Per my discussion with Adam last night, I'm posting a link to Fantasy Brain here: http://bit.ly/Ld4byD

Fantasy Brain, analogous to Klout for Fantasy Sports, imports your existing fantasy leagues and uses algorithms that analyze every move for efficiency: every draft pick, start/sit decision and transaction based on the scoring rules and roster composition in that league.

The result is a universal ranking for both the player and the league, normalized data that allows anyone to compete against anyone else regardless of what league, format or provider they use.

There is other unique social functionality we provide as a result of creating a social graph for fantasy. Would love any thoughts or feedback from the Fantasy guys at PandaWhale!

This app is a stat lover's dream.

Pulling all the fantasy league play Stats into one place is awesome.

Bloomberg has also joined the fantasy football analytics game:

https://www.bloombergsports.com/decision-maker

Yes, there's a big market opportunity here. Bloomberg also brings data analytics to fantasy sports, but they've just created their own algorithm that they apply to the athletes and then spit out a projection. It's a paid app and most users have found the projections to be fairy inaccurate.

Fantasy Brain is applying algorithms to the users themselves. In testing, we've found that the top 5%-10% of fantasy players are significantly better than everyone else (and the next 80% are very similar in terms of results). By mining expertise data for trends, we think we'll be able to provide the most accurate data products on the market and yet most of our product is and will always be free.

1) I love that your product is free.

2) I love that your users can become BETTER at fantasy football by using your product.

True dat! But let's not forget the real motivation...there's an estimated additional $2B in side bets/pots/gambling for league winners, or an average of $900 per league. There will be premium data subscriptions down the road too (which converted in large numbers at $25/year in customer development testing), but we are giving a lot of data away for free. The product with the best data helps you win CASH!!

My impression is that a lot of fantasy football players do bet.

$2 billion in side bets is nothing short of astounding, JJ.

That comes from the Fantasy Sports Trade Association's claim that the average league has a $75/team ante, and at 30M users, that comes out to $2.25B. In our customer development research, 92% of customers interviewed had a league ante and that was a median of $75 and an average of $175. So if any of the above data is correct, $2B could be conservative.

This is, of course, on top of fantasy sports already being a $2B market in products, services and advertising.

Is it mostly football, or are other sports as big?

American football is by far the biggest as of now. Baseball is actually in decline (also a different target market since you need to change starting pitchers 162 times to play) but basketball is on the rise, as is NASCAR. Fantasy Cricket is exploding in India.

Then there is Fantasy Soccer (futbol). It's big, but not that big because the game hasn't had the statistical relevance to support a compelling data-driven game (1 goal, 1 assist, 15 saves doesn't work). But this is changing quickly as moneyball and new real-time data infrastructure comes to soccer (futbol). New statistics could turn fantasy soccer into a global phenomenon and turn a $2B market into a $10B market within 5-10 years.

http://www.wired.com/playbook/2012/09/major-league-socccer-micoach/all/

Fascinating. Thank you for the rundown!

How do you deal with custom scoring or even just the different scoring schemes across different leagues?

For example, I give more points for yardage than the default scoring system and this would affect player recommendations and rankings of users within your system.

Without giving too much of the secret sauce away, your league scoring rules and roster composition are taken into account when analyzing your moves. Yes it's all taken into account.