What's new
Fantasy Football - Footballguys Forums

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

2006 FBG Preseason Projections (1 Viewer)

I can do some follow-up analyses, but I just looked at how the FBGs ranked their top-20 QBs. Using Yardage as my measure (and, I could do TDs or other variables), the correlation was only .20 (and not statistically significant).Another test (paired sample t-test) suggested that their predictions were way off (t=2.23, p<.05).I can look at some other data in a bit (or tomorrow). But, at least in terms of how the top-20 was predicted to do on Passing Yardage, the results ain't good. Not suggest that any other site would've done better (I'm guessing they'd be just as unreliable). That's just the reality of making these kinds of predictions, imo.

Rk Player. . . . . . . . ..Predicted. . .Actual1 Peyton Manning. . . . 4015. . .43972 Carson Palmer. . . . .3636. . .40363 Tom Brady. . . . . . . 3798. . .35334 Donovan McNabb. . .3640. . .26125 Daunte Culpepper. . 3566. . .9296 Matt Hasselbeck. . . .3504. . .24427 Jake Delhomme. . . .3492. . .28058 Marc Bulger. . . . . . .3685. . .43019 Michael Vick. . . . . . .2477. . .247410 Eli Manning. . . . . . . 3528. . .324411 Trent Green. . . . . . .3548. . .134212 Jake Plummer. . . . . 3240. . .199413 Kurt Warner. . . . . . .3586. . .137714 Chris Simms. . . . . . .3312. . .58515 Brett Favre. . . . . . . .3620. . .388516 Steve McNair. . . . . . 3232. . .305017 Jon Kitna. . . . . . . . . 3163. . .420818 Philip Rivers. . . . . . .3312. . .338819 Mark Brunell. . . . . . 3112. . .178920 Drew Brees. . . . . . . 3197. . .4424
Click to expand...
 
Last edited by a moderator:
I think this is something FBG should do and provide on its own. I think it will show how good the FBG folks are. And if there are any weaknesses, it will motivate improvements. At the least, I'd like to see them participate in Rotosource.com's rankings.

I know one site specifically changed its process for projections for a position because it did poorly in the Rotosource.com rankings. Kudos to that site for acknowledging a weakness and working to improve it.

Now, maybe FBG is doing the same, just not making the results public.

 
gheemony said:
I think this is something FBG should do and provide on its own. I think it will show how good the FBG folks are. And if there are any weaknesses, it will motivate improvements. At the least, I'd like to see them participate in Rotosource.com's rankings.I know one site specifically changed its process for projections for a position because it did poorly in the Rotosource.com rankings. Kudos to that site for acknowledging a weakness and working to improve it. Now, maybe FBG is doing the same, just not making the results public.
Click to expand...
I think there are two parts to this. One, I think it would be interesting (alarming, even) to see how bad these projections are. Again, I don't think that's a FBG issue. I think it's just a losing proposition to think that these projections actually will look like reality at the end of the day.But, the second thing (and, this would take some serious work) is how FBG projections look compared to other sites that do their projections.Still, there seems to be very little variability across sites. We're all lemmings following lemmings. Kurt Warner was predicted to do pretty well by most sites last year, and Drew Brees was not. This was the case across the board. So, I doubt we could read a whole lot into what little variability that might actually exist.
 
One more...RB YARDAGECorrelation: .355 (not significant)Paired Sample t-test: 1.13, p=.273So, one could say the correlation was closer. The t-test that compared means didn't come out significant (i.e., reject the null hypothesis that the predictions and actual outcomes were different), which is what we want. But, still...it's not terribly uplifting, either.

RNK NAME Predicted Actual1 Larry Johnson 1558 17892 LaDainian Tomlinson 1430 18153 Shaun Alexander 1610 9064 Rudi Johnson 1333 13085 Ronnie Brown 1283 10056 Steven Jackson 1262 15287 Willie Parker 1283 14948 LaMont Jordan 1117 4349 Cadillac Williams 1355 79810 Willis McGahee 1240 99011 Edgerrin James 1174 115912 Clinton Portis 1183 52313 Brian Westbrook 817 121714 Reggie Bush 917 55815 Warrick Dunn 1115 114016 Reuben Droughns 1146 75817 DeShaun Foster 943 89718 Kevin Jones 943 68919 Julius Jones 1060 108420 Frank Gore 1008 1695
Click to expand...
 
cobalt_27 said:
I can do some follow-up analyses, but I just looked at how the FBGs ranked their top-20 QBs. Using Yardage as my measure (and, I could do TDs or other variables), the correlation was only .20 (and not statistically significant).Another test (paired sample t-test) suggested that their predictions were way off (t=2.23, p<.05).I can look at some other data in a bit (or tomorrow). But, at least in terms of how the top-20 was predicted to do on Passing Yardage, the results ain't good. Not suggest that any other site would've done better (I'm guessing they'd be just as unreliable). That's just the reality of making these kinds of predictions, imo.

Rk Player. . . . . . . . ..Predicted. . .Actual1 Peyton Manning. . . . 4015. . .43972 Carson Palmer. . . . .3636. . .40363 Tom Brady. . . . . . . 3798. . .35334 Donovan McNabb. . .3640. . .26125 Daunte Culpepper. . 3566. . .9296 Matt Hasselbeck. . . .3504. . .24427 Jake Delhomme. . . .3492. . .28058 Marc Bulger. . . . . . .3685. . .43019 Michael Vick. . . . . . .2477. . .247410 Eli Manning. . . . . . . 3528. . .324411 Trent Green. . . . . . .3548. . .134212 Jake Plummer. . . . . 3240. . .199413 Kurt Warner. . . . . . .3586. . .137714 Chris Simms. . . . . . .3312. . .58515 Brett Favre. . . . . . . .3620. . .388516 Steve McNair. . . . . . 3232. . .305017 Jon Kitna. . . . . . . . . 3163. . .420818 Philip Rivers. . . . . . .3312. . .338819 Mark Brunell. . . . . . 3112. . .178920 Drew Brees. . . . . . . 3197. . .4424
Click to expand...
Click to expand...
You really need to work in a per game basis for this to have meaning other than that FBGs can't predict injuries any better than anyone else.
 
cobalt_27 said:
I can do some follow-up analyses, but I just looked at how the FBGs ranked their top-20 QBs. Using Yardage as my measure (and, I could do TDs or other variables), the correlation was only .20 (and not statistically significant).Another test (paired sample t-test) suggested that their predictions were way off (t=2.23, p<.05).I can look at some other data in a bit (or tomorrow). But, at least in terms of how the top-20 was predicted to do on Passing Yardage, the results ain't good. Not suggest that any other site would've done better (I'm guessing they'd be just as unreliable). That's just the reality of making these kinds of predictions, imo.

Rk Player. . . . . . . . ..Predicted. . .Actual1 Peyton Manning. . . . 4015. . .4397......... +2 Carson Palmer. . . . .3636. . .4036 .........+3 Tom Brady. . . . . . . 3798. . .3533 .......... off slightly4 Donovan McNabb. . .3640. . .2612 ........... injury5 Daunte Culpepper. . 3566. . .929 ............ injury 6 Matt Hasselbeck. . . .3504. . .2442.............injury7 Jake Delhomme. . . .3492. . .2805.............injury + incompetence8 Marc Bulger. . . . . . .3685. . .4301............+9 Michael Vick. . . . . . .2477. . .2474...........dead on10 Eli Manning. . . . . . . 3528. . .3244.............off slightly11 Trent Green. . . . . . .3548. . .1342.............injury12 Jake Plummer. . . . . 3240. . .1994.............lost job (foreshadowed)13 Kurt Warner. . . . . . .3586. . .1377.............lost job (expected at some point)14 Chris Simms. . . . . . .3312. . .585..............injury15 Brett Favre. . . . . . . .3620. . .3885............+16 Steve McNair. . . . . . 3232. . .3050.............off slightly17 Jon Kitna. . . . . . . . . 3163. . .4208.............++18 Philip Rivers. . . . . . .3312. . .3388............. dead on19 Mark Brunell. . . . . . 3112. . .1789..............lost job20 Drew Brees. . . . . . . 3197. . .4424.............+++
Click to expand...
Click to expand...
Actually if you remove the results of the injured - McNabb, Culpepper, Hasselbeck, Green, and Simms this list is not so bad. The plusses out weigh the minuses especially when you factor in the known possible changes in Arizona and Denver.
 
cobalt_27 said:
I can do some follow-up analyses, but I just looked at how the FBGs ranked their top-20 QBs. Using Yardage as my measure (and, I could do TDs or other variables), the correlation was only .20 (and not statistically significant).Another test (paired sample t-test) suggested that their predictions were way off (t=2.23, p<.05).I can look at some other data in a bit (or tomorrow). But, at least in terms of how the top-20 was predicted to do on Passing Yardage, the results ain't good. Not suggest that any other site would've done better (I'm guessing they'd be just as unreliable). That's just the reality of making these kinds of predictions, imo.

Rk Player. . . . . . . . ..Predicted. . .Actual1 Peyton Manning. . . . 4015. . .43972 Carson Palmer. . . . .3636. . .40363 Tom Brady. . . . . . . 3798. . .35334 Donovan McNabb. . .3640. . .26125 Daunte Culpepper. . 3566. . .9296 Matt Hasselbeck. . . .3504. . .24427 Jake Delhomme. . . .3492. . .28058 Marc Bulger. . . . . . .3685. . .43019 Michael Vick. . . . . . .2477. . .247410 Eli Manning. . . . . . . 3528. . .324411 Trent Green. . . . . . .3548. . .134212 Jake Plummer. . . . . 3240. . .199413 Kurt Warner. . . . . . .3586. . .137714 Chris Simms. . . . . . .3312. . .58515 Brett Favre. . . . . . . .3620. . .388516 Steve McNair. . . . . . 3232. . .305017 Jon Kitna. . . . . . . . . 3163. . .420818 Philip Rivers. . . . . . .3312. . .338819 Mark Brunell. . . . . . 3112. . .178920 Drew Brees. . . . . . . 3197. . .4424
Click to expand...
Click to expand...
Per game averages and team totals would be much "fairer" measures IMHO. That being said, I honestly have no idea how my or others projections stack up in those areas relative to whatever baseline you're comparing them to :banned:
 
Jason Wood said:
cobalt_27 said:
I can do some follow-up analyses, but I just looked at how the FBGs ranked their top-20 QBs. Using Yardage as my measure (and, I could do TDs or other variables), the correlation was only .20 (and not statistically significant).Another test (paired sample t-test) suggested that their predictions were way off (t=2.23, p<.05).I can look at some other data in a bit (or tomorrow). But, at least in terms of how the top-20 was predicted to do on Passing Yardage, the results ain't good. Not suggest that any other site would've done better (I'm guessing they'd be just as unreliable). That's just the reality of making these kinds of predictions, imo.

Rk Player. . . . . . . . ..Predicted. . .Actual1 Peyton Manning. . . . 4015. . .43972 Carson Palmer. . . . .3636. . .40363 Tom Brady. . . . . . . 3798. . .35334 Donovan McNabb. . .3640. . .26125 Daunte Culpepper. . 3566. . .9296 Matt Hasselbeck. . . .3504. . .24427 Jake Delhomme. . . .3492. . .28058 Marc Bulger. . . . . . .3685. . .43019 Michael Vick. . . . . . .2477. . .247410 Eli Manning. . . . . . . 3528. . .324411 Trent Green. . . . . . .3548. . .134212 Jake Plummer. . . . . 3240. . .199413 Kurt Warner. . . . . . .3586. . .137714 Chris Simms. . . . . . .3312. . .58515 Brett Favre. . . . . . . .3620. . .388516 Steve McNair. . . . . . 3232. . .305017 Jon Kitna. . . . . . . . . 3163. . .420818 Philip Rivers. . . . . . .3312. . .338819 Mark Brunell. . . . . . 3112. . .178920 Drew Brees. . . . . . . 3197. . .4424
Click to expand...
Click to expand...
Per game averages and team totals would be much "fairer" measures IMHO. That being said, I honestly have no idea how my or others projections stack up in those areas relative to whatever baseline you're comparing them to :confused:
Click to expand...
I agree. That would possibly be a more interesting analysis. However, that's a different question than the one that the OP posed. If we remove 35% of the individuals that were projected to do this or that...and didn't because they got injured or lost their job...I'm sure the correlations between prediction and reality will be much more favorable.Still, we don't know what the base rates are for accurate predictions. For this to be truly interesting, we'd want to compare various sources, including FBGs, to see if we (you) are more accurate than the competition. It's a great project, and I would be happy to run the analyses. But, the data collection would be a pain. If anyone is bored and wants to do that and send me the data, I'd volunteer to throw it into SPSS and see how it all comes out.
 
cobalt_27 said:
Jason Wood said:
cobalt_27 said:
I can do some follow-up analyses, but I just looked at how the FBGs ranked their top-20 QBs. Using Yardage as my measure (and, I could do TDs or other variables), the correlation was only .20 (and not statistically significant).Another test (paired sample t-test) suggested that their predictions were way off (t=2.23, p<.05).I can look at some other data in a bit (or tomorrow). But, at least in terms of how the top-20 was predicted to do on Passing Yardage, the results ain't good. Not suggest that any other site would've done better (I'm guessing they'd be just as unreliable). That's just the reality of making these kinds of predictions, imo.

Rk Player. . . . . . . . ..Predicted. . .Actual1 Peyton Manning. . . . 4015. . .43972 Carson Palmer. . . . .3636. . .40363 Tom Brady. . . . . . . 3798. . .35334 Donovan McNabb. . .3640. . .26125 Daunte Culpepper. . 3566. . .9296 Matt Hasselbeck. . . .3504. . .24427 Jake Delhomme. . . .3492. . .28058 Marc Bulger. . . . . . .3685. . .43019 Michael Vick. . . . . . .2477. . .247410 Eli Manning. . . . . . . 3528. . .324411 Trent Green. . . . . . .3548. . .134212 Jake Plummer. . . . . 3240. . .199413 Kurt Warner. . . . . . .3586. . .137714 Chris Simms. . . . . . .3312. . .58515 Brett Favre. . . . . . . .3620. . .388516 Steve McNair. . . . . . 3232. . .305017 Jon Kitna. . . . . . . . . 3163. . .420818 Philip Rivers. . . . . . .3312. . .338819 Mark Brunell. . . . . . 3112. . .178920 Drew Brees. . . . . . . 3197. . .4424
Click to expand...
Click to expand...
Per game averages and team totals would be much "fairer" measures IMHO. That being said, I honestly have no idea how my or others projections stack up in those areas relative to whatever baseline you're comparing them to :goodposting:
Click to expand...
I agree. That would possibly be a more interesting analysis. However, that's a different question than the one that the OP posed. If we remove 35% of the individuals that were projected to do this or that...and didn't because they got injured or lost their job...I'm sure the correlations between prediction and reality will be much more favorable.Still, we don't know what the base rates are for accurate predictions. For this to be truly interesting, we'd want to compare various sources, including FBGs, to see if we (you) are more accurate than the competition. It's a great project, and I would be happy to run the analyses. But, the data collection would be a pain. If anyone is bored and wants to do that and send me the data, I'd volunteer to throw it into SPSS and see how it all comes out.
Click to expand...
It's agreed that no one can accurately predict injuries and over a set of experts, those occurances should offset each other over the course of the season. For example, everybody was off on McNabb.But one could argue that you can predict lost jobs and the rise of a rookie QB over the course of the season - it's done every year and many predicted that VY would come on at the end of last year. Actually, I was looking at them in terms of comparing the FBG's to each other for use in the PD. Overall, who was the most accurate? At each position? I have a better idea on how to do this, and maybe I'll start on them myself.Joel
 
You must log in or register to reply here.

Similar threads

BassNBrew
Staff Projections Are Back On The Player Pages
Replies
1
Views
192
Joe Bryant
Joe Bryant
Scoresman
  • Poll Poll
Shark Pool Player Rankings Contest - Interest Check
Replies
48
Views
775
Scoresman
Scoresman
TheWinz
Question about QB sack projections
Replies
0
Views
328
TheWinz
TheWinz

Users who are viewing this thread

Total: 1 (members: 0, guests: 1)
Back
Top