Crowd-Sourced Rookie Rankings

[junk_monkey]

Via the FiveThirtyEight website I came across a tool published and run by Princeton for free. It allows you to enter a long (or short) list of things and have people pick between two randomly selected items from that list at a time. As more results come in, it gradually gives each a score out of 100 and ranks them in order of preference. You can use it for a whole bunch of decision making purposes, but the thing that sprung to my mind was crowdsourcing FF rankings.

To get the ball rolling, I have set one up for non-PPR rookies for the 2018 draft. I grabbed the list from the DLF list of rookies - there are 63 I believe. You can start adding your results here: http://www.allourideas.org/2018DLFRookieRankingsNonPPR

I did a dry run of this a while back with people from my Dynasty league just with QBs. I think I had about 40 names in there and it took about 1500-2000 results before the list felt definitive. With 63 names I'm going to guess at needing 3-4000 results. It only takes a few seconds to do each one - you're just selecting one of two buttons on the screen.

If it goes down well, I (or someone else) can set one up for PPR leagues, for Dynasty overall, position specific... whatever you want.

Let me know what you think.

[dynastyninja]

Tried to contribute where I could. Lot's of "I don't know" in there for me

[clarion contrarion]

is it a bottomless pit ? I stopped at like 308 pairings - fun exercise but too many silly comparisons like jordan villamon or derrius guice really
is that necessary ?

[Pac_Eddy]

I think DLF had or has something similar. Like Clarion said, it gets a bit dull when there are so many slam dunk easy pairings.

I'll give it a try after the NFL draft. I just don't know enough right now.

[junk_monkey]

Tried to contribute where I could. Lot's of "I don't know" in there for me

Yes, I'm in the same boat to be honest. Maybe this wasn't the ideal group to test it on.

I'll give it a try after the NFL draft. I just don't know enough right now.

Fair point and perhaps an appropriate way forward, to do it when we have much more knowledge.

is it a bottomless pit ? I stopped at like 308 pairings - fun exercise but too many silly comparisons like jordan villamon or derrius guice really
is that necessary ?

Yes, inevitably you'll have a bunch of lopsided comparisons - when I did the QB one I got a lot of Rodgers vs McCown type ones, but of course the system doesn't differentiate between them - it doesn't have any knowledge of the people (or whatever) you're ranking - so needs to ask those questions. And obviously the more people are listed for ranking the more times you have to answer things in order to get a decent order.

And yes, you can just keep answering for as long as you want. The more data entered the better the quality of the list will be. There's no upper limit where it stops you entering more information, it just comes that the new info will have less impact on moving players up and down.

Perhaps it would be better to do a full dynasty start-up ranking list (top 50/100/200 players) on the basis that people would know a lot more about the players so everyone could contribute adequately. Wouldn't stop the lopsided nature of some of the questions, of course.

[junk_monkey]

All that said, we've had over 2000 rankings entered so far. I might give it another day or so and then compare our list with the DLF list and see if there are any big discrepancies...

[IBall2]

I added in about 50-60 votes. All I could do for now. Maybe I'll hop back on when I have more time

[dynastyninja]

It's already a lot better than it was yesterday

[Valhalla]

Interesting concept but probably too early in the process. Many haven’t studied enough yet to know much past our current top 20 or so players, and even that will change.

[hockeyBjj]

I like this idea. You should re-make it after the nfl draft when we know more and I think it'd be a good comparision for the "masses" opinion versus a select few experts

Currently it's this, so not horrible
Saquon Barkley 99
Derrius Guice 95
Sony Michel (did people miss no-ppr?) 91
Nick Chubb 90
Ronald Jones 89
Calvin Ridley 87
Courtland Sutton 83 (Interesting Ridley and Sutton switched at the moment. Not sure if that's due to sample size or not)
James Washington 81
DJ Moore 80
Rashaad Penny 78

After the NFL draft, you should post links for non-ppr, ppr, and 2QB and share them here. I bet these forums could end up with a pretty good list. Right now, I was hitting "Don't know" on any that didn't involve a top 10 guy. This may be nicer for the 2nd through 4th rounds to see, as its gets muddy there

[dynastyninja]

I like this idea. You should re-make it after the nfl draft when we know more and I think it'd be a good comparision for the "masses" opinion versus a select few experts

Currently it's this, so not horrible
Saquon Barkley 99
Derrius Guice 95
Sony Michel (did people miss no-ppr?) 91
Nick Chubb 90
Ronald Jones 89
Calvin Ridley 87
Courtland Sutton 83 (Interesting Ridley and Sutton switched at the moment. Not sure if that's due to sample size or not)
James Washington 81
DJ Moore 80
Rashaad Penny 78

After the NFL draft, you should post links for non-ppr, ppr, and 2QB and share them here. I bet these forums could end up with a pretty good list. Right now, I was hitting "Don't know" on any that didn't involve a top 10 guy. This may be nicer for the 2nd through 4th rounds to see, as its gets muddy there

It's hard after Guice. I (kinda) think Michel is (probably) the most "talented" at that point, so his spot makes sense. He, Chubb, and Jones are all close.

[Friction]

The thing with Michel is, where we he be ranked if he did not have a big game against OU? I did not hear anybody mention him near the top of the 1st before that, and I was trying to talk myself into him being a sneaky prospect at the end of the round. Not denying his talent or anything, and I know we will never know his ranking if his performance in the CFP did not happen. The same happened with Scarborough last year, except he did not enter the draft (and his following season was subpar). I know most will say they had his talent pegged before that, but I honestly did not hear or see it anywhere until after the game on a national stage. I actually don't like my picks that fell between 3-5, as I think a big tier of rookies after that are similar to me. At this stage at least. I am sure the combine and draft will vault a couple/few into the second tier.

[dynastyninja]

The thing with Michel is, where we he be ranked if he did not have a big game against OU? I did not hear anybody mention him near the top of the 1st before that, and I was trying to talk myself into him being a sneaky prospect at the end of the round. Not denying his talent or anything, and I know we will never know his ranking if his performance in the CFP did not happen. The same happened with Scarborough last year, except he did not enter the draft (and his following season was subpar). I know most will say they had his talent pegged before that, but I honestly did not hear or see it anywhere until after the game on a national stage. I actually don't like my picks that fell between 3-5, as I think a big tier of rookies after that are similar to me. At this stage at least. I am sure the combine and draft will vault a couple/few into the second tier.

I usually try not to follow CFB hype and just scout prospects after the season. Michel looks like he belongs in that Chubb/Jones tier.

[BigJoeWall72]

junk_monkey... This is a great find and an awesome idea.

I agree with one of the above posts, it would probably be beneficial to start 3 new sets of rankings once the combine and draft are over so people have a better handle on how they feel about players - but this is a great tool for now.

This is an excellent contribution to the forum - well done!

[junk_monkey]

junk_monkey... This is a great find and an awesome idea.

I agree with one of the above posts, it would probably be beneficial to start 3 new sets of rankings once the combine and draft are over so people have a better handle on how they feel about players - but this is a great tool for now.

This is an excellent contribution to the forum - well done!

Thanks very much!

I only found it late in the summer last year, which was a bit late for sharing with everyone really, so I've been sitting on it for 6 or 7 months.

I'm happy to set up three rankings setups after the draft, but anyone else can create an account and setup their own version, position specific, rule specific, non-FF related if you should so choose.

I also set one up last year to troll the league, asking them to vote on which team they thought was the strongest, which was good fun.