My 2008 accuracy analysis is up on the Fifth Down! Check it out, Ranking the Rankings in Fantasy Football, 2008 and let me know what you think. Does my system make sense? Will you be thinking about this at all when you decide which rankings to look at next summer? What else would you like me to do with this?
If you're wondering which sites I looked at, here's a full list of all 23 sites in alphabetical order:
In case you were wondering about the difference between how all of these sites actually did, you can see the battle was pretty tight based on the numbers for the Top 10 sites below. If you take the square root of the average number listed below, you would get the average amount of ranking spots "off" that each site was per player across all positions. I'm not showing that number here since I think knowing how "off" a site was per player is really most valuable when looking at specific positions, so you'll see that number in the details for the position rankings below. As you can read in my Fifth Down post, I ranked sites based on the sum of squared differences between rankings as averaged by the number of rankings for that position; the square root number is just being shown since I think it's easier for most of us to wrap our minds around than an obscure average number.
Top 10 sites across all positions:
Top 10 sites for QBs:
Top 10 sites for RBs:
Top 10 sites for WRs:
Top 10 sites for TEs:
A few things to note - you can see that on average the experts were far more "off" on their QB rankings than other positions. I attribute much of this to the Tom Brady effect. I left him in the rankings (and all other injured players) so to some extent it skews how off a ranking was on QBs, but it should have penalized all sites in roughly the same way.
I also think it's fascinating how close the WR predictions were - sites #4 through #9 are separated by less than 1 ranking spot. Very, very close.
Anyway, I could go on at great length here about all of the interesting things I found in looking at the details (like Laurence Maroney's difference in rankings single-handedly wrecked one site's RB rankings) but I think I'll pause and maybe give you all a breather from my rankings analysis talk.
Back soon with links to more content!
If you're wondering which sites I looked at, here's a full list of all 23 sites in alphabetical order:
- ADP (via FF Calculator)
- Ask the Commish
- CBS Sports
- ESPN
- Fantasy Hulks
- Football Outsiders
- FantasyFootball.com
- Fantasy Football Nerd
- Fantasy Football Sherpa
- Fantasy Football Starters
- Fantasy Football Trader
- Fantasy Football Xtreme
- FFToday
- FFToolbox
- Fantazzle Fantasy Sports Games
- FootballDocs
- FOX Sports
- KFFL
- Robocoach
- Rotoworld
- Sports Illustrated
- SportsBuff
- Yahoo! Fantasy Sports
In case you were wondering about the difference between how all of these sites actually did, you can see the battle was pretty tight based on the numbers for the Top 10 sites below. If you take the square root of the average number listed below, you would get the average amount of ranking spots "off" that each site was per player across all positions. I'm not showing that number here since I think knowing how "off" a site was per player is really most valuable when looking at specific positions, so you'll see that number in the details for the position rankings below. As you can read in my Fifth Down post, I ranked sites based on the sum of squared differences between rankings as averaged by the number of rankings for that position; the square root number is just being shown since I think it's easier for most of us to wrap our minds around than an obscure average number.
Top 10 sites across all positions:
SITE | AVERAGE | |
1 | Fantazzle | 55.025 |
2 | Yahoo | 56.793 |
3 | Fantasy Trader | 57.033 |
4 | ADP | 58.293 |
5 | FFToday | 60.843 |
6 | FantasyFootball.com | 61.475 |
7 | FOX Sports | 63.205 |
8 | ESPN | 63.253 |
9 | Ask the Commish | 63.802 |
10 | Rotoworld | 64.815 |
Top 10 sites for QBs:
SITE | SQ ROOT | |
1 | FFStarters | 9.249 |
2 | CBS Sportsline | 9.309 |
3 | Football Outsiders | 9.463 |
4 | ESPN | 9.558 |
5 | Robocoach | 9.674 |
6 | FFXtreme | 9.742 |
7 | FFToolbox | 9.750 |
8 | Fantasy Hulks | 9.805 |
9 | FOX Sports | 9.887 |
10 | Yahoo | 9.907 |
Top 10 sites for RBs:
SITE | SQ ROOT | |
1 | KFFL | 5.638 |
2 | Fantazzle | 6.477 |
3 | FantasyFootball.com | 6.876 |
4 | FFHulks | 6.930 |
5 | Yahoo | 7.063 |
6 | SI | 7.123 |
7 | FOX Sports | 7.208 |
8 | Fantasy Trader | 7.293 |
9 | FFDocs | 7.334 |
10 | FFToday | 7.353 |
Top 10 sites for WRs:
SITE | SQ ROOT | |
1 | Fantasy Trader | 6.309 |
2 | Yahoo | 6.327 |
3 | FFNerd | 6.374 |
4 | Fantazzle | 6.414 |
5 | FantasyFootball.com | 6.421 |
6 | FOX Sports | 6.445 |
6 | Rotoworld | 6.445 |
8 | FFToday | 6.446 |
9 | FFXtreme | 6.470 |
10 | Ask the Commish | 6.501 |
Top 10 sites for TEs:
SITE | SQ ROOT | |
1 | FantasyTrader | 5.007 |
2 | CBS Sportsline | 5.018 |
3 | ADP | 5.249 |
4 | FFStarters | 5.555 |
5 | Fantazzle | 5.852 |
6 | Yahoo | 6.264 |
7 | FFToday | 6.655 |
8 | Ask the Commish | 7.519 |
9 | FantasyFootball.com | 7.595 |
10 | Football Outsiders | 7.736 |
A few things to note - you can see that on average the experts were far more "off" on their QB rankings than other positions. I attribute much of this to the Tom Brady effect. I left him in the rankings (and all other injured players) so to some extent it skews how off a ranking was on QBs, but it should have penalized all sites in roughly the same way.
I also think it's fascinating how close the WR predictions were - sites #4 through #9 are separated by less than 1 ranking spot. Very, very close.
Anyway, I could go on at great length here about all of the interesting things I found in looking at the details (like Laurence Maroney's difference in rankings single-handedly wrecked one site's RB rankings) but I think I'll pause and maybe give you all a breather from my rankings analysis talk.
Back soon with links to more content!
Comments
Ryan Parr
http://www.fantazzle.com
CEO
Fantazzle Fantasy Sports Games
Rick Perkins.
Fantasy Football Trader.
It's interesting to see that ADP placed #4 this year. Thanks for including it in the analysis. The long-term trends will be interesting to see as well.
Kevin Day
Fantasy Football Calculator