Numbers vs. Intuition

Reevaluating stats through fantasy baseball and why more teams aren't doing this.

For the past four years one thing I’ve always looked forward to, besides the start of the baseball season, was the start of the fantasy baseball season.  The first time I played this extremely complicated version of a guessing game, I had no idea what to do.  The only point was getting the players that I knew I liked and using them.  Ignore the fact that they were on the DL, or were in a slump or they were just plain awful.  If I liked them, they were on my team.

I still remember the first lousy trade I made: I sent Cardinals RHP Chris Carpenter in exchange for then Dodgers-closer Eric Gagné, fresh out of surgery and heading for the DL.  My only thought was, I like Gagné, let’s get him on the team regardless.  Needless to say, I ended last in my league.  I obviously didn’t follow the business mentality of weighing pros and cons.  I just acted on impulse like any new person in the game. 

The next year I decided a different approach.  I signed up for MLB’s Fantasy League and played their three types of leagues: live draft, list draft and random draft.   I would try one team I had no control over and two others I had a degree of control.  My strategy for the live draft league was: pick a player you like, and then switch and trade in accordance to trends every two-three weeks.  The list and random drafts followed the same strategy but I had no idea who got chosen.

I lost all three leagues.

I was beginning to feel defeated.  How could so many other players be successful?  Was there a trick to master behind drafting, maybe scouting or just watching games?  There had to be some advantage that I didn’t have.

The immediate answer was that, other than watching a Wednesday or Sunday Night ESPN broadcast, I wasn’t watching any games.  I had no idea what was going on.  So, during the offseason I changed my strategy once again.

I decided on playing three live drafts, buying the 2010 Baseball of America Almanac and getting MLB.tv for the first time.  I scheduled my drafts for the last day of spring training and watched every game possible.  I reviewed stats but, at the moment, wasn’t really into sabermetrics so I had no idea what to do with most stats.  My rule was, players should have no lower than a .275 batting average and pitching staffs should have an ERA no higher than 3.60.  Random numbers but they helped me start a scouting pattern.

I ended up 6^th in two leagues and 3^rd in the last one.  It was an improvement but still not enough to help me win the league.

Then 2011 came.  I voided buying the 2011 Baseball Almanac.  If I just needed batting averages and ERAs, the Internet would provide them.  Again I crashed and burned.  The improvement I built upon held me back once again.  The formula that worked one year failed to deliver the next. 

Why?

Players are constantly changing.  Some are getting older and more powerful.  Others become more frustrated.  The famous but not so good are being replaced by the incredibly unknown.  A revolution was going on in baseball that I had failed to notice if it weren’t for a Michael Lewis’ Moneyball and the many stats that were introduced by Bill James and other sabermetricians.  I’m skeptical to say all this because I may look like a fool and another nut that’s jumped on the movie bandwagon, but the truth is that it was not the movie but the book that sold me on research. 

The movie gave me the will to want to enter the business side and front office world of baseball and gave me the personal goal to try and become a GM.  But the book showed me the value of numbers, scouting, analysis and research that has to go into every single player.  I thus decided that 2012 would be a year of more tinkering and experimenting.

The past summer (I live in Peru so summer goes from December to March) I started reading on three subjects: economics, finance and sabermetrics.  I started noticing patterns between the three topics that made me more eager to see if my experiment would work and thus motivate me more into moving in that GM direction. 

I evaluated about ten players per position, plus ten more candidates for the utility role to get an idea of stats and projections beyond batting average.  I analyzed stats for on base percentage, slugging; predicted hits and total bases.  Got a feel for what the probable averages would be. 

The same for pitching staffs.  I looked at their numbers from previous years and used any and every formula to determine their probable ERAs, hits allowed, walks, strikeout ratios. 

I had my staff: my 2012 Baseball Almanac, MLB.tv, and Internet.  I had a sense of who I would draft in what order.  If one player wasn’t available, who was the next likely candidate, independent of whether I liked him or not.  If his numbers were good, he was good enough for me.  I drafted.  Actually, I couldn’t draft the first time because of a server problem and couldn’t log on to the war room, so I created a second draft account and went from there. 

So far, the results have been spectacular.  Not only I am first in my league, I’m first in almost every category, except, curiously, for batting average.

If I could draft based on numbers, why aren’t more GMs doing what Billy Beane did?  Why aren’t managers paying attention to stuff like OBP or SLG or dERA instead of wasting their time on good names and pretty faces?  Why aren’t evaluations done to players?  Why aren’t more people employed to investigate and research and evaluate every single player?

These people are employed, though in very small numbers. 

Numbers work.  Stats tend to show an average of how a player will perform, and there are formulas to predict whether there will be or there won’t be any major fluctuations.  Of course there are other factors, but the main idea is that baseball is not just a sport, it’s a numbers game.  It’s the only professional sport where numbers DO matter and they DO tell everybody involved something. 

But I rant on.  The focus was just to show that numbers, stats and research work in baseball and help us get an idea of what might happen and when. 

Now I’m thinking: I have very few resources at hand when it comes to all this investigation and scouting process that goes behind deciding who is a better player.  One or two books.  The chance of watching up to four games at a time.  Access to stats and the ability to generate them as the necessary numbers come in.  How come baseball organizations, which hold more resources, ignore something as fundamental as this. 

In my opinion, it is a fear of change.  Change is something that many people are afraid of, regardless of whether the outcome will be for better or worse.  If the system is tampered with, then changes may occur.  And since the results of change are unpredictable, then one becomes afraid. 

Think of it along the lines of chaos theory.  Even if you can try and predict what the outcome of change may be, there is no degree of certainty that that will be the end result.  For example, owners may extend the walls of a ballpark to make it more pitcher friendly.  Predicted result: less home runs, more pitching duels.  Unpredicted results: lower ERAs, less total bases and lower slugging percentages.  And even more unpredicted results: batters start to become more patient and look for balls they can drive. 

The same happens if managers and general managers start building their teams in accordance to numbers and not names.  The immediate effect is that sales will drop because the big household names no longer appear on the roster.  That’s completely predictable. 

Now, change all the big names that have low numbers and insert the unknowns with incredible numbers.  Change happens.  You can predict that the team will get better, post better numbers and, ergo, win more games.  Managers won’t like this.  They don’t believe in numbers because they weren’t brought up to think about numbers.

Baseball is a culture, much as any religion, ideology or culture. And cultures don’t change.  If there is a line of thought, it’s meant to be followed.  Here is where the front office clashes with the guys in the dugout.  People in the dugout look for something different.  Abilities, talent, versatility, duress, poise, etc.  Front offices look to sell tickets and create a team capable of winning (i.e. Los Angeles Angels in 2012 with the signings of big names like Albert Pujols and C.J. WIlson).

If front offices really want to field championship teams, they should look for productive players with the stats to support their decision.  If there is empirical evidence to back their decision, it should be acknowledged, not ignored like managers tend to do. 

But if managers were slowly introduced into the numbers game, then they can, easily adapt to the new environment that sabermetrics is providing within baseball.  Instead of forcing change, they should be adapted into it. 

Let’s look at an example.  Billy Beane, in 2003, traded Carlos Peña because Art Howe wouldn’t use Scott Hatteberg at first base.  That’s forced change.  Never mind that Howe ignored Beane’s insistence on using numbers to evaluate players.  Howe was afraid of changing the baseball mentality because he did not believe in Beane’s ideology.  He didn’t want to compromise what he was brought up to believe against what was being said on paper.  The classic battle of a priori concepts against a posteriori evidence introduced by Kant in The Critique of Pure Reason appears under another banner. 

Any person with common sense would follow evidence.  Then again, many times we follow intuition even though the evidence we hold is more thought out.  We thus have two points of view: the game/art of baseball and the science of baseball.  It pans out just like the cases of religion vs. science, but narrowed down to a sport.

If people listened to numbers more often, then it would take the excitement out of baseball, it would probably kill the sport and the idea that anything can happen.  But, let’s go back to chaos theory.  This is not necessarily what is going to happen.  Yes, the game can become more competitive.  Yes, it can become even more predictable due to the fact that we only listen to numbers.  But within its predictability, it becomes more unpredictable. 

Numbers give us a sense of what might happen, not of what’s going to happen.   Though it may look like it, baseball is not a science.  We may think, as sabermetricians, that it is, but it is not.  Numbers are fun and help us try and discover the secret element of the game.  It’s faith versus science. 

What I’m trying to say is not that we should abandon the household names and the beauty that is the unpredictability of baseball.  The point is that, if organizations have the numbers and the resources to evaluate every single player out there, then they have a statistical advantage that allows them to try and field the most competitive team possible stats-wise. 

Forget Moneyball.  We are not trying to create a competent team with as little money as possible.  We are trying, year-in and year-out, to create a team capable of winning the World Series.  And for that, the first step, I believe is evaluating the stats of every player in order to understand who can help you improve on the field and get that ring at the end of the year.