Predicting 2017 rookie RB production (with results)

[dlf_jules]

Which rookie RBs have the highest likelihood of producing down the stretch in 2017? Submit your answers here and see what the community thinks.

I couldn't upload my attachment, but you can find the results here: https://twitter.com/brianmaloneff/statu ... 7012259842

[tstafford]

Tough poll. Gave my best guesses. When will you post the results?

[dlf_jules]

Tough poll. Gave my best guesses. When will you post the results?

Thanks, Tim. Results will probably be up tomorrow.

[dlf_jules]

One and only bump.

[nathanq42]

Gosh I feel like I was waaaay too optimistic hahaha!

[Valhalla]

Deleted as I don't want to adversely affect the poll

[kylejargon]

Nice poll. I am eager to see the results.

[dlf_jules]

I couldn't upload my attachment, but you can find the results here: https://twitter.com/brianmaloneff/statu ... 7012259842

[Valhalla]

Wow those are high % across the board IMO. If that's all true, there will be a lot of rookies in the top 12...
There are still a LOT of veteran RBs in the NFL, aren't there? Did I miss a book of obituaries?

[bdynomite]

Wow those are high % across the board IMO. If that's all true, there will be a lot of rookies in the top 12...
There are still a LOT of veteran RBs in the NFL, aren't there? Did I miss a book of obituaries?

Only Fournette and McCaffrey were for top 12 I believe, with about a 50% chance apiece. The rest were top 18 with only Mixon cracking a 40% chance. Personally I didn't meant to say that they would ALL crack the top 18 but I do expect one or two to sneak in.

[Valhalla]

I think people would get a better perspective on this if they did the same thing for all the vet RBs...and the % totals affected each other. I don't know how to set that up in an excel or google doc, but I imagine it can be done. Some type of format that calculates out the total probabilities that can be allocated out. So, if you allocate Carlos Hyde a 60% chance of top 12, it then affects the total % chance available to allocate out to the rest of the RB pool.
Basically, it's impossible to have 10 or so players at 80% chance, then 10 or so more players at 50%, then 15 or so more players at 20%, then a large number of handcuffs and rookies at 10%, then all the other random players that do have a small chance. We'd end up with 50 top 12 RBs. RBs that are thought little of succeed all the time, so even small % allocations have to be divied out to end of roster players. There are so many RBs in the NFL. There are many that are talented enough to get the job done with an opportunity. It doesn't add up to put all these rookies at 10%+ chances each.
I'd love it if someone could put together a google spreadsheet with all the NFL RBs with a top 12%, top 18% chances...and to cap the allocations somehow. If you, for instance, have 5 RBs that you are pretty certain have an 80% chance of being top 12, it leaves less room for % chances to the rest of the RB pool. I think it would force people to be more honest if they had to allow for other guys. If you, for instance, start your allocation percentages to all these rookies and the top 10 or so RBs, and the spreadsheet says the cap on all remaining RBs needs to range on average from 0-1%, then you look at guys like CJ Anderson, Tevin Coleman, Doug Martin, and Spencer Ware and realize "oh bleep...they are definitely more likely to make it than this rookie I gave 30% to," the program would force our hands into being more honest in the global picture, instead of looking at each back through a hype lens. Suddenly we'd be forced to, for example, lower that 30% odds rookie down to a more true 8% so we could raise Spencer Ware from 0.5% up to 14%. I don't know what equations to use for this or how to set it up, but I'm sure it can be done.

[dlf_jules]

It's weird, but because of the way this is set up, the percentages actually should add up to more than 12 top-12 RBs.

It's complicated, but it's because players who don't play six games aren't scored. Suppose there are 48 RBs with a non-negligible chance to finish in the top 12. How many of them will get hurt and thus not qualify? (Around 10, maybe.) If those 10 have an average of 25% chances to finish top 12 when healthy, our totals should yield 14.5 top-12 RBs.

[CK_]

It's weird, but because of the way this is set up, the percentages actually should add up to more than 12 top-12 RBs.

It's complicated, but it's because players who don't play six games aren't scored. Suppose there are 48 RBs with a non-negligible chance to finish in the top 12. How many of them will get hurt and thus not qualify? (Around 10, maybe.) If those 10 have an average of 25% chances to finish top 12 when healthy, our totals should yield 14.5 top-12 RBs.

Yeah that's how I understood it because it was weeks 8-16. So I believe I gave McNichols and Hunt a higher grade than most purely based off of opportunity and the likelihood of Martin being injured and chances of Ware fumbling again and being benched. I liked how some were top 18 and some were top 12. Honestly wish I gave McNichols 50% chance on his rookie season. Only gave 1 a 50% chance. Rest were 10-40%. Just see a RB having a better chance due to the high turnover rate and McNichols seems like he would have a great opportunity at being a top 18 back between weeks 8-16.

[dlf_jules]

@Valhalla, I didn't really answer your main point. I agree we'd probably get more accurate results if veterans were included.

That said, I don't think the cumulative results are crazy. If we sum the mean predictions (which are mislabeled as the median), we should expect a total of 2.15 top 18 performances from Mixon, Cook, Hunt, Kamara, Perine, Foreman, and McNichols, plus 1.02 top-12 performances from Fournette and McCaffrey. And that's assuming they all play enough games to qualify.