WR's who break out in Year 1 are GOLD JERRY, GOLD!

General talk about Dynasty Leagues.
User avatar
JJRules
Pro Bowler
Pro Bowler
Posts: 1148
Joined: Thu Jun 01, 2017 6:06 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby JJRules » Wed Oct 28, 2020 10:05 am

PR0v3 wrote: Wed Oct 28, 2020 9:28 am Past results are not indicative of future performance.
No, but repeatable, long-term patterns are worth paying attention to.
10-team Superflex, 0.5 PPR, 15-man rosters, pseudo-dynasty
Keep any number of players (0-15), lose same number of draft picks
2010, 2015, 2020 - 🏆, 2013 - 🥈

QB, 2RB, 2WR, TE, SF, FLEX, D/ST

QB: Dak, Watson
RB: Swift, Robinson Jr, Singletary, Brown, Hubbard, Zeus, Herbert
WR: Wilson, Olave, Pickens, Aiyuk
TE: Pitts, LaPorta
D/ST:

skinfanjon
Role Player
Role Player
Posts: 351
Joined: Fri Jan 12, 2018 9:52 am

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby skinfanjon » Wed Oct 28, 2020 12:04 pm

PR0v3 wrote: Wed Oct 28, 2020 9:28 am
skinfanjon wrote: Wed Oct 28, 2020 6:42 am
jjleurquin wrote: Tue Oct 27, 2020 10:44 pm

This is one resource of many. While this isn't the end all be all it should give you insights. Context is important as Davante Adams had 2 solid vets in front of him both seasons in Nelson/Cobb and James Jones/Cobb. The rookie that is in the most similar situation this season is Gabe Davis as he has not only 2 but 3 solid vets there in front of him. Claypool is probably in a similar spot too. Do of that what you will.

I just don't see how seeing this upsets people and makes them point out every player who didn't perform well right away. These odds are not opinions and they don't really care about how anyone feels about them. This is a nice resource for everyone.
Agree with the sentiment here but its important to note these are not ODDS. They are recent results. Big difference.
Yep. This is my problem with the analytics crowd, they do the same thing with dominator rating + draft capital. Past results are not indicative of future performance. You can’t just take X% results and attribute that as the odds of the same thing occuring in the future. That’s not how statistics work.

The boxes these players are being put in are not odds of future performance, they are the historical results of different players given selected variables. The “odds” for current players looking forward are unique for each player/individual given their specific circumstances.
This exactly. I don't object to looking at things like this, they are useful, but they tend to tell us what we already know. If you have a great rookie season you tend to have a better career than a player who doesn't. If you breakout early in college, dominate your offense statistically, and get drafted early, you have a pretty good chance at success.

The pending correction for Claypool highlights one of the short comings though. So because he is averaging one less yard per game, his % of success drops by what, 25 or 30% (whatever the number, not scrolling back to find it)? Kinda silly. This is why statistical models are not *predictive*, which is what it means to say such and such has an X% chance of 3 1k seasons in their career. It's good to know what the results of past seasons have to say about it, but putting any kind of percentage on the likelihood of it happening in a single instance is something the model is not capable of.

Anyway, not trying to crap on the information. It's useful to know.

PR0v3
Captain
Captain
Posts: 781
Joined: Sat Oct 27, 2018 2:58 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby PR0v3 » Wed Oct 28, 2020 12:37 pm

skinfanjon wrote: Wed Oct 28, 2020 12:04 pm
PR0v3 wrote: Wed Oct 28, 2020 9:28 am
skinfanjon wrote: Wed Oct 28, 2020 6:42 am

Agree with the sentiment here but its important to note these are not ODDS. They are recent results. Big difference.
Yep. This is my problem with the analytics crowd, they do the same thing with dominator rating + draft capital. Past results are not indicative of future performance. You can’t just take X% results and attribute that as the odds of the same thing occuring in the future. That’s not how statistics work.

The boxes these players are being put in are not odds of future performance, they are the historical results of different players given selected variables. The “odds” for current players looking forward are unique for each player/individual given their specific circumstances.
This exactly. I don't object to looking at things like this, they are useful, but they tend to tell us what we already know. If you have a great rookie season you tend to have a better career than a player who doesn't. If you breakout early in college, dominate your offense statistically, and get drafted early, you have a pretty good chance at success.

The pending correction for Claypool highlights one of the short comings though. So because he is averaging one less yard per game, his % of success drops by what, 25 or 30% (whatever the number, not scrolling back to find it)? Kinda silly. This is why statistical models are not *predictive*, which is what it means to say such and such has an X% chance of 3 1k seasons in their career. It's good to know what the results of past seasons have to say about it, but putting any kind of percentage on the likelihood of it happening in a single instance is something the model is not capable of.

Anyway, not trying to crap on the information. It's useful to know.
Yep, these models or studies are not too much different than flipping a coin 10 times, landing heads 7 times, and saying the odds of flipping heads is 70%. That's not how it works. The mathematical proof of a coin flip being 50/50 is what is predictive, the past coin flip results are meaningless going forwards.

Obviously we don't have a way too proof out each individuals player's mathematical odds of success, so people have sort of fallen into this faux-statistical process. In reality, whatever numbers Corey Coleman, Calvin Johnson, Julio Jones, or Darrius Heyward-Bey posted are completely irrelevant to the outcomes of Ceedee Lamb, Aiyuk, or Claypool. The only thing that matters is what these individual guys can do on the field.

Obviously it's nice to know what success/failures looked like in the past, but there's not really anything predictive about it.
12 Team .5 PPR - 1 QB, 2 RB, 2 WR, 1 TE, 1 WR/RB 20 man rosters, 5 man taxi est. 2018
QB: Dak, Pickett
RB: CMC, Taylor, Gibson, Dillon, Akers, Penny
WR: JJeff, Hollywood, Olave, Toney, Aiyuk, Jeudy, C. Davis, Boyd, C. Samuel,
TE: Njoku, Gesicki
2023 picks: 1.x, 2.x, 3.x, 4.x

Sriracha
Ring of Fame
Ring of Fame
Posts: 3698
Joined: Tue Jun 11, 2019 2:38 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby Sriracha » Wed Oct 28, 2020 12:44 pm

PR0v3 wrote: Wed Oct 28, 2020 12:37 pm
skinfanjon wrote: Wed Oct 28, 2020 12:04 pm
PR0v3 wrote: Wed Oct 28, 2020 9:28 am

Yep. This is my problem with the analytics crowd, they do the same thing with dominator rating + draft capital. Past results are not indicative of future performance. You can’t just take X% results and attribute that as the odds of the same thing occuring in the future. That’s not how statistics work.

The boxes these players are being put in are not odds of future performance, they are the historical results of different players given selected variables. The “odds” for current players looking forward are unique for each player/individual given their specific circumstances.
This exactly. I don't object to looking at things like this, they are useful, but they tend to tell us what we already know. If you have a great rookie season you tend to have a better career than a player who doesn't. If you breakout early in college, dominate your offense statistically, and get drafted early, you have a pretty good chance at success.

The pending correction for Claypool highlights one of the short comings though. So because he is averaging one less yard per game, his % of success drops by what, 25 or 30% (whatever the number, not scrolling back to find it)? Kinda silly. This is why statistical models are not *predictive*, which is what it means to say such and such has an X% chance of 3 1k seasons in their career. It's good to know what the results of past seasons have to say about it, but putting any kind of percentage on the likelihood of it happening in a single instance is something the model is not capable of.

Anyway, not trying to crap on the information. It's useful to know.
Yep, these models or studies are not too much different than flipping a coin 10 times, landing heads 7 times, and saying the odds of flipping heads is 70%. That's not how it works. The mathematical proof of a coin flip being 50/50 is what is predictive, the past coin flip results are meaningless going forwards.

Obviously we don't have a way too proof out each individuals player's mathematical odds of success, so people have sort-of fallen into this faux-statistical process. In reality, whatever numbers Corey Coleman, Calvin Johnson, Julio Jones, or Darrius Heyward-Bey posted are completely irrelevant to the outcomes of Ceedee Lamb, Aiyuk, or Claypool. The only thing that matters is what these guys can do on the field.
If the sample size was 10 you'd have an apt comparison.

The chance that you flip at least 7 coins in 10 tries is 17.19%, the odds of flipping 126 (or more) heads in 180 tries (70%) is so small the number would break your calculator.

Sure, the chance that it is the result of random chance is there. But so is the chance that reality randomly poofs within the next second.

skinfanjon
Role Player
Role Player
Posts: 351
Joined: Fri Jan 12, 2018 9:52 am

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby skinfanjon » Wed Oct 28, 2020 12:53 pm

Sriracha wrote: Wed Oct 28, 2020 12:44 pm
PR0v3 wrote: Wed Oct 28, 2020 12:37 pm
skinfanjon wrote: Wed Oct 28, 2020 12:04 pm

This exactly. I don't object to looking at things like this, they are useful, but they tend to tell us what we already know. If you have a great rookie season you tend to have a better career than a player who doesn't. If you breakout early in college, dominate your offense statistically, and get drafted early, you have a pretty good chance at success.

The pending correction for Claypool highlights one of the short comings though. So because he is averaging one less yard per game, his % of success drops by what, 25 or 30% (whatever the number, not scrolling back to find it)? Kinda silly. This is why statistical models are not *predictive*, which is what it means to say such and such has an X% chance of 3 1k seasons in their career. It's good to know what the results of past seasons have to say about it, but putting any kind of percentage on the likelihood of it happening in a single instance is something the model is not capable of.

Anyway, not trying to crap on the information. It's useful to know.
Yep, these models or studies are not too much different than flipping a coin 10 times, landing heads 7 times, and saying the odds of flipping heads is 70%. That's not how it works. The mathematical proof of a coin flip being 50/50 is what is predictive, the past coin flip results are meaningless going forwards.

Obviously we don't have a way too proof out each individuals player's mathematical odds of success, so people have sort-of fallen into this faux-statistical process. In reality, whatever numbers Corey Coleman, Calvin Johnson, Julio Jones, or Darrius Heyward-Bey posted are completely irrelevant to the outcomes of Ceedee Lamb, Aiyuk, or Claypool. The only thing that matters is what these guys can do on the field.
If the sample size was 10 you'd have an apt comparison.

The chance that you flip at least 7 coins in 10 tries is 17.19%, the odds of flipping 126 (or more) heads in 180 tries (70%) is so small the number would break your calculator.

Sure, the chance that it is the result of random chance is there. But so is the chance that reality randomly poofs within the next second.
The difference is you're talking about a quarter and all quarters are the same. Players and their situations vary wildly.

Sriracha
Ring of Fame
Ring of Fame
Posts: 3698
Joined: Tue Jun 11, 2019 2:38 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby Sriracha » Wed Oct 28, 2020 1:04 pm

skinfanjon wrote: Wed Oct 28, 2020 12:53 pm
Sriracha wrote: Wed Oct 28, 2020 12:44 pm
PR0v3 wrote: Wed Oct 28, 2020 12:37 pm

Yep, these models or studies are not too much different than flipping a coin 10 times, landing heads 7 times, and saying the odds of flipping heads is 70%. That's not how it works. The mathematical proof of a coin flip being 50/50 is what is predictive, the past coin flip results are meaningless going forwards.

Obviously we don't have a way too proof out each individuals player's mathematical odds of success, so people have sort-of fallen into this faux-statistical process. In reality, whatever numbers Corey Coleman, Calvin Johnson, Julio Jones, or Darrius Heyward-Bey posted are completely irrelevant to the outcomes of Ceedee Lamb, Aiyuk, or Claypool. The only thing that matters is what these guys can do on the field.
If the sample size was 10 you'd have an apt comparison.

The chance that you flip at least 7 coins in 10 tries is 17.19%, the odds of flipping 126 (or more) heads in 180 tries (70%) is so small the number would break your calculator.

Sure, the chance that it is the result of random chance is there. But so is the chance that reality randomly poofs within the next second.
The difference is you're talking about a quarter and all quarters are the same. Players and their situations vary wildly.
100%, of course the player will succeed with or without the analytical data; but we don't know everything about these players or exactly what determines who fails or succeeds at the NFL level. We can have inclinations that make us believe they will succeed (size/speed/tape/dynamism etc); but for the vast majority of us who aren't privy to this info it's a form of hubris to ignore analytical trends when forecasting player's career paths.

jjleurquin
Captain
Captain
Posts: 962
Joined: Tue May 19, 2020 4:10 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby jjleurquin » Wed Oct 28, 2020 5:46 pm

PR0v3 wrote: Wed Oct 28, 2020 9:28 am
skinfanjon wrote: Wed Oct 28, 2020 6:42 am
jjleurquin wrote: Tue Oct 27, 2020 10:44 pm

This is one resource of many. While this isn't the end all be all it should give you insights. Context is important as Davante Adams had 2 solid vets in front of him both seasons in Nelson/Cobb and James Jones/Cobb. The rookie that is in the most similar situation this season is Gabe Davis as he has not only 2 but 3 solid vets there in front of him. Claypool is probably in a similar spot too. Do of that what you will.

I just don't see how seeing this upsets people and makes them point out every player who didn't perform well right away. These odds are not opinions and they don't really care about how anyone feels about them. This is a nice resource for everyone.
Agree with the sentiment here but its important to note these are not ODDS. They are recent results. Big difference.
Yep. This is my problem with the analytics crowd, they do the same thing with dominator rating + draft capital. Past results are not indicative of future performance. You can’t just take X% results and attribute that as the odds of the same thing occuring in the future. That’s not how statistics work.

The boxes these players are being put in are not odds of future performance, they are the historical results of different players given selected variables. The “odds” for current players looking forward are unique for each player/individual given their specific circumstances.
This is the same perspective as thinking you're smarter than the casino. Maybe you can beat them by a little if you're very savvy. At the end of the day the casino will always profit in the long run. Casinos base their odds off of historical data so by your argument you're saying that the oddsmakers don't know what they're doing. That's fine if you want but I wouldn't bet against the oddsmakers.

Yarnith
All Pro
All Pro
Posts: 1869
Joined: Sun May 10, 2020 10:31 am

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby Yarnith » Wed Oct 28, 2020 5:59 pm

I always find it useful information. Usually it really weighs when I am in a position of something like "do I drop Jefferson or Jeudy?". If I am in a position that I need to make that choice I would give this some weight while debating who stays.
Cavaliers 12 team standard, 1QB, 1RB, 1WR, 2WR/TE, 1flex
QB: J. Allen, S. Howell, Z.Wilson
RB: A. Kamara, J. Taylor, CEH, K. Gainwell, P. Strong
WR: A.J. Brown, A. Cooper, J. Dotson, C. Ridley, E. Moore, W. Robinson, J. Meyers, T. Marshall,
TE: G. Kittle, J. Johnson, J. Woods, H. Henry
'23 1.10, 2.10, 3.10 '24 1, 2, 3
Recent Championships '19,'21,'22

Fosters Home for Imaginary Friends- 1QB, 2RB, 3WR, 1TE, 2Flex
QB: Mahomes, Purdy, Lance
RB: Pierce, Miller, K. Herbert, E. Hull
WR: Chase, Watson, Davis, JSN, Metchie, Pierce, V. Jefferson, Claypool, Thornton, Westbrook-Ikhine
TE: Pitts, Woods, Latu

User avatar
Dynasty DeLorean
Degenerate
Degenerate
Posts: 8855
Joined: Tue Feb 24, 2015 8:45 am

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby Dynasty DeLorean » Wed Oct 28, 2020 6:16 pm

I think some of you are overthinking this.

User avatar
Dynasty DeLorean
Degenerate
Degenerate
Posts: 8855
Joined: Tue Feb 24, 2015 8:45 am

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby Dynasty DeLorean » Wed Oct 28, 2020 6:27 pm

.
Last edited by Dynasty DeLorean on Wed Oct 28, 2020 9:25 pm, edited 1 time in total.

PR0v3
Captain
Captain
Posts: 781
Joined: Sat Oct 27, 2018 2:58 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby PR0v3 » Wed Oct 28, 2020 6:35 pm

jjleurquin wrote: Wed Oct 28, 2020 5:46 pm
PR0v3 wrote: Wed Oct 28, 2020 9:28 am
skinfanjon wrote: Wed Oct 28, 2020 6:42 am

Agree with the sentiment here but its important to note these are not ODDS. They are recent results. Big difference.
Yep. This is my problem with the analytics crowd, they do the same thing with dominator rating + draft capital. Past results are not indicative of future performance. You can’t just take X% results and attribute that as the odds of the same thing occuring in the future. That’s not how statistics work.

The boxes these players are being put in are not odds of future performance, they are the historical results of different players given selected variables. The “odds” for current players looking forward are unique for each player/individual given their specific circumstances.
This is the same perspective as thinking you're smarter than the casino. Maybe you can beat them by a little if you're very savvy. At the end of the day the casino will always profit in the long run. Casinos base their odds off of historical data so by your argument you're saying that the oddsmakers don't know what they're doing. That's fine if you want but I wouldn't bet against the oddsmakers.
No, it’s not the same perspective. Casino games are built on odds. The slot machines have built in odds to the games. For example, for every $1.00 put into the slot machine, it is programmed to give back $.98. This is written in code, it is predictable, it is the odds. Over time, this is what will happen. There may be some jackpots hit every now and then that cost the casino money on that pull, but with a large enough sample size they will make $.02 on every play. The odds are in the casino’s favor. Over time you cannot win.

What people are doing with fantasy football “analytics” and “models” is different. We do not know the odds of the game, we only know the results of the past slot machine pulls. If I play the slot machine, feed it $100 and hit a million dollar jackpot, I made a losing play. My EV on a $100 pull is $98. What the analytic crowd is doing is taking the $1M jackpot, comparing it to the $100 pull, and saying that when I go to the casino next door and play a completely different slot machine that my odds of winning $1M is 100%, since, you know, it just hit that way last time I pulled a slot. That is not how it works.
12 Team .5 PPR - 1 QB, 2 RB, 2 WR, 1 TE, 1 WR/RB 20 man rosters, 5 man taxi est. 2018
QB: Dak, Pickett
RB: CMC, Taylor, Gibson, Dillon, Akers, Penny
WR: JJeff, Hollywood, Olave, Toney, Aiyuk, Jeudy, C. Davis, Boyd, C. Samuel,
TE: Njoku, Gesicki
2023 picks: 1.x, 2.x, 3.x, 4.x

Ruggenater
All Pro
All Pro
Posts: 1750
Joined: Tue Aug 23, 2016 7:36 am

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby Ruggenater » Wed Oct 28, 2020 6:46 pm

PR0v3 wrote: Wed Oct 28, 2020 6:35 pm
jjleurquin wrote: Wed Oct 28, 2020 5:46 pm
PR0v3 wrote: Wed Oct 28, 2020 9:28 am

Yep. This is my problem with the analytics crowd, they do the same thing with dominator rating + draft capital. Past results are not indicative of future performance. You can’t just take X% results and attribute that as the odds of the same thing occuring in the future. That’s not how statistics work.

The boxes these players are being put in are not odds of future performance, they are the historical results of different players given selected variables. The “odds” for current players looking forward are unique for each player/individual given their specific circumstances.
This is the same perspective as thinking you're smarter than the casino. Maybe you can beat them by a little if you're very savvy. At the end of the day the casino will always profit in the long run. Casinos base their odds off of historical data so by your argument you're saying that the oddsmakers don't know what they're doing. That's fine if you want but I wouldn't bet against the oddsmakers.
No, it’s not the same perspective. Casino games are built on odds. The slot machines have built in odds to the games. For example, for every $1.00 put into the slot machine, it is programmed to give back $.98. This is written in code, it is predictable, it is the odds. Over time, this is what will happen. There may be some jackpots hit every now and then that cost the casino money on that pull, but with a large enough sample size they will make $.02 on every play. The odds are in the casino’s favor. Over time you cannot win.

What people are doing with fantasy football “analytics” and “models” is different. We do not know the odds of the game, we only know the results of the past slot machine pulls. If I play the slot machine, feed it $100 and hit a million dollar jackpot, I made a losing play. My EV on a $100 pull is $98. What the analytic crowd is doing is taking the $1M jackpot, comparing it to the $100 pull, and saying that when I go to the casino next door and play a completely different slot machine that my odds of winning $1M is 100%, since, you know, it just hit that way last time I played. That is not how it works.
Sounds like you’re firmly in the Bayesian camp. History may not be perfectly predictive of the future, but historical data is very regularly used as a best available proxy of what can be expected to happen going forward in predictive modeling.

I don’t think DD is trying to say that someone who falls into the top tier is a 100%, can’t miss, guaranteed long-term stud. Just that history shows players in that tier have had great success rates, and if you’re buying rookies right now, they look like good bets.
12 Team Superflex - PPR, 0.25 PPC - QB/2RB/3WR/TE/Flex/Superflex
QB: L Jackson, Tagovailoa, Rodgers, Pickett, Tannehill
RB: Swift, Pacheco, Sanders, Hubbard, Spears, Dillon, Herbert, McLaughlin, Chandler, Dowdle
WR: DeVonta, Waddle, Aiyuk, Nacua, McLaurin, Hopkins, M Williams, Mingo, Wan’Dale, Hyatt
TE: Kelce, Okonkwo, Schoonmaker

PR0v3
Captain
Captain
Posts: 781
Joined: Sat Oct 27, 2018 2:58 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby PR0v3 » Wed Oct 28, 2020 6:51 pm

Ruggenater wrote: Wed Oct 28, 2020 6:46 pm
PR0v3 wrote: Wed Oct 28, 2020 6:35 pm
jjleurquin wrote: Wed Oct 28, 2020 5:46 pm

This is the same perspective as thinking you're smarter than the casino. Maybe you can beat them by a little if you're very savvy. At the end of the day the casino will always profit in the long run. Casinos base their odds off of historical data so by your argument you're saying that the oddsmakers don't know what they're doing. That's fine if you want but I wouldn't bet against the oddsmakers.
No, it’s not the same perspective. Casino games are built on odds. The slot machines have built in odds to the games. For example, for every $1.00 put into the slot machine, it is programmed to give back $.98. This is written in code, it is predictable, it is the odds. Over time, this is what will happen. There may be some jackpots hit every now and then that cost the casino money on that pull, but with a large enough sample size they will make $.02 on every play. The odds are in the casino’s favor. Over time you cannot win.

What people are doing with fantasy football “analytics” and “models” is different. We do not know the odds of the game, we only know the results of the past slot machine pulls. If I play the slot machine, feed it $100 and hit a million dollar jackpot, I made a losing play. My EV on a $100 pull is $98. What the analytic crowd is doing is taking the $1M jackpot, comparing it to the $100 pull, and saying that when I go to the casino next door and play a completely different slot machine that my odds of winning $1M is 100%, since, you know, it just hit that way last time I played. That is not how it works.
Sounds like you’re firmly in the Bayesian camp. History may not be perfectly predictive of the future, but historical data is very regularly used as a best available proxy of what can be expected to happen going forward in predictive modeling.

I don’t think DD is trying to say that someone who falls into the top tier is a 100%, can’t miss, guaranteed long-term stud. Just that history shows players in that tier have had great success rates, and if you’re buying rookies right now, they look like good bets.
I definitely appreciate the work, and it’s nice to compare what guys are currently doing compared to what guys have done in the past. For me it will just always come down to what is happening/what happened with each specific individual player. I definitely enjoy the thread and reading people’s thoughts, but I think it goes a little overboard sometimes.
12 Team .5 PPR - 1 QB, 2 RB, 2 WR, 1 TE, 1 WR/RB 20 man rosters, 5 man taxi est. 2018
QB: Dak, Pickett
RB: CMC, Taylor, Gibson, Dillon, Akers, Penny
WR: JJeff, Hollywood, Olave, Toney, Aiyuk, Jeudy, C. Davis, Boyd, C. Samuel,
TE: Njoku, Gesicki
2023 picks: 1.x, 2.x, 3.x, 4.x

jjleurquin
Captain
Captain
Posts: 962
Joined: Tue May 19, 2020 4:10 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby jjleurquin » Wed Oct 28, 2020 9:31 pm

PR0v3 wrote: Wed Oct 28, 2020 6:35 pm No, it’s not the same perspective. Casino games are built on odds. The slot machines have built in odds to the games. For example, for every $1.00 put into the slot machine, it is programmed to give back $.98. This is written in code, it is predictable, it is the odds. Over time, this is what will happen. There may be some jackpots hit every now and then that cost the casino money on that pull, but with a large enough sample size they will make $.02 on every play. The odds are in the casino’s favor. Over time you cannot win.

What people are doing with fantasy football “analytics” and “models” is different. We do not know the odds of the game, we only know the results of the past slot machine pulls. If I play the slot machine, feed it $100 and hit a million dollar jackpot, I made a losing play. My EV on a $100 pull is $98. What the analytic crowd is doing is taking the $1M jackpot, comparing it to the $100 pull, and saying that when I go to the casino next door and play a completely different slot machine that my odds of winning $1M is 100%, since, you know, it just hit that way last time I pulled a slot. That is not how it works.
I'm not talking about the damn slot machines. I'm talking about the oddsmakers that decide what the lines are going to be every sunday. While their model is not perfect they will always come out ahead by using historical data to determine the odds and they'll move the line to even out the bets to get their 10% commission. Is it possible to beat them in the long run? Yes. but it's not likely.

PR0v3
Captain
Captain
Posts: 781
Joined: Sat Oct 27, 2018 2:58 pm

Re: WR's who break out in Year 1 are GOLD JERRY, GOLD!

Postby PR0v3 » Wed Oct 28, 2020 10:40 pm

jjleurquin wrote: Wed Oct 28, 2020 9:31 pm
PR0v3 wrote: Wed Oct 28, 2020 6:35 pm No, it’s not the same perspective. Casino games are built on odds. The slot machines have built in odds to the games. For example, for every $1.00 put into the slot machine, it is programmed to give back $.98. This is written in code, it is predictable, it is the odds. Over time, this is what will happen. There may be some jackpots hit every now and then that cost the casino money on that pull, but with a large enough sample size they will make $.02 on every play. The odds are in the casino’s favor. Over time you cannot win.

What people are doing with fantasy football “analytics” and “models” is different. We do not know the odds of the game, we only know the results of the past slot machine pulls. If I play the slot machine, feed it $100 and hit a million dollar jackpot, I made a losing play. My EV on a $100 pull is $98. What the analytic crowd is doing is taking the $1M jackpot, comparing it to the $100 pull, and saying that when I go to the casino next door and play a completely different slot machine that my odds of winning $1M is 100%, since, you know, it just hit that way last time I pulled a slot. That is not how it works.
I'm not talking about the damn slot machines. I'm talking about the oddsmakers that decide what the lines are going to be every sunday. While their model is not perfect they will always come out ahead by using historical data to determine the odds and they'll move the line to even out the bets to get their 10% commission. Is it possible to beat them in the long run? Yes. but it's not likely.
I’m not really sure what your point is then and how it relates to my original post?
12 Team .5 PPR - 1 QB, 2 RB, 2 WR, 1 TE, 1 WR/RB 20 man rosters, 5 man taxi est. 2018
QB: Dak, Pickett
RB: CMC, Taylor, Gibson, Dillon, Akers, Penny
WR: JJeff, Hollywood, Olave, Toney, Aiyuk, Jeudy, C. Davis, Boyd, C. Samuel,
TE: Njoku, Gesicki
2023 picks: 1.x, 2.x, 3.x, 4.x


Who is online

Users browsing this forum: Ahrefs [Bot], Bing [Bot], BlackOmega, moishetreats, Shcritters, trc and 29 guests