Subscriber Contest (1 Viewer)

[Ignoratio Elenchi]

Correct, and they are at the exact same disadvantage as the Finley entries. Romo entries are no better off. The winning entry is not likely to be the one with devastating injuries along the way.

Define "devastating." If we use Dodds's top ten TEs for the remainder of the season, 311 of the remaining Finley owners have at least one of them on their roster. Again, you never want to lose a $21 player to a season-ending injury, but it's not necessarily "devastating" to many of these entries. Just seems totally random to boldly claim that a Finley owner won't win the contest, when there are lots of Finley entries left (which have managed to survive the past five weeks without Finley) and lots of other bigtime injuries that have happened since Finley went down.

 

[eaglezzz]

What if you had Finley and your back up TE's are Scheffler and Gronkowski? Are you doomed then?

 

[TheChairman]

What if you had Finley and your back up TE's are Scheffler and Gronkowski? Are you doomed then?

Yes. Doomed. DOOMED!Muahahhahahahah

 

[Touchdown There]

Correct, and they are at the exact same disadvantage as the Finley entries. Romo entries are no better off. The winning entry is not likely to be the one with devastating injuries along the way.

Define "devastating." If we use Dodds's top ten TEs for the remainder of the season, 311 of the remaining Finley owners have at least one of them on their roster. Again, you never want to lose a $21 player to a season-ending injury, but it's not necessarily "devastating" to many of these entries. Just seems totally random to boldly claim that a Finley owner won't win the contest, when there are lots of Finley entries left (which have managed to survive the past five weeks without Finley) and lots of other bigtime injuries that have happened since Finley went down.

Bold

 

[Ignoratio Elenchi]

If an 18-player team wins the whole thing, that won't change the statistics. It will have won despite having less of a chance.

Well, we don't know that for sure. We're pretty confident at this point that more players = better chance of surviving the first 13 weeks. The smaller rosters may in fact have an offsetting advantage in the playoffs (I don't think they do, and no one's yet shown that they do, but they might). Whether or not a small roster wins this year won't really answer that question. If someone did want to answer that question, one way they might try is to show that a roster densely packed with expensive players (so densely packed that basically the only way to afford them all is by using the smallest possible roster) is most likely to score a ton of points over a three-week span at the end of the season. It's not enough to just show that having Rodgers, Chris Johnson, and Andre Johnson gives you an advantage, for example, because there are 60 such rosters remaining, only 27 of which are 18-man rosters. Big rosters have studs, too; all those 18-man rosters who think they've got a step on the competition because they own the most expensive QB, RB, and WR will still find themselves competing against bigger rosters that have all those same core players. You would need to show that the best possible roster for the playoffs is one that you couldn't afford if you took more than 18 or 19 players. This might involve correlations between preseason projections and late-season fantasy points, etc.

I don't think anyone's really going to do that, but that's basically what the stud-related arguments come down to for me. You can't just say studs give you a better chance of winning, because I can build a bigger roster with studs. I need to see that a small roster has some kind of advantage just by virtue of being small. The higher-variance argument Doug brought up earlier is along these lines, although I'm not personally convinced it's enough. I think they're likely to have a lower average score than larger rosters, which mitigates their variance advantage somewhat. And to whatever extent variance does provide a real potential advantage in the final three weeks, I don't think it's enough of an advantage to make a small roster the right decision in August.

 

[Giantseasonticketholder]

I'm going to nickname my team "No Chance" since I only have 18 guys.

---------------------------------------------------------------------------------------------------------------------------------------------

Philip Rivers $19 27.40 32.70 32.95 24.05 34.95 17.45 22.70 26.55 37.05 bye

Carson Palmer $16 29.15 8.25 14.55 30.55 19.75 0.00 38.60 20.10 23.40

Derek Anderson $6 21.75 6.05 17.10 1.20 0.30 0.00 4.80 15.70 16.55

---------------------------------------------------------------------------------------------------------------------------------------------

Adrian Peterson $38 11.60 27.10 33.50 0.00 11.30 13.60 24.20 22.70 28.40

Ryan Mathews $29 8.20 6.50 0.00 13.30 6.40 9.10 3.20 12.90 7.00 bye

Ahmad Bradshaw $18 16.30 8.90 20.30 21.30 10.40 15.30 14.80 0.00 23.20

Arian Foster $13 42.30 15.30 14.60 32.20 3.70 24.70 0.00 27.20 33.70

---------------------------------------------------------------------------------------------------------------------------------------------

Marques Colston $25 11.20 11.70 6.50 7.60 16.70 10.30 27.20 19.60 14.50 bye

Steve Smith $24 9.30 7.50 19.30 7.00 20.90 13.00 25.10 0.00 14.60

Anquan Boldin $24 18.00 8.50 39.90 14.10 1.80 16.20 21.20 0.00 4.80

Greg Camarillo $3 3.90 1.30 3.60 0.00 0.00 9.50 0.00 0.00 10.60

Patrick Crayton $3 0.00 2.50 8.70 0.00 3.20 17.70 15.20 8.60 10.00 bye

---------------------------------------------------------------------------------------------------------------------------------------------

Visanthe Shiancoe $12 19.60 17.60 2.20 0.00 5.80 0.00 7.50 9.80 18.60

Aaron Hernandez $5 6.00 19.10 16.80 10.40 0.00 13.90 12.90 6.30 24.30

---------------------------------------------------------------------------------------------------------------------------------------------

Jay Feely $3 5.00 1.00 8.00 7.00 15.00 0.00 4.00 5.00 6.00

Graham Gano $2 9.00 11.00 10.00 5.00 13.00 7.00 7.00 10.00 0.00

---------------------------------------------------------------------------------------------------------------------------------------------

New York Giants $5 14.00 3.00 1.00 16.00 7.00 8.00 7.00 0.00 6.00

New Orleans Saints $4 3.00 10.00 4.00 5.00 6.00 3.00 1.00 7.00 14.00 bye

---------------------------------------------------------------------------------------------------------------------------------------------

TOTAL 180.45 160.50 200.05 159.45 131.65 146.85 193.20 144.35 206.35

CUTOFF 125.90 141.75 145.00 129.75 127.10 139.90 123.50 132.55 151.70

 

[Balco]

My whole strategy of picking 30 players was that 20 of them would pan out.out of those cheep type of players I have Tolbert $1, Bess $4, Branch $3, Murphy $3, T. Jones $7, Williams $8 and Gronkowski $3, that have all been good. Even the guys that haven't took off yet are still on the roster and could get a TD any week, or could move into a starting role. Scott $6, Washington $8, Choice $4, Stroughter $4, Camarillo $3, E. Sanders $2, Finneran $1, Scheffler $7All these guys to go with my studsRoethlisberger/Rivers ComboBradshawFosterCalvin/B Marsh Comboand guys likeGarcon and Spiller who still have a chance to break out big time. My roster is not very unique at all but out of 30 roster spots I have only lost 1 to injury Finley. So I have 29 guys that can get me points each week, that is huge advantage going down the stretch in my opinion.

I basically had the same line of reasoning, but me choosing 30 was basically because of the injury/bye week probability.

 

[Doug Drinen]

The more I think about it, the less confident I am that well-constructed large rosters have a better chance of winning (at the start of the season) than well-constructed small rosters.

Yes, the small rosters took a beating last year, and they continued to take a beating in week one. The more I think about it, though, the more I think this really doesn't mean much at all. Why? Because those results only reflect the results in the actual season that occurred, and not the gajillion other seasons that might have occurred instead.

It's a basic rule in huge contests like this that any strategy that gives you a better chance of finishing first also gives you a better chance of finishing last.

Consider variance, for instance. As has been discussed ad nauseum on this board, the QB/WR hookup increases your variance. This increases your chances of a great week, but it also increases your chances of a terrible week. That's great if you're an underdog and bad if you're a favorite. The key point is that, in a contest like this, everyone is an underdog. A HUUUUGE underdog. If you're an underdog and you pursue a high-variance strategy (as underdogs should), the most likely result is a loss. A bad loss. But that doesn't mean that what you did didn't give you a better chance of winning. It just means it didn't work out.

I guess what I'm saying is this: we do not have a sample size of 26,000 on which to base our hypotheses about big vs small rosters. We have a sample size of one. After this season, the sample size will be two. And that means nothing.

To summarize:

1. I have been a longtime advocate of the majority opinion here: that larger rosters are better. But I've now moved to an agnostic position.

2. I'm not saying this is the case, but I think it's quite possible that a small roster of studs has a better chance of winning it all than a large roster. The small roster also has a greater chance of flaming out early, but that's not relevant to the discussion.

3. I do not believe that the aggregate data on survival rates of different-sized rosters tells us anything about the best a priori strategy.

[DISCLAIMER: all this assumes that what we're interested in is maximizing our chances of finishing #1. I suspect many people, including me, get a lot of gratification from surviving as long as possible, even if they don't win. In that case, the discussion would be much different.]

Care to re-vist this Doug? We're on pace for under 30 eighteen man rosters making it to the final 250.

OK, now I'm ready to re-visit.My main argument was here:

Yes, the small rosters took a beating last year, and they continued to take a beating in week one. The more I think about it, though, the more I think this really doesn't mean much at all. Why? Because those results only reflect the results in the actual season that occurred, and not the gajillion other seasons that might have occurred instead.

The ownership rates of certain players is highly correlated to roster size. At the beginning of the contest, Arian Foster was owned by 81% of large (28+) rosters and only 54% of small (18--19) rosters. Andre Johnson was owned by 28% of small rosters and only 6% of large rosters. My thinking was, if Foster had turned out to be a dud, and Johnson had turned out to be great, things could have looked very different.So here's what I did. I re-ran the whole season, with the following changes:

1. I switched out Andre Johnson's 2010 stats with his 2009 stats. Same for Chris Johnson, Miles Austin, and Aaron Rodgers.

2. I replaced Foster's stats with C.J. Spiller's, and Mike Williams' with Jerricho Cotchery's.

Here are the actual survival rates (I threw out staff, so this may look slightly different from what's been posted):

18 0.09621419 0.14916020 0.17215621 0.19166722 0.23183023 0.29411824 0.29774125 0.33500026 0.36274527 0.40343328 0.38333329 0.38297930 0.417293Here are the survival rates in my pretend world:

18  0.13850419  0.16123920  0.15793421  0.18541722  0.21553923  0.22712424  0.24435325  0.23000026  0.26470627  0.29184528  0.27777829  0.32624130  0.338346

That flattens things out quite a bit, and still leaves the stud-laden-rosters-have-a-better-chance-in-the-final-three-weeks argument, and the junk-entries-are-disproportionally-small argument, but it's enough to move me out of the agnostic camp and back to my original position that larger rosters are very likely (not definitely, but very likely) the way to go.Finally, I decided to get ridiculous and give Foster, Williams, Ahmad Bradshaw, Thomas Jones, and Aaron Hernandez goose-eggs and give Johnson, Johnson, Austin, and Rodgers 3 points per game above what they did in 2009. Almost, but still not quite enough:

18  0.17100619  0.17962220  0.16018021  0.17916722  0.17669223  0.16830124  0.17659125  0.17750026  0.21568627  0.18884128  0.23888929  0.19858230  0.240602

I'm going to try to talk Dodds into compressing the prices even more next season...

 

[Ignoratio Elenchi]

Wow, nice work Doug.

 

[firstseason1988]

I'm going to try to talk Dodds into compressing the prices even more next season...

That would make things more difficult.

 

[firstseason1988]

Case closed.

See post 3513 above. CLOSED.

 

[H8fulHuck]

I'm going to nickname my team "No Chance" since I only have 18 guys.---------------------------------------------------------------------------------------------------------------------------------------------Philip Rivers $19 27.40 32.70 32.95 24.05 34.95 17.45 22.70 26.55 37.05 bye Carson Palmer $16 29.15 8.25 14.55 30.55 19.75 0.00 38.60 20.10 23.40 Derek Anderson $6 21.75 6.05 17.10 1.20 0.30 0.00 4.80 15.70 16.55 ---------------------------------------------------------------------------------------------------------------------------------------------Adrian Peterson $38 11.60 27.10 33.50 0.00 11.30 13.60 24.20 22.70 28.40 Ryan Mathews $29 8.20 6.50 0.00 13.30 6.40 9.10 3.20 12.90 7.00 bye Ahmad Bradshaw $18 16.30 8.90 20.30 21.30 10.40 15.30 14.80 0.00 23.20 Arian Foster $13 42.30 15.30 14.60 32.20 3.70 24.70 0.00 27.20 33.70 ---------------------------------------------------------------------------------------------------------------------------------------------Marques Colston $25 11.20 11.70 6.50 7.60 16.70 10.30 27.20 19.60 14.50 bye Steve Smith $24 9.30 7.50 19.30 7.00 20.90 13.00 25.10 0.00 14.60 Anquan Boldin $24 18.00 8.50 39.90 14.10 1.80 16.20 21.20 0.00 4.80 Greg Camarillo $3 3.90 1.30 3.60 0.00 0.00 9.50 0.00 0.00 10.60 Patrick Crayton $3 0.00 2.50 8.70 0.00 3.20 17.70 15.20 8.60 10.00 bye ---------------------------------------------------------------------------------------------------------------------------------------------Visanthe Shiancoe $12 19.60 17.60 2.20 0.00 5.80 0.00 7.50 9.80 18.60 Aaron Hernandez $5 6.00 19.10 16.80 10.40 0.00 13.90 12.90 6.30 24.30 ---------------------------------------------------------------------------------------------------------------------------------------------Jay Feely $3 5.00 1.00 8.00 7.00 15.00 0.00 4.00 5.00 6.00 Graham Gano $2 9.00 11.00 10.00 5.00 13.00 7.00 7.00 10.00 0.00 ---------------------------------------------------------------------------------------------------------------------------------------------New York Giants $5 14.00 3.00 1.00 16.00 7.00 8.00 7.00 0.00 6.00 New Orleans Saints $4 3.00 10.00 4.00 5.00 6.00 3.00 1.00 7.00 14.00 bye ---------------------------------------------------------------------------------------------------------------------------------------------TOTAL 180.45 160.50 200.05 159.45 131.65 146.85 193.20 144.35 206.35 CUTOFF 125.90 141.75 145.00 129.75 127.10 139.90 123.50 132.55 151.70

Hey, Hey Hey! My team's nickname is "NC, No Class" with the likes of Michael Vick and Ben Roethlisberger.

 

[H8fulHuck]

I'm going to try to talk Dodds into compressing the prices even more next season...

That would make things more difficult.

Keep shaking things up Doug. You wouldn't want us to run out of things to debate.

 

[QuizGuy66]

That's what I love about this thread - we are building up alternate universes and tearing 'em down and dissecting 'em and debating and and and

Best thread out there

-QG

 

[Fullback Fro]

I'm going to try to talk Dodds into compressing the prices even more next season...

Compressing the rosters even more just levels the playing filed between the people that put a ton a time and thought into their team and those that throw it together the last minute. Why do you want to do this?

 

[WisWolvrns]

I believe he means prices. I don't think the roster limits would change. But if players were closer in value, then who you pick becomes more and more a factor.

Right now the people who are surviving are the ones which lucked into picking the correct $3 cheapo players. ( or $7 ones in some WR cases ) The price is also the reason Foster was on so many rosters. He was the cheapest "starting" RB out there.

Obviously to be able to still have varying roster lengths, there would still need to be some tiered pricing, but I think the comment relates to more similar prices for players in each tier. So you can thus pick any of 10-15 players, who all cost the same in their tier. The people who pick the "right" players would then make the cuts and the ones with the "wrong" (or more likely injured..._) players would miss the cuts.

 

[Wide Right]

I'm going to try to talk Dodds into compressing the prices even more next season...

Compressing the rosters even more just levels the playing filed between the people that put a ton a time and thought into their team and those that throw it together the last minute. Why do you want to do this?

I agree - the whole point of the contest is to go through the process of - How much do I want to spend for a stud versus 2, 3 or more role players? For example - why would someone pick a mid-tier RB when for a dollar or two more they can get Chris Johnson or ADP? If someone want those of type players they should need to shell out the $10-15 more it takes to get them and take the salary cap hit. If the prices get compressed even more, you might as well just get rid of them completely and just see who is best at predicting who will be the biggest studs of the upcoming season. I think it would take a fun aspect out of the contest.

 

[Doug Drinen]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*

If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?

* - not really

 

[ctriopelle]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

A: Not as many as today, but a disproportionally large number compared to the general population, due to the likely stronger attempt to construct a decent team. Similar to the fact that Finley owners are still surviving at a better than expected rate given the loss of productivity for an expensive player.eta: I'll just take immunity in the form of getting a decent (honest) chance from the sim

 

[H8fulHuck]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

A: Not as many as today, but a disproportionally large number compared to the general population, due to the likely stronger attempt to construct a decent team. Similar to the fact that Finley owners are still surviving at a better than expected rate given the loss of productivity for an expensive player.eta: I'll just take immunity in the form of getting a decent (honest) chance from the sim

It's like the Freddy Taylor factor. He has done nothing to help teams win, but those selecting him have survived at a higher rate than average.

 

[H8fulHuck]

Check that...Make it the Washington report. Leon that is.

Survival rate for ALL teams: 17.73 percent

Survival rate for teams with Leon Washington: 23.93 percent

Definitely a good thing to have him on your roster right?

ETA: 62 of 236 teams have BOTH Leon Washington and Fred Taylor on their rosters for a 26.27% survival.

It's actually better when teams have both of these studs!!!

 

[Modog814]

Yes but this doesn't always have to be true.

You can say that about anything. We're talking about averages.Out of 13,000 entries the survival stats still show (again this year) that more players = good.

If an 18-player team wins the whole thing, that won't change the statistics. It will have won despite having less of a chance.

I was refering to the point at which diminishing returns kicks in. This year it could be 30, next it could be 27, the year after 26, then 28. Yes, you're looking at 13,000 entries, but it's over a single scenario path. Next year "could" produce different results. Drawing a conclusion based on that anaylsis based on a single year is faulty.

 

[Modog814]

If an 18-player team wins the whole thing, that won't change the statistics. It will have won despite having less of a chance.

Well, we don't know that for sure. We're pretty confident at this point that more players = better chance of surviving the first 13 weeks. The smaller rosters may in fact have an offsetting advantage in the playoffs (I don't think they do, and no one's yet shown that they do, but they might). Whether or not a small roster wins this year won't really answer that question. If someone did want to answer that question, one way they might try is to show that a roster densely packed with expensive players (so densely packed that basically the only way to afford them all is by using the smallest possible roster) is most likely to score a ton of points over a three-week span at the end of the season. It's not enough to just show that having Rodgers, Chris Johnson, and Andre Johnson gives you an advantage, for example, because there are 60 such rosters remaining, only 27 of which are 18-man rosters. Big rosters have studs, too; all those 18-man rosters who think they've got a step on the competition because they own the most expensive QB, RB, and WR will still find themselves competing against bigger rosters that have all those same core players. You would need to show that the best possible roster for the playoffs is one that you couldn't afford if you took more than 18 or 19 players. This might involve correlations between preseason projections and late-season fantasy points, etc.

I don't think anyone's really going to do that, but that's basically what the stud-related arguments come down to for me. You can't just say studs give you a better chance of winning, because I can build a bigger roster with studs. I need to see that a small roster has some kind of advantage just by virtue of being small. The higher-variance argument Doug brought up earlier is along these lines, although I'm not personally convinced it's enough. I think they're likely to have a lower average score than larger rosters, which mitigates their variance advantage somewhat. And to whatever extent variance does provide a real potential advantage in the final three weeks, I don't think it's enough of an advantage to make a small roster the right decision in August.

very

 

[Modog814]

2. I replaced Foster's stats with C.J. Spiller's, and Mike Williams' with Jerricho Cotchery's.

Very interesting and great anaylsis. But why use Foster? He was owned by a large percentage of 18-man teams as well. Why not use someone like Tomlinson who I would think would show the differences between a small roster and larger roster much clearer. You've selected 2 guys that were pretty highly hyped in the preseason, so I'd expected these guys to be on a disproprotionately high amount of smaller rosters compared to other guys in their price range. I'd think Nate Washington is probably a better choice than Mike Williams too. Just saying that an 18-man team that's looking for a $12-$13 RB is probably going to choose Foster over LT, so by removing Foster you may be biasing your results against the smaller teams. Same with Williams vs. Washington (I dont' have the owership stats in front of me so I could be wrong on this comparison).

 

[bonesman]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

Hmmmmm.... I would've survived every week besides last... I have to think it would be pretty high. <7% of what it is nowSo.... 67 or 66%?

 

[Ignoratio Elenchi]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

19.76%

 

[LittlePhatty]

I'm going to try to talk Dodds into compressing the prices even more next season...

And expand the max roster size to 40. Or 50.And set the minimum to 10.

 

[LittlePhatty]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?

Zero. Trick question. Brooks Foster (WR, Miami) was not available in this contest.

 

[H8fulHuck]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*

If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?

* - not really

I'll guess 100%. Those with Foster have a higher mortality rate due to sweating it out week after week in this contest.

 

[Touchdown There]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

Overall Contest Survival Rate: 17.73Current Foster Survival Rate: 26.68ACL Tear Guess: 8.16

 

[Marvelous]

I'm going to try to talk Dodds into compressing the prices even more next season...

Compressing the rosters even more just levels the playing filed between the people that put a ton a time and thought into their team and those that throw it together the last minute. Why do you want to do this?

This is a pretty strong statement from DD that:1. larger rosters with some cheaper players is the way to go2. By compressing the prices of the players, he would be leveling the playing field between studs and value playersI am strongly in the camp for 30, but compressing the prices could make me go lower, or even get a stud on my roster.

 

[Marvelous]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

Overall Contest Survival Rate: 17.73Current Foster Survival Rate: 26.68ACL Tear Guess: 8.16

I'll guess around 24%. I am assuming a lot of the non-Foster owners had poorly constructed teams.

 

[bostonfred]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

Overall Contest Survival Rate: 17.73Current Foster Survival Rate: 26.68ACL Tear Guess: 8.16

I'll guess around 24%. I am assuming a lot of the non-Foster owners had poorly constructed teams.

You're doing it wrong. If that's your logic, you should be guessing that they'd do worse than the overall survival rate.

 

[bostonfred]

Most teams that had Foster didn't take him as their #1 or #2 running back so they probably had other options. Most teams with Foster and Grant, or a similar combination like Foster and DeAngelo, would be long gone. But the rest of them should do just slightly worse than average. Foster was a cheap player, so they still have most of their cap space avalable to play for them, and were just as likely to have found sleepers as any other team. The average roster would be hitting at a 17% clip right now. The average non-Foster roster would be hitting at about 18%, while the average Foster roster would be hitting at about 13%.

 

[Ignoratio Elenchi]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?* - not really

Overall Contest Survival Rate: 17.73Current Foster Survival Rate: 26.68ACL Tear Guess: 8.16

I'll guess around 24%. I am assuming a lot of the non-Foster owners had poorly constructed teams.

You're doing it wrong. If that's your logic, you should be guessing that they'd do worse than the overall survival rate.

He's saying that the survival rate for Foster owners wouldn't drop much, because even if Foster was putting up zeroes every week, they'd still be stronger than rosters that didn't have Foster to begin with.

 

[Touchdown There]

Doug, can I be a pain and ask if you can run a sim for Finley owners survival in week 10? They have been surviving at a higher than average rate without any Finley points.

 

[Ignoratio Elenchi]

Most teams that had Foster didn't take him as their #1 or #2 running back so they probably had other options. Most teams with Foster and Grant, or a similar combination like Foster and DeAngelo, would be long gone. But the rest of them should do just slightly worse than average. Foster was a cheap player, so they still have most of their cap space avalable to play for them, and were just as likely to have found sleepers as any other team. The average roster would be hitting at a 17% clip right now. The average non-Foster roster would be hitting at about 18%, while the average Foster roster would be hitting at about 13%.

You have to take into consideration how many Foster owners there were to begin with. If only 13% of Foster owners were still alive, then 26% of non-Foster owners would have to be alive to get up to 2300 remaining entries.

 

[LittlePhatty]

Well, we don't know that for sure. We're pretty confident at this point that more players = better chance of surviving the first 13 weeks. The smaller rosters may in fact have an offsetting advantage in the playoffs (I don't think they do, and no one's yet shown that they do, but they might). Whether or not a small roster wins this year won't really answer that question.

Instead of figuring out Survival Percentage based on roster size, someone could separate all entries by roster size then calculate Average Score and see if the small rosters are outscoring the large ones (include scores from all teams, even eliminated ones). Or wait to do that in Weeks 11 - 16 when there are no byes.Last year the 23s and 24s outperformed the 20s and 21s in the final weeks.

The highest finish for a 20 was in spot #10:

1. 24

2. 23

3. 24

4. 24

5. 24

6. 24

7. 24

8. 24

9. 24

10. 20

22. 20

23. 20

27. 20

32. 20

33. 20

34. 20

37. 20

42. 20

47. 20

50. 20

 

[Ignoratio Elenchi]

Well, we don't know that for sure. We're pretty confident at this point that more players = better chance of surviving the first 13 weeks. The smaller rosters may in fact have an offsetting advantage in the playoffs (I don't think they do, and no one's yet shown that they do, but they might). Whether or not a small roster wins this year won't really answer that question.

Instead of figuring out Survival Percentage based on roster size, someone could separate all entries by roster size then calculate Average Score and see if the small rosters are outscoring the large ones. Or wait to do that in Weeks 11 - 16 when there are no byes.

I've done that previously in the thread. Larger rosters have a higher average score than smaller rosters. Two arguments for why that might not exactly be useful information are (a) small roster scores are hurt more by the bye weeks and (b) there were disproportionately more "junk" entries among the smaller rosters. I'll keep tracking it and we'll look at it in weeks 14-16 and see if that trend is reversed. That still won't definitively answer the question, though - some players are just owned more by smaller rosters, so if that player blows up at some point during the playoffs, average small-roster scores will increaase relative to average large-roster scores. That player could just as easily have not scored a lot of points, and we'd see the opposite effect. That's the gist of "the whole season is one trial, not 13,061 trials" idea. And it's a valid point.

That's why the only way to really definitively answer this question is to somehow show that small rosters have some kind of built-in advantage in the playoffs, regardless of how the playoffs actually play out this year. It's like if you flip a coin, and it comes up heads three times in a row - that doesn't mean it was more likely to come up heads, that's just what happened this time. Independently of these results, we can use other methods to actually calculate how likely it really is to come up heads on any given flip.

Same idea here - we can't just look at the playoff results to determine whether or not small rosters are more likely to win, because anything can happen this one season. Someone would need to show that a small roster is inherently better than a larger roster over a three-week span at the end of the year.

 

[TheChairman]

Imagine a 9 hand poker table. The hands are already dealt and all in preflop, one hand is face up with Aces showing. You can choose the hand with the Aces or the other 8 hands. Which one do you choose?

 

[LittlePhatty]

Instead of figuring out Survival Percentage based on roster size, someone could separate all entries by roster size then calculate Average Score

I've done that previously in the thread. Larger rosters have a higher average score than smaller rosters.....

Someone would need to show that a small roster is inherently better than a larger roster over a three-week span at the end of the year.

How about Average Score by roster size over the first 3 weeks of this year? Before byes and before (most) injuries.

Would that take you long to do, since you did the other average score calculations?

Still wouldn't be conclusive, but I'm just curious.

 

[TheChairman]

How about Average Score by roster size over the first 3 weeks of this year? Before byes and before (most) injuries.Would that take you long to do, since you did the other average score calculations?Still wouldn't be conclusive, but I'm just curious.

It's been done, 18 man rosters get hammered. Their excuse was 'junk entries'Weeks 4-10, same thing, excuse is 'byes affect them more'Weeks 11-16 will be the same thing, excuse will be 'not large enough sample size left'

 

[Ignoratio Elenchi]

How about Average Score by roster size over the first 3 weeks of this year? Before byes and before (most) injuries.

Would that take you long to do, since you did the other average score calculations?

Still wouldn't be conclusive, but I'm just curious.

Here are the average cumulative scores for weeks 1-3, but again, this doesn't really answer the question, for various reasons:
[*]There were arguably more "junk" entries among the small rosters, dragging their average down. That's hard to define but I think it might be true.

[*]This is just one possible outcome, out of many that could have occurred (i.e. perhaps players that were heavily owned by large rosters happened to be really good in the first three weeks this year, but that didn't have to be the case). That's definitely true.

[*]Most importantly, the argument isn't necessarily that small rosters will score more points on average. It's more that their occasional amazing weeks result in more points than the best weeks put up by larger rosters. That may be valid. I think if the playoffs were one week long, it would be more valid. Over three weeks I think I might rather have a higher-average, lower-variance team than a lower-average, higher-variance team.

Size	Entries	Avg Score18	5416	485.3819	1905	491.3220	1338	495.2421	 961	500.0422	 799	501.2223	 615	504.0224	 491	508.9225	 402	509.5626	 308	507.8627	 233	510.8128	 180	505.6829	 142	503.1930	 271	507.56

 

[Modog814]

Instead of figuring out Survival Percentage based on roster size, someone could separate all entries by roster size then calculate Average Score

I've done that previously in the thread. Larger rosters have a higher average score than smaller rosters.....

Someone would need to show that a small roster is inherently better than a larger roster over a three-week span at the end of the year.

How about Average Score by roster size over the first 3 weeks of this year? Before byes and before (most) injuries.

Would that take you long to do, since you did the other average score calculations?

Still wouldn't be conclusive, but I'm just curious.

Does average score really tell us anything (serious question, not trying to be a jerk)? I mean, if I have a 18-man roster, I don't really care what the other 18-man rosters have done. In fact, I think the argument isn't that 18-man rosters as a whole have an advantage over larger rosters. It's that an individual 18-man roster "could" have an advantage over any other single larger roster.

I have a theory on logically why smaller rosters could have an inherent advantage over larger rosters, but I'm still working it out in my head. But even if it's logically capable, doesn't mean it's reality either.

 

[Ignoratio Elenchi]

For argument's sake, rather than trying to define "junk" and weed them out, as we did earlier this year, let's just say that teams that are still alive at this point aren't junk. It would be hard to seriously call any entry junk if it survived the first 9 weeks of the season, right? So here are the average week 1-3 scores for just the 2316 teams that are still alive now:

Size	Entries	Avg Score18	522	524.5319	284	526.0920	231	525.4321	185	528.6322	186	527.2423	183	523.5124	145	531.1325	136	529.5126	112	524.0427	 94	532.9128	 69	531.7129	 55	523.6430	114	531.15

 

[Modog814]

How about Average Score by roster size over the first 3 weeks of this year? Before byes and before (most) injuries.

Would that take you long to do, since you did the other average score calculations?

Still wouldn't be conclusive, but I'm just curious.

Here are the average cumulative scores for weeks 1-3, but again, this doesn't really answer the question, for various reasons:
[*]There were arguably more "junk" entries among the small rosters, dragging their average down. That's hard to define but I think it might be true.

[*]This is just one possible outcome, out of many that could have occurred (i.e. perhaps players that were heavily owned by large rosters happened to be really good in the first three weeks this year, but that didn't have to be the case). That's definitely true.

[*]Most importantly, the argument isn't necessarily that small rosters will score more points on average. It's more that their occasional amazing weeks result in more points than the best weeks put up by larger rosters. That may be valid. I think if the playoffs were one week long, it would be more valid. Over three weeks I think I might rather have a higher-average, lower-variance team than a lower-average, higher-variance team.

Size Entries Avg Score18 5416 485.3819 1905 491.3220 1338 495.2421 961 500.0422 799 501.2223 615 504.0224 491 508.9225 402 509.5626 308 507.8627 233 510.8128 180 505.6829 142 503.1930 271 507.56

Another request for you Iggy...would you be able to produce variances for these as well? And how does the distribution look amongst the individual roster sizes? Is it safe to assume Normal?I think we've kind of assumed that smaller rosters would have a higher variance, but as I think about it, just the pure fact that having a larger number of possible scores each week and still only selecting 10 of them should lead to a higher variance, so I'm not certain that a 18-man roster would have a much higher variance.

 

[Modog814]

For argument's sake, rather than trying to define "junk" and weed them out, as we did earlier this year, let's just say that teams that are still alive at this point aren't junk. It would be hard to seriously call any entry junk if it survived the first 9 weeks of the season, right? So here are the average week 1-3 scores for just the 2316 teams that are still alive now:

Size	Entries	Avg Score18	522	524.5319	284	526.0920	231	525.4321	185	528.6322	186	527.2423	183	523.5124	145	531.1325	136	529.5126	112	524.0427	 94	532.9128	 69	531.7129	 55	523.6430	114	531.15

These are pretty close, especially between 21 and 30.

 

[Ignoratio Elenchi]

Size Entries Avg Score stDev Max Score18 5416 485.38 49.21 643.6519 1905 491.32 47.91 628.0020 1338 495.24 46.16 628.1521 961 500.04 44.68 632.3522 799 501.22 46.10 652.6523 615 504.02 43.75 647.0524 491 508.92 43.79 615.7025 402 509.56 43.40 626.5526 308 507.86 43.46 629.0027 233 510.81 42.27 655.7528 180 505.68 46.36 640.1029 142 503.19 47.60 629.5030 271 507.56 49.55 627.60
Interesting (or not): The highest scores in each of weeks 1, 2, and 3 were by 18-man, 19-man, and 18-man rosters respectively. However, the highest cumulative score for weeks 1-3 was a 27-man roster.

(Keep in mind these are counting all 13061 entries - even the eliminated ones and what they would have scored, since if you have a bad score in week 14, you're not out of the playoffs. For example, the 19-man roster that I'm reporting had the highest score in week 2 was actually eliminated in week 1, so if you look at the actual standings you won't see a 19-man team at the top that week.)

 

[JKL]

Trivia question. Free immunity in weeks 10 through 13 to whoever comes closest to the answer*If Foster had torn his ACL on the first play of the season, what percentage of Foster owners would still be alive?

I'm going to say that is still higher than the overall survival rate because he was still cheaper and not usually someone's first or second options, rosters with Foster were probably better constructed on average, were larger, and probably also had other cheap options who have contributed (Tomlinson, T. Jones, McFadden, Tolbert, Jackson).I'll go 18.63%

 

[H8fulHuck]

For argument's sake, rather than trying to define "junk" and weed them out, as we did earlier this year, let's just say that teams that are still alive at this point aren't junk. It would be hard to seriously call any entry junk if it survived the first 9 weeks of the season, right? So here are the average week 1-3 scores for just the 2316 teams that are still alive now:

Size	Entries	Avg Score18	522	524.5319	284	526.0920	231	525.4321	185	528.6322	186	527.2423	183	523.5124	145	531.1325	136	529.5126	112	524.0427	 94	532.9128	 69	531.7129	 55	523.6430	114	531.15

These are pretty close, especially between 21 and 30.

Very close indeed. Even the 18s out produce the 23, 26, and 29s with the group we have left. We have ourselves a race!Someone projected the 18 man rosters would drop to around 40 by the time we get to the final 250. Can this be projected for all roster sizes? I'm very curious about the % chance to make it to the end for each.