|
Forever_Peace posted:Levitan has an article up on overdrafted players and I found myself agreeing with nearly every single thing he says. I don't think miller or obj are being drafted that high honestly. I probably won't own many shares but I don't think their adp is outrageous.
|
#
?
Jul 22, 2015 17:14
|
|
|
|
| # ? May 29, 2024 10:15 |
|
|
Kevin White top 80?
|
|
#
?
Jul 22, 2015 17:38
|
|
|
Papes posted:I don't think miller or obj are being drafted that high honestly. I probably won't own many shares but I don't think their adp is outrageous. OBJ going ahead of Dez and Julio is bananas. He is going to regress and I can't fathom taking him ahead of the 1/2 turn.
|
|
#
?
Jul 22, 2015 17:50
|
|
|
Papes posted:I don't think miller or obj are being drafted that high honestly. I probably won't own many shares but I don't think their adp is outrageous. I've done about a dozen MFL10s, and OBJ has never fallen into the second round from what i've seen, 7 sounds about right (based on my anecdotal experience.) Miller is usually mid-late third round, but falls to the fourth sometimes. And Kevin White goes wayy too early, usually the second rookie WR off the board just before Agholor and Perriman, who are both usually taken in the 7th/8th. I'm a Chicago homer but won't touch him at current ADP due to the FoxFactor.
|
|
#
?
Jul 22, 2015 17:51
|
|
|
Papes posted:I don't think miller or obj are being drafted that high honestly. I probably won't own many shares but I don't think their adp is outrageous. obj's adp is 7. Ahead of known studs like Dez. That's pretty outrageous. The most interesting part of that article to me was that the Rotoworld guys are starting a podcast.
|
|
#
?
Jul 22, 2015 17:51
|
|
|
Azhais posted:obj's adp is 7. Ahead of known studs like Dez. That's pretty outrageous. I think it's a DFS-only podcast that Levitan and Evan Silva will be hosting every Thursday. Evan Silva is already on the Fantasy Feast podcast with Ross Tucker, so if you like his stuff, give it a listen.
|
|
#
?
Jul 22, 2015 17:53
|
|
|
Forever_Peace posted:OBJ going ahead of Dez and Julio is bananas. He is going to regress and I can't fathom taking him ahead of the 1/2 turn. Dez and Julio should probably both be gone by pick 6 imo. Regardless, obj could regress by 30% of his per game average last season and still put up wr1 numbers(unless wr numbers go up across the board) Also you have to at least consider the possibility that he doesn't regress. While extremely unlikely, he could actually be one of the best wrs to ever play and put up a historically great season. Last year Antonio Browns tds were supposed to be unsustainable and would surely regress, instead he had a career high. I wouldn't take him before dez, Julio, or brown....but a mid-late 1st seems fair to me. However I am extremely down on most of the first round rbs this year, so my draft board is very wr heavy at the top.
|
|
#
?
Jul 22, 2015 18:06
|
|
|
Papes posted:Dez and Julio should probably both be gone by pick 6 imo. The OBJ hype is warranted given his rookie season, but I think the bolded part is being underappreciated by anyone drafting him mid 1st. I'd take all of Charles, Lacy, Bell, AP, Brown, Dez, Julio, DT, Gronk, Lynch, Foster, Calvin, Green, Jordy, and probably Cobb, before OBJ. I'd put OBJ right with CJ Anderson and Jeffery, ahead of Forte. OBJ will probably be a low end WR1 after regressing a decent bit from a historic rookie season, and he should be drafted that way, after all the guaranteed WR1s and top end RBs and Gronk (and Rodgers and Luck, if it's 6 pt passing TD or high pts for passing yardage). You don't take the risky guy that put up an extremely outlier 12 game stretch, especially when injury to Cruz and Eli having his best season in many ways are obvious contributing factors. That's just asking to whiff on your 1st round pick, that should be a rock solid high end starter for you.
|
|
#
?
Jul 22, 2015 18:56
|
|
|
I don't worry about regression with OBJ so much as injury. He has a history of soft tissue problems and I worry about him missing time. I don't think his numbers were flukey at all if you watched him play. The guy knows how to find the hole in a zone and has the agility and speed to just blow past coverage. Even when teams tried shutting him down, he was finding ways to make monster plays. I'd be amazed if he regressed without injuries being a factor.
|
|
#
?
Jul 22, 2015 18:58
|
|
|
I'd probably take OBJ over Megatron but not over Dez/Julio/Green. Megatron isn't done but IMO he's out of the first.
|
|
#
?
Jul 22, 2015 19:03
|
|
|
LmaoTheKid posted:I'd probably take OBJ over Megatron but not over Dez/Julio/Green. Megatron is going to be on a lot of championship teams this year because of how he falls.
|
|
#
?
Jul 22, 2015 19:04
|
|
|
Sataere posted:I don't worry about regression with OBJ so much as injury. He has a history of soft tissue problems and I worry about him missing time. I don't think his numbers were flukey at all if you watched him play. The guy knows how to find the hole in a zone and has the agility and speed to just blow past coverage. Even when teams tried shutting him down, he was finding ways to make monster plays. I'd be amazed if he regressed without injuries being a factor. The thing is that putting up historically good numbers means you are almost guaranteed to regress, ya know? It's not about him being flukey or anything, it's about that production being almost unsustainable for anyone. Brown will very likely regress some too, but given his track record, I'd bet on him regressing less. E: I'm curious, but phone posting. Anyone bored and want to waste time at work?  How did the player with highest ppg overall or for their position do the following season? Did their ppg go up, down? What about positional ranking? It doesn't take into account the specifics of what OBJ did (rookie, historic ppg, Cruz injury, etc), but it'd at least give a baseline for how we expect season leading guys to do the next year (generally regress, generally don't, etc). Or use a ppg cutoff: how often do guys that score 25+, 22+, ppg regress or maintain it improve?
sourdough fucked around with this message at 19:25 on Jul 22, 2015 |
|
#
?
Jul 22, 2015 19:05
|
|
|
Sataere posted:Megatron is going to be on a lot of championship teams this year because of how he falls. I guess, he's still a great WR but unless I see 2nd round value I'm not touching him.
|
|
#
?
Jul 22, 2015 19:06
|
|
|
LmaoTheKid posted:I guess, he's still a great WR but unless I see 2nd round value I'm not touching him. That is my point. He is going in the second round a lot.
|
|
#
?
Jul 22, 2015 19:33
|
|
|
I think people are also gonna take Gronk too early and screw themselves this year. I can't justify taking a TE in the 1st no matter who it is.
|
|
#
?
Jul 22, 2015 19:36
|
|
|
I've created a second goon MFL10 best ball league. I'll update this when it starts. http://football20.myfantasyleague.com/2015/public?O=12&ACTION=LIST&PASSWORD=goonsquad2015 What do you guys think of Duke Johnson?
|
|
#
?
Jul 23, 2015 22:22
|
|
|
Also, I'd love your guys take on which players are significantly helped/hurt by the best ball format. I thought about trying to do some "big game" math but gave up.
|
|
#
?
Jul 23, 2015 22:34
|
|
|
Its Miller Time posted:What do you guys think of Duke Johnson? He's no David Cobb
|
|
#
?
Jul 23, 2015 22:36
|
|
|
Its Miller Time posted:Also, I'd love your guys take on which players are significantly helped/hurt by the best ball format. I thought about trying to do some "big game" math but gave up. You want to load up on high-variance guys. Look for deep threats (heyday DJax types) and endzone targets over otherwise equivalent slot guys. Avoid "bye-week guys" and plodders (TRich, donald brown) and get lottery tickets instead (Randle etc). Take an extra quarterback. Wait on TEs and pick up a few redzone guys. Duke Johnson is a talented back in a crowded backfield with other equally talented backs in a run-heavy hot-hand offense that can't pass for poo poo. Lots of upside, but he could very well be invisible all season. I'd rather have Crowell in that offense, and I'd rather have Gordon, Abdullah, Yeldon, and Coleman out of the rookie RBs.
|
|
#
?
Jul 23, 2015 23:00
|
|
|
Its Miller Time posted:Also, I'd love your guys take on which players are significantly helped/hurt by the best ball format. I thought about trying to do some "big game" math but gave up. "Best ball format helps this players value" is the new "good play in a gpp". While you do want players who can have big weeks, the idea is so loosely thrown around to justify suboptimal picks that it begins to lose meaning,
|
|
#
?
Jul 24, 2015 00:24
|
|
|
Damnit, I'm trying to think of a great way of determining draft order this year. Maybe some kind of video game simulation?
|
|
#
?
Jul 24, 2015 01:53
|
|
|
89 posted:Damnit, I'm trying to think of a great way of determining draft order this year. Maybe some kind of video game simulation? At your kid's birthday party have a potato sack race. Everyone gets assigned a kid and however they finish determines the draft order.
|
|
#
?
Jul 24, 2015 02:01
|
|
|
89 posted:Damnit, I'm trying to think of a great way of determining draft order this year. Maybe some kind of video game simulation? Drunken obstacle course.
|
|
#
?
Jul 24, 2015 02:10
|
|
|
89 posted:Damnit, I'm trying to think of a great way of determining draft order this year. Maybe some kind of video game simulation?
|
|
#
?
Jul 24, 2015 02:11
|
|
|
89 posted:Damnit, I'm trying to think of a great way of determining draft order this year. Maybe some kind of video game simulation? Play a game of horse
|
|
#
?
Jul 24, 2015 02:12
|
|
|
I'm thinking I might throw us in a battle royal on a WWE game. Where's the sweet spot for a 10 man draft as far as draft position goes this year? I get to pick since I'm league champ (except for 1st and 2nd, end of last season determines those)....I was thinking maybe 4th or 6th pick. 4th would give me a guaranteed top 5 RB and then a later 2nd and early 3rd. Maybe end up with say Lynch/Jordy/Julio? (wishful thinking). It's a 2 RB, 3 WR, 1 QB, 1 TE league with NO flex. But, pick 6 is reliant on there being somebody biting on a non RB. Maybe pick 5 to be safe. 89 fucked around with this message at 03:12 on Jul 24, 2015 |
|
#
?
Jul 24, 2015 03:07
|
|
|
89 posted:Damnit, I'm trying to think of a great way of determining draft order this year. Maybe some kind of video game simulation? If you use FAAB, let them submit email blind bids for each pick with the idea that the money permanently comes out of their budget if they win. Ties are determined by whoever the commissioner thinks has the best team name.
|
|
#
?
Jul 24, 2015 03:12
|
|
|
Silly Burrito posted:If you use FAAB, let them submit email blind bids for each pick with the idea that the money permanently comes out of their budget if they win. Ties are determined by whoever the commissioner thinks has the best team name. What about the commish bidding though? or does he graciously take last?
|
|
#
?
Jul 24, 2015 04:44
|
|
|
First place last season gets first choice, last place last season gets last choice, 2 - 7 are determined by an all-AI stock Smash Wii U game where each person chooses their player. Spots are chosen by place your AI player finishes.
|
|
#
?
Jul 24, 2015 04:49
|
|
|
1. use the random function to select an initial wikipedia page 2. using only hyperlinks, use the smallest amount of clicks to get to "X" where X is charlie sims / pornography / some nerd poo poo / whatever managers must retrace their steps and submit their solution to the commish. draft order is decided by lowest clicks / fastest solution --- 1. everyone submits the name of a football player to the commish 2. without knowing who anyone else has picked, everyone guesses how many rings or MVPs or arrests or some poo poo the entire pool of players has collectively won draft order is decided by accuracy of guess
|
|
#
?
Jul 24, 2015 06:14
|
|
|
Teletubbies endurance marathon (Netflix has 1 season of 13 episodes so you may need to repeat a few cycles), no food or drink permitted other than Xanax and alcohol, reverse draft order determined by sequence of who falls asleep or leaves. I recommend securing a luxury hotel suite for this so the contest can be waged on neutral ground.
|
|
#
?
Jul 24, 2015 06:29
|
|
|
spacejung posted:Teletubbies endurance marathon (Netflix has 1 season of 13 episodes so you may need to repeat a few cycles), no food or drink permitted other than Xanax and alcohol, reverse draft order determined by sequence of who falls asleep or leaves. I recommend securing a luxury hotel suite for this so the contest can be waged on neutral ground. What is wrong with you?
|
|
#
?
Jul 24, 2015 06:31
|
|
|
Metapod posted:What is wrong with you? This is why I keep 'pod in the squad
|
|
#
?
Jul 24, 2015 06:35
|
|
|
MacheteZombie posted:What about the commish bidding though? or does he graciously take last? He writes down his bids first before sending the email. He can take a picture with a timestamp if shenanigans are called.
|
|
#
?
Jul 24, 2015 12:59
|
|
|
spacejung posted:Teletubbies endurance marathon (Netflix has 1 season of 13 episodes so you may need to repeat a few cycles), no food or drink permitted other than Xanax and alcohol, reverse draft order determined by sequence of who falls asleep or leaves. I recommend securing a luxury hotel suite for this so the contest can be waged on neutral ground. I like this guy
|
|
#
?
Jul 24, 2015 15:26
|
|
|
Teletubbies would probably be a fun show to watch while high.
|
|
#
?
Jul 24, 2015 16:04
|
|
|
Yeah, make it mushrooms and Requiem for a Dream and then we'll see how tough everyone really is
|
|
#
?
Jul 24, 2015 19:59
|
|
|
Alright, weeks ago I told somebody I would make a more sophisticated zero-rb strategy to plug into my draft simulator and run against a few thousand 2014 drafts. My initial basic zero-rb approaches simply just didn't take rbs for the first two rounds. As you might expect, this was a bit of an arbitrary constraint and didn't fare so well (and did particularly bad from the end of the draft order in 10-team leagues, where they passed up on Demarco Murray and Lynch). I've finally gotten around to finishing up a more sophisticated bot. Included in the changes are two primary sources of modified player valuations: 1) An alternate "positional need" curve (weighting WRs on the front end of the draft and RBs at the back end of the draft), and 2) A function that weights running backs by their variance (with preference given to rbs with a wide range of rankings). Without getting too far into the weeds, we value late-round rbs by the estimated 75th percentile of rankings for that player instead of the mean rankings for that player (i.e. we draft primarily for upside), then center everything back to the normal non-inflated rankings scale so they can be compared to other positions. Like before, I ran 250 simulations for each of the even-numbered spots in a 10-man league (running this zero-rb strategy vs. a field of pure rank-based drafters), in addition to 250 baseline simulations where everybody was a rank-based drafter. Then, I calculated the point totals of the best-performing season-long starting teams for each of the 1500 drafts. And finally, I calculated the average difference between the basline drafter from each position and the zero-rb drafter from that same position. Here's that data: quote:Baseline So again, doesn't look too promising. HOWEVER, one thing really popped out from the zero-rb drafts: they were significantly more reliable. I expected a small effect from the lowered bust rate for WRs, but the decrease in standard deviation of "best drafted team" was really notable: quote:2: -12.88 Further, I'm noticing one additional qualitative benefit: rb2s that are SO BAD, it should be possible to get a better one from the waiver wire. This is a significantly better situation from having a mediocre team across the board - with a few studs at some positions and terrible god-awful players at others, it becomes a pretty straightforward matter to replace the terrible ones and keep the studs. For instance, here's some representative drafts from the 4 position: quote:Regular strategy: For the regular draft, after Peterson busted, the rb core of Ellington/Bradshaw/Sankey/Bush was going to be a liability every week, but are going to be productive enough that they may be hard to replace from the WW. In contrast, from the same position, the zero-rb approach went DT/Jordy/Peyton in the first three rounds, but had a load of hot trash at RB (Hill excepted). I'd say the chances are better of working the WW in a 10-team when one of the lottery tickets taken in the draft don't work out (remember, it emphasizes upside). So now the questions becomes whether the ability to replace trash RBs through the wire can overcome the disadvantage on the drafted team from the zero-rb strategy. Personally, I'm still not particularly convinced that a zero-rb approach reaches parity with some of the other better strategies I've examined, but it looks viable enough that I can't really find a good argument against it either. It might depend on how much risk you prefer (which was clearly the major difference in outcomes). Forever_Peace fucked around with this message at 20:29 on Jul 24, 2015 |
|
#
?
Jul 24, 2015 20:21
|
|
|
I just pick the best player available at whatever position I need unless a player available at another position is at a huge value. Seems to well for me.
|
|
#
?
Jul 24, 2015 20:32
|
|
|
|
| # ? May 29, 2024 10:15 |
|
|
jabro posted:I just pick the best player available at whatever position I need unless a player available at another position is at a huge value. Seems to well for me. Well that's the idea behind most of the strategies I've programmed and simulated, anyways! The main issue is how to define "best" (mean rank? ADP? upside? VORP?) and how to define "need" (e.g. all other things equal, how much do you need a 3rd RB? A Fourth? The 1st TE? The second QB?). The simulator sort of came from a desire to see how these different choices actually effect your draft (by looking at last year's outcomes or this year's projections). I also was curious how valuations changed according to the particulars of the league scoring or size.
|
|
#
?
Jul 24, 2015 20:53
|
|