This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
43% chance. It was nice while it lasted. Pretty sure that estimate is a little high for my team... Romo on bye, Lefty benched, Stafford injured... I will need a lot of huge games from my lowball guys to stick around.Doug Drinen said:
Made it through last week with 172 and no QB.
I did miss that part. I'll go back and look for the post. If that's the case, then it would appear 24 man rosters do indeed have a much better chance. My whole point was that you can't say 24 > 20 just because more of them are advancing. This information is what's more important to look and correlates much better with giving a 24 man roster a significantly higher chance to win. I didn't know that information was available so thanks for the post.sinatravolta said:
Yeah, last week was much better for being without a QB. The cut ended up being at 110. This week it is looking like a cut around 140 is likely. Tough to do without a QB. I am sitting at 95%, not sure I feel that good though. 140 is a lot of points, and I have been below that twice already.
Actually, I think Turk's average cut estimate is 131. Even that "feels" high to me given the fact its a bye week and we are still cutting under 20%. I suppose it somewhat depends on the amount of high powered offenses going this week, and only 1 of the 4 teams on bye is a decent offensive unit. I'd suspect that a this means relatively few rostered players are actually on bye this week, increasing everyone's potential to score. Suddenly I don't feel so good about Witten being off for me (my only key bye this week), since my backup TE may not score at all due to lack of targets.
In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
Interesting. But the only difference between 22 and 24 players is $2, right? How can that make a difference? That $2 could have been spent on an extra cheap TE or RB that could ultimately take you deep into the contest.It seems like success in this contest is determined by two major factors: 1) the ability to pick talented players that will succeed, and 2) the luck of picking a roster that doesn't get decimated by injuries or benchings. Looking at guys like Nate Burleson, Mark Sanchez, Cedric Benson, and Brent Celek, the amount of money spent is no fair barometer of a player's value. Ultimately, whether you pick 20 men or 24, you still have to pick good players that represent value. But it seems like that with 24, the risk factor is lessened. I compare this to football betting. I've always fared far better with four $10 bets than I have with one $40 bet. The 24 man roster is akin to the four $10 bets. You are lessening the chances of being hurt by being dead wrong about a player.I've enjoyed watching/reading the 20 vs. 24 debate as well, but I started looking at the percentages a little bit differently. It seems that there is a general shift in agreement that at least the teams with a 24 man roster have a better shot to advance week-to-week. While the long-term success of the rosters is still (technically) debatable there seems as if there is another note to be taken from the data.We've discussed before that at some point there would have to be a breaking point... a point where you're spreading your money so thin in order to acquire as many players as possible that you begin to miss out on all of the steady performances that the stud players provide. I am going t suggest that the breaking point may be within the range we have already: 22.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
I think that has been the debate all along. Personally I think that number might be around 30.I've enjoyed watching/reading the 20 vs. 24 debate as well, but I started looking at the percentages a little bit differently. It seems that there is a general shift in agreement that at least the teams with a 24 man roster have a better shot to advance week-to-week. While the long-term success of the rosters is still (technically) debatable there seems as if there is another note to be taken from the data.We've discussed before that at some point there would have to be a breaking point... a point where you're spreading your money so thin in order to acquire as many players as possible that you begin to miss out on all of the steady performances that the stud players provide. I am going t suggest that the breaking point may be within the range we have already: 22.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
I would drop Driver and upgrade collins to manning or something. You will get so many more points there than Driver would ever contribute above that group of WRs
Yeah, that was the term that kept escaping me. Thanks for the catch.To jdoggydogg's point about what does $2 do, I agree. In a vacuum it does very little. It all depends on how you use it. Yes, you can invest in some one dollar fliers or you can upgrade positions by one notch, but it's all about who you get. Someone who upgrades their $1 WR flier to a $2 Darius Heyward-Bey isn't reaping the same rewards as someone who upgraded to Kenny Britt or Mario Manningham.In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
Yeah, that was the term that kept escaping me. Thanks for the catch.To jdoggydogg's point about what does $2 do, I agree. In a vacuum it does very little. It all depends on how you use it. Yes, you can invest in some one dollar fliers or you can upgrade positions by one notch, but it's all about who you get. Someone who upgrades their $1 WR flier to a $2 Darius Heyward-Bey isn't reaping the same rewards as someone who upgraded to Kenny Britt or Mario Manningham.In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
This is what is boils down to. To me, the 20 vs 24 debate won't be settled until the end, if it gets settled at all. It's on who you spend your money on, not the cost amount and not the total players in your roster. Personally, I picked players that I thought were going to do well this season and I tried to get as many of them on my team as I could. I ended up with 21 players. And yes I could have shifted a few people and came up with an extra $3 to fill out a full 24, but I didn't want to drop from Grant to Derrick Ward ($2 diff), or Schaub to Orton ($1) or even Roethlisberger ($1) b/c by the end of the year I think Schaub will have scored more points. I didn't want to drop from Boldin to Owens ($1 diff) or Colston to Royal ($2). Or Daniels to Keller ($1). Yes, those 3 players could have turned out to be Collie, Britt and Manningham, but they could have turned into Heyward-Bey, Sam Hurd and Craig Davis just as easily. So I think the ideal strategy is to take the players you think will perform the best and IF you have money left over, you start taking those $1 or $2 guys. Am i right? I don't know yet, but if I traded a few of those guys I probably wouldn't still be alive in this thing either.Yeah, that was the term that kept escaping me. Thanks for the catch.To jdoggydogg's point about what does $2 do, I agree. In a vacuum it does very little. It all depends on how you use it. Yes, you can invest in some one dollar fliers or you can upgrade positions by one notch, but it's all about who you get. Someone who upgrades their $1 WR flier to a $2 Darius Heyward-Bey isn't reaping the same rewards as someone who upgraded to Kenny Britt or Mario Manningham.In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
This is what is boils down to. To me, the 20 vs 24 debate won't be settled until the end, if it gets settled at all. It's on who you spend your money on, not the cost amount and not the total players in your roster. Personally, I picked players that I thought were going to do well this season and I tried to get as many of them on my team as I could. I ended up with 21 players. And yes I could have shifted a few people and came up with an extra $3 to fill out a full 24, but I didn't want to drop from Grant to Derrick Ward ($2 diff), or Schaub to Orton ($1) or even Roethlisberger ($1) b/c by the end of the year I think Schaub will have scored more points. I didn't want to drop from Boldin to Owens ($1 diff) or Colston to Royal ($2). Or Daniels to Keller ($1). Yes, those 3 players could have turned out to be Collie, Britt and Manningham, but they could have turned into Heyward-Bey, Sam Hurd and Craig Davis just as easily. So I think the ideal strategy is to take the players you think will perform the best and IF you have money left over, you start taking those $1 or $2 guys. Am i right? I don't know yet, but if I traded a few of those guys I probably wouldn't still be alive in this thing either.
I agree. I myself have a roster of 22. It's wasn't by design either, but basically I did the same thing you did. Fill out the core unit with players I felt were at the appropriate value. Personally, I grabbed Roddy White and Greg Jennings as I expected them to perform close to the upper level of receivers while saving me $7-$8. I tried to find the cheapest RBs that would get consistent work (Grant, TJones, Parker) and whatever money was left... that's what I used to try and spread around on fliers.For the record, I'm not trying to make the argument for the 22 man roster because that's what I had... I just thought the 20 vs. 24 debate could use a new wrinkle after looking at the elimination numbers 5 week in.
I think where that extra $1 should go is to getting that extra kicker and defense. (I only drafted 2 of each). I think teams should draft 3 $1 kickers and 3 $ 1 defenses. Those two groups, kickers especially, can garner a good amount of points. And at $1 too. Plenty of people will tell you how great the Saints have done at $1. And Dan Carpenter has given me solid production including a 17 point game for $1.Yeah, that was the term that kept escaping me. Thanks for the catch.To jdoggydogg's point about what does $2 do, I agree. In a vacuum it does very little. It all depends on how you use it. Yes, you can invest in some one dollar fliers or you can upgrade positions by one notch, but it's all about who you get. Someone who upgrades their $1 WR flier to a $2 Darius Heyward-Bey isn't reaping the same rewards as someone who upgraded to Kenny Britt or Mario Manningham.In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
I couldn't agree with this more (although this year I used a lot of those extra dollars to upgrade at certain positions), but I think 4 $1 or $2 kickers and defenses will out perform the "studs" at the position. Every week some random kicker boots 5 field goals and some random defense blows up (49ers a couple weeks back and Seattle last week). In best ball, I think quantity at those positions is better than quality since you have more good matchups to exploit.I think where that extra $1 should go is to getting that extra kicker and defense. (I only drafted 2 of each). I think teams should draft 3 $1 kickers and 3 $ 1 defenses. Those two groups, kickers especially, can garner a good amount of points. And at $1 too. Plenty of people will tell you how great the Saints have done at $1. And Dan Carpenter has given me solid production including a 17 point game for $1.Yeah, that was the term that kept escaping me. Thanks for the catch.To jdoggydogg's point about what does $2 do, I agree. In a vacuum it does very little. It all depends on how you use it. Yes, you can invest in some one dollar fliers or you can upgrade positions by one notch, but it's all about who you get. Someone who upgrades their $1 WR flier to a $2 Darius Heyward-Bey isn't reaping the same rewards as someone who upgraded to Kenny Britt or Mario Manningham.In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
Def agree with the K position- having 3 1$ kickers is certainly better than 1 3$ kicker.I couldn't agree with this more (although this year I used a lot of those extra dollars to upgrade at certain positions), but I think 4 $1 or $2 kickers and defenses will out perform the "studs" at the position. Every week some random kicker boots 5 field goals and some random defense blows up (49ers a couple weeks back and Seattle last week). In best ball, I think quantity at those positions is better than quality since you have more good matchups to exploit.I think where that extra $1 should go is to getting that extra kicker and defense. (I only drafted 2 of each). I think teams should draft 3 $1 kickers and 3 $ 1 defenses. Those two groups, kickers especially, can garner a good amount of points. And at $1 too. Plenty of people will tell you how great the Saints have done at $1. And Dan Carpenter has given me solid production including a 17 point game for $1.Yeah, that was the term that kept escaping me. Thanks for the catch.To jdoggydogg's point about what does $2 do, I agree. In a vacuum it does very little. It all depends on how you use it. Yes, you can invest in some one dollar fliers or you can upgrade positions by one notch, but it's all about who you get. Someone who upgrades their $1 WR flier to a $2 Darius Heyward-Bey isn't reaping the same rewards as someone who upgraded to Kenny Britt or Mario Manningham.In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.Something to chew on.
Could 2010 be the year we meet 32 Kicker Guy?
Hindsight is 20/20, though. Not once when I was building my roster did I feel like I was sacrificing quality by going with 24 instead of 20 players. This is a best ball format so first and foremost, I wanted value. Once I found my value picks, then I built my roster around them. I did not feel safe splurging on a huge money running back or receiver, I felt it was putting too many eggs in one basket, especially at running back, where injuries are so common, especially those 2-3 week nagging ones. So I went with quantity value players at running back (obviously, this is subjective). Same with receivers although I spent a little more on my top two because I felt they were great value - Chad Ochocinco and Anquan Boldin. Having 24 players has allowed me to get contributions from many players, as well as being able to weather some injuries (parker, laurent robinson, hasselback) and ineffectiveness (Mark Bradley, Leon Washington, Lendale White).
And I would love to trade Brady (?) for Shaun hill (11) and Carson Palmer (21)- so that I wouldn't go with zero pts from my QB this week (I actually have the Palmer/Hill combo but whatever). Unless you drafted a near perfect team there would always be players/combos you want to swap. If L Robinson hadn't broken his leg and Sims walker had your suddenly not so thrilled about that trade.
But that is the point - the winner will be decided by the "right" players - not the number of players on your roster.
Here's the problem with that argument: if you DOWNGRADE $7 from Chris Henry to Nate Burleson, you'd have a better chance of winning this contest. Having $4 extra to spend on your 20 man team presumes upgrading a player or two presumes that upgrade carries with it a guarantee of production. If you look at how well cheap, key players perform every year, to me that re-enforces the need to spread your money around even more to decrease risk.Yeah, that was the term that kept escaping me. Thanks for the catch.To jdoggydogg's point about what does $2 do, I agree. In a vacuum it does very little. It all depends on how you use it. Yes, you can invest in some one dollar fliers or you can upgrade positions by one notch, but it's all about who you get. Someone who upgrades their $1 WR flier to a $2 Darius Heyward-Bey isn't reaping the same rewards as someone who upgraded to Kenny Britt or Mario Manningham.In economics that is called diminishing returns. At some point the increase in return of having more players, will be outweighed by the advantage of having better quality players.The pct_alive from 20 to 22 man roster rises a remarkable 16.79%, but from 22 to 24 you're only seeing an additional 8.41% gain. Yes, without a doubt, a higher percentage is better. After all, we all want to maximize our chances of moving on. However, my point is that a roster of 22 still produces the large security net that owners are coveting while providing their owners to upgrade at a couple of positions.
Something to chew on.
Well said. However, would you rather have Jason Witten for $27 or spend even less on the combined team of Celek, Carlson, and Finley? That's just what I did. While I never dreamed that those three guys were better or more valuable than Witten, it turns out that one of those three have outscored Witten almost every week. I can see your logic: spend the right amount of money on the players you think will succeed. But the dilemma there is that your risk is higher with a smaller roster. We all make bad calls on certain players. I thought that Chris Henry would be good value at $12. I was wrong. But Burleson at $5 saved that bad choice.
The bolded was especially true for WRs this year. While the RB corps was very inflated, I felt there were a ton of good values at the WR position.Balco said:
I would still argue for taking Witten (or any top 3/4 TE) plus two fliers is the right way to go.With TE getting 1.5 per recpt, that is a huge bonusJason Witten $27 14.60 16.80 21.20 9.10 12.20 Todd Heap $5 20.90 8.40 10.10 10.60 14.60 Jermichael Finley $3 2.10 11.60 0.00 27.80 0.00 That is two scorers from the TE position every week - and solid scoring, very consistent scoring.
I've done this contest for fours years now, if I am correct. The first three years, I purchased a top guy or two at a couple of positions. And I have regretted it each year. Yes, it's possible that guys who drafted Adrian Peterson this year will be in the money. But that is just too much money at one position for me. You have very little margin for error if Peterson has a bad week or gets hurt.
OK, so I dumped a bunch of data and ran some regressions.Let's start with kickers. Here's the plan: look at every team --- even the ones that have been eliminated --- and record the following pieces of information about them:
Definitely sig-worthy
Cool info Doug, but do you REALLY have to spell out the scientific ideal strategy for everyone? That's almost worse that posting bye weeks next to the player costs on the entry form...
Seeing as you'd only be giving up 2-3 points per week to even the best teams, it appears the optimal strategy is to draft 0 kickers and spend that money elsewhere.
Keeping in mind the caveats about the thin data, this does seem to confirm what I and many others have probably always believed - that spending any more than $1-2 on a kicker is pointless. It seems like every year there are a some "elite" kickers priced in the $4-7 range and I really never saw the point. For the same reason you don't draft a kicker until the last round in a regular serpentine draft, you don't ever spend $7 on a kicker in this contest.
I know you were kidding (sort of), but I'll take this opportunity to point out that the above study was a lot of work, but it only scratches the surface of a tiny, tiny piece of the contest strategy. Even if this sort of study could be completely nailed down over a period of 20 years worth of data, it doesn't even touch major pieces of strategy like bye week allocation and such. In other words, this is a pretty cool contest.Cool info Doug, but do you REALLY have to spell out the scientific ideal strategy for everyone? That's almost worse that posting bye weeks next to the player costs on the entry form...
Except the Top TE is cheaper than the top qb/rb or wr - in fact significantly cheaper, while providing a higher average score, and typically more consistent.
Yes it is - in combo with the boards, who da thunk that a draft once and forget it competition could provide so much fun and entertainment?I know you were kidding (sort of), but I'll take this opportunity to point out that the above study was a lot of work, but it only scratches the surface of a tiny, tiny piece of the contest strategy. Even if this sort of study could be completely nailed down over a period of 20 years worth of data, it doesn't even touch major pieces of strategy like bye week allocation and such. In other words, this is a pretty cool contest.Cool info Doug, but do you REALLY have to spell out the scientific ideal strategy for everyone? That's almost worse that posting bye weeks next to the player costs on the entry form...
It seems like the effect of the 2nd, 3rd, and 4th should go up if it when you run it on 2008 data because all the BYE weeks will be in the data.
No, the point is that the optimal roster of 24 "right" players will, on average, beat an optimal roster of 20 "right" players. And with 13,000 entries a fair number of people will pick a roster that's close to optimal.
Great work, i still need to digest this a bit. I guess for the flex spot you need to figure out based on the numbers of WR you have, how likely it is to have a WR as your flex. Then you have to figure out how much additional points that was over the player he beat out. It can get complex.
I don't see anybody who came close to picking an optimal lineup. It would be so easy to pick a lineup right now that is far superior to anything out there. Even the top teams have lots of mistakes on them.
Well there's no way to have any idea what the optimal team looks like yet. That won't be determined until after week 16.ETA: And re: the post you responded to, the number of possible roster combinations is at least in the quadrillions, so it's not a sure bet that any one of the 13,000 entries has constructed anything very close to what the optimal roster will turn out to be. Only 250 of the 13,000 even have a shot at making that claim, since in order to be considered optimal you have to at least make it to week 14.
That's exactly why I'm going to keep my opinions about his post, plus my thoughts and personal K strategy to myself... but DD's analysis is fascinating and fun stuff to read.Cool info Doug, but do you REALLY have to spell out the scientific ideal strategy for everyone? That's almost worse that posting bye weeks next to the player costs on the entry form...
Absolutely. I agree. Though I'd still argue that taking any of the top five guys at any position maximizes your risk.
