Wins are an Imperfect Way to Measure Who is Best

Bill Parcells once uttered these words which have become a silly mantra.

You are what your record says you are.

Why is this silly?  I think there is a kernel of truth in Parcells statement.  Our current state is a product of what we have done, so in a way we are what we were.  However, this is a statement that is fraught with potential error.  What if the way in which we define ourselves and others is an imperfect way of defining ourselves and others?  What if what we use to measure does not translate accurately what we are?  Is your record truly defining you.

We know that any statistic in baseball, wins included, is an approximation of ability.  You truly are not measuring ability, you are measuring the impact of ability.  Stats are the imprint of the hand, not the hand.  That imprint can have errors.  This is a known fact.  OK, maybe some do not know this to be true as I somehow spent much of Monday discussing why statistics are not perfect definitions of ability.

Maybe a picture will help.

Converting Performance into Wins

What we have in the above diagram are really three components: ability, fortune, and result.  Ability is the grey circle full of components that make up how well a team will perform.  Fortune is represented by a magnet as it contains elements like where balls go in the field or how events stack up on each other, which can alter how well ability will translate into an outcome.  The result is simply that outcome.

Let’s step through this graphic piece by piece.


This category can go by a lot of names, so we should not be too hung up on whether I chose to name this the right thing.  What I mean to put in here is basically the sum result of active elements, items that we can be informed about, items that we are able to change by choice.  This includes traditional concepts of team like talent, training, health, and coaching; but also items that perhaps people do not think about as much, such as scheduling.  All of these elements affect how well a team can perform.  That level of performance is what we are trying to find when we look to name a team the best at something.  However, this is a very difficult thing to measure.  It may well be impossible to measure, so we try to find a way to determine who has the best ability.


The way we try to figure this out is by converting ability into wins by the way of runs scored and runs given.  This is how we define the outcome of a single game.  With a sport where the best team won 60% of their games and the worst won 34% of their games, it takes a pretty long season to figure out which team is the best.  This is not like football where you can have a team flirt with perfection one way or another.  In baseball, you need a larger sample size than 16 to figure it all out.

Right?  Well, that is because wins are an imperfect measure of which team has the most ability.  Why is a single win not a sufficient way to measure how good a team is?


This is another section to not get too bog down into what it is called.  You can call this component fortune, luck, chance, randomness, etc.  Basically, this grouping is what makes wins an imprecise measure of which team is the best because it will affect how ability is converted into wins.

An example, Balls in Play.

This is something we have all experienced multiple times playing this game or watching it.  One team is striking guys out left and right while putting up an extra base hit every inning while the other team somehow manages to have balls hit right to them when the opponent is in scoring position and managing to have a wind aided ball carry just over the fence.  With our eyes, we see that the better team lost.

There is nothing that can be trained or taught that can shift a hard hit line drive straight to a second baseman versus five feet to the right.  That is the same amount of ability from the batter, but one ends in an out and the other might turn into a double.  That is an error in conversion.  The hope is that by playing 162 games that the noise that the conversion error generates will be muted…that the signal (who is better) emerges from the messy noise of wins and losses.

But What About the Post-season?

Well, after using the entire season to figure out who the best teams are, we do not have many options left.  We accept a shorter schedule that will award one of the best teams with a championship.  The championship is not designed to determine which team is truly the one with the best ability, merely the best possible way to figure out who might be.  One game series will not figure out which is the better team, nor will seven game series.

That is OK.  We do not need the World Series to definitively determine which team is best.  It works merely in determining who wins the most, which can be different from which team has the most ability.

Restating It All

So, the basic fact of this is that wins are an imperfect measure of the best team.  Ability does not neatly translate over to wins because there are elements beyond ability that affect that conversion.  That is OK.  With a long enough season, differentiation can be good enough to set up a playoff among many of the best teams.  A shortened post-season is fine because trimming away the lesser teams and letting the best teams duke it out is a sound approach.  However, we know that the team with the most wins is not automatically the best team.  They might be, they might not.  And that is OK.

Share this post on
  • Facebook
  • Yahoo! Buzz
  • Twitter
  • Google Bookmarks
  • email
  • Google Buzz
  • Posterous
  • Tumblr

About the author

Jon Shepherd  

Jon Shepherd founded the Baltimore Orioles blog Camden Depot in 2007. In addition to Baltimore Orioles analysis, the blog also focuses on qualitative and quantitative approaches to assessing baseball in general as well as providing mainstream reviews and commentary on substances alleged to performance enhancing. Dr. Shepherd’s writing has been featured on ESPN, and his blog has been part of the ESPN Sweetspot Network since May 2011. He has made radio and podcast appearances for Orioles’ centered programs.

This entry was posted in Baltimore Orioles. Bookmark the permalink.

14 Responses to Wins are an Imperfect Way to Measure Who is Best

  1. dan says:

    I’ve repeatedly asked you what is the goal of defining the “best” “better” or “lesser” teams, and i haven’t yet heard an answer. What is the goal of this exercise? Bragging rights for losing teams? Or does it have to do with justifying incorrect projections? Our is it just a fun exercise that in the end is meaningless in the real world? For certainly teams may find individual evaluations useful, but how are team evaluations useful if that team fails to win games, but is still, in your system considered the best?

  2. dan says:

    I think the problem is, from the start you are throwing out some big concepts. How do we define ourselves…well, that is never going to be perfect to an objective observer because we are not objective beings, everything in our life is subjective. So is there a perfect way to define anything? Not unless it exists in a vacuum we created. Since a team does not exist in a vacuum, there will not be a perfect way to define that team, but there certainly is an agreed upon way to judge the teams accomplishment, and that is their win/loss record. Since the game has well defined parameters and rules the teams that are the best at playing the game within those well defined parameters are the ones with the most wins. Since the game is played on earth the laws of physics matter, and fortune, which broken down is still a slave to physics us negated. Does skill matter when one ball goes to the second baseman and one goes past? No. But timing does. Is it a repeatable skill? Possibly not. But that shouldn’t matter. Better should include environmental factors. In battle certainly you want to choose the high ground, but that is not always an option. In life, it is easier to be born into a wealthy family, but you have no option. If these environmental factors are out pod the control of players, then they should be ignored. Otherwise, points should be allotted to whomever had the most difficult childhood, or players didn’t have the same opportunities that the others did growing up

  3. dan says:

    Sorry, posted too soon (on my phone). My point is that the goal of the game is to won. It’s not to win by 20 runs one day, and lose by one rum the next three days. The teams that win the most games, by definition of the sport, are the best teams. The reason they are the best teams is that they win the most games. Trying to redefine what it means to be “good” to me seems rather silly, unless there is a reason for it which you have failed to mention.

  4. Jon Shepherd says:

    What is the goal?
    The goal simply is to state that wins are approximations of “ability”. There is no goal beyond that other than to evaluate a basic statistic. It is not about giving bragging rights or providing solace or to say we need a better way. None of those. It is not a fun, pointless exercise. It simply is to state that random events occur and can impact the translation of ability to a win.

    “Your system”?
    I do not know what you mean by my system. There is no system here. The diagram above is merely a description of things that happen that impact winning and losing.

    Subjectiveness of life?
    I think you have some good ideas in the second part, but I think you are letting those ideas overwhelm you. The point is not to define everything, figure out who has the best ability or who has been affected the most by chance. The point is to simply acknowledge that wins are imperfect because there are susceptible to accumulations of randomness. That is as far I as went in the article and as far as I needed to go. All the article is about is how records are not perfect descriptions of the team’s ability.

    Redefining good is silly because games are designed to figure out who won?
    I think first of all, there is no redefinition occurring. Second, I disagree that games are defined by wins. Wins came about because people were trying to figure out who was best and to have rules to decide that is the simplest and most straightforward way to do it. I think it is an important thing to remember that wins are an imprecise measurement even if someone defined them at one time to say this is how we measure who is best. A rule does not make reality, but a rule may define reality the best way possible.

    So, to restate, the point of the original response over wins is about how they are imperfect. Your background is not always a definition of self. It can simply be a description according to well-intended and well-used rules that are imprecise. And, that is OK.

  5. dan says:

    “It simply is to state that random events occur and can impact the translation of ability to a win.”

    That’s it? I don’t think anyone would agrue that. I think you saying the the “best” team doesn’t always win is where the arguement happens. You are stating the the “best” team can lose due to factors outside of their control (wind, umpires, etc.), where as I think the best team is the one that wins and overcomes these enviromental factors no matter the underlying ability.

    But in the end, I don’t think I disagree with anything you are saying once you state that, “The point is to simply acknowledge that wins are imperfect because there are susceptible to accumulations of randomness.” I mean, there’s nothing to argue with there, that’s pretty obvious. I’m not sure why you stated it, I don’t think people have been going around claiming that “wins are perfect”, i mean, it’s just a game. Wins are the result of the game. Number of wins is how we judge who is the best. You don’t disagree with either of that, right? I thought you were trying to redefine who was a better team based on something that wasn’t wins, but just stating something is not perfect, yeah, if that’s all this is about, I would of course have no problem with that. I mean, everything is imperfect.

  6. Jon Shepherd says:

    So, we both agree that wins are imperfect.

    Is it really too much of a logic leap to say that because we, as a population, declare the team with the most wins as the best that this criteria is also imperfect?

    That is not to make the claim that there is a more perfect way, but that there is some imperfection in determining who is best, which makes that determination potentially wrong?

  7. dan says:

    Ahh…i think this is where we will disagree, i don’t think that just because something is imperfect it means it is wrong. The game is played in a world we cannot control. Perfect isn’t the goal of the game, competition is. Not just competition against another team, but competition against the elements, against everything. So, i still believe the best team is always the team that wins. Do teams have players with more skill and lose? Of course, skill alone does not make the best team.

  8. Jon Shepherd says:

    The way imperfect is being applied here is that the capability does not directly translate to wins due to randomness. If you consider randomness to be part of a team’s ability, then you should think wins are perfect. If it is not part of a team’s ability, then the stat is imperfect. When you have that mistranslation, you open up the possibility of a team with lesser ability winning.

    It seems your definition of best includes randomness as a factor. I disagree with that, so our perspectives cannot come to an agreement as that is unresolvable. I think each step of the way we agree. We simply disagree on where to make the final determination. I think of performance outside of randomness. You prefer defining players by their abilities and the randomness of events involving them.

  9. dan says:

    Yes, you are dead on there. Since life doesn’t exist in a vacuum the environment that people play in, and the outside forces that affect the players certainly matter. Dealing with wind or weather in general, dealing with a umpire that doesn’t like you, whatever it may be, overcoming that is part of winning. Can these things be measured? Probably not.

    I think the term “ability” is where the problem occurs. I am thinking of ability as something very close to skill, but might you be using it more of as “opportunity”? If so, then this might be a very different discussion, but if you are using it as skill then yeah, we agree to disagree.

  10. Jon Shepherd says:

    There are ways to measure randomness, but it can be dirty.

    Based on your statements though…player-wise…you’d like Win Probability Added. Only issue you might have with it is that it considers all parties to be average when you have actual specific people involved with varying ability levels.

  11. dan says:

    Yeah, I’m not really into stats much, I tried for a while, and I find them interesting, but I just prefer to watch the games.

  12. Jon Shepherd says:

    I prefer to watch the games as well, but it enables me to drill down into what I saw and to challenge my own preconceptions. It also lets me *see* games I could not watch.

  13. dan says:

    Wanna hear a crazy stat, Chris Davis has already seen more 3-0 counts this year than in any other year in his career…crazy right?

    I mean, I still love looking up stats, that’s been a fun part of baseball since I was a kid (though it was more about runs, RBIs, HR, and BA in those days)…stats are always a part of baseball. And I love listening to games on the radio, cause then I see it in my head, the whole game kinda slows down when you listen on the radio, and I totally get why people love stats and trying to challenge their preconceptions. But alot of the advanced metrics/stats, I’m not really into them…

    Unless they say something good about Chris Davis.

  14. dan says:

    Which wpa apparently does (like chris davis).

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>