Diversification for your fantasy football team (1 Viewer)

[sheastadium]

This topic is in regards to the QB/WR combo that a lot of fantasy owners enjoy. There was a topic on this earlier in the forum, but it focused more on which combo was good and bad, instead of the theory behind using combos. I have a lot of questions with this model, but I wanted to throw it out there and see what you guys thought.

As I was reading the QB/WR thread, I was wondering if there was a way to quantify how viable a QB/WR combo is by using correlations and return/risk metrics. The first thing that came to mind is "diversification" to lessen the overall risk of your fantasy team. In finance, they call this mean-variance optimization, coined by Harry Markowitz who won a Nobel Prize for it.

Mean-variance optimization, in fantasy football terms, basically says, you prefer players with low correlations for fantasy points so you're more diversified. High correlation players would include QB/WR combos because they would share TDs/yards/catches most of the time and hence score a lot of fantasy points at the same time. Very highly correlated. This is a bad thing according to the model because it heightens your risk, risk defined as the volatility of your "portfolio" of players. So to lessen risk, you actually want to pick players that have little correlation to each other but still have high expected return and low volatility. This has some important implications during your draft.

For example, assume you drafted Drew Brees and now was thinking about a wide receiver. Also assume, for the sake of this post, that you expect Terrell Owens and Marques Colston to score the same number of points and they had the same "volatility". A lot of owners in this situation would say "Fantastic Colston/Brees combo that's awesome!". But in theory, it's better to pick TO.

What do you guys think?

 

[sheastadium]

Sorry one extra post, wanted to enable email notification for this topic to make sure I see all your replies.

 

[Trig2U]

Not to say that I disagree with the concept as you're describing it, but I'm not sure a fantasy football team and a stock portfolio are analogous enough to make a direct comparison.

It's generally smart to diversify stocks because the investor wants to conserve small gains accumulated over the long term, which in theory will take years to realize. That way a sudden loss in a particular sector of the economy won't destroy all of the slowly accrued gains.

Also you're not competing with other investors, at least not in a direct, head-to-head way. In theory, many investors could follow the same strategy and they could all make money.

In fantasy football on the other hand there's a relatively brief and predetermined field of action (the NFL football season, weeks 1-16). And each week is a contest in and of itself, with a winner and a loser. A fantasy team that can win two out of three times over the course of 16 weeks has a great chance to go all the way, even if the games lost were by very large margins.

To put it another way, if we had a 16 week contest to see who could choose stocks that gained the most percentage points on the NY Stock Exchange, and if the contest was determined on a week by week basis in head-to-head competition with other investors; the guy who put together a diverse, well-balanced portfolio would not necessarily come out on top two-thirds of the time. Riskier strategies might have an advantage in such a short contest, especially where there is only one ultimate winner and a bunch of also-rans.

Acting "conservatively" in fantasy football might be appropriate sometimes, but depending on what your team's situation is and who your opponents are, it might not.

So, if say every season Team A drafted a QB/WR NFL hook-up and Team B drafted QBs and WR's from different NFL teams, I imagine Team B might have some advantage over the long run (like say 10 or 15 years). But if in a given year the hook-up happened to be something like Steve Young to Terrell Owens in 1998, or Brett Favre to Antonio Freeman, or Kurt Warner to Larry Fitzgerald last year, or a number of other particular examples I could think of, Team A would destroy Team B more often than not.

Not that I'm an economist or anything, but that's my two cents, anyway...

 

[Hagger]

Think of having the Brady to Moss combo of '07. I'm guessing those owners who had that combo dominated their leagues. Of course, the trick is predicting which combo will have a year like that. Last year I banked on Derek Anderson and Braylon Edwards. Yeah.

I actually like the strategy, despite getting burnt by it last year. I also think certain combos are relatively safe: Brady-Moss, Warner-Fitzgerald, Manning-Wayne are probably the three safest. I also like Brees-Colston, Rodgers-Jennings, Ryan-White but am less certain about them.

I understand the idea of not putting all of your eggs in one basket but I also think it's a risk that can really pay off if chosen correctly. Of course, just look at me last year so...

 

[Doug Drinen]

This is one of my favorite topics.

Hard to believe it's been 11 years since this article was written.

Here is one from last year that discusses diversification as it pertains to QB/RB, RB/WR, and WR/WR combos.

 

[Banger]

To me it makes sense if both guys are good enough to stand on their own. I don't necessarily go looking for a QB/WR combo but I've had them in the past where they've very successful...Warner/Bruce, Warner/Holt, Cutler/Marshall, etc. I really just think it depends on the players...I'd take Wayne/Manning, Brady/Moss, without hesitation.

 

[Eminence]

I'm a fan of:

Roddy White / Michael Turner - Only 4 weeks with under 20 combined points.

Scoring:

1 point per 10 yards.

6 points per TD.

Week 1:

White - 2 Receptions / 54 Yards

Turner - 220 yards / 2 TDs

Standard Scoring: 39 Points

Week 2:

White - 4 Receptions / 59 Yards

Turner - 42 Yards / No TDs

Standard Scoring: 9 Points

Week 3:

White - 5 Receptions / 119 Yards / 1 TD

Turner - 104 Yards / 3 TDs

Standard Scoring: 45 Points

Week 4:

White - 7 Receptions / 90 Yards

Turner - 56 Yards / 0 TDs

Standard Scoring: 14 Points

Week 5:

White - 8 Receptions / 132 Yards / 1 TD

Turner - 121 Yards / 1 TD

Standard Scoring: 37 Points

Week 6:

White - 9 Receptions / 112 Yards / 1 TD

Turner - 54 Yards / 0 TD

Standard Scoring: 22 Points

Week 7:

White - N/A

Turner - N/A

Standard Scoring: 0 Points

Week 8:

White - 8 Receptions / 113 Yards / 2 TDs

Turner - 58 Yards / 0 TDs

Standard Scoring: 28 Points

Week 9:

White - 5 Receptions / 54 Yards / 0 TDs

Turner - 139 Yards / 0 TDs

Standard Scoring: 18 Points

Week 10:

White - 5 Receptions / 68 Yards / 1 TD

Turner - 96 Yards / 1 TD

Standard Scoring: 27 Points

Week 11:

White - 5 Receptions / 102 Yards / 0 TD

Turner - 81 Yards / 2 TD

Standard Scoring: 30 Points

Week 12:

White - 4 Receptions / 70 Yards

Turner - 117 Yards / 4 TDs

Standard Scoring: 42 Points

Week 13:

White - 6 Receptions / 112 Yards / 0 TD

Turner - 120 Yards / 0 TD

Standard Scoring: 23 Points

Week 14:

White - 10 Receptions / 164 Yards / 0 TD

Turner - 61 Yards / 1 TD

Standard Scoring: 28 Points

Week 15:

White - 4 Receptions / 61 Yards / 0 TD

Turner - 182 Yards / 1 TD

Standard Scoring: 30 Points

Week 16:

White - 3 Receptions / 24 Yards / 0 TD

Turner - 70 Yards / 1 TD

Standard Scoring: 15 Points

Week 17:

White - 3 Receptions / 48 Yards / 1 TD

Turner - 208 Yards / 1 TD

Standard Scoring: 36 Points

-

Highest Week: 45 Points

Lowest Week: 9 Points

Average: 27 Points

Lowest to Highest:

9, 14, 15, 18, 22, 23, 27, 28, 28, 30, 30, 36, 37, 39, 42, 45.

Stats speak for themselves.

 

[Banger]

I'm a fan of:

Roddy White / Michael Turner - Only 4 weeks with under 20 combined points.

Scoring:

1 point per 10 yards.

6 points per TD.

Week 1:

White - 2 Receptions / 54 Yards

Turner - 220 yards / 2 TDs

Standard Scoring: 39 Points

Week 2:

White - 4 Receptions / 59 Yards

Turner - 42 Yards / No TDs

Standard Scoring: 9 Points

Week 3:

White - 5 Receptions / 119 Yards / 1 TD

Turner - 104 Yards / 3 TDs

Standard Scoring: 45 Points

Week 4:

White - 7 Receptions / 90 Yards

Turner - 56 Yards / 0 TDs

Standard Scoring: 14 Points

Week 5:

White - 8 Receptions / 132 Yards / 1 TD

Turner - 121 Yards / 1 TD

Standard Scoring: 37 Points

Week 6:

White - 9 Receptions / 112 Yards / 1 TD

Turner - 54 Yards / 0 TD

Standard Scoring: 22 Points

Week 7:

White - N/A

Turner - N/A

Standard Scoring: 0 Points

Week 8:

White - 8 Receptions / 113 Yards / 2 TDs

Turner - 58 Yards / 0 TDs

Standard Scoring: 28 Points

Week 9:

White - 5 Receptions / 54 Yards / 0 TDs

Turner - 139 Yards / 0 TDs

Standard Scoring: 18 Points

Week 10:

White - 5 Receptions / 68 Yards / 1 TD

Turner - 96 Yards / 1 TD

Standard Scoring: 27 Points

Week 11:

White - 5 Receptions / 102 Yards / 0 TD

Turner - 81 Yards / 2 TD

Standard Scoring: 30 Points

Week 12:

White - 4 Receptions / 70 Yards

Turner - 117 Yards / 4 TDs

Standard Scoring: 42 Points

Week 13:

White - 6 Receptions / 112 Yards / 0 TD

Turner - 120 Yards / 0 TD

Standard Scoring: 23 Points

Week 14:

White - 10 Receptions / 164 Yards / 0 TD

Turner - 61 Yards / 1 TD

Standard Scoring: 28 Points

Week 15:

White - 4 Receptions / 61 Yards / 0 TD

Turner - 182 Yards / 1 TD

Standard Scoring: 30 Points

Week 16:

White - 3 Receptions / 24 Yards / 0 TD

Turner - 70 Yards / 1 TD

Standard Scoring: 15 Points

Week 17:

White - 3 Receptions / 48 Yards / 1 TD

Turner - 208 Yards / 1 TD

Standard Scoring: 36 Points

-

Highest Week: 45 Points

Lowest Week: 9 Points

Average: 27 Points

Lowest to Highest:

9, 14, 15, 18, 22, 23, 27, 28, 28, 30, 30, 36, 37, 39, 42, 45.

Stats speak for themselves.

that's interesting but the addition of Gonzo is sure to water down the concentration on those two.

 

[Eminence]

that's interesting but the addition of Gonzo is sure to water down the concentration on those two.

That's what I was thinking. But:-Ryan will be more familiar with the offense, and put up bigger numbers. White = Go to Guy.

-White will see less double coverage with the addition of Tony Gonzalez.

-The Chiefs had an obscurely low 9 Rushing TD's last season. (3 from Thigpen himself.) The team passed way more than necessary. Giving Tony G. more rezone attempts.

-In a 12 Year career, Tony G has only had 4 seasons with 9 or more Touchdowns. In his 13th year, he's regressing.

-Tony Gonzalez is one hell of a blocker.

-With two recieving threats, I can see fewer 8-man fronts for Turner. But also less goal line opportunities. If this allows him to bust off a few more long gainers, his numbers shouldn't suffer too badly.

-9 of Tony's 10 TDs from last season came from within the Redzone. But we all know how much of a beast Turner is there.

Breakdown:

Whereas Tony G. will spread the yardage and TD's away from Turner and White, I believe that Matt Ryan will put up far better numbers this season. 16 TDs / 11 Ints / 3,440 Yards last season. I expect about 23 / 12 Ints / 3,800 Yards. Which would still allow White to put up about 7+ TDs, and most likely increase his yardage from the prior year. Tony G will get his own. But I HIGHLY doubt he sees 10 TDs like last season. That Chiefs team was always playing from behind.

Even if Turner drops a few TDs, I expect his overall yardage to increase. That was his first year in the system, dontcha know.

In short, I'm very high on the Atlanta offense next season.

 

[Jayrod]

I'm a fan of:

Roddy White / Michael Turner - Only 4 weeks with under 20 combined points.

I believe that would be a RB/WR combo, which has a much lower correlation than WR/QB which was the discussion up to this point.

 

[Eminence]

I'm a fan of:

Roddy White / Michael Turner - Only 4 weeks with under 20 combined points.

I believe that would be a RB/WR combo, which has a much lower correlation than WR/QB which was the discussion up to this point.

Oops.

 

[CalBear]

You're begging the question--what should fantasy football owners' attitudes be towards risk? You make the assumption that fantasy football owners should be risk-averse, but I don't think that's clear. In a stock portfolio, you can always compare your potential return against the return on a "safe" investment--treasury bills or some such--the idea being that you could always put your money in something earning 3% and take on no risk. You can measure the risk of your investment against the risk of a T-bill, and compare the returns, and figure out whether you're being properly compensated for your risk.

But that's not the way fantasy football works. First of all, you're not starting at zero; you're starting down $100 (or whatever your entry fees are). Second, your payout is zero unless you finish at the top of all money managers (in the top 10-30%). If you take a completely risk-averse strategy, and guarantee that your team will finish at 7-6, you lose. In fact, all scenarios in which you miss the playoffs have exactly equal value--it's no better to finish at 7-6 than it is to finish at 0-13. It's like if you were managing a fund where your contract stipulates that you have to get 10% annual returns, or you lose all the money. In such a scenario, taking on risk is a requirement; the "safe" play is a guaranteed loser.

Think about a fantasy league where only the Super Bowl winner gets paid. The return on finishing first is 1200%; the return on everything else is -100%. In this league, any move which increases your chance of winning is the correct move, even if it increases your chance of finishing 0-13. If it raises your chance of winning from 10% to 11%, and raises your chance of finishing last from 10% to 50%, you should still do it.

 

[VonRomig]

Just grab the best player as you see it, at the draft spot you're on, and ignore the concept of combos. A top QB will get you the same points whether you have his WR or not. While that same WR, a top pick himself, will still get the points he will get regardless of the fact you have his QB or not.

As for the idea that injuries would have a double penalty for a team who has a QB/WR combo, it's not much more likely than if the players were on seperate teams. Afterall, one of the reasons the two players are considered studs is due to their durablity. So a long lasting injury is rare. (Brady being one of those rare times)

I would not pass on a player just because his selection would give me a combo, and I would not try to get a combo. I just take the best player at that spot...

As for those who would say that getting Brady/Moss of '07 is a reason to like this concept, I say... Every year there are good combos, even several combos per season, but in most cases you don't have much chance to get both. Unless you actually aim for it and/or reach to do it. This is because the two players to make the combo are both ranked so high that for one team to be able to draft them both would be unlikely. The counterpart being taken by someone else before you get to grab him. The Brady/Moss combo was just so huge that it overshadows the reality of it's worthiness.

Combo selections are not worth going for IMHO, as you never can tell what combo will be worth the change in draft strategies that forces you to bypass other players.

Von

 

[jackdubl]

Injuries aren't more likely, no, but if one occurs it now affects two players in your lineup instead of one. That's the double penalty.

 

[big0mar]

Diversification in fantasy football is a good idea if you want to finish 5th.

 

[Sabertooth]

I tend to diversify across a few leagues. I don't like having the same players on both my fantasy teams.

 

[Mystery Achiever]

Just grab the best player as you see it, at the draft spot you're on, and ignore the concept of combos. A top QB will get you the same points whether you have his WR or not. While that same WR, a top pick himself, will still get the points he will get regardless of the fact you have his QB or not.

This has always been my view. Plus, as has been said before, you have double the upside in a good game wiped out by the chance of double the downside with a bad game.I also sometimes see parallels with the market. But, consider this. In a portfoliof 20-30 stocks (ok, slightly larger than ff roster), there is nothing wrong with having two tech stocks or two health care stocks., esp. if it is a hot sector.

 

[5-ish Finkle]

I thought this was gonna be a post about drafting more Asian players onto fantasy teams.

I'm disappointed I was wrong.

"Obey the Hiroki Nakamura, plebs!!"

**End Transmision**

 

[Ridgelake]

It also matters about your league structure. You may want to diversify if your league is based upon annual totals.

Head-to-head leagues are a bit different. Each week becomes a different head-to-head matchup against a perceived stronger or weaker or equal opponent. If I believe I have the stronger team, I want to diversify my risk and be more likely that the depth of my team will beat my opponent. Conversely, if I feel I have a weaker team than my opponent, I'm more likely to "swing for the fences" and increase my expected volatilty, and thus chance to win.

 

[tytyty]

You're begging the question--what should fantasy football owners' attitudes be towards risk? You make the assumption that fantasy football owners should be risk-averse, but I don't think that's clear. In a stock portfolio, you can always compare your potential return against the return on a "safe" investment--treasury bills or some such--the idea being that you could always put your money in something earning 3% and take on no risk. You can measure the risk of your investment against the risk of a T-bill, and compare the returns, and figure out whether you're being properly compensated for your risk.

But that's not the way fantasy football works. First of all, you're not starting at zero; you're starting down $100 (or whatever your entry fees are). Second, your payout is zero unless you finish at the top of all money managers (in the top 10-30%). If you take a completely risk-averse strategy, and guarantee that your team will finish at 7-6, you lose. In fact, all scenarios in which you miss the playoffs have exactly equal value--it's no better to finish at 7-6 than it is to finish at 0-13. It's like if you were managing a fund where your contract stipulates that you have to get 10% annual returns, or you lose all the money. In such a scenario, taking on risk is a requirement; the "safe" play is a guaranteed loser.

Think about a fantasy league where only the Super Bowl winner gets paid. The return on finishing first is 1200%; the return on everything else is -100%. In this league, any move which increases your chance of winning is the correct move, even if it increases your chance of finishing 0-13. If it raises your chance of winning from 10% to 11%, and raises your chance of finishing last from 10% to 50%, you should still do it.

man i dont want to play anymore!!!!

 

[Ridgelake]

This is one of my favorite topics.

Hard to believe it's been 11 years since this article was written.

Here is one from last year that discusses diversification as it pertains to QB/RB, RB/WR, and WR/WR combos.

Just read them both. OUTSTANDING analysis!Thank you!!