"Cancelling" your own QB (1 Viewer)

[CalBear]

Last week there was a thread about changing lineups for Monday night based on Sunday's scoring, which got sidetracked by lame discussions about whether it was really a disguised WDIS thread (which it did not seem to be). A similar situation came up for my team this week, and I will note in advance that, while the question came up based on my score going into Monday night, my team does not have any of the other players I'm going to mention, so even if I wanted to make a lineup change, I couldn't. And frankly, I wouldn't listen to your advice anyway. So it is not a WDIS thread, and if you want to be sanctimonious instead of contributing to the discussion, please go elsewhere.

That being said...

Let's say you're heading into Monday Night with a 20-ish point deficit; your opponent is done, and you have a good QB-WR pair like Vick and D.Jackson left to play. [That is my team's actual position--the rest from here is speculation]. Your QB scoring two TDs would be enough to get you 20+ points and win your game. You have another, lesser option at WR from the other team (like Santana Moss). Do you start Jackson or Moss?

The interesting thing about this question is that it depends on the frequency of the scenarios. In the scenario where Vick scores no TDs, Jackson is unlikely to score a TD, and you'll probably not get 20+ points. (Jackson isn't a perfect example, because he does have some chance of running in a TD, which I'll ignore for the purposes of this discussion). In the scenario where Vick gets exactly one TD, you win if it goes to Jackson, and you might win if it doesn't. In the scenario where Vick gets two or more TDs, you always win.

Scenario 1: Vick scores no TDs, you win 10% of the time if you start Jackson

Scenario 2: Vick scores 1 TD, you win 50% of the time if you start Jackson

Scenario 3: Vick scores 2+ TDs, you win no matter who you start

In this scenario, might it make more sense to start Moss? He increases your chances of winning in Scenario 1, because he can score even if Vick/Jackson don't. He might decrease your chances of winning in Scenario 2, because he's not as good as Jackson. In Scenario 3 it doesn't matter. It looks to me like it only makes sense to start Jackson if Scenario 2 is significantly more likely than Scenario 1. (Is it? I'm not sure.)

Or how about the reverse scenario, where you have Jackson at WR, and can start Vick or McNabb at QB? My gut suggests that the point differential between the two QBs is more significant than the point differential between the WRs--so you should start Vick even if you have the option to start McNabb.

 

[Jersey35]

Come on Calbear... a WDIS? Shame on you!

This is a really interesting example of what we discussed (very little) last week.

I don't have a lot of time at the moment, but my initial gut tells me to hedge my bet a little bit by spreading out my chances (going with two guys who are NOT correlated).

I think, in this case, it does come down a bit more to the names being considered. Vick has a much higher chance of scoring 2 TD's than say, Colt McCoy. Likewise, I'm more inclined to start Santana Moss over Desean as compared to Darius Heyward-Bey or some such. The matchup would come into play quite a bit as well. As well as PPR or no.

But I think there would be times I definitely would start the opposing guy SPECIFICALLY for this reason.

 

[rizzler]

Vick and Jackson

-- yardage matters as well. and it's always "doubled up" when the two connect.

go with skill, talent and potency of offense.

 

[Chaka]

I don't this scenario negates the whole "Start the players you think will score the most points." argument.

In this case DeSean >>>>>> Santana so roll with DeSean

 

[LittlePhatty]

Cue the dude who will say "start the guy who will score the most points"

Anyway, I'd start Vick over McNabb no matter what, no thinking about it. And the only reason I'd start Moss is it makes MNF more fun to watch when you have guys potentially scoring on both offenses and you don't have to wait if one team is eating up the clock.

 

[CalBear]

I don't this scenario negates the whole "Start the players you think will score the most points." argument.In this case DeSean >>>>>> Santana so roll with DeSean

Desean is averaging 2.8 points per game more than S.Moss in my league. That's ">", but certainly not ">>>>>>". Moss is on track for 96 receptions and 1200 yards; he's a real receiver.

 

[flc735]

its very simple for me.

when you start a qb/wr from the same team, it decreases your consistency but increases your total potential points (less consistency, lower floor, higher ceiling)

when you start a qb/wr from different teams, it increases your consistency but decreases your total potential points (more consistency, higher floor, lower ceiling). aka diversification

so the answer depends on how many points you need.

if total points needed is less than the projected points per game, than you will win this scenario more often with the qb/wr from different teams

if total points needed is more than the projected points per game, than you will win this scenario more often with the qb/wr from the same team

more often - is comparing against the alternative scenario, not necessarily over 50%

this assumes desean and moss are projected out for the same amount of points.

 

[footballnerd]

Vick is the one that carries you in this one regardless who you start at WR, and if he does throw, its a high probability its to DJax. I'd start both. Even if he doesn't get a TD you could win with yards, as Vick is likely to get rushing yds as well.

Now if this was a Peyton/Garcon or Eli/Manningham scenario where the WR at play is not the QB's go to option it would be a tougher decision.

 

[CalBear]

its very simple for me.

when you start a qb/wr from the same team, it decreases your consistency but increases your total potential points (less consistency, lower floor, higher ceiling)

when you start a qb/wr from different teams, it increases your consistency but decreases your total potential points (more consistency, higher floor, lower ceiling). aka diversification

so the answer depends on how many points you need.

if total points needed is less than the projected points per game, than you will win this scenario more often with the qb/wr from different teams

if total points needed is more than the projected points per game, than you will win this scenario more often with the qb/wr from the same team

more often - is comparing against the alternative scenario, not necessarily over 50%

this assumes desean and moss are projected out for the same amount of points.

The way I view the scenario is, Desean is projected for more points than Moss (and reasonably can be expected to score more on average), but the total points needed is less than the projected points per game. So if they were projected equally, you would start the QB/WR from different teams, but because Desean is projected for more, you might start him anyway. At what point does the better WR overcome the scenario advantage?

 

[Jersey35]

its very simple for me.

when you start a qb/wr from the same team, it decreases your consistency but increases your total potential points (less consistency, lower floor, higher ceiling)

when you start a qb/wr from different teams, it increases your consistency but decreases your total potential points (more consistency, higher floor, lower ceiling). aka diversification

so the answer depends on how many points you need.

if total points needed is less than the projected points per game, than you will win this scenario more often with the qb/wr from different teams

if total points needed is more than the projected points per game, than you will win this scenario more often with the qb/wr from the same team

more often - is comparing against the alternative scenario, not necessarily over 50%

this assumes desean and moss are projected out for the same amount of points.

This is a very

 

[DexterDew]

if total points factor anywhere down the line, start the WR that you think will score the most points and double down on PHI's passing game.

if you just want to improve your odds on winning, diversify and start the WR on the other team (as long as it is not a scrub)

i did a variation of this last night being up with a QB (Vick or Roth) to go, and my opponent had Ward. the only way he was going to beat me was if Ward went off and Vick had a really bad game. So I put Roth to cancel out any Ward TDs and some of his yardage. It was moot, of course, but I would do it again.

 

[Jeff Pasquino]

Vick and Jackson-- yardage matters as well. and it's always "doubled up" when the two connect.go with skill, talent and potency of offense.

The names aren't the question here though - certainly for a Sunday game you'd have both the QB and WR going.Now with new information (most games done) you know that if the QB has a bad game, it's virtually assured that his WR does little.The odds of winning are likely better to take the other WR if the scoring you need is not too great. If you need a big score then I'd start the combination as the QB explosion will also likely coincide with the big WR game. If you need minimal points and the QB gets nada, the other WR still gives you a chance.

 

[PolishNorbi]

Normally, double down is nice. However I think the smart play would be Vick/Moss. There are games where QBs just don't do anything.

Take a look at Peyton, normally you can count on him to put up numbers and a duo of their top two is usually nasty. However, yesterday 185/0 for Peyton and 73/0 Tamme. #### happens sometimes.

 

[reprisal]

Interesting question. In this case I think I'd definitely go with Vick and Jackson. With the eagles duo its fairly easy to see either one, or both, scoring 20 points individually. You'll usually want those points as a tie breaker. I'd say its more interesting if you have two lesser options, with similar projections and similar standard deviations. Say you need 25 points and have Aaron Rodgers and either a health Donald Driver or Santana Moss? Otherwise its usually going to come down to either skill, match up or risk profile/upside and not which team your player is on.

edit: Man, I was slow responding and a number of solid replies in the mean time.

 

[CalBear]

Normally, double down is nice. However I think the smart play would be Vick/Moss. There are games where QBs just don't do anything.Take a look at Peyton, normally you can count on him to put up numbers and a duo of their top two is usually nasty. However, yesterday 185/0 for Peyton and 73/0 Tamme. #### happens sometimes.

That's a good example of the scenario you're trying to avoid with the straddle play. If you had Manning and Wayne, surely you'd start Wayne over Ochocinco; but if that game were Monday night, and you needed 20 points, you would have been screwed this week. Manning=6, Wayne=6 in PPR, while Ochocinco would have won your game for you (Ochocinco=21). Last week, Wayne outscored Ochocinco by a lot, but Manning/Ochocinco would still have won you your 20-point game (Manning=19, Ochocinco=2).

 

[reprisal]

Normally, double down is nice. However I think the smart play would be Vick/Moss. There are games where QBs just don't do anything.Take a look at Peyton, normally you can count on him to put up numbers and a duo of their top two is usually nasty. However, yesterday 185/0 for Peyton and 73/0 Tamme. #### happens sometimes.

That's a good example of the scenario you're trying to avoid with the straddle play. If you had Manning and Wayne, surely you'd start Wayne over Ochocinco; but if that game were Monday night, and you needed 20 points, you would have been screwed this week. Manning=6, Wayne=6 in PPR, while Ochocinco would have won your game for you (Ochocinco=21). Last week, Wayne outscored Ochocinco by a lot, but Manning/Ochocinco would still have won you your 20-point game (Manning=19, Ochocinco=2).

Ocho isnt the best example. He has had 3 great games and 6 mediocre or worse. Wayne has been much more consistent. If we take that as statistically relevant then it seems pretty likely that even though you would have won this week you would have made the wrong play and reduced the probability of winning.

 

[CalBear]

Normally, double down is nice. However I think the smart play would be Vick/Moss. There are games where QBs just don't do anything.

Take a look at Peyton, normally you can count on him to put up numbers and a duo of their top two is usually nasty. However, yesterday 185/0 for Peyton and 73/0 Tamme. #### happens sometimes.

That's a good example of the scenario you're trying to avoid with the straddle play. If you had Manning and Wayne, surely you'd start Wayne over Ochocinco; but if that game were Monday night, and you needed 20 points, you would have been screwed this week. Manning=6, Wayne=6 in PPR, while Ochocinco would have won your game for you (Ochocinco=21). Last week, Wayne outscored Ochocinco by a lot, but Manning/Ochocinco would still have won you your 20-point game (Manning=19, Ochocinco=2).

Ocho isnt the best example. He has had 3 great games and 6 mediocre or worse. Wayne has been much more consistent. If we take that as statistically relevant then it seems pretty likely that even though you would have won this week you would have made the wrong play and reduced the probability of winning.

I'm not sure that's true. Wayne is clearly better than Ocho as a play when all else is equal. But in this scenario, it's not. In this scenario, if your QB (Manning) scores 20+ points, you win regardless of who you play at WR. Because Manning usually scores 20+ points, and when he doesn't score 20+ points, Wayne is unlikely to score much, your risk is higher with the Manning/Wayne combo.It is entirely possible (and in fact, is probably true this year, though that's a small data set) that Manning+Wayne will average more points, but that Manning+Ocho will score 20+ points a greater percentage of the time than Manning+Wayne. Certainly that fact must be true for some sets of QBs and WRs. That's the core of this discussion.

 

[flc735]

vick to desean, 88 yard td

 

[LittlePhatty]

OR...if you started Vick and DeSean then you probably got 20 points on the very first play.

 

[flc735]

its very simple for me.

when you start a qb/wr from the same team, it decreases your consistency but increases your total potential points (less consistency, lower floor, higher ceiling)

when you start a qb/wr from different teams, it increases your consistency but decreases your total potential points (more consistency, higher floor, lower ceiling). aka diversification

so the answer depends on how many points you need.

if total points needed is less than the projected points per game, than you will win this scenario more often with the qb/wr from different teams

if total points needed is more than the projected points per game, than you will win this scenario more often with the qb/wr from the same team

more often - is comparing against the alternative scenario, not necessarily over 50%

this assumes desean and moss are projected out for the same amount of points.

The way I view the scenario is, Desean is projected for more points than Moss (and reasonably can be expected to score more on average), but the total points needed is less than the projected points per game. So if they were projected equally, you would start the QB/WR from different teams, but because Desean is projected for more, you might start him anyway. At what point does the better WR overcome the scenario advantage?

if projected points is greater than points needed for either combination, you are winning this game more than 50% of the time no matter which combo you go with. since the main goal is accomplished with either combo, you are no longer concerned with "highest ceiling". it doesn't matter if you win by 5 points or 15 points. so this problem simplifies to

vick/desean - <strike>less consistency</strike>, lower floor, <strike>higher ceiling</strike>

vick/moss - <strike>more consistency</strike>, higher floor, <strike>lower ceiling</strike>

you take the higher floor over the lower floor

 

[Your Mom]

Hope you started DJ

 

[reprisal]

I'm not sure that's true. Wayne is clearly better than Ocho as a play when all else is equal. But in this scenario, it's not. In this scenario, if your QB (Manning) scores 20+ points, you win regardless of who you play at WR. Because Manning usually scores 20+ points, and when he doesn't score 20+ points, Wayne is unlikely to score much, your risk is higher with the Manning/Wayne combo.

It is entirely possible (and in fact, is probably true this year, though that's a small data set) that Manning+Wayne will average more points, but that Manning+Ocho will score 20+ points a greater percentage of the time than Manning+Wayne. Certainly that fact must be true for some sets of QBs and WRs. That's the core of this discussion.

Yes but with Ocho he's put up less than 5 points 5 times so he has essentially been boom or bust. Meaning that you are risking substantially decreasing the chance of winning when Manning doesnt carry you. When Manning only has a pretty good day you may still lose (especially if your 4pt per TD, -2pt int etc). In this exact example youre avoiding the correlation by playing a player who wont work in tandem with your QB much at all. This is why I suggested earlier that it would be easier to discuss this with out using elite talents(so you most likely need something out of both) and using similar types of players (so we dont get side tricked by a debate about the risk/reward of different guys ).

 

[CalBear]

Hope you started DJ

As I said, I don't have Moss or McNabb on my team; this was not a WDIS. (And besides, I can't bench a Golden Bear).We have a $10 bonus for yearly player high score, might get that with Vick.I have an interesting idea of how to look at this question; I'll be back later with some stats.

 

[CalBear]

OK, an interesting trio came up in this thread; Manning, Wayne, and Ochocinco. They are fairly well matched for this discussion, as they've played a number of years and have had few injury issues, except for Ochocinco in 2008. Since 2004, they've been in similar tiers, but Wayne has outscored Ochocinco five of six years, so he should be considered a better fantasy play, and Manning+Wayne outscored Manning+Ochocinco over that time period. The question is, did they do better scoring more than a baseline number of points?

I set the baseline as 20 points for non-PPR, 25 points for PPR. How many weeks did the combos fail to reach those targets? (Eliminating the 2008 season, and also games where Manning threw fewer than 10 passes because Indy had clinched).

Failures

2004: One each

2005: Wayne 4 times (3 PPR), Ochocinco once (2 PPR)

2006: Wayne once (PPR only), Ochocinco once

2007: Wayne twice (3 PPR), Ochocinco once

2009: Wayne 3 times, Ochocinco twice (3 PPR)

2010 YTD: Wayne twice, Ochocinco three times (2 PPR)

Totals: Wayne 11 times (12 in PPR), Ochocinco 8 times (9 in PPR)

Given these constraints, Wayne scored 989.7 points (1499.7 PPR), while Ochocinco scored 985.0 (1467 PPR). So the hypothesis is confirmed for this larger (but still relatively small) data set: Manning+Ochocinco scored the baseline number of points more often than Manning+Wayne did, even though Manning+Wayne scored more total points (barely).

This result should not be surprising, as Wayne's production is at least loosely coupled to Manning's, and Ochocinco's isn't. If you were faced with this decision every week, and you had chosen to start Ochocinco every week in the time period, you would have won three more games than the guy who chose to start Wayne.

I don't have an easy way to extend the data set without a lot of manual work, but I'm sure the finding would remain the same.

 

[Jersey35]

OK, an interesting trio came up in this thread; Manning, Wayne, and Ochocinco. They are fairly well matched for this discussion, as they've played a number of years and have had few injury issues, except for Ochocinco in 2008. Since 2004, they've been in similar tiers, but Wayne has outscored Ochocinco five of six years, so he should be considered a better fantasy play, and Manning+Wayne outscored Manning+Ochocinco over that time period. The question is, did they do better scoring more than a baseline number of points?

I set the baseline as 20 points for non-PPR, 25 points for PPR. How many weeks did the combos fail to reach those targets? (Eliminating the 2008 season, and also games where Manning threw fewer than 10 passes because Indy had clinched).

Failures

2004: One each

2005: Wayne 4 times (3 PPR), Ochocinco once (2 PPR)

2006: Wayne once (PPR only), Ochocinco once

2007: Wayne twice (3 PPR), Ochocinco once

2009: Wayne 3 times, Ochocinco twice (3 PPR)

2010 YTD: Wayne twice, Ochocinco three times (2 PPR)

Totals: Wayne 11 times (12 in PPR), Ochocinco 8 times (9 in PPR)

Given these constraints, Wayne scored 989.7 points (1499.7 PPR), while Ochocinco scored 985.0 (1467 PPR). So the hypothesis is confirmed for this larger (but still relatively small) data set: Manning+Ochocinco scored the baseline number of points more often than Manning+Wayne did, even though Manning+Wayne scored more total points (barely).

This result should not be surprising, as Wayne's production is at least loosely coupled to Manning's, and Ochocinco's isn't. If you were faced with this decision every week, and you had chosen to start Ochocinco every week in the time period, you would have won three more games than the guy who chose to start Wayne.

I don't have an easy way to extend the data set without a lot of manual work, but I'm sure the finding would remain the same.

Great work CB. This is just awesome.