)37thchamber Wrote: IRV is a decent option. I still think the beatpath method would be better, though, partly because IRV also forces a strict order (edit: and assumes transitive property). When a voter thinks two candidates are equally deserving, they should be allowed to reflect that opinion in their ballot. If you can get me something that will make it easy for GMs to vote and easy for me/whoever is in charge to tabulate it, I'm all for it! The main reason I'm going with 3 is because it requires the least amount of change on the backend for me putting the Photoshop data sets together in python. |
|