Fantasy Football Discussion [Both Leagues]

[-Jrabbit]

I think QBs should be dominant (if any position is). That said, weeks like this always make it look more out of whack than it actually is. It's important to keep in mind how unusual games like Brady's today actually are.

FWIW, 50 yd/pt is going way too far IMO.

[SlowwHand]

It is what it is. I waited way late to take QB's. All have to have 2 WR and 2 RB and a WR/RB each week, before Byes strike.
Take a QB in the first round, but to me, it's a huge gamble.

[Zopperoni]

Brady Moss got me over 112

JR had the massive displeasure this week at FFZ AFC of facing Walrus who fielded Brady+Moss+Welker+T. Jones = 173.30 pts

[Ben Kenobi]

I waited way late to take QB's. All have to have 2 WR and 2 RB and a WR/RB each week, before Byes strike.

Take a QB in the first round, but to me, it's a huge gamble.

We've had ridiculous RB weeks as well. I think I'm one of the only teams without at 40.

Edit, me Slowwy and teh love wow are the only ones without a 40!

[Japher]

Thanks for your advice guys. I was able to maximize my scoring, though it may not be enough thanks to Forte laying a goose egg.

[snoopy369]

Looks like the only suspense is whether I can get 35 points off the Bolts tomorrow.

Congrats to Sloww, snoopy, Berz, koya, Zopper and EPW for their wins.

I'm not running to the winner's circle yet, 13+knowshon should hopefully be enough vs. sproles, but that man is capable of the occasional surprise...

[Ben Kenobi]

He needs about 24 points or so to be a threat. It's possible. I only need 27 points off Rivers, LT, Gates, Jackson and Kaeding.

[snoopy369]

I think QBs should be dominant (if any position is). That said, weeks like this always make it look more out of whack than it actually is. It's important to keep in mind how unusual games like Brady's today actually are.

FWIW, 50 yd/pt is going way too far IMO.

I agree. 50yd/pt is absurd. I think some people [ahem Berz] seem to forget that it's not the total # of points, it's the average differential, that matters. I'd be willing to bet that taking it to only TDs and 50yd/pt would magnify QB scoring differentials, because often what keeps mediocre QBs in the game is either their completions (for infrequent but accurate passers) and yardage (for frequent passers). Making it essentially TD-only not only makes it far from realistic [so a QB that has 400 yards passing and no TDs is only worth 8 points, while a QB with 100 yards and 2 short TDs is worth 10-14?] but it probably makes the QB position MORE important to draft, not less.

In general, I prefer this sort of scoring, where QB is relevant, more than default Y! scoring, where QBs are practically an afterthought (no comp/inc points). The only thing that I think I'd tweak would be to take TDs down to 4 points; otherwise leave it as is. Taking TDs down to 4 points would slightly lessen the effect of monster days like Brady's.

But then again... he did have a monster day. Why shouldn't we reward fantasy owners for that? And it obviously wasn't THAT imbalancing, as neither Brady owner won this week, and only one of them was facing a huge QB day ...

[snoopy369]

He needs about 24 points or so to be a threat. It's possible. I only need 27 points off Rivers, LT, Gates, Jackson and Kaeding.

You seem to assume that projections mean something He needs 15 points to be a threat. Knowshon isn't a lock to score any points... I hope he's worth 10-15, thus making your projection accurate, but I don't assume much.

[Ben Kenobi]

I have no problem with letting the Brady owners rack up the points for the TDs and winning outright. 6 TDs, you should win! There have been 41 games so far, and that has happened twice, and owners are 1-1. You aren't likely to run into an opponent who gets 75+ points off their QB.

[-Jrabbit]

Hey, I got 62 from Brees in a game yesterday and got waxed by 80 points. In our league, I'm gonna score about 140 but lose by 40, while Berz scored 136 and wins by 30. In the end, FF success is as much about the random occurence of your weekly in-league matchup than anything else.

That's what makes Brady, Peyton, Brees and a few select others worthy of a high pick. When they are on fire, they can really light it up. But remember, those days are the exception. Brees has had days of 24 and 16 pts already. Brady has been inconsistent (including a 20 and a 32) before exploding yesterday. But yes, when their offenses are on fire, they can indeed go very large.

Conversely, there will be days when mediocre QBs get big numbers as well, especially when they fall behind. It's actually more correlated to QB attempted passes than anything else, and that is dictated by the flow of the game. God help any QB whose coach's game plan is controlling the clock with the ground game.

I can see tweaking our settings a bit, but am more inclined toward an increase in yardage per point as opposed to the 4-pt TD pass.

[snoopy369]

I'd be happy to run my model for various different scoring settings (now or at the end of the year), if you like. It's trivial to adjust the scoring settings and show you what the distribution becomes [though at the moment, it's just 'in total' and not week by week, because that's how the stats were input]. Without being at my home PC at the moment, I think you'll find that the following are true:
* Points per yard is negatively correlated with QB differential [ie, more points per yard, or fewer yards per point, means overall less difference between QBs] and moderately postively correlated with QB performance effect on FFL game [ie, makes your QB's performance to expected more important]
* Points per TD is positively correlated with QB differential, and highly positively correlated with effect
* Completions/Incompletions magnitude is positively correlated with QB differential, though not highly correlated with points per TD (ie, 2/1 vs 1/0.5, keeping proportion constant) and is moderately positively correlated with effect
* Completions/Incompletions proportion is negatively correlated or not highly correlated with QB differential (ie, 1/0.5 vs 1/0.25) and is highly positively correlated with effect
* INT and SACK scoring is less correlated to QB differential than most other stats, though not uncorrelated (ie, because of the positive correlation of INTs and sacks with pass attempts, only the QBs who are actually really good [Manning/Brees/Brady, ie] do very well by this stat, while QBs who attempt a lot of passes [Hasselbeck, Palmer, etc.] and thus score highly, will tend to lose points here. Probably somewhat negatively correlated to effect.
So it's a matter of what do you want to accomplish:
* increase or decrease the effect a good pick at QB has on winning/losing fantasy games: increase/decrease the magnitude of the 'positive correlated to differential', or decrease/increase the 'negative'
* increase or decrease the effect of a QB's performance in general on winning/losing fantasy games [the week-to-week differential, basically]: increase or decrease 'highly' positively correlated stats [like TDs or Completions/Incompletions proportion]
* Change how the points are distributed to more closely/less closely reflect reality: increase/decrease the appropriate 'more realistic' or 'less realistic' statistics, while changing a stat in the reverse to compensate [for example, increase points per yard, which negatively correlates to differential and , and then decrease INT points, which would then positively correlate to differential, and also counteracts the QB effect correlation.

In general, if a change had to be made, I'd vote for decreasing the week-to-week differential while otherwise leaving things alone. That either means dropping TD scoring some [which limits the 'high' end of QB scoring] or increasing completion/incompletion magnitude [which would tend to make nights like Brady's less important, as he didn't complete all that many passes in the scheme of things]. Lowering the points per yard [ie, raising yards per point] would tend to have little effect on 'big days' like Brady's or Brees' but would actually hurt middle QBs on offenses like, say, Dallas or Atlanta, who don't tend to score very many passing TDs [last night aside ] but instead have a lot of passing that then later makes a rushing TD possible [ie, a lot of west coast offenses]. Lowering the value of those QBs doesn't seem to me that it would have a significantly positive impact on the distribution of QB scoring, and probably would have a negative one.

[Ben Kenobi]

Whatever decisions are made, can we stow them till the end of the season? I don't think it's fair to change the rules of the game halfway through.

[-Jrabbit]

No worries, Ben. That would never happen. (In fact, I don't think it can.)

I think the main question to be answered is this: Are we trying to flatten the distribution among QBs, or are we trying to reduce the value of QBs relative to the other positions? I'm not convinced either is necessary, but it might be interesting to run some scenarios after the season.

Right now, QB and defense (aggregate) are the most powerful forces in our settings, followed by RBs and WRs, then Kickers and TEs. I think.

We arrived at these settings through a fairly random combination of trying things over the years. The power of the QB relative to other offensive positions was a deliberate reaction to the RB-centric settings of the original "standard Yahoo league." We went to PPR, then increased QB yardage points. We dialed back the QBs a bit by incorporating the subtraction for incomplete passes (and now, sacks, although that's quite minor).

So it's been a system of evolution rather than metrics. It seems like every year, there are one or two weeks like this (with multiple massive QB points), followed by the inevitable whining of the have-nots. Generally, all is forgotten by the end of the year, as it's pretty rare for any QB to keep up a pace like that for an extended period (Brady's 54?-TD year being the main exception). Even today, the gold standard for a great QB season is 30 TDs and maybe 3500 yards -- something few attain each year. Yet that translates to only around 35 points/week (sorry, very rough guesstimating here, 20/34-200-2).

What gets lost in the tumult is the simple fact that one bad day by a QB during the playoffs is a killer. At that time of year, big days are less likely (due to weather etc). It can really pay to have a reliable 35-pt performer (with some reliable W/Rs) for those big games.

I know I'm rambling a bit here, but I'm taking a break (from writing, ironically enough) and thought I'd let my mind wander. OK, back to billable hours...

[Ben Kenobi]

The power of the QB relative to other offensive positions was a deliberate reaction to the RB-centric settings of the original "standard Yahoo league." We went to PPR, then increased QB yardage points. We dialed back the QBs a bit by incorporating the subtraction for incomplete passes (and now, sacks, although that's quite minor).

Thanks JR!

I don't mind getting thrashed by a solid QB (Manning destroyed me outright in week 3). I like the settings of this league quite a bit. The only suggestion I'd have to add is the minus for a safety conceded, and bonus points if a defensive player forces a safety. It doesn't happen all that often.