FANALYTICS: One Month—Bringing it home, part 1

We just stocked up our September one-month leagues, fielding 317 teams. That's not a bad turnout after July's 444-team trial given that the deadline was during fantasy football draft weekend. There are big rollout plans for 2014.

Still, I continue to get pushback from long-time traditional fantasy leaguers who claim that one month is far too short a time period to have any relevance. Certainly, a full season competition is more telling of our proficiency at skills evaluation; it's also more of a testament to commitment. But I have been parading around a piece of research that does provide some validation to the one-month concept:

Of the teams that will eventually go on to win their full-season league, 80% of them will finished the month of April in 1st, 2nd, 3rd or 4th place. Expressed otherwise, 80% percent of the time, a league's eventual winner will already be sitting in a money spot after one month.

This research is from a 2003 article by Scott Wilderman (who now runs OnRoto.com) and Matt Watson in the now-defunct Sports Weekly Fantasy Hot Sheet entitled, "Do I Still Have a Chance? Recovering from a Slow Start." Within the piece is a table labeled, "% of time eventual winner was in each actual standings place." Using their commissioner service archives, it charted where the league winners were in the standings on the 1st and 15th of each month during the season. 

And they showed that, on average, your league's eventual winner will already be in a money spot after one month 80% of the time.

A few weeks ago, I received the following email from one of the monthly league participants:

"I have been wondering about your 1-month statistics about winners. I wonder what it would be for two months. The reason I ask is that I also believe one month is too short. It seems that three 2-month leagues per year might be a good compromise. It seems that, in the leagues I was involved in this year, there was a very strong team that emerged after two months."

Intuitively, the two month time period has always made sense to me too. In fact, I wrote a piece a few years ago that proposed that the major league season (and hence, the fantasy season) should be split into 50-game chunks. We'd take a one week off in between each period that would be dedicated to intense analysis (similar to pre-Super Bowl week) and conducting new fantasy drafts.

Fifty games was a time period that made sense to me.

As it turns out, after one month those extra games don't have as much impact as you might think. Here are the stats for the entire season, by month:

          % eventual winner
ON DATE   is in top 4 spots
=======   =================
May 1           80%
June 1          88%
July 1          89%
Aug 1           98%
Sept 1          99.9%

The incremental improvement month-to-month after April is tiny, relatively speaking. A good portion of the season is already determined after one month.

In fact, this phenomenon mostly holds even if we look only at the top two spots in the standings:

          % eventual winner
ON DATE   is in top 2 spots
=======   =================
May 1           60%
June 1          64%
July 1          78%
Aug 1           85%
Sept 1          94%

To reinforce the dramatic implications of these results, let's flip the analysis and look at how tough it is to come back from a bad start. Here are your odds of winning your league if you are in the bottom third of the standings at the beginning of each month:

          % chance of winning
ON DATE   if in bottom third
=======   ===================
May 1             9%
June 1            5%
July 1            3%
Aug 1         Less than 1%  
Sept 1       Football time

Over the years, I've heard about and experienced many anecdotal incidences of teams coming back from huge deficits to take league titles. Admittedly, these are isolated outliers. You always remember the one team that storms back. That's the guy who gets an article written about him. That's the story we tell to our grandkids.

Nobody remembers the other half dozen stragglers in the same league that failed.

Last week, I was wondering if this phenomenon even holds up in the major league standings. We always trumpet the need for each team to play it out for six full months. After all, where would the Dodgers be if they packed it in after their slow start this season? On May 1 last year, the eventual World Champion Giants were barely at .500 and 4.5 games out.

As it turns out, these are just a few more isolated outliers. The truth is, major league standings track fairly closely to the fantasy league research.

Looking back at all seasons from 2002 to 2012, the teams that were in first place in their respective divisions on May 1 ended up winning their division 44% of the time. Okay, nothing notable here.

However, of the teams within three games of first place on May 1, the eventual division winner arose from that group 77% of the time.

As we play it out in 2013, it looks like that percentage will continue to nudge up a bit. Here are the in-process results for this season:

Division      1st on May 1   Within 3 games  
==========    ============   ==============
AL East          *BOS             NYY
AL Central        KC         *DET, MIN, CLE 
AL West          *TEX             OAK
NL East          *ATL            none
NL Central        STL        *PIT, MIL, CIN
NL West           COL           SF, ARI

* Current leader

If this 83% success rate (5 of 6 divisions) holds, the overall success rate since 2002 would increase to 78%.

For all of us who are fans of teams not on this list, the inverse percentage play is pretty discouraging, frankly. Simply, if your team is more than three games out of first place on May 1, you have only a 23% chance of winning your division.

April turns out to be a pretty important month.

And yes, it's just one month.

Thank goodness for the wild cards. Right, Bud?
 

 


Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.