How are the Predictions from WhatIfSports Looking Now?

I wrote a post back in September at the beginning of the NFL season relating the predictions of WhatIfSports with regard to the division winners, wild card teams and Super Bowl participants and winner. Since ten weeks have transpired, let's take a peek at how accurate WhatIfSports' predictions are appearing to be.

They claim that in the AFC the division winners would be the Jets, Ravens, Chargers and Texans, with the wild cards being Tennessee and Indianapolis. The Jets and Ravens seems like decent picks to win their divisions, but the Chargers seem to be behind the 8-ball with regard to winning the AFC West and the Houston Texans are at the bottom of their division. The prediction that New England and Pittsburgh would NOT make the playoffs in the AFC also seems off base as well as their prediction that the top 3 teams in the AFC East - New York, New England and Miami would all finish with 8-8 records. Both New York and New England already have 8 wins.

In the NFC, WhatIfSports predicted that the division winners would be Dallas, Green Bay, San Francisco and New Orleans. Of this set, Dallas is totally out and it looks tough for San Francisco. The other two teams have a good chance but have strong competition in their division races. The wild card picks, Minnesota and Carolina only have 4 wins between both teams - they're not going anywhere.

Overall, the accuracy of WhatIfSports predictions is seriously lame. If they used their typical methodologies that they use for their core business to make these preditions, I think you know all you need to know about these guys. Maybe WhatIfSports needs to be renamed WTF Sports?

It's NFL Kickoff Weekend, But Someone Knows the Results Now!

Yes, indeed. Although the Week 1 games are still in progress, WhatIfSports thinks they know who will make the playoffs and even win the Super Bowl. Here's the teams WhatIfSports think will go to the playoffs, win the conference championships and win the Super Bowl:

AFC East - New York Jets
AFC North - Baltimore Ravens
AFC West - San Diego Chargers
AFC South - Houston Texans
Wild Card - Tennessee Titans
Wild Card - Indianapolis Colts

NFC East - Dallas Cowboys
NFC North - Green Bay Packers
NFC West - San Francisco 49ers
NFC South - New Orleans Saints
Wild Card - Minnesota Vikings
Wild Card - Carolina Panthers

AFC Champion - San Diego Chargers
NFC Champion - Green Bay Packers

Super Bowl - Green Bay Packers

They predict the New England Patriots don't even make the playoffs. We'll see. The company is touting its simulation technology, but I'm not convinced that it is that useful. I've done posts in the past which predicted fantasy rankings, based on the premise that stating what happened last year would happen this year. Individual performances may not vary significantly, but teams' performances can vary significantly from year to year. The recent Super Bowl loser slump is a great example.

Do you think WhatIfSports is correct? Do you think the top 3 teams in the AFC East will all finish with 8-8 records? I have doubts as to the usefulness of this approach with regard to team performance.

Horse Racing Looking to Fantasy Sports?

BloodHorse magazine in their July 24, 2010 issue had an opinion piece titled "Fantasy Sports: An Opportunity - and a Competitor." The article described the size and growth of the fantasy sports market, the typical demographic, a warning about the potential threat and the potential opportunity. I think the author pointing towards fantasy sports isn't that bad, but his findings are off.

The quote from the article takes recent information from the Fantasy Sports Trade Association's research. There are approximately 27 million Americans and approximately 3 million Canadians that play fantasy sports. That's a pretty good size group of people. With regard to the American player, he is male, 37, college-educated, and has a household income of $94,000 - a great demographic that the horse racing industry would love. As is well known by the typical horseplayer demographic, you would think you need an AARP card in order to place a bet.

Here's where the author is off the mark. The author lays out a decent overview of TVG's Fantasy Horseracing product, which allows players to create a league of horses, trainers and jockeys. He thinks that horse racing would be a great fantasy sports vehicle, which would attract fantasy sports players, who may then become horseplayers to some extent. Wrong, wrong, wrong. No chance.

Why? Super simple.

As anyone who knows fantasy sports (i.e. me) can tell you, fantasy enthusiasts are primarily FANS OF THE SPORT. If you are a fantasy baseball player, odds are you are a baseball fan, then you got into fantasy baseball. The same is true for football, basketball, NASCAR, etc., etc. Fantasy sports players play fantasy sports to get more out of the sport they enjoy. Only the most geeky would have statistics be an attractant. For those, baseball is king...and in my opinion cricket probably isn't far behind - but cricket isn't as popular in the US.

What will not happen is horse racing fantasy being a big draw and sucking in new fans like Washington DC sucking in your tax dollars. Nope. Fantasy horse racing players are horse racing fans who got into fantasy horse racing. If you're not into horse racing, you're not going to play fantasy horse racing. This is where the author is way off.

The author is correct that the demographic would be very attractive to horse racing. They are younger, affluent and intelligent - exactly the kind of new handicappers the industry needs. I've got an idea of how you get these folks, but it's not this way. More on that in a future post.

2009 Fantasy Football QB Prediction Results

In a previous post, I provided some fantasy football QB rankings and compared them to a couple of fantasy football magazines. The intent was to test the hypothesis that basically saying what happened last year would happen this year wasn't radically different from the "deep analysis" provided by fantasy football information sources, which you need to pay for. To recap the methodology stated in last year's post:

"I will take last year's QB rankings and use the premise that the rankings from last year will be the same as this year. There will be a couple of adjustments. If for example, a QB moves from one team to the other and will be the starter, I'll keep that QB at the same ranking, even though they changed teams. QBs coming back from injury will be replaced with the highest rank position from the best of their replacements (i.e. Brady for Cassel). They will be projected at the ranking level that their substitute had last year. Finally, I put in Mark Sanchez in the slot held by Gus Frerotte (#29). Frerotte is not in the league at this point and Favre is now in Minnesota.

Could that cause some inaccuracy? Yes, but to keep the level of my "expert analysis" to a minimum, I'll keep the adjustments paltry and simplistic. You will also see by doing this just the few instances where you can intuitively (i.e. for free) make your own adjustment and not pay $7.99 to read something you basically already know. I made one adjustment to be fair to the magazines, dealing with Brett Favre. They didn't have him in their top rankings so I assumed that they would placed Favre in the position of the top rated Minnesota QB (#27). "

"The stats were based on total performance (passing and rushing), not just TDs. The league I was in counted passing TDs for 3 pts instead of 6 pts, so QBs that maybe didn't pass as well but got some stats rushing may show up higher in this list than the rankings your league has. The ranking lists of the magazines also were based on a performance model, not just TDs, to keep the comparison consistent."

Here's the results for my projections and fantasy football magazines A and B (names omitted):

YouGaming Blog
Correct within 5 or less spots - 12 (40%)
Correct within 10 or less spots - 21 (70%)
Incorrect by 11 or more spots - 9 (30%)

Fantasy Football Magazine A
Correct within 5 or less spots - 15 (50%)
Correct within 10 or less spots - 24 (80%)
Incorrect by 11 or more spots - 6 (20%)

Fantasy Football Magazine B
Correct within 5 or less spots - 18 (60%)
Correct within 10 or less spots - 23 (77%)
Incorrect by 11 0r more spots - 7 (23%)

The fantasy football magazines were able to beat the performance of simply stating what happened last year would happen this year, but not by very much. Overall results to get rankings within 10 spots is pretty much the same. So, at least compared to the magazines considered, you could have simply taken last year's QB stats rankings as your cheat sheet and likely not done much worse. What this analysis doesn't take into account is the actual numbers produced by the quarterbacks. It could be that even though a QB is off by several spots in ranking, the actual difference in performance stats may not be drastically different, particularly after you get past the top tier of quarterbacks.

What is fairly consistent but not reported in detail is that after the top 20 quarterbacks, no one did very well in terms of accuracy. If a prognosticator could do a very good job of predicting quarterback rankings from 15 to 30, that would be a great value as often those players would be your bye week QB or your QB that can keep your team solid if your top QB on your roster has a nagging injury or a tough weekly matchup.

I may consider this year looking at actual projected stats to see how well the fantasy football magazines do in that area.