Challenge to the Shark Pool (1 Viewer)

[BassNBrew]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.

Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.

Team A is starting

QB: Warner 22

RB: CJ 20, Rice 19

WR: AJ 15, Welker 13, Moss 13

TE: Davis 12

D: BAL 11

Your roster has the following choices.

QB: Brady 22, Schuab 22, Romo 22

RB: Moroney 11, Forte 10, Hightower 9

WR: Boldin 12, Austin 11, Rice 10, Crabtree 10

TE: Watson 4, Scheffler 4

D: SD 7, SF 6

Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed. Your goal is to win this head to head matchup. This is a pass/fail excercise. No bonus points for a close loss.

 

[Stinkin Ref]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.Team A is startingQB: Warner 22RB: CJ 20, Rice 19WR: AJ 15, Welker 13, Moss 13TE: Davis 12D: BAL 11Your roster has the following choices.QB: Brady 22, Schuab 22, Romo 22RB: Moroney 11, Forte 10, Hightower 9WR: Boldin 12, Austin 11, Rice 10, Crabtree 10TE: Watson 4, Scheffler 4D: SD 7, SF 6Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed.

hypothetically asking....why did you pick these particular players for this thread in this forum.....

 

[PatsFanCT]

Brady

Maroney, Forte

Bolden, Austin, Rice

Watson

SF

 

[BassNBrew]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.Team A is startingQB: Warner 22RB: CJ 20, Rice 19WR: AJ 15, Welker 13, Moss 13TE: Davis 12D: BAL 11Your roster has the following choices.QB: Brady 22, Schuab 22, Romo 22RB: Moroney 11, Forte 10, Hightower 9WR: Boldin 12, Austin 11, Rice 10, Crabtree 10TE: Watson 4, Scheffler 4D: SD 7, SF 6Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed.

hypothetically asking....why did you pick these particular players for this thread in this forum.....

These players because some of them are loosely linked and Dodd's has them projected in the same neighborhood of each other. I'm curious how people tackle this.

 

[BassNBrew]

I'll include the link that generated this topic, but let's keep the discussion there and lineup selections here.

http://forums.footballguys.com/forum/index...howtopic=511081

 

[Stinkin Ref]

I'll include the link that generated this topic, but let's keep the discussion there and lineup selections here.

http://forums.footballguys.com/forum/index...howtopic=511081

so hypothetically we are going to have the same discussion here.....just different players/scenerio?

 

[ratbast]

QB: Brady

RB: ??? toss up at first glance I'd probably bench Forte but I would really want to play him.

WR: Boldin, Austin, Rice

TE: Watson

D: either, you'll know better than me.

I think I'd be a little nervous about the trifecta you possibly have with NE and SF/Ariz (same game)

Many argue that combos make no difference. This situation might. Its 1 and done here and you have a lot of eggs wrapped up in 2 baskets. However, you are certainly the underdog so you need to look at ceiling and forget about floor. This is hypothetical so replace "you" with "player x".

 

[Chaka]

Where is the hypothetical question here?

 

[buymeagoat]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

 

[BassNBrew]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

This is correct. All my leagues this year were draft and go. I'm giving you a set of tools and a problem to solve. Interested in the solutions presented.

 

[buymeagoat]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

This is correct. All my leagues this year were draft and go. I'm giving you a set of tools and a problem to solve. Interested in the solutions presented.

Any chance you can include matchups and where the games are played?

 

[Dragon1952]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.Team A is startingQB: Warner 22RB: CJ 20, Rice 19WR: AJ 15, Welker 13, Moss 13TE: Davis 12D: BAL 11Your roster has the following choices.QB: Brady 22, Schuab 22, Romo 22RB: Moroney 11, Forte 10, Hightower 9WR: Boldin 12, Austin 11, Rice 10, Crabtree 10TE: Watson 4, Scheffler 4D: SD 7, SF 6Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed.

hypothetically asking....why did you pick these particular players for this thread in this forum.....

Because he thought the other thread posing the same exact question wasn't enough?

 

[Stinkin Ref]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

This is correct. All my leagues this year were draft and go. I'm giving you a set of tools and a problem to solve. Interested in the solutions presented.

my cars broke....here's my tool box.....thanks....

 

[Kool-Aid Larry]

Because he thought the other thread posing the same exact question wasn't enough?

haha....I think he's just honestly trying to get a bunch of people to pick the players, which is pretty much a lost cause in the other thread.I understand his interest in the thought experiment, but it's a little misguided as you can't learn anything from the actual results.once the point totals roll in, it's obvious who you 'should have' played, the only merit is in the weighing of the percentages and strategies beforehand.I think maybe he even made mention of the fact that good strategies sometimes produce poor results, while bad strategies can be successful.it's like betting on a die roll.if we bet even money, and I get the numbers 1-4, that'd be a good bet even if the die comes up 5.and for people who haven't read the linked thread, this isn't at all a WDIS, but rather a hypothetical illustration of the strategy of 'countering' one player with another.

 

[Chaka]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

This is correct. All my leagues this year were draft and go. I'm giving you a set of tools and a problem to solve. Interested in the solutions presented.

So you just substituted Warner and Boldin for Rivers and VJax?Well according to the projections it doesn't matter who team B starts.

 

[Kool-Aid Larry]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

This is correct. All my leagues this year were draft and go. I'm giving you a set of tools and a problem to solve. Interested in the solutions presented.

So you just substituted Warner and Boldin for Rivers and VJax?Well according to the projections it doesn't matter who team B starts.

I think this is the crux of the problem --- a lot of people seem to mistake projections for reality.in the other thread I suggested we post a list of 20 players and see how many matched their projections --- maybe that's what we need to do.

 

[lyon812]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.Team A is startingQB: Warner 22RB: CJ 20, Rice 19WR: AJ 15, Welker 13, Moss 13TE: Davis 12D: BAL 11Your roster has the following choices.QB: Brady 22, Schuab 22, Romo 22RB: Moroney 11, Forte 10, Hightower 9WR: Boldin 12, Austin 11, Rice 10, Crabtree 10TE: Watson 4, Scheffler 4D: SD 7, SF 6Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed. Your goal is to win this head to head matchup. This is a pass/fail excercise. No bonus points for a close loss.

QB: Brady over Schaub. I trust Brady more against a tough defense, and Schaub's been average since Daniels went out and their running game was shown to be pathetic. RB: Hightower and Maroney. SF's tough against the run but weak against the pass. I expect Hightower to be much more useful than Wells in this game. Maroney has a better matchup, but the return of Morris and Faulk have eaten significantly into his PT, as he was only involved in 43% of the plays last week. I think the three-person committee detracts more from his value than the one in AZ, since Hightower still gets the majority timeshare. However, Forte I simply wouldn't start. Green Bay is a strong defense, period. Avoid, avoid.WR: Boldin, Rice, and Austin. Cinci pass D is good. SD's has struggled and SF is woeful. Rice should get the lion's share of Favre's attention, especially if Harvin's out. TE: Watson, but I don't like either. D: SD has a better chance of stopping Dallas. Romo and Co. are more inconsistent.It should also be said that Team A is much better on paper.

 

[Jeff Pasquino]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.

Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.

Team A is starting

QB: Warner 22

RB: CJ 20, Rice 19

WR: AJ 15, Welker 13, Moss 13

TE: Davis 12

D: BAL 11

Your roster has the following choices.

QB: Brady 22, Schuab 22, Romo 22

RB: Moroney 11, Forte 10, Hightower 9

WR: Boldin 12, Austin 11, Rice 10, Crabtree 10

TE: Watson 4, Scheffler 4

D: SD 7, SF 6

Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed. Your goal is to win this head to head matchup. This is a pass/fail excercise. No bonus points for a close loss.

I'll play.I believe the crux of the argument is three-fold:

1. Player / Team A has a lineup in and you know what he's doing. You know some info in advance of the contest.

2. You have players that are comparable in projections.

3. You have decent crossover at RB-WR-TE to his QB and vice versa.

So given 1 and 2, the correct answer if you think you are in a close matchup and if you are a diehard believer in the projections and/or you are torn on who to start, you should pick:

QB: Brady or Schaub. Reasoning - if Schaub explodes, it would counteract AJ in Player A's lineup (who would be the beneficiary of a Schaub big game). Also a strong and similar argument for Brady and Welker and Moss. Best answer is probably Brady.

RB: One of them should be Hightower. Reasoning - if Kurt Warner checks down often in a PPR league, you get the points. Even without PPR, if he does well through the air, so does Warner, so again you counter Warner.

WR: Boldin is one choice. Similar reasoning to using Hightower - big game by Warner, likely big game by Boldin.

Now, based on these projections, you're not the favorite. You're even or down at every spot.

So, if you want to try and hit a home run and win the game, you probably want to swing for the fences and go with the best matchups based on your own opinion. While Boldin is still a good choice, going Brady/Maroney or Romo/Austin gives bigger upside with your lineup, so - based on a projected loss - you'll need to really hope for a Brady-Maroney TD pass or (more likely) a Romo-Austin hookup.

The same argument goes to Brady-Watson.

Based on the projections above, I'd choose:

Brady (have to try and counteract both Welker and Moss)

Austin/Boldin/Crabtree

After those I would simply take my own best guess of the rest of the lineup.

I think the only factor here in this case, since you're "behind" before you even start, is that you really need to use Brady instead of Romo or Schaub because a big game by Brady would likely mean a losing day for you if you don't start him as it also would mean Moss and Welker had big days.

It's a good question.

 

[rotorooter]

Team A is starting

QB: Warner 22

RB: CJ 20, Rice 19

WR: AJ 15, Welker 13, Moss 13

TE: Davis 12

D: BAL 11

Your roster has the following choices.

QB: Brady 22, Schuab 22, Romo 22

RB: Moroney 11, Forte 10, Hightower 9

WR: Boldin 12, Austin 11, Rice 10, Crabtree 10

TE: Watson 4, Scheffler 4

D: SD 7, SF 6

not sure exactly what the intent of this exercise is...it would help to see who is playing what team and where so I can see if I agree with the projections.

You obviously have team B as a significant "dog" by 35-38 pts.

I would go : schaub, moroney, forte, boldin, austin, s. rice , watson, SD

hard for me to start Brady with team A having moss/welker.

i could max bet romo-austin connection but would likely bust.So I would have to count a big from schuab without going to AJ; hope Warner throws to Boldin all day and pray that CJ gets stuffed...

 

[Kool-Aid Larry]

It should also be said that Team A is much better on paper.

that was his intent as part of the exercise.

 

[Chaka]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

This is correct. All my leagues this year were draft and go. I'm giving you a set of tools and a problem to solve. Interested in the solutions presented.

So you just substituted Warner and Boldin for Rivers and VJax?Well according to the projections it doesn't matter who team B starts.

I think this is the crux of the problem --- a lot of people seem to mistake projections for reality.in the other thread I suggested we post a list of 20 players and see how many matched their projections --- maybe that's what we need to do.

Believe me, no one doubts projections more than I do. I think they are a pointless exercise and I have suggested repeatedly to believers that they track the projections vs. reality for players that they use projections to decide between.In a very cursory, non-peer reviewed, quick examination of that question I found that projections were less than a 50% proposition. I would like to do an extensive review on this topic if I decide to actually make the time, however considering the fact that I never use projections I am not sure I will actually follow through on that notion.

 

[lyon812]

It should also be said that Team A is much better on paper.

that was his intent as part of the exercise.

Yes, obviously. But it also skews things to an extreme degree, because his team is clearly superior at both RB and WR; for example, in my league he'd have the #1 and 3 RBs and the #1, 4, and 5 WR, and the #3 TE. Conversely, I have RB 16, 18, and 34. and WR 8, 10, and 19, and TE 22 and 23. In other words, it's a purposefully exaggerated exercise which seems biased in its attempt to demonstrate a point. It might be more interesting, and useful, to 1) not simply give them the best players at every position, and 2) not make every player match up exactly. For example, keep Team B the same, but have Team A:QB: Smith 20RB: CJ 20, Addai 14WR: AJ 15, Welker 13, Moss 13TE: Keller 5D: SD 6

 

[Ariakis]

In one of my leagues my opponent tomorrow starts Driver and Jennings while I start Rodgers. I'm just going to hope for a big day for Rodgers and that some of his production falls to other targets rather than play an inferior QB and get to root against the Packers in general.

One other interesting thing about this weekend is I face a player in one league that awards 3 for a TD and another I own that scores 6. I'll be rooting for a low yardage 1-2 TD game to give me a slightly better chance to win both leagues.

Similarly, I face Rodgers in another league that gives out a 300 yard bonus so I'll be rooting that while he has a great day he finishes below 300 yards.

 

[GregR_2]

I think BassNBrew is trying to figure out how to tackle a projections issue using players based on how Dodds as ranked them. I don't think this is a case where he's asking who to start as I believe the players picked aren't representative of his own players. Is that accurate?

This is correct. All my leagues this year were draft and go. I'm giving you a set of tools and a problem to solve. Interested in the solutions presented.

So you just substituted Warner and Boldin for Rivers and VJax?Well according to the projections it doesn't matter who team B starts.

I think this is the crux of the problem --- a lot of people seem to mistake projections for reality.in the other thread I suggested we post a list of 20 players and see how many matched their projections --- maybe that's what we need to do.

Believe me, no one doubts projections more than I do. I think they are a pointless exercise and I have suggested repeatedly to believers that they track the projections vs. reality for players that they use projections to decide between.In a very cursory, non-peer reviewed, quick examination of that question I found that projections were less than a 50% proposition. I would like to do an extensive review on this topic if I decide to actually make the time, however considering the fact that I never use projections I am not sure I will actually follow through on that notion.

When I read this I found myself wondering what "found that projections were less than 50% proposition" meant exactly.If projections are "a pointless exercise" then they would have to be less accurate than other means of selecting starters. Otherwise they would have a point to doing them. Accounting for things like injuries, offensive game plan, weather, quality of opponent's defense against pass or rush, etc, is why you make a projection. So there isn't a whole lot to compare it to besides doing something like blindly following the PPG for the season to date.Since the ranking of, say, Fitzgerald and Kevin Walter, is likely going to be correct in both or wrong in both on most weeks, I'd expect the difference between the systems to be much less than 50% anyway if we did a comparison for a good sized sample of players at each position.

 

[Pickle Rody]

Team A is starting:

QB: Warner 22

RB: CJ 20, Rice 19

WR: AJ 15, Welker 13, Moss 13

TE: Davis 12

D: BAL 11

I'd counter with:

QB: Schuab 22

RB: Forte 10, Hightower 9

WR: Boldin 12, Austin 11, Rice 10

TE: Watson 4

D: SD 7

As far as Team A: I would hope that only Moss or Welker gets some points, not both of them.

I'm countering:

Schaub to AJ.

Boldin & Hightower to Warner.

Liking Rice a lot but even more so with Harvin probably out.

Forte should have a solid game. Can't count on that from Maroney.

Team A probably still wins this matchup, but I'm thinking it will be close.

 

[BassNBrew]

It should also be said that Team A is much better on paper.

that was his intent as part of the exercise.

Yes, obviously. But it also skews things to an extreme degree, because his team is clearly superior at both RB and WR; for example, in my league he'd have the #1 and 3 RBs and the #1, 4, and 5 WR, and the #3 TE. Conversely, I have RB 16, 18, and 34. and WR 8, 10, and 19, and TE 22 and 23. In other words, it's a purposefully exaggerated exercise which seems biased in its attempt to demonstrate a point. It might be more interesting, and useful, to 1) not simply give them the best players at every position, and 2) not make every player match up exactly. For example, keep Team B the same, but have Team A:QB: Smith 20RB: CJ 20, Addai 14WR: AJ 15, Welker 13, Moss 13TE: Keller 5D: SD 6

Actually that is part of the point. I wanted A to be a clear favorite and some players whose play would compliment each other.

 

[Chaka]

When I read this I found myself wondering what "found that projections were less than 50% proposition" meant exactly.If projections are "a pointless exercise" then they would have to be less accurate than other means of selecting starters. Otherwise they would have a point to doing them. Accounting for things like injuries, offensive game plan, weather, quality of opponent's defense against pass or rush, etc, is why you make a projection. So there isn't a whole lot to compare it to besides doing something like blindly following the PPG for the season to date.Since the ranking of, say, Fitzgerald and Kevin Walter, is likely going to be correct in both or wrong in both on most weeks, I'd expect the difference between the systems to be much less than 50% anyway if we did a comparison for a good sized sample of players at each position.

When I first started questioning the value of projections it was in the context of using them to decide between players on the waiver wire. For example if you are looking for a one week starter at WR. I pulled out all the data for WRs that were selected on the WW in my league this season and the week they were picked up. I compared all the WRs projected scores for that week with the actual outcomes from those weeks.From this very cursory, and limited, analysis there were 11 potential comparisons and the projections correctly identified the higher scoring player for the week in question four times and failed to do so seven times.Not nearly enough data points to draw far reaching and indisputable conclusions but enough for me to stand by my position until better data presents itself.

 

[BassNBrew]

These are the starters that I'm chosing...

QB: Brady 22, Schuab 22, Romo 22

RB: Moroney 11, Forte 10, Hightower 9

WR: Boldin 12, Austin 11, Rice 10, Crabtree 10

TE: Watson 4, Scheffler 4

D: SD 7, SF 6

We'll see how it plays out.

 

[Pickle Rody]

Cool.......keep us posted......

 

[BassNBrew]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.

Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.

Team A is starting

QB: Warner 22

RB: CJ 20, Rice 19

WR: AJ 15, Welker 13, Moss 13

TE: Davis 12

D: BAL 11

Your roster has the following choices.

QB: Brady 22, Schuab 22, Romo 22

RB: Moroney 11, Forte 10, Hightower 9

WR: Boldin 12, Austin 11, Rice 10, Crabtree 10

TE: Watson 4, Scheffler 4

D: SD 7, SF 6

Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed. Your goal is to win this head to head matchup. This is a pass/fail excercise. No bonus points for a close loss.

I'll play.I believe the crux of the argument is three-fold:

1. Player / Team A has a lineup in and you know what he's doing. You know some info in advance of the contest.

2. You have players that are comparable in projections.

3. You have decent crossover at RB-WR-TE to his QB and vice versa.

So given 1 and 2, the correct answer if you think you are in a close matchup and if you are a diehard believer in the projections and/or you are torn on who to start, you should pick:

QB: Brady or Schaub. Reasoning - if Schaub explodes, it would counteract AJ in Player A's lineup (who would be the beneficiary of a Schaub big game). Also a strong and similar argument for Brady and Welker and Moss. Best answer is probably Brady.

RB: One of them should be Hightower. Reasoning - if Kurt Warner checks down often in a PPR league, you get the points. Even without PPR, if he does well through the air, so does Warner, so again you counter Warner.

WR: Boldin is one choice. Similar reasoning to using Hightower - big game by Warner, likely big game by Boldin.

Now, based on these projections, you're not the favorite. You're even or down at every spot.

So, if you want to try and hit a home run and win the game, you probably want to swing for the fences and go with the best matchups based on your own opinion. While Boldin is still a good choice, going Brady/Maroney or Romo/Austin gives bigger upside with your lineup, so - based on a projected loss - you'll need to really hope for a Brady-Maroney TD pass or (more likely) a Romo-Austin hookup.

The same argument goes to Brady-Watson.

Based on the projections above, I'd choose:

Brady (have to try and counteract both Welker and Moss)

Austin/Boldin/Crabtree

After those I would simply take my own best guess of the rest of the lineup.

I think the only factor here in this case, since you're "behind" before you even start, is that you really need to use Brady instead of Romo or Schaub because a big game by Brady would likely mean a losing day for you if you don't start him as it also would mean Moss and Welker had big days.

It's a good question.

Disagree with the bolded part Jeff. Team A is about a 40 point favorite. If Moss and Welker go off I think you're looking at a loss regardless of whether or not you start Brady. I think the play to make up a 40 point expected deficit is to decouple from Team A. I'm hitching wagon chances of a victory to Romo/Austin having a better than projected day and Brady/Welker/Moss having a less than projected day and hoping that benefits Maroney.

 

[BassNBrew]

Team A is starting

QB: Warner 22

RB: CJ 20/37, Rice 19/28

WR: AJ 15/31, Welker 13/11, Moss 13/2

TE: Davis 12

D: BAL 11/5

Your roster has the following choices.

QB: Brady 22/11, Schuab 22/24, Romo 22/20

RB: Moroney 11/11, Forte 10/7, Hightower 9

WR: Boldin 12, Austin 11/19, Rice 10/10, Crabtree 10

TE: Watson 4/10, Scheffler 4/0

D: SD 7/1, SF 6

Preliminary update with scoring from the early games (follows the Dodds projections). Looks like no lineup will over take Team A as projected. Decoupling Brady from Moss and Welker at least gives an outside shot to the teams that started Schuab or Romo if he can do something in the second half and Crabtree and Hightower light it up.

Updated through the day games.

Team A 114 + Warner + Davis. If Rice and CJ had performed to expectations, Team A would be at 88 points.

My selections have 70 + Hightower + Crabtree + SF D to go.

The start your best players team has 59 + Boldin + (10 for Watson or 0 for Scheffler).

 

[GregR_2]

When I read this I found myself wondering what "found that projections were less than 50% proposition" meant exactly.

If projections are "a pointless exercise" then they would have to be less accurate than other means of selecting starters. Otherwise they would have a point to doing them. Accounting for things like injuries, offensive game plan, weather, quality of opponent's defense against pass or rush, etc, is why you make a projection. So there isn't a whole lot to compare it to besides doing something like blindly following the PPG for the season to date.

Since the ranking of, say, Fitzgerald and Kevin Walter, is likely going to be correct in both or wrong in both on most weeks, I'd expect the difference between the systems to be much less than 50% anyway if we did a comparison for a good sized sample of players at each position.

When I first started questioning the value of projections it was in the context of using them to decide between players on the waiver wire. For example if you are looking for a one week starter at WR. I pulled out all the data for WRs that were selected on the WW in my league this season and the week they were picked up. I compared all the WRs projected scores for that week with the actual outcomes from those weeks.From this very cursory, and limited, analysis there were 11 potential comparisons and the projections correctly identified the higher scoring player for the week in question four times and failed to do so seven times.

Not nearly enough data points to draw far reaching and indisputable conclusions but enough for me to stand by my position until better data presents itself.

I wouldn't stand by that interpretation.Sample size aside... if the other decision maker we have, points per game on the season, only successfully chose the right player 1 time in 11, then those projections would be a pretty useful tool. Or alternatively, if PPG chose the right player 7 times in 11, then projections might be a poor tool. Or if PPG chose the right player 4 times in 11 as well, it might not be any better or worse.

Or we could compare it to random chance. If 3 WRs are being taken every week then projections are coming out slightly better than random chance. If 5 are being taken then it's doing a fair amount better than random chance. (Again, sample size aside.)

I guess I'm saying it doesn't sound like it was a look at the issue that would make me think one could draw anything meaningful from it yet. There isn't even a context yet by which to judge whether 4 in 11 is good or bad.

 

[Chaka]

When I read this I found myself wondering what "found that projections were less than 50% proposition" meant exactly.

If projections are "a pointless exercise" then they would have to be less accurate than other means of selecting starters. Otherwise they would have a point to doing them. Accounting for things like injuries, offensive game plan, weather, quality of opponent's defense against pass or rush, etc, is why you make a projection. So there isn't a whole lot to compare it to besides doing something like blindly following the PPG for the season to date.

Since the ranking of, say, Fitzgerald and Kevin Walter, is likely going to be correct in both or wrong in both on most weeks, I'd expect the difference between the systems to be much less than 50% anyway if we did a comparison for a good sized sample of players at each position.

When I first started questioning the value of projections it was in the context of using them to decide between players on the waiver wire. For example if you are looking for a one week starter at WR. I pulled out all the data for WRs that were selected on the WW in my league this season and the week they were picked up. I compared all the WRs projected scores for that week with the actual outcomes from those weeks.From this very cursory, and limited, analysis there were 11 potential comparisons and the projections correctly identified the higher scoring player for the week in question four times and failed to do so seven times.

Not nearly enough data points to draw far reaching and indisputable conclusions but enough for me to stand by my position until better data presents itself.

I wouldn't stand by that interpretation.Sample size aside... if the other decision maker we have, points per game on the season, only successfully chose the right player 1 time in 11, then those projections would be a pretty useful tool. Or alternatively, if PPG chose the right player 7 times in 11, then projections might be a poor tool. Or if PPG chose the right player 4 times in 11 as well, it might not be any better or worse.

Or we could compare it to random chance. If 3 WRs are being taken every week then projections are coming out slightly better than random chance. If 5 are being taken then it's doing a fair amount better than random chance. (Again, sample size aside.)

I guess I'm saying it doesn't sound like it was a look at the issue that would make me think one could draw anything meaningful from it yet. There isn't even a context yet by which to judge whether 4 in 11 is good or bad.

If the best projections can offer me is 4 out of 11, I'll stick with blindly throwing darts behind my back at last years magazines to make my decisions.

 

[BassNBrew]

Team A is startingQB: Warner 22RB: CJ 20/37, Rice 19/28WR: AJ 15/31, Welker 13/11, Moss 13/2TE: Davis 12D: BAL 11/5Your roster has the following choices.QB: Brady 22/11, Schuab 22/24, Romo 22/20RB: Moroney 11/11, Forte 10/7, Hightower 9WR: Boldin 12, Austin 11/19, Rice 10/10, Crabtree 10TE: Watson 4/10, Scheffler 4/0D: SD 7/1, SF 6Preliminary update with scoring from the early games (follows the Dodds projections). Looks like no lineup will over take Team A as projected. Decoupling Brady from Moss and Welker at least gives an outside shot to the teams that started Schuab or Romo if he can do something in the second half and Crabtree and Hightower light it up.Updated through the day games.Team A 114 + Warner + Davis. If Rice and CJ had performed to expectations, Team A would be at 88 points.My selections have 70 + Hightower + Crabtree + SF D to go. The start your best players team has 59 + Boldin + (10 for Watson or 0 for Scheffler).

Still an extreme longshot, but this has been a clear demonstration of why techniques like decoupling and countering can be a good idea as an underdog.I may give this another go next week with a new set of players.

 

[Bri]

This is hypothetical, so hopefully the mods will grant me some leeway here as it's a spinoff from a strategy thread.Your opponent this week has submitted the following lineup. The numbers are Dodds weekly projections.Team A is startingQB: Warner 22RB: CJ 20, Rice 19WR: AJ 15, Welker 13, Moss 13TE: Davis 12D: BAL 11Your roster has the following choices.QB: Brady 22, Schuab 22, Romo 22RB: Moroney 11, Forte 10, Hightower 9WR: Boldin 12, Austin 11, Rice 10, Crabtree 10TE: Watson 4, Scheffler 4D: SD 7, SF 6Start 1/2/3/1, std FBG scoring. Pick your lineup to beat Team A. If you don't like any of the choices or your roster, feel free to substitute anyone from Dodd's rankings that are within 10% or one point of the players listed. Your goal is to win this head to head matchup. This is a pass/fail excercise. No bonus points for a close loss.

Anyhow Harvin was out sunday AM so I'm unsure if Dodds #s reflected that. The status made me like Rice more.I don't like either TE but Sanders being out would have had me choosing (incorrectly) Scheffler in that tie-breaker.Schaub was hurt week before, Romo and the Boys stink in December so Brady who rocks in December.Nothing would make me start Maroney short of him playing elsewhere. Chargers had numerous defenders injured

 

[BassNBrew]

Team A is starting

QB: Warner 22/5

RB: CJ 20/37, Rice 19/28

WR: AJ 15/31, Welker 13/11, Moss 13/2

TE: Davis 12/9

D: BAL 11/5

Your roster has the following choices.

QB: Brady 22/11, Schuab 22/24, Romo 22/20

RB: Moroney 11/11, Forte 10/7, Hightower 9/2

WR: Boldin 12/4, Austin 11/13, Rice 10/10, Crabtree 10/13

TE: Watson 4/10, Scheffler 4/0

D: SD 7/1, SF 6/18

Projections/Actual Pts

Team A - 128 actual

My selections - 97

Start the better players team - 57 or 67

This challenge was an excersise in how to pick a lineup to pull off an upset. Team A was coming in a 40 pt favorite based on it's superior RBs, TE, and D. Rice/CJ/AJ/Davis themselves made any combo selected a loser this weekend. That said, some things became obvious...

1. Starting Boldin was a losing proposition. Any upset would required a bad day from Warner. That would carry over to Boldin. These players had to be decoupled.

2. Needing Warner to lay an egg, means that the potential for points to flow to the SF exists. This makes the coin SD/SF a no brainer SF start. Also enhancing this choice is the fact that my starting QB was facing SD D.

3. Under no circumstances can you start Brady. The only prayer for an upset is an off day by Moss/Welker. If that happens it will roll down to Brady to a large extent.

4. Potentially some of the usual points not going to Moss/Welker go elsewhere. Start Maroney/Watson and hope they flow that direction.

5. In this example Schuab would be the best start from a points standpoint. Well those went to AJ who decimated a realistic chance at an upset. The anatomy of an upset required AJ to be quiet this week which likely would have flowed to Schuab. This along with point 3 made the third best QB Romo a must start. Austin dropped a TD at the end of the Dallas game which would have closed the gap. The hookup potential was key to an upset.

Before the bashing begins, keep in mind that I would only visit lineup decisions for these reason when I'm a significant underdog.

Look for another version next week to continue to kick this topic around.

 

[Bri]

This challenge was an exercise in how to pick a lineup to pull off an upset.

You say this in the last post?

 

[BassNBrew]

This challenge was an exercise in how to pick a lineup to pull off an upset.

You say this in the last post?

Not sure I follow you???No lineup overtakes team A this go around. I believe Drinnen expected an extra few wins every decade using some of these concepts. I'll continue to play with this, but hopefully some of the potential effectiveness of the concepts peeked out.