Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Locked thread
Amergin
Jan 29, 2013

THE SOUND A WET FART MAKES

Fight Club Sandwich posted:

work leagues are full of retards, go for safe picks instead of reaching

Gearman posted:

No need to get cute at pick #4. Go with Brown. If any of Charles-Bell-Peterson is there, take them instead.

That's what I was thinking, thanks for the input!

Adbot
ADBOT LOVES YOU

Matt Zerella
Oct 7, 2002

Norris'es are back baby. It's good again. Awoouu (fox Howl)
PPR, who goes after Brown? Julio, Dez, DT?

I'm sitting at 8 in PPR and our league loves QBs and RBs.

Kind of leaning Julio but Dez owns.

Gearman
Dec 6, 2011

I'd put Julio/Gronk/Thomas/Dez all right there. Whoever you feel better about. If It's keeper, I'd probably go Dez, redraft I'd flip a coin for either Julio or Thomas. If you want to basically put a WR1 in your TE slot, Gronk is right there too.

Matt Zerella
Oct 7, 2002

Norris'es are back baby. It's good again. Awoouu (fox Howl)

Gearman posted:

I'd put Julio/Gronk/Thomas/Dez all right there. Whoever you feel better about. If It's keeper, I'd probably go Dez, redraft I'd flip a coin for either Julio or Thomas. If you want to basically put a WR1 in your TE slot, Gronk is right there too.

Gronk's TD prowess is ridiculous so I may have to go with him even if we don't have a dedicated TE slot (WR/TE only flex).

We're a keeper league but only after the 6th round so that part doesn't matter.

Papes
Apr 13, 2010

There's always something at the bottom of the bag.
Julio at 4 is reasonable , I've already done it once this year and would again

The SituAsian
Oct 29, 2006

I'm a mess in distress
But we're still the best dressed
Wasn't someone going to make a dedicated DFS thread at one point?

Prophecy120
Feb 4, 2003

God Bless the Enclave! God Bless America!

Silly Burrito posted:

No, I'll add that to the list of ones to try, thanks!


Just listened to this on the way home, and I enjoyed the hell out of it. If the quality stays the same in the next few podcasts, this is going to be my #1 daily football podcast. Still have the other ones to try.

I particularly like their mock draft episodes. They pick a format and draft strategy and run with it pick by pick and go into each pick in depth as a group.

Michael Corleone
Mar 30, 2011

by VideoGames
Mr. Beer, I am trying to put my own values together, but also looking at your stuff. I play in a 12 team 6pt PTD league, with 2 RB 3WR 2RB/WR 1 TE/WR and 1 TE standard scoring. Um, what is the best sheet I should look at? I have looked at a lot of them and I am figuring an ideal lineup would be 1 QB 3-4 RB and 3-4 WR, so I am thinking the best bet is whatever has an even amount of RB and WR spots would be my best bet. Also a keeper starting now so will adjust my values 'with my gut'. Last thing, am I dumb for trying to go off auction value to determine worth when I am going into a draft? I think it gives a certain, different value of a guy and I am working on a spreadsheet to try to do what you are doing, but basically it is the same thing, I just want the best auction prices to take from you to put in my stuff. TIA, and ya, you are awesome!

e; maybe I can work a ratio and go from the 14 or 16 team projected price, and ya, I am snake drafting trying to work my mojo from auction prices. WTF is wrong with me?!

Michael Corleone fucked around with this message at 00:08 on Aug 22, 2015

Azhais
Feb 5, 2007
Switchblade Switcharoo

Spoeank posted:

Have you tried CBS Fantasy Football Podcast? They're pretty jovial.

Is that the one that is co-hosted by an actual stand up comedian that occasionally can't make the podcast because he has gigs?

Silly Burrito posted:

I'm probably in the minority here, but I wouldn't mind finding a new fantasy football podcast that is decently funny, even if it does go into morning zoo routines at the expense of total football news/coverage. I've got a 50 minute drive to and from work, and most of the "serious" podcasts have been lulling me to sleep lately.

I do listen to the Fantasy Focus podcast for this reason, but they're still feeling out the Yates/Berry team. Any similar suggestions?

The footballguys on the couch podcast is also usually pretty chatty

There's also Fantasy Taz

http://www.blogtalkradio.com/fantasyfootballpowerhour

I didn't like him because I found the schtick a little too, I dunno, loud, but maybe that'd work for you Burrito

Gobias Ind.
Apr 5, 2007

If your girlfriend says hey to me that's our girlfriend now idc

LmaoTheKid posted:

PPR, who goes after Brown? Julio, Dez, DT?

I'm sitting at 8 in PPR and our league loves QBs and RBs.

Kind of leaning Julio but Dez owns.

I'm in the exact same position - although it's 0.75 PPR because the commissioner is dumb. I'm expecting Dez to be gone, but even if he's there I'm kind of liking the idea of Julio at 8.

Silly Burrito
Nov 27, 2007

SET A COURSE FOR
THE FLAVOR QUADRANT

Azhais posted:

Is that the one that is co-hosted by an actual stand up comedian that occasionally can't make the podcast because he has gigs?


The footballguys on the couch podcast is also usually pretty chatty

There's also Fantasy Taz

http://www.blogtalkradio.com/fantasyfootballpowerhour

I didn't like him because I found the schtick a little too, I dunno, loud, but maybe that'd work for you Burrito

It's worth a shot, thanks! Unless I'm looking at a different link, that one and the CBS one haven't updated since last year, so I'll keep a watch for them.

89
Feb 24, 2006

#worldchamps
I know I'm probably annoying at this point...but I don't know if I should draft Gronk and JJ Watt as my first two picks at the 7 spot in a 10 man league or get a higher rated RB or elite WR cause I'm pretty sure JJ Watt is gonna go before my 3rd round pick!! :AND THE DRAFT IS TOMORROW AT NOON :cry: :cry: :cry:

Metapod
Mar 18, 2012
Do what your heart desires

89
Feb 24, 2006

#worldchamps

Metapod posted:

Do what your heart desires

Andrew Luck and Antonio Brown are pretty set to go in the first 5 picks, I'm pretty sure...which means I might have a top 4 RB fall into my lap..

Best case scenario? Top RB > Gronk > JJ Watt slips to me in the third. That's just BALLSY.

My heart also desires Top WR > Gronk > Great WR and roll out an elite aerial crew since this is a no flex league with 3 WR spots, 1 TE, and 2 RB amongst others.

This is why my heart never has a girlfriend.

89 fucked around with this message at 03:24 on Aug 22, 2015

Tiptoes
Apr 30, 2006

You are my underwater, underwater friends!

LmaoTheKid posted:

PPR, who goes after Brown? Julio, Dez, DT?

I'm sitting at 8 in PPR and our league loves QBs and RBs.

Kind of leaning Julio but Dez owns.
ODB :smuggo:

365 Nog Hogger
Jan 19, 2008

by Shine
Looking at Beersheets and other custom cheatsheets I'm wondering what purpose they really serve compared to a rigorous tier list, I might not completely understand how they function though... What am I missing?

Forever_Peace
May 7, 2007

Shoe do do do do do do do
Shoe do do do do do do yeah
Shoe do do do do do do do
Shoe do do do do do do yeah

365 Nog Hogger posted:

Looking at Beersheets and other custom cheatsheets I'm wondering what purpose they really serve compared to a rigorous tier list, I might not completely understand how they function though... What am I missing?

You can customize it to your league settings (particularly scoring, league size, and starting positions), while rankings cannot.

It also lets you do value-based comparisons across positions,

sourdough
Apr 30, 2012

365 Nog Hogger posted:

Looking at Beersheets and other custom cheatsheets I'm wondering what purpose they really serve compared to a rigorous tier list, I might not completely understand how they function though... What am I missing?

Beer customizes his sheets for league settings. How do Boris Chen's tiers suggest you draft if you are in a typical 1/2/3/1 starters league vs. 2QB league vs. a 1RB, 1WR, 2 WR/RB flex, 1 WR/TE flex league vs. 6 point passing TD and 0.25 points per completion league?

365 Nog Hogger
Jan 19, 2008

by Shine
Cool, thanks!

Metapod
Mar 18, 2012
Has a bizarro league beersheet been made?

Beer4TheBeerGod
Aug 23, 2004
Exciting Lemon

Michael Corleone posted:

Mr. Beer, I am trying to put my own values together, but also looking at your stuff. I play in a 12 team 6pt PTD league, with 2 RB 3WR 2RB/WR 1 TE/WR and 1 TE standard scoring. Um, what is the best sheet I should look at? I have looked at a lot of them and I am figuring an ideal lineup would be 1 QB 3-4 RB and 3-4 WR, so I am thinking the best bet is whatever has an even amount of RB and WR spots would be my best bet. Also a keeper starting now so will adjust my values 'with my gut'. Last thing, am I dumb for trying to go off auction value to determine worth when I am going into a draft? I think it gives a certain, different value of a guy and I am working on a spreadsheet to try to do what you are doing, but basically it is the same thing, I just want the best auction prices to take from you to put in my stuff. TIA, and ya, you are awesome!

e; maybe I can work a ratio and go from the 14 or 16 team projected price, and ya, I am snake drafting trying to work my mojo from auction prices. WTF is wrong with me?!

Remind me tomorrow and I will make you a sheet.

Spoeank
Jul 16, 2003

That's a nice set of 11 dynasty points there, it would be a shame if 3 rings were to happen with it

Azhais posted:

Is that the one that is co-hosted by an actual stand up comedian that occasionally can't make the podcast because he has gigs?

Nah, they are all full-time fantasy sports guys. They rotate out a bunch of guys: Adam Aizer is usually the host, and he's usually with Dave Richard and Jamey Eisenberg. Heath Cummings and Chris Towers also rotate in. It's a lot of fun and informative as well.


Edit: I just saw you couldn't find CBS Silly Burrito here you go: http://www.cbssports.com/podcast

Spoeank fucked around with this message at 05:20 on Aug 22, 2015

Beer4TheBeerGod
Aug 23, 2004
Exciting Lemon
CBS is what I listen to most often.

Azhais
Feb 5, 2007
Switchblade Switcharoo

Silly Burrito posted:

It's worth a shot, thanks! Unless I'm looking at a different link, that one and the CBS one haven't updated since last year, so I'll keep a watch for them.

Ah, Taz apparently moved to Sirius XM Fantasy

Papes
Apr 13, 2010

There's always something at the bottom of the bag.

89 posted:

I know I'm probably annoying at this point...but I don't know if I should draft Gronk and JJ Watt as my first two picks at the 7 spot in a 10 man league or get a higher rated RB or elite WR cause I'm pretty sure JJ Watt is gonna go before my 3rd round pick!! :AND THE DRAFT IS TOMORROW AT NOON :cry: :cry: :cry:


You've posted this like 10 times...just do it. It's ok to not start the draft in a traditional matter. It's obviously the strategy that you want to do so just do it.

Spoeank
Jul 16, 2003

That's a nice set of 11 dynasty points there, it would be a shame if 3 rings were to happen with it
There's no Konami Code to win your draft. Follow your heart.

Beer4TheBeerGod
Aug 23, 2004
Exciting Lemon
Yeah I would do that. 10 team leagues are great for depth so go nuts on starters.

Matt Zerella
Oct 7, 2002

Norris'es are back baby. It's good again. Awoouu (fox Howl)

Hahahah. He's kept so that worry is out of my head.

89
Feb 24, 2006

#worldchamps
The thing that's messing with my head on getting Gronk is that Olsen's value just shot up but his ADP hasn't with Benjamin done

Silly Burrito
Nov 27, 2007

SET A COURSE FOR
THE FLAVOR QUADRANT

Azhais posted:

Ah, Taz apparently moved to Sirius XM Fantasy

Ah well, I appreciate the suggestion. :)

Spoeank posted:

Nah, they are all full-time fantasy sports guys. They rotate out a bunch of guys: Adam Aizer is usually the host, and he's usually with Dave Richard and Jamey Eisenberg. Heath Cummings and Chris Towers also rotate in. It's a lot of fun and informative as well.


Edit: I just saw you couldn't find CBS Silly Burrito here you go: http://www.cbssports.com/podcast

Thanks. I don't know what link I originally found for CBS, but that one only had episodes from December. I'll check it out.

Matt Zerella
Oct 7, 2002

Norris'es are back baby. It's good again. Awoouu (fox Howl)
PFF premium. Any good? I kind of don't want to do 4for4 this year and I haven't don't FBG in 3 because I kind of hate it.

CompeAnansi
Feb 1, 2011

I respectfully decline
the invitation to join
your hallucination

LmaoTheKid posted:

PFF premium. Any good? I kind of don't want to do 4for4 this year and I haven't don't FBG in 3 because I kind of hate it.

Haven't used their premium, but they have two of the top rated experts in Jeff Ratcliffe and Mike Clay, so it's probably worth a shot.

Zombie Tsunami
Jun 22, 2006



:getin:

Azhais
Feb 5, 2007
Switchblade Switcharoo



:colbert:

Doltos
Dec 28, 2005

🤌🤌🤌

The other guy's teamname is better for the selection.

Metapod
Mar 18, 2012
Please rank: evans hilton cooks 1 ppr

Papes
Apr 13, 2010

There's always something at the bottom of the bag.

Metapod posted:

Please rank: evans hilton cooks 1 ppr

Evans > Hilton > cooks

Beer4TheBeerGod
Aug 23, 2004
Exciting Lemon

Forever_Peace posted:

Alrighty, here's the updated excel sheet. Columns I added are highlighted. You should be able to copy-paste any new data array into here and have it work, but you will need to sort the data (i.e. the unlighted bits) by "Average" (column S).

I currently have it set to include players within 2/3 of a standard deviation of the mean projection for that tier before moving on. This is the point at which a random projection moves from being more likely to have that player inside the tier to being more likely to have that player outside the tier, either above it or below it. You can change the size of the tiers as you see fit by changing the two instances of "2/3" in the formula in cell A3 (don't forget to change both).

Again, here's what I'm doing:
- Calculate the pooled variance of the current tier (this weights the variances by the sample sizes to give more emphasis to the players you have more projections for), and turn it into a pooled standard deviation.
- Calculate the mean of the current tier.
- Add the current player to the existing tier if it is within the cutoff from the mean (ie 2/3 SDs) and if the top player of the tier has not moved outside of the same distance from the mean from the lowered average.

It's a quick and dirty approximation but it should help fix the current emphasis on the exact values for the top player in each tier.

So I tried it out and it looks like it's going to work like a charm. I think I'll use it next year so that people don't freak out about tweaks I make in the middle of the season, but I just wanted to say thanks. This is really loving cool.

Any thoughts on using skew?

Forever_Peace
May 7, 2007

Shoe do do do do do do do
Shoe do do do do do do yeah
Shoe do do do do do do do
Shoe do do do do do do yeah

Beer4TheBeerGod posted:

So I tried it out and it looks like it's going to work like a charm. I think I'll use it next year so that people don't freak out about tweaks I make in the middle of the season, but I just wanted to say thanks. This is really loving cool.

Any thoughts on using skew?

Glad it works for ya!

I agree that it really feels like skew should be important to the tiers. For example, if a player projection distribution has a long right tail, it seems like we ought to be less confident about distinguishing him from the tier above (where most of the variance is), and more confident in distinguishing him from the tier below (where there's very little variance).

Neither the pooled variance technique or the Kernel Density Estimation take this into account. But I've been poking around some possibilities that could help here.

One thing I've been looking into is resampling the individual projections directly rather than working with the projection means. In these "bootstrapping" methods, the individual projections being sampled from keep their "natural" skewed shape, but because of the central limit theorem, the resampled means are normally distributed. I found a recent paper on bootstrapped discretization that I can access once I'm back at work that looks promising.

If that doesn't work, the other idea I had was to fit a kernel density estimation curve to discretize the projection means (like I did a few posts back), then fit additional curves to find the expected distribution of projections for each individual player, so that I can sample from those individual player curves and calculate the proportion of samples which end up inside each of the original tiers. What I might return for each player is the most likely tier, along with percentage of projections that would be expected to have him placed in the tiers above and below. Note: this method is very much bullshit that I just pulled out of my rear end and I have no idea how valid it is.

I'll keep you apprised as I try to teach myself this poo poo. It's good practice for me.

Forever_Peace fucked around with this message at 18:06 on Aug 22, 2015

Adbot
ADBOT LOVES YOU

Beer4TheBeerGod
Aug 23, 2004
Exciting Lemon

Forever_Peace posted:

Glad it works for ya!

I agree that it really feels like skew should be important to the tiers. For example, if a player projection distribution has a long right tail, it seems like we ought to be less confident about distinguishing him from the tier above (where most of the variance is), and more confident in distinguishing him from the tier below (where there's very little variance).

Neither the pooled variance technique or the Kernel Density Estimation take this into account. But I've been poking around some possibilities that could help here.

One thing I've been looking into is resampling the individual projections directly rather than working with the projection means. In these "bootstrapping" methods, the individual projections being sampled from keep their "natural" skewed shape, but because of the central limit theorem, the resampled means are normally distributed. I found a recent paper on bootstrapped discretization that I can access once I'm back at work that looks promising.

If that doesn't work, the other idea I had was to fit a kernel density estimation curve to discretize the projection means (like I did a few posts back), then fit additional curves to find the expected distribution of projections for each individual player, so that I can sample from those individual player curves and calculate the proportion of samples which end up inside each of the original tiers. What I might return for each player is the most likely tier, along with percentage of projections that would be expected to have him placed in the tiers above and below. Note: this method is very much bullshit that I just pulled out of my rear end and I have no idea how valid it is.

I'll keep you apprised as I try to teach myself this poo poo. It's good practice for me.

What about running a simple test? This might be something you're talking about above, but what happens if we check out the skew on a given projection and then if the skew is high enough to warrant further interest (I use a value of 1) then you run a calculation to see what the mean/variance would be if the lowest/highest projection was removed? If doing so causes a tier bump then you add a note to the individual player's line to indicate that they might be worth picking up ahead of the others in that tier.

  • Locked thread