Tell you what, I guarantee Hasselbeck to be a top 5 QB week 7, if not i will delete my account here and never come back, however if your numbers for him are wrong you do the same, I mean since your so confident. How about it? Deal jjiggler?
Makaveli wrote:Tell you what, I guarantee Hasselbeck to be a top 5 QB week 7, if not i will delete my account here and never come back, however if your numbers for him are wrong you do the same, I mean since your so confident. How about it? Deal jjiggler?
That's a very permanent bet. I never said I was confident in my rankings, I was just explaining why my formula said what it said.
I honestly think that Hasselbeck will not be a Top-5 QB this week, but wouldn't bet my account on it.
Pick a paperbag avatar for me to sport for a month if I'm wrong. I'll pick one for you.
I got to agree with jigglers, or at least just put words in his mouth.
The point of coming up with a formula is to indicate what prior events would give us reason to believe about the future. It's not always going to be right, but if you can get the accuracy of the picks/formula to be consistent, then it really should be making you realize some things you otherwise might not.
One example is Hass: It seems like he should tear DAL apart, at least in my mind. But the formula spit this out and I'm looking back at the raw numbers: Hasselbeck has thrown only 8 TDs in 6 games (0 rushing TDs, too) while the Cowboys have yielded 8 passing TDs in 6 games. Granted, it's a nutty offense in Seattle (SA could score 4 TDs just as easily as he could score 0) ... but if you're just looking at numbers so far, this one's easy to expect 1.33 TDs from Hass on average against the Cowboys (there are other weighs on this in the formula, but I'll just stick with the obvious here). It's fun to make predictions and just say Hass will get 3 TDs. But there's an adage in gambling (and statistics) - don't bet against a streak just cause you think somebody's "due."
I'd rather have Hass than McCown, obviously, but a formula like this is just screaming that McCown is the better play. The formula will be wrong; but if it's only wrong about that kind of a recommendation less than 15% of the time over the long haul, it seems helpful.
Also to PLindsey - smackthefirst beat me to it, but I've been thinking about this too and I've also thought that the biggest thing that makes an offensive position more difficult to predict than defenses is that anything can happen around the goal-line - the objective here is to find a solid combination of stats and proper weighting that makes more accurate predictions than just looking at the best QBs and worst defenses against the pass. Who knows how possible it is. I still need to figure out what I even think "success" is for this algorithm. I'll start with better than Hector and Victor (or a monkey with a dart, same difference) and move up from there.
... in my opinion, of course - but that should always go without saying.
www.bigfellas.net (home of the golf rap "On The Green")
Are your inputs the fantasy scores by the QBs, DEFs, etc? or the actual yards/TDs/TOs? I think the latter would be a farm more effective way to do it, but I've never tried something like this.
Secondly, I know you're including the QBs history and the DEF's history, but are you including the history of the QBs the DEFs have played against? It would probably be a good deal more work, but it would certainly make your predictions more accurate.
Anyhow, it seems like it should have a reasonable shot of working. GL and HF with it.
2) Passing strength of Def's prior opponents, as well as Pass D's strength of QB's prior opponents hasn't been incorporated yet, but that's the next thing to be added. I've looked at this historically as well as I can and it seems to have a good correlation so I'm planning on including it. Right now I'm just trying to figure the best way to do it.
... in my opinion, of course - but that should always go without saying.
www.bigfellas.net (home of the golf rap "On The Green")
Are your inputs the fantasy scores by the QBs, DEFs, etc? or the actual yards/TDs/TOs? I think the latter would be a farm more effective way to do it, but I've never tried something like this.
Secondly, I know you're including the QBs history and the DEF's history, but are you including the history of the QBs the DEFs have played against? It would probably be a good deal more work, but it would certainly make your predictions more accurate.
Anyhow, it seems like it should have a reasonable shot of working. GL and HF with it.
The formula I'm using calculates strength of schedule modifiers, so if the QB has played against easy Defenses, his stats are toned down, and some goes for Defenses.
I calculate completions (for those with completion leagues), yards, TDs, and INTs, and then tally the points afterwards.
This update doesn't change much, just uses this week's injury reports instead of last week's.
Drew Bledsoe vs. SEA 18.31
Donovan McNabb vs. SDG 14.83
Josh McCown (?) vs. TEN 14.37
Eli Manning vs. DEN 13.19
Carson Palmer vs. PIT 13.01
Mark Brunell vs. SFO 12.82
Gus Frerotte vs. KAN 12.28
Kurt Warner (?) vs. TEN 11.55
Matt Hasselbeck vs. DAL 10.69
Brett Favre vs. MIN 10.32
Peyton Manning vs. HOU 10.03
Jake Plummer vs. NYG 9.07
Aaron Brooks vs. STL 8.89
Kelly Holcomb vs. OAK 8.61
Daunte Culpepper vs. GNB 7.96
Trent Dilfer vs. DET 6.60
Jamie Martin vs. NOR 6.19
Drew Brees vs. PHI 6.04
Vinny Testaverde vs. ATL 5.64
Steve McNair vs. ARI 5.52
Kerry Collins vs. BUF 5.10
Trent Green vs. MIA 4.31
Ben Roethlisberger vs. CIN 4.20
Michael Vick vs. NYJ 4.01
Anthony Wright vs. CHI 1.71
Joey Harrington vs. CLE 1.48
David Carr vs. IND 1.17
Kyle Orton vs. BAL -0.05
Alex Smith vs. WAS -4.90
... in my opinion, of course - but that should always go without saying.
www.bigfellas.net (home of the golf rap "On The Green")
Cafe Home
Fantasy Baseball
Fantasy Basketball
Fantasy Hockey
by Makaveli » Wed Oct 19, 2005 12:49 am 
