Sunday, June 27, 2010

Multi-year results for accuracy with TE rankings

We're rolling on with the multi-year accuracy results - for those who are new to the blog, since 2007 I've gathered and analyzed preseason rankings from fantasy football experts and noted which expert is the most accurate with their rankings. You can see the results of 2009 here, 2008 here, and 2007 here. However, I think many of you would agree that an aggregate look at which site has been the most accurate across all three years is of even greater importance. So I posted the results for multi-year accuracy with QBs and WRs last week and tonight we've got TEs. For details on how the accuracy is calculated, just scope out one of the previous year's results, and to see the details on how/why I calculated a percentile average (remember: the lower the average, the better!) for the multi-year approach, take a look at the QBs post from last week. If you'd like a list of all 54 sites included in this analysis, scope out the WR post. Ten of those 54 sites have participated all 3 years, another 13 sites for 2 years, and the remaining 31 sites for just 1 year.

As with the QBs and WRs, I'm posting the top results for sites that have participated for different lengths of time separately since it's really not fair to compare those sites to one another. Sites should only be ranked relative to other sites that have participated for the same length of time, hence why you see below that I've broken it down into the best of sites that have been involved for all 3 years, top results for those who have only been in it for 2 years, and a few mentions for accurate sites with just 1 year of participation. Clearly I don't believe they can be compared directly but for some relative sense of accuracy, take a look at the percentile averages on the right side. Again, if your average is lower than your peers, that means you have performed with greater accuracy.

Sites that have 3 years of participation:
Ranking Site Average
1 Ask The Commish 0.29
2 Yahoo! 0.31
3 FOX Sports 0.34
4 0.37
5 Rotoworld 0.38

Sites that have 2 years of participation:
Ranking Site Average
1 FFStarters 0.16
2 FFCalculator (ADP_ 0.18
3 FFToday 0.20
4 FFTrader 0.21
5 Fantazzle 0.25

Sites that have 1 year of participation:
Ranking Site Average
1 Fantasy Mojo 0.04
2 MyOfficePool 0.08
3 Fantistics 0.15


Anonymous said...

ask the commish seems to do pretty well, i wonder if that's cause he's awesome or because he looks at the popular sites(yahoo,roto, and such that do well) and augments their picks? either way he's being added to my own ADP list!

Anonymous said...

I'm curious to find out what these lists would look like if you ignored say the top 50% - 66% of starters (we all know who the usual suspects will be barring injury). Meaning a 12 team league that starts 3 WR has 36 starters...throw out the consensus preseason top 18 - 24 WR's and see how the sites ranked. Ranking the lower players drafted in the middle and lower rounds accurately wins championships... as does playing the waiver wire and trades. Having Andre Johnson as the 1st WR is better than having him ranked 3rd last year... but what really helps is the prognosticator that had Miles Austin ranked, as say the 35th WR, where other sites didn't have him ranked in their top 50 if at all. Another example would be a site that had Bernard Berrian waaay further down the list compared to their competition.