Geeks Rule?

by KerryWhisnant

Moneyball, the 2003 book by Michael Lewis about how the Oakland A’s searched out undervalued players using unconventional (at the time) metrics to gain a competitive advantage, first spotlighted the increasing trend towards sports analytics. Since then more teams have taken the plunge and are now using sports analytics to improve their product both on and off the field. This was particularly evident by the attendance figure for the fourth annual MIT Sloan Sports Analytics Conference, at more than a thousand strong, and more than twice as big as last year’s meeting.

The conference schedule had four parallel tracks, with three panel discussions and a research paper presentation going on at any one time. The only exception was the early afternoon panel entitled “What Geeks Don’t Get: The Limits of Moneyball,’’ which was the only event in its time slot. Moderated by Michael Lewis, the panel included Mark Cuban, owner of the Dallas Mavericks, Jonathan Kraft, president of the Kraft Group, which owns the New England Patriots, Daryl Morey, general manager of the Houston Rockets, Bill Polian, president of the Indianapolis Colts, and Bill Simmons, columnist for ESPN. Morey was also co-organizer for the conference.

Despite the title of the session, which might lead one to believe that sports analytics has reached a plateau, the overriding theme of the panel was that there was much more that analytics could accomplish and much more that could be learned. Not that the geeks are taking over, either, but rather they are being absorbed by forward thinking sports teams — and woe to the teams that either don’t follow suit, or are slow to do so. It was clear from the team representatives on the panel that they thought they had a competitive advantage over teams who had not embraced analytics. They also thought that they could tell which teams were not using analytics from their play on the field or on the court.

The fact that there was no baseball person on the main panel was indicative of the fact that sports analytics has gone beyond the sport of its infancy, baseball, and has thoroughly infiltrated basketball and football as well. In fact, in addition to a Baseball Analytics panel, there were also panels on Basketball, Coaching and Emerging Analytics, the latter including representatives from football (both the American version and, interestingly, soccer).

The Baseball Analytics panel was moderated by Rob Neyer of ESPN, and included John Abbamondi, assistant general manager of the St. Louis Cardinals, John Dewan, owner of Baseball Info Solutions, Dan Duquette, owner of Duquette Sports Academy and former general manager of the Boston Red Sox and Montral Expos, Shiraz Rehman, director of baseball operations for the Arizona Diamondbacks, and Tom Tippett, director of Baseball Information Services for the Boston Red Sox. The team representatives all indicated that they are heavily into analytics, but that it is blended with traditional scouting to give overall ratings of players. Collecting and synthesizing the data in an easily understandable and accessible way is also a challenge.

When asked what advancement in analytics they would most like to see, both Abbamondi and Rehman said they would like to be able to better quantify a player’s character, or what they called his “make-up.” (Similar themes were evident on other panels, where the increasing importance of psychological testing was stressed.) John Dewan and Dan Duquette mentioned catcher defense as an area that needs a lot of additional study.

Everyone also agreed that there is more work to do on evaluating defense throughout the rest of the lineup, and that Sportvision’s new Field F/X systems that are being installed in major league ballparks will fuel a revolution in the way fielders are rated. These camera systems will record the positions and motion of the batted ball, runners and fielders. John Abbamondi said – facetiously, he later admitted – that he wasn’t sure it was a good thing that every team would be getting all of this new information, because it might level the playing field. But as is true in almost any endeavor, some people will be able to use the information more effectively, and thereby obtain a competitive advantage. The teams that are at the forefront of analytics now have a head start in that race, but they certainly can’t become complacent.

Another area in need of advancement is that of understanding pitcher durability, especially being able to predict and promote it. Dewan wants a formula that could predict pitcher injury, and Duquette wished that there was a good way to produce 20-game winners. The ability to better project career tracks, for both pitchers and position players, was mentioned by Tippett.

In addition to the team connections of many of the panel members, many other teams were represented — among baseball teams, the Indians, Mets and Pirates had attendees. But basketball had the biggest contingent – by my count, twelve NBA teams had representatives at the conference.

The coaching analytics session was moderated by Ric Bucher of ESPN and included Brent Barry, former NBA player and currently with NBA TV, Avery Johnson, ESPN basketball analyst and former head coach of the Dallas Mavericks, Kevin Kelley, head coach and athletic director at Pulaski Academy High School, Buck Showalter, ESPN baseball analyst and former manager of the Yankees, Diamondbacks and Rangers, and Nate Silver, developer of the PECOTA forecasting system.

Johnson discussed how he had started as a more traditional coach until he was quickly converted to an analytics-savvy coach by Mavericks owner Mark Cuban. Showalter commented that not only is it important that analytics people convey their information in an understandable way, but they must also convince managers and general managers that they are credible. Having more analytics-leaning owners such as Cuban would certainly speed up the process, as would the continued success of teams using analytics.

Another problem mentioned was how a manager can convince players to accept stats that can help team performance. (I would have added: how do you convince players to not believe stats that aren’t especially useful in helping team performance, such as saves or batting average, particularly when those stats might still be viewed as important come contract time?)

The conference wasn’t just about sabermetrics and its cousins in other sports. There were also panel discussions on the business and ethical aspects of sports, such as Developing the Athlete’s Brand, International Expansion, Innovations at Arenas and Stadiums, Social Media Marketing, the Future of Sports Journalism, and Performance Enhancement.

So it’s safe to say that the geeks are here to stay, but they’ll make the biggest impact if they can learn to communicate with the more traditional leaders in the field and convince them that they have something to offer. Judging from this conference, that day is rapidly approaching.

50 Responses to “Geeks Rule?”

  1. Rick Says:

    Every intelligent business employs new technology or information to gain a competitive edge – but the information must remain relevant.

    For baseball, this means more than mere “communication” of ideas but their actual usefulness. Herein lies the problem. For “geeks” to remain significant, they’ll need to produce new, applicable material – no team’s going to pay someone to tell them that batting average isn’t important.

    So as sabermetricians produce more work, its use will grow more and more marginal. Unless someone revolutionizes the field, diminishing returns on investment will result.

    In other words, sabermetrics will not grow increasingly more relevant. It will crest at some point, having made its contribution.

    And traditional scouting will remain.

  2. ThomasWayne Says:

    From Kerry’s text:

    So it’s safe to say that the geeks are here to stay, but they’ll make the biggest impact if they can learn to communicate with the more traditional leaders in the field and convince them that they have something to offer. Judging from this conference, that day is rapidly approaching.

    My Response:

    Couldn’t agree more. The only problem is the geeks need to realize that just because they come up with something that may be useful it does not have to be at the expense of ideas and people who don’t immediately buy into their set of ideas or standards.

    In other words….you may be smart…you may be able to crunch the numbers in a way that may impress some people…but in the end…if you REALLY KNEW THE GAME you were trying to infiltrate, you’d understand that what you are offering is just a small part of a bigger picture…not the picture itself.

    I’ve said it before and I will say it again…the problem with sabermetrics isn’t the numbers, applications, or formulas…its the sabermetricians who believe that just because they have these numbers they deserve a place at the table.

    Nice article Kerry, keep up the good work.

    Thomas Wayne

  3. Hossrex Says:

    Rick: “So as sabermetricians produce more work, its use will grow more and more marginal. Unless someone revolutionizes the field, diminishing returns on investment will result.”

    How would this sound: “So as physicists produce more work, its use will grow more and more marginal. Unless someone revolutionizes the field, diminishing returns on investment will result.”

    Of course, someday… given an infinite length of time… physicists will understand everything there is to know about how everything works. How long will that take? I don’t know. You don’t know. No one knows. So until then, physicists will have a high value.

    So…

    Of course, someday… given an infinite length of time… sabermetricians will understand everything there is to know about how everything works. How long will that take? I don’t know. You don’t know. No one knows. So until then, sabermetricians will have a high value.

    The only way those two points could be claimed as dissimilar would be if you were to say sabermetricians were close to quantifying baseball through statistics.

    Considering sabermetrics (a phrase I loathe) is in it’s infancy, I’m reminded of the 1899 quote from Charles H. Duell, Commissioner of the United States Patent and Trademark Office, regarding his suggestion that the patent office be closed.

    Charles H. Duell: “Everything that can be invented has been invented.”

    It’s shortsighted, naive, and arrogant.

  4. Rick Says:

    Hossrex,

    I ask that you consider some of the work that’s earning PhD’s today, if you’re aware. If not, take some time to research it. More people are earning advanced degrees; the effort to develop original work is producing more work of marginal value.

    Is sabermetrics in its infancy? Your premise is poorly thought out. The largest strides in sabermetrics have been made, each stride will get smaller until the science of baseball has been mastered, and near exhaused.

    Comparing baseball to the massive scientific endeavors of the 20th century is childish.

  5. Patrick Says:

    The Field F/X system is probably something we have all “invented” in our heads when trying to conceive of a proper fielding rating system. That’s why I always said “why bother” with UZR or Rtot? They have to lump too many variables into one basket, so the end result is somewhere between untrustworthy and ridiculous.

    UZR supporters might argue their system was an important step that led to Field F/X but I would say that computer technology and simple baseball common sense led to it, whether UZR ever existed or not. Wrong technology is worse than no technology when it comes to analyzing something as simple as baseball

    To Rick’s and Hoss’ debate. I think what Field F/X will prove is that baseball people get it close to right with the naked eye. It doesn’t mean you stop trying to understand as much as you can but as soon as you know 2 + 2 = 4(F/X + Field Pct = Fielding Rating) you stop working on it. At that point you have at least quantified 99% of what goes into rating a fielder. So indeed, it really is an example of diminishing returns, especially if my “naked eye” assumption proves correct.

    The same could be said for improvements on OPS. In most cases, OPS states an accurate offensive value, so any improvements on that basic guideline will be minimal. I guess the law of diminishing returns has baseball executives looking past the obvious and are now trying to quantify things such as character and durability.

    Kerry, great article.

    PS; BA is the 2nd most important component to run scoring, behind only SLG, In my opinion.

  6. Hossrex Says:

    Rick: “I ask that you consider some of the work that’s earning PhD’s today, if you’re aware.”

    Because “some of the work” clearly invalidates “the body of work”.

    Well thought out.

    Rick: “More people are earning advanced degrees; the effort to develop original work is producing more work of marginal value.”

    Yup. That’s probably true.

    If you have an exclusive club of ten super genius’ working on theoretical physics, they’ll probably put together a very high standard of work.

    However if you have a less exclusive club of ten THOUSAND “regular” people working on theoretical physics, probably you’ll see their “work is producing more work of marginal value.”

    That’s a completely misleading point, which has absolutely no barring on anything, since the sum total of the work being produced is VASTLY greater than it ever has been before, but I agree… I have no problem with the idea that more marginal work is being done.

    Rick: “Is sabermetrics in its infancy?”

    You’re playing semantics with the term “infancy”. My definition is a field of study which has nearly infinite room to grow… while yours seems to be “hell… if she’s got grass on the lawn, I’m ready to mow ‘er!”

    Rick: “Your premise is poorly thought out.”

    At least I have one.

    Rick: “The largest strides in sabermetrics have been made”

    The ignorance in that statement is astounding.

    Statistics still tell us relatively NOTHING about baseball, and yet you have the gall to say “the largest strides… have been made.”

    Ignorance to the extreme.

    Please refer to my Charles H. Duell quote above, as I now think it’s even more fitting than I’d first surmised. Glad I preemptively quoted it, since it’s the basic point you’ll be arguing for the next half dozen posts, while never understanding it’s significance.

    Rick: “each stride will get smaller until the science of baseball has been mastered, and near exhaused. “

    I agree. The question then is one of scale.

    How much smaller? How long until it’s exhausted?

    You don’t have a clue, and yet you imply it’ll be soon.

    Why?

    Rick: “Comparing baseball to the massive scientific endeavors of the 20th century is childish.”

    Sabermetrics is a science.

    This is a fact. By definition and by colloquial use, sabermetrics are inarguably a science.

    I’m not, nor did I assign a HUMANISTIC VALUE to the sciences, but they are both sciences, and hence they follow the same rules/structure.

    Saying its childish to compare the two is frankly moronic.

  7. Patrick Says:

    While it’s certainly not childish to compare baseball to a physical science, it’s important to note that explaining baseball through statistics (sabermetrics) has gone hand in hand since the beginning. So it’s definitely not in it’s infancy.

    Just look at how easy it is to get an accurate picture of what components lead to runs using only 2 traditional stats, BA and SLG;

    Team/BA+SLG/RUNS/League place in both BA+SLG and R

    NYY/.755/915 1st
    LAA/.726/883 2nd
    BOS/.724/872 3rd
    PHI/.705/820 5th and 4th
    MIN/.703/817 7th and 5th
    COL/.702/804 8th and 6th
    TBR/.702/803 8th and 7th
    TOR/.706/798 4th and 8th
    MIL/.691/785 10th and 9th
    TEX/.705/784 5th and 10th

    Additionally, the top 10 teams in runs are also the top 10 teams in OPS and SLG, though the orders are a little more skewed than that of combining BA+SLG.

    The one category that tells you nothing on it’s own is OBP. 3 of the top 10 teams in OBP aren’t in the top 10 in runs. Philadelphia is last in their division in OBP but 1st in runs by a long shot! That may be the stat of the year, ATL, FL, NYM and WASH, all ahead of the Phillies in out avoidance. The Mets are 12th in OBP and 25th in runs. Texas was 24th in OBP and 10th in runs.

    I’m not really sure how much more can be known that will significantly change the game or give an “analytical” team a decided advantage against a more traditional team.

  8. Jeff Says:

    “They have to lump too many variables into one basket, so the end result is somewhere between untrustworthy and ridiculous.”

    The number of variables has no relation to the trustworthiness or ridiculousness of an algorithm. Whether the variables relate to what you’re trying to measure, well, that’s a separate issue.

    “I think what Field F/X will prove is that baseball people get it close to right with the naked eye. It doesn’t mean you stop trying to understand as much as you can but as soon as you know 2 + 2 = 4(F/X + Field Pct = Fielding Rating) you stop working on it.”

    Just because we can “get something close to right” doesn’t mean we stop working on it. We can get a get a pretty idea of whether we have a fever by feeling our head, but we wouldn’t conclude that oral thermometers were a useless innovation.

    If you have the time to watch every MLB game, then you probably don’t have much use for fielding metrics. For those of us who don’t, they can be useful despite their flaws.

    “The same could be said for improvements on OPS. In most cases, OPS states an accurate offensive value, so any improvements on that basic guideline will be minimal.”

    If by “OPS states an accurate offensive value” you mean it accurately measures base percentage plus slugging percentage, then yes. If you’re saying it accurately measures a player’s contributions to run scoring, no.

    “BA is the 2nd most important component to run scoring, behind only SLG, In my opinion.”

    There’s plenty of research on this. Stats like wOBA, and OPS to a lesser extent, correlate much more highly with run scoring. BA doesn’t correlate very well at all.

  9. Patrick Says:

    Jeff, the variables DO relate to what you’re trying to measure. Fielding ratings have the field cut into pie slices and 3 types of ball speeds, slow, med and hard. That’s throwing different variables into the same basket. There are infinite speeds and vectors to consider. Plus, unlike F/X, current fielding metrics don’t account for where the fielder positions himself. Current fielding ratings will be soon obsolete due to F/X and with apologies to the fielding metric “pioneers”, I welcome the accuracy.

    Concerning offense, you have to look at accumulative team totals when analyzing how certain events translate into actual runs. Slugging is the main component that translates to runs, not OBP. The higher the pct of hits that make up OBP, the higher the run total. See post #7 or look it up yourself.

  10. Kerry Says:

    Hossrex has already rebutted Rick, so I won’t add anything to that other than to reply to one of Rick’s comments.

    Rick: “The largest strides in sabermetrics have been made, each stride will get smaller until the science of baseball has been mastered, and near exhaused.”

    Granted that offense has been pretty well understood, but the baseball people (repeat, baseball people) on the panel thought that defense was only 60% understood, and catcher defense 25% (if I remember the numbers correctly). Field F/X will likely close the gap on all but catching fairly fast, but there’s still a lot of work to do on projecting players into the future, too.

    Hossrex: “So until then, physicists will have a high value.”

    Why thank you!

    Patrick: “BA is the 2nd most important component to run scoring, behind only SLG, In my opinion.”

    I hate to disagree (and at the risk of starting a long argument), but OBP correlates more than AVG.

  11. Patrick Says:

    It won’t be a long argument. Hits are obviously the preferred component to OBP, but regardless of hit fueled vs. walk fueled, OBP is virtually useless in the absence of adequate slugging (Mets) but the reverse is not true. Adequate slugging without adequate OBP still produces adequate runs (Rangers).

  12. Kerry Says:

    That’s a little strong, but, yes, SLG is a much better driver of scoring than OBP. When I looked at it, AVG had a 77% correlation to runs scored, OBP 85% and SLG 93% (that was just for one year, but it’s probably similar in other years).

    Patrick, it wasn’t you that I was worried about regarding the long argument… :-)

  13. Cameron Says:

    Meh, the sabermetrics and stats movement’s been a blessing and a curse in my opinion.

    On one hand, it’s made analyzing the game a lot easier, and has made tracking the performance of players much easier, immensely helping baseball offices and other people involved with the game (video game designers are up there, which I thank), and have opened the possibility of a better game for the future due to these methods of statistical analysis.

    …On the other hand…

    It has given rise to the “only thing that matters are the stats” brand of jackasses. These people kill so many conversations about baseball I could open a graveyard tonight, I swear. It’s made baseball near-unbearable to talk about with anyone except the casual fan. For me, that means going outside and talking to people. Granted, I live in a pretty relaxed baseball environment, so I’m kinda lucky…

    But honestly, now you have people tracking the most inane shit and I just want it to stop. I want baseball to be fun again.

    …As much as I hate to rag on a fellow Kansan, Bill James killed a lot of the fun baseball had.

    The older fans seem to care less about the stats. I have old teachers and an 83 year old grandmother who I love to talk about the game with. Some current stuff, and some historical, but that’s the most fun I have, hearing about the old greats firsthand from the fans.

    The current generation of fans has fallen into the Moneyball frame of mind, where stats make the world go ’round. As much as Joe Morgan is laughable as a commentator, I respect him in the fact that stats aren’t everything about the game.

    Do numbers create memories? No. To me, they create headaches.

    …God help the future of baseball if this is what the fans are going to become.

    -Cameron Nelson

  14. Tonus Says:

    If I own a sports team, then I have sunk several million (possibly tens or even hundreds of millions) of dollars into an investment. I would be looking at every possible way of getting the most out of that investment. That means embracing analytics but NOT rejecting traditional methods of performance evaluation. Which means that I want to staff my organization with people who do not reject either of those fields out of hand. People who will find ways to make those work together.

    I certainly hope that owners of major sports organizations aren’t getting their guidance from people posting pro/anti-sabermetric ramblings on the internet. Or from grumpy self-styled old-schoolers who balk at anything new. Or from snarky know-it-alls who would scrap everything that doesn’t involve a complex math equation. That’s all well and good for internet forums and blog comment sections, but it scares me to think that in this day and age, owners would still kneecap themselves by reflexively rejecting new ways to evaluate talent (or blindly accepting them). Not when there are anywhere from seven to nine figures (or even ten!?!) worth of overall value involved.

  15. Cameron Says:

    Not saying that stat analysis isn’t helpful to the owners. It’s made things a lot better on owners and will be able to help owners style their teams into what they want. Doesn’t always work out (most of Billy Beane’s signings post, what, 2004?) or if the owner doesn’t seem to know jack about the game (Dayton Moore, John Coonely).

    The fans just get annoying sometimes. Mention WAR and the shit hits the fan.

  16. Kerry Says:

    Tonus: “… I want to staff my organization with people who do not reject either of those fields out of hand…. I certainly hope that owners of major sports organizations aren’t getting their guidance from people posting pro/anti-sabermetric ramblings on the internet.”

    The owners and GMs at the conference certainly understood that a blend of the old and the new ways of evaluation (not to mention the possible future, such as more psychological testing) is the way to go.

    Cameron: “God help the future of baseball if this is what the fans are going to become.”

    I think there are a lot of fans in a lot of sports who don’t care about the nitty gritty details that really go into being good at that sport, and the fact that a subset of fans does care doesn’t hurt anything. (How many NASCAR fans actually work on their own cars?)

  17. Cameron Says:

    For the record, I’m not saying the casual fan is dead. I’m just saying the stat supremacists are becoming quite a bit more prevalent in recent days. Hopefully a large part will still be casual, though.

  18. Raul Says:

    I would like to encourage the “geeks” to actually suit up and play baseball.

    And that’s all I have to say, for now.

  19. brautigan Says:

    LOL @ Raul: Raul, I’m actually one of those that suited up and turned into a geek! If geeks suited up, I am afraid it would like that craptacular movie, “The Benchwarmers”.

    Cameron: “Bill James killed a lot of the fun baseball had”. I had the reverse effect. I bought my first abstract in ‘82 and my initial reaction was a mix between horror and curiosity. James killed a lot of preconceptions, and he actually made baseball a lot more interesting in many ways (for me). He also brought up things I had not seen or considered (such as the weird stats put up by Wally Moses), and he had me laughing my ass off with his offensive comments on Enos Cabell. James is a lot of things, but he never killed any baseball enjoyment from me, he enhanced it a great deal.

    One thing I have noted by going to PGE park regularly: there are a lot of older fans keeping score with pencils and their scorecards. They are now being regularly joined by younger fans with their pencils and scorecards.

  20. Patrick Says:

    Yeah, Raul. I’m player first, manager second and geek third. I don’t have to play on Bill James’ team, do I?

  21. Cameron Says:

    braut, I actually don’t hate Bill James all too much. I love the Abstract and own a copy myself, although it does seem terribly dated, any word on a revision that will include this decade?

    It’s just that he’s the one who popularized the movement. Opened the floodgates of SABR hell, so to speak. …I may be shooting the messenger here, but if there’s a better person to blame, you tell me.

    I’m not as negative as it seems, I just have a flair for vitriolic ranting.

  22. Jeff Says:

    Patrick:
    “Current fielding ratings will be soon obsolete due to F/X and with apologies to the fielding metric “pioneers”, I welcome the accuracy.”

    Oh, if you’re just saying F/X will improve upon the current fielding metrics, then I agree.

    “Concerning offense, you have to look at accumulative team totals when analyzing how certain events translate into actual runs. Slugging is the main component that translates to runs, not OBP. The higher the pct of hits that make up OBP, the higher the run total. See post #7 or look it up yourself.”

    The research does look at team totals. The fact is that SLG and OBP each correlate to runs much more highly than BA does, and stats like eqa and wOBA even moreseo. Eyeballing team stats for individual years doesn’t change anything.

    Cameron: “It has given rise to the “only thing that matters are the stats” brand of jackasses.”

    I see some variation of this statement just about every time I come to this site. In reality, I think I’ve encountered maybe 1 “stats-are-everything” person for every 100 people I’ve seen make the statement above. Where do you find these people? The ESPN message boards?

  23. Hossrex Says:

    Patrick: “While it’s certainly not childish to compare baseball to a physical science, it’s important to note that explaining baseball through statistics (sabermetrics) has gone hand in hand since the beginning. So it’s definitely not in it’s infancy.”

    You can look at the disagreement about statistics in this thread, and say sabermetrics isn’t still firmly entrenched in it’s infancy?

    People still don’t agree over which is more useful between batting average, and on base percentage… and those are two FUNDAMENTAL statistics. That would be like physicists who couldn’t agree about the basic laws of planetary motion.

    If any other science were as unrepresentative of the thing they were attempting to quantify, it would be impossible to say any major inroads had been made.

  24. James Kunz Says:

    By the way Hossrex, as wonderfully apt as that Charles Duell quote is, it’s apocryphal

  25. brautigan Says:

    Hossrex:

    “Of all the inventions of man I doubt whether any was more easily accomplished than that of a Heaven.” G.C. Ligtenberg

    I tend to think that we may have come to the end of the line with “baseball stats” and what else they can tell us. How many other variations of the same sets of numbers can we tinker with? How many more “W.H.A.R.F. R.A.T.” numbers can we create?

  26. Jeff Says:

    Much more interesting than the player evaluation stats that have been tirelessly debated here is the bigger picture of analytics that Kerry writes about.

    This includes things like quantifying the relationship between various arm slots and arm injuries for pitchers. Or anything else you can attempt to measure using objective data. Some of this stuff could be fascinating, and we’ve hardly seen the tip of the iceberg yet.

  27. Patrick Says:

    Jeff, when I say BA is more important than OBP all I’m really saying is that the higher the ratio of H/BB contributing to the OBP the better, because BA is the main component of OBP.

    Hossrex, the points that get contested are a very small part of baseball. I’m sure that the two most polar opposite fans on this site still agree on 95% of the sum total of baseball. We’re splitting hairs with most of our debates.

    I agree with Jeff that we haven’t seen the tip of the iceberg but most of the advances will be computer generated like F/X, and not statistical in a traditional sense. More like a scouting tool really, accurate bat speed measurements, reaction times and probably even the speed of balls hit. They’ll probably have computer overlays that show things like opening your hips too early or dropping your shoulder.

  28. Hossrex Says:

    James Kunz: “By the way Hossrex, as wonderfully apt as that Charles Duell quote is, it’s apocryphal”

    Frankly, as I was typing it originally, that possibility occurred to me. Quotes like that usually are.

    Braut: “I tend to think that we may have come to the end of the line with “baseball stats” and what else they can tell us. How many other variations of the same sets of numbers can we tinker with? How many more “W.H.A.R.F. R.A.T.” numbers can we create?”

    You’re basically saying “I can’t think of anything else, so it must be impossible.”

    Thankfully, scientists do not think this way.

    Jeff: “This includes things like quantifying the relationship between various arm slots and arm injuries for pitchers. Or anything else you can attempt to measure using objective data. Some of this stuff could be fascinating, and we’ve hardly seen the tip of the iceberg yet.”

    Beautifully put. I agree completely.

    Jeff: “Hossrex, the points that get contested are a very small part of baseball. I’m sure that the two most polar opposite fans on this site still agree on 95% of the sum total of baseball. We’re splitting hairs with most of our debates.”

    Even accepting that number is true (I think it’s quite high, but I’ll grant it to you for the sake of argument), quantifying the first 95% of ANYthing is the easy part.

    Patrick: “I agree with Jeff that we haven’t seen the tip of the iceberg but most of the advances will be computer generated like F/X, and not statistical in a traditional sense.”

    Traditional sense? I’m not sure what you mean by that. Of course they wont be as “traditional” as hits divided by at-bats, but that doesn’t make them any less statistical, or any less useful.

    Patrick: “More like a scouting tool really”

    All stats are scouting tools.

    In conclusion, it’s human nature (at least amongst people not scientifically minded) to assume that the current level of cultural understanding is the apex.

    Again, thankfully scientists don’t think that way.

    Think about what you did today. The electronic tools you used at work. The electronic tools you used for entertainment.

    Now think back to 20 years ago (for those of us capable).

    Would you have thought computers would be what they are today? Would you have thought your radical Walkman would be replaced by gadget the size of a cassette tape capable of holding 10,000 songs?

    Did you have any concept WHATSOEVER about the global community represented by the internet?

    Of course not.

    Yet still.

    STILL.

    You think what we have today is basically the apex of human development. Sure, you think things will get smaller, or faster, and the newest iPhones might have a couple new apps… but you wouldn’t expect there to be any sort of future GROUNDBREAKING developments like the internet, or (to go back slighty further than 20 years) the microprocessor.

    Why?

    Because it’s human nature.

    Luckily for us all, you’re wrong.

  29. Jeff Says:

    Patrick: “Jeff, when I say BA is more important than OBP all I’m really saying is that the higher the ratio of H/BB contributing to the OBP the better, because BA is the main component of OBP.”

    Yes, hits are better than walks.

    “I agree with Jeff that we haven’t seen the tip of the iceberg but most of the advances will be computer generated like F/X, and not statistical in a traditional sense. More like a scouting tool really, accurate bat speed measurements, reaction times and probably even the speed of balls hit. They’ll probably have computer overlays that show things like opening your hips too early or dropping your shoulder.”

    So you’re saying there will be more advances in data gathering and less in data analysis. Maybe, because there are so many things we haven’t even begun to figure out how to measure. For example, I envision an x-ray that doctors can point at a pitcher and zoom in on his elbow while he’s throwing, and measure what exactly happens to the ligaments. But the analysis would be an important part of that too — eg, how many cells die or fibers get frayed or whatever when he throws at a 38 degree arm angle.

  30. brautigan Says:

    Hoss: I am sure there are more permutations out there, but there is a finite limit. And frankly, I am stumped as to what that information will tell us that we do not already know. The trend will be as to what inefficiencies (are left) to exploit, and/or what changes can we anticipate in the relationship of numbers and perspective?

    And I’m really trying to not bust out laughing, but all I need to say is “Fangraphs” + “Ben Zobrist”.

  31. Hossrex Says:

    Braut: “I am sure there are more permutations out there, but there is a finite limit.”

    Of course there is.

    But where?

    You don’t know, and yet you allude to it being found soon.

    Why?

    Braut: “I am stumped as to what that information will tell us that we do not already know.”

    I.E. “I can’t think of anything, so I don’t think anyone else can either.”

    Braut: “And I’m really trying to not bust out laughing, but all I need to say is “Fangraphs” + “Ben Zobrist”.”

    1: I agree Fangraphs overrates Zobrist to comedic levels.
    2: Zobrist is a hell of a baseball player.
    3: Just because one statistic, by one organization is flawed… doesn’t mean in any way that all statistical analysis is flawed.

    I’ve been on a “judge with your eyes, not with your calculator” kick lately… but that’s BECAUSE I think statistical analysis hasn’t advanced far enough YET… not because I think it CAN’T advance far enough.

  32. Chuck Says:

    “All stats are scouting tools.”

    In what sense?

  33. Raul Says:

    “all stats are scouting tools”

    When a scout goes to an American Legion game and finds a 16 year old kid who hit .530, I guarantee you that scout doesn’t give a damn about his batting average.

    You guys have to understand that scouting isn’t about what a player has done. It’s about what a player can, and hopefully will do.

    I’m not saying stats don’t matter. Sure, you’ll see a scout with a stopwatch and maybe a radar gun, but stats are virtually the last thing he or she cares about when evaluating a prospect.

  34. Patrick Says:

    What I meant by the stat advances not being traditional stats was that they won’t pertain to a player’s game as much as they will his tools, approach and body, like Jeff has already stated so well. Also, those are more scouting tools. Scouts don’t say so and so is getting on base at a .400 clip so he must be good, they say he has good plate discipline and coverage, good contact, quick bat, so he should be successful at getting on base at a high rate.

    It’s still a stat if the computer indicates that you’re dropping your shoulder 3% more than last week or your arm slot is dropped 2.8% and you’re putting 13% more pressure on your rotator cuff. Bat speed and reaction time(0 to 60, so to speak) are also stats.

    As far as baseball stats, there is definitely still work to do but we’re almost there. Once the great computer starts spitting out speeds and vectors of every ball hit plus the results of the play, including arm strength to the 100th of mph over X amount of distance, and that gets calculated into actual runs saved or allowed, it will produce a more accurate WAR type number and thankfully, Pujols will be better than Zobrist again.

    Of course, there is also still work to do on the offensive end but that only requires quantifying the 24 base-out states into the already known data and not calculating only against the league average, but also what an individuals stats are worth in a vacuum. Meaning a walk is worth .3 runs because of the existence of power hitters. Field a team of Marco Scutaro’s and you will see he isn’t worth what WAR claims he is, offensively anyway.

    Yes, the first 95% of anything is the easiest to understand but I wasn’t saying that only 95% is known, I was saying that polar opposites still can agree on 95%. One of those polar opposites is wrong and one is right about most of the remaining 5%.

    So I think the collective understanding of offensive baseball stats are close to 99%.

    When I get in a debate with Shaun about out-avoidance, it’s not that I disagree with the importance of a high OBP. We disagree about the percentages.

    Kerry wrote; “That’s a little strong, but, yes, SLG is a much better driver of scoring than OBP. When I looked at it, AVG had a 77% correlation to runs scored, OBP 85% and SLG 93% (that was just for one year, but it’s probably similar in other years).”

    That looks right to me with a minor disclaimer that BA drives OBP about 3:1. Demeaning the importance of BA while extolling the value of OBP is like saying John, Paul and George had little value but the Beatles were the most valuable band in the world. That’s been my point for these past few years, that SLG is the most important component of scoring runs and BA and HR drives SLG, especially in modern baseball. All of the analysis in the world isn’t going to change that simple fact.

  35. Kerry Says:

    Patrick, I think we do mostly agree and what we are “arguing” about is partly semantics. However, even though AVG makes up a big part of OBP, it is just plain missing some information (i.e., walks and HBP), and for some players that can be an important part of their value. So why rely on an incomplete measure when there’s a better one? There’s a reason we use OPS, not APS :-)

    Case in point: Carlos Pena in 2009. You look at his AVG (.227) and think he’s not very good, even with a decent SLG (.537). Or at the least you think he has plenty of room for improvement. But his OBP is .356 (he had 87 walks and 107 hits), which is actually above average by more than 20 points, and when combined with SLG makes him look pretty darn good (OPS+ of 130). Now would you like for him to trade some walks for singles? Sure, but he’s plenty good as is.

    If we could get the casual fan to embrace OBP and SLG (and OPS) rather than AVG, I would be happy. And as Patrick mentioned, it would be nice if they understood that SLG is the more important part of OPS. But they don’t necessarily need to accept the more sophisticated measures that wring out the last few percent of run value. (OTOH, the teams themselves do, because they want any competitive advantage they can get.)

  36. Patrick Says:

    Kerry, of course I agree. Sometimes it may seem like I champion something like the value of RBI rate or BA as a lone indicator but I’m only trying to make sure that they’re INCLUDED in the big picture. I’m usually only responding to a statement that says RBI rate and BA are poor indicators of value. They’re simply not.

    Pena is a good example. An .893 OPS is productive regardless of how you look at it but it’s important to distinguish the elements that make up the .893.

    A .300/.356/.537 for a 130 OPS+ is more productive than a .227/.356/.537 for an equal 130 OPS+, though it’s important to note that the .300 BA version has a lower Isolated Power. The other spect of valuing the difference is the .300 hitter has substantially more AB’s and the .537 is sustained for a substantially larger amount of attempts, because a walk is a non entity towards the slugging equation.

    So when Barry Bonds slugs .700 but walks 230 times, the slugging value is diminished because it’s only based on 400 of his PA’s instead of the 550 most players would have.

    I do agree about OPS and the casual fan, just understanding that probably puts you to 95% accuracy of value. When you compare team OPS with actual runs scored, it runs very similar. From there, it’s all about adjusting for the components of the OPS (BA:BB:TB:PA) and add in base running prowess.

    Someday, physics may find worm holes that actually allow travel through dimensions. That ain’t gonna happen with baseball. It’s almost understood.

  37. Raul Says:

    An OBP of .370 means dogsh*t if the hitter is batting .355

  38. brautigan Says:

    Hoss, I know what you mean. And of course I am prone to sweeping generalizations, so give me some slack. I think Baseball stats are like toilet paper, how much more improvement can you have? (Or find) Because I can’t think of it, doesn’t mean it does not exist. It’s just that since Bill James, Pete Palmer and John Thorn, there have been oodles of intelligent and pragmatic souls out there scouring the numbers for relevance. I just think we may be coming to the end of the line. I may be right, and of course, I may be wrong.

  39. Patrick Says:

    Shameless plug. RPR also tells you a lot. Texas had a .320 OBP(24th) and 1.005 OPR(28th) but were 2nd in the majors with a .121 RPR, thus, they were 10th in overall runs despite less chances to score. The Cubs on the other hand were 4th in OPR (1.105) but 28th in RPR (.098) so they only finished 22nd in overall runs.

  40. Raul Says:

    Brautigan,

    The only thing these advanced stats do is they’re going to tell us what happened (perhaps) with greater detail/accuracy.

    The stats have never, and will never, improve nor develop the game in a better manner than has already been done for over 100 years.

    The real focus and growth within baseball should be on the scientific and medical aspect of player development, injury, and recovery.

  41. brautigan Says:

    Raul:

    I agree. You would think someone would have implemented a 4 man pitching rotation by now. But, tradition is sometimes an ugly beast. We shall see…..

  42. Raul Says:

    Brautigan,

    I’d love to see what a team with a 4-man rotation would look like in this age.

    My fear is that teams would have 4 starters who each pitched 4 and 2/3 innings, and then bring in 6 relievers for the final 4 and 1/3 of the game.

    Mike Mussina started his career in 1991 and is 66th on the All-Time Innings Pitched list with 3,562.2.

    The leading active pitcher who’s started his career after Mussina is Javier Vazquez (he’ll be 34 in June) and he’s at 2,490…..1,072 innings off.

    I doubt that any pitcher who started his career after Moose will reach that many innings. They’re geared to pitch less and are taken out of trouble very quickly these days.

  43. Shawn Says:

    Very well said Raul. Brautigan I also agree. I hate the set in stone five man rotations, 100 pitch count, and using six relievers in two innings way that managers have decided is the norm. That and closers who can’t go more than one inning. You’re not a real pitcher if you can’t pitch more than one inning to me.

  44. Patrick Says:

    I could see Javier Vazquez reaching Moose in IP. He’s an old fashioned work horse.

    I’d love to see someone return to the 4 man rotation. Look at the Yanks. CC, AJ, Vazquez and Pettite for 5, 6 or 7 IP every 5th day. They could have both Joba and Hughes in the pen setting up Mo. It’d be perfect.

    Besides boxing, I can’t think of any other athletic endeavors that require 4 days of rest before doing it again.

  45. Chuck Says:

    I can see a four man rotation at some point down the road. The changes Nolan Ryan instilled in the Rangers have already spidered throughout baseball, with Seattle, Kansas City and Tampa being three teams who have adapted his methods.

    Having four solids and a fifth, swing type guy, like Bob Stanley used to be with the Red Sox is the ultimate rotation design.

    Four guys making 35-38 starts, (ideally), with the fifth guy getting maybe 15-20 starts and another 60-70 innings as a long reliever.

    The amount of innings the remainder of the bullpen would have at their disposal wouldn’t change all that much, if you use them correctly.

    No reason why a guy couldn’t pitch two innings today, one batter tomorrow, a day off, another inning or so, and so on.

    We’re all big BR or Retrosheet users here, go back and look at the ‘71 or ‘72 Orioles and how many innings they threw despite having four 20 game winners.

  46. brautigan Says:

    Chuck: Spot on.

  47. Hossrex Says:

    I think you’re going to have a difficult time convincing pitchers used to making 33 starts per year to instead make 38.

    Don’t get me wrong though, I’d love to see it. I’d love to see a team with four starters each make 40-41 starts.

    I think CC and Halladay could probably do it… at least CC could have two years ago, and Halladay could have 5 years ago.

    But I just don’t see pitchers voluntarily shortening their careers (whether that would happen or not, that would be the perception, regardless of the GOOD reasons it might not).

    The second thing I’d do if I had a time machine would be to watch the 1884 Brooklyn Atlantics play opening day at Washington Park.

    http://superdickery.com/index.php?option=com_content&view=article&id=516:worst-explanation-of-time-travel-i-ever-saw&catid=33:weird-science-index&Itemid=37

  48. Raul Says:

    I should have known as soon as the website said “superdickery.com”

  49. Chuck Says:

    The Rays led the Majors last season with 970 starter IP.

    That’s six innings per start, which is really low.

    Stretch it to seven innings, the total is 1134.

    Say you have four regular starters, a fifth, swing guy and maybe get starts from three other guys and they split the extra 162 innings, that’s 20 additional, per season, per guy.

    Over 35 starts, that’s what, an extra 2/3 of an inning?

    Doesn’t seem like alot, does it?

  50. Raul Says:

    It doesn’t seem like a lot.

    And if I were a player on that team, it would “feel” like a nice psychological boost to know that my relievers just have to pitch the 8th inning (assuming the closer only comes in for the 9th).

Leave a Reply


RSS
Categories
Fan Duel
FanDuel - Daily Fantasy Baseball
YardBarker
Advertisement