@estang said:
I don't begin to understand dWAR and I really don't care to delve into it.
I go back to the overall WAR number and that Ozzie Smith walked about once more every 4 games and his overall WAR dwarfs Larry Bowa.
The three best shortstops of the 1970s (off the top of my head) were Concepcion, Bowa and Bill Russell. They played for winning teams and made all-star teams. Ozzie came along late in the decade & flashed greatness in the field, while flipping around like one of those wind-up toys.
I just cannot look at the overall WAR between Smith & Bowa and draw a conclusion from that single number that Ozzie was more than 3 times better than Bowa. It's absurd and doesn't pass the sniff test...
That's the kind of stuff I was looking for in this discussion. There simply is no way Ozzie Smith was saving that many more balls than Larry Bowa(who also had range), made every play a shortstop can make, and was as sure handed as they come.
Bowa was certainly not a statue like Luzinksi, and while I understand why that comparison was made, it does not apply to Bowa in any way. Bowa had more lateral quickness range than someone like Cal Ripken who is viewed as an excellent defender. Bowa was not slow or limited in range in any way.
It is all the other factors I mentioned that made those 'metrics' view Ozzie as being that much better, when the reality was that Ozzie was simply getting more playable balls to field than Bowa got...and Bowa made less errors to make up for the five spectacular plays a season that Ozzie 'may' have made that Bowa 'might' not have made.
There simply is no way Ozzie Smith was saving that many more balls than Larry Bowa(who also had range), made every play a shortstop can make, had a plus arm, and was as sure handed as they come.
Bowa was certainly not a statue like Luzinksi, and while I understand why that comparison was made, it does not apply to Bowa in any way. Bowa had more lateral quickness range than someone like Cal Ripken who is viewed as an excellent defender. Bowa was not slow or limited in range in any way.
Bowa had speed, quickness, and agility. Bowa stole 318 bases in his career, so he could most certainly run. For comparison Belanger only had 167. While that isn't a pure determination of speed, we all saw that Bowa was not slow in any way. He could run and was quick.
So this notion that Bowa did not have physical foot speed range in the field is absurd. Bill James only came to that conlcusion based on the fielding statistic he created, the same one that simply does not account for the actual number of playable balls hit their way, and DOES INDEED also rely on assist and putout totals to manufacture his number...all of which are skewed and not equal among players.
I don't know if Bill James needed glasses or depth perception goggles or not because if he watched Bowa physically move, he would not come to the conclusion that Bowa was physically limited in his range at SS...and certainly not a statue.
It is all the factors I mentioned that made those 'metrics' view Ozzie as being that much better, when the reality was that Ozzie was simply getting more playable balls to field than Bowa got...and Bowa made less errors to make up for the five spectacular plays a season that Ozzie 'may' have made that Bowa 'might' not have made.
It sounds like what you want is a stat that takes into account how many strikeouts the pitching staff had, in addition to the basic stats like A, PO, DP, and E. And for it to be more accurate it would also take into account how many base runners were already on base. And for it to be really accurate, it would take into account how many innings were being pitched by LHP and RHP, since more innings pitched by LHP means more RH hitters and more balls hit to the left side. And once you'd done all your calculations for one SS, you'd want to compare that to how other SS were doing on the same basis. Problem is, you'd need a massive database to do something like that, and a whole lot of time. I've never done it and I'm never going to. You've never done it, and you're never going to. So where does that leave us?
I won't speak for you, but where it leaves me is relying heavily on the one person who did assemble that database, and did do all of those calculations for every SS in history and explained it all in painstaking detail in a series of books that I have read and understood. And there may be a nit to pick here and there in some of his defensive evaluations, but there's literally nothing with which anyone could take serious exception. And he rates Ozzie Smith an A+ SS - as does every other person on the planet besides you who ever saw him play. He rates Bowa a C shortstop, which is pretty much where I thought everyone rated him until this thread.
Again, I don't want to speak for you but my suspicion is that the variable you are missing is innings pitched by LHP. I didn't review their entire careers but I looked at 1972 (Bowa's best year) and 1982 (one of several of Ozzie's phenomenal years). And the Phillies pitched 1,400 innings with 701.2 by LHP (50%) while the Cardinals pitched 1,465.1 innings with 293.2 by LHP (20%). The Phillies pitchers did strike out 200+ more hitters, which Bill James absolutely accounts for, but with 1,000 or so non-K outs recorded with a LHP on the mound, there were a whole lot more outs for the Phillies SS and 3B to make, too. If you've read Win Shares, I'd be happy to pick nits on his methods with you. If you haven't read it, then you should. Everyone should.
And for the record, I was always a fan of Larry Bowa. Being an average major league SS for 2,000+ games probably makes him one of the 30 or 40 greatest SS of all time. If he could have hit a lick he'd probably be in the HOF. If I could take out Jack Morris and put Bowa in I'd do it in a heartbeat. But there is nothing in the universe that would make me believe he was as good a SS as Ozzie Smith or that Vic Wertz was as good a hitter as Mantle. They are equally preposterous statements, and until this thread I's have laughed if someone told me there was even a single person who agreed with either statement.
Regarding Templeton from someone else's comments, I remember thinking "THIS (Ozzie Smith) is all we could get for such a great player??". But my next thought was "but then, what is the market value of an a$$hole?". Templeton had to go, on that point there was no debate. But the whole league knew that, and knew we were beggars, not choosers. Templeton did grow up after being traded for a .220 hitter, which I assume bruised his ego tremendously, but the Cardinals still came out miles ahead on that trade which was a huge and pleasant surprise.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
I think win shares is deliciously arcane. I would love to have someone explain it. Past the headlines. Like explain the team and game context data and how it was developed! Don’t explain the linear normalization or the pool hierarchy. Explain the stories that make the spreadsheet work.
Explain it. No one ever will because it’s so arbitrary in spots. Trust and faith.
I think we need to take a longer look at advanced pre participle nano distribution theory in a contiguous multi transference hypothesis relative to a flowing continuum to better measure a players defensive value.
@bgr said:
I think win shares is deliciously arcane. I would love to have someone explain it. Past the headlines. Like explain the team and game context data and how it was developed! Don’t explain the linear normalization or the pool hierarchy. Explain the stories that make the spreadsheet work.
Explain it. No one ever will because it’s so arbitrary in spots. Trust and faith.
This is an odd comment. I mean, there's a book called, appropriately, "Win Shares", that explains what Win Shares are in excruciating detail. If you haven't read it then, well, read it. Bill James holds the readers hand and walks him through how Win Shares are calculated from start to finish, and at every position for fielding. And if you have read it and still find them arcane then nobody else is going to be able to explain them better than the man that invented them.
As I said, they are hardly Gospel but only one person in history has attempted to do what Bill James did, and while he surely made some mistakes and more surely made some assumptions that are different than other people might have made, he is 100% transparent about what he did so we are all free to debate or nitpick what he did from a level playing field. I can't imagine anything more ignorant than dismissing what he did. Well, the pinnacle of ignorance would be dismissing what he did without even reading his book. That's willful ignorance, and I've got no time for debating the willfully ignorant.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
But that’s the point. I understand his work also and I subscribed online years ago. I haven’t checked his site in many years but it hasn’t been updated for many years last I had checked.
You want people to believe because he compiled a lot of information that it’s meaningful. Because you’ve reviewed it. You understand his basic math and variable selection and normalization method.
It’s telling that he realized that win shares required negative win shares. More so that an acolyte won’t defend their masters hallucinations - I liked how you used that and it has the feels.
He hasn’t updated his site since I was last a member it appears. No new data to siphon.
It’s tough to defend a system which has been obsoleted. I just wanted you to frame how Bill constructs his… value. I guess people just have to trust you. End of the day it was his own rigid stubbornness which launched WAR into general acceptance.
Books. And handbooks? Because a lot changed since 2002 in how he approached the problem. Are you talking about before he basically did what WAR does or after?
I love how you just assert you’re correct and that I’m too stupid to understand it. Basic.
If you've read his book - and you never said you did - and still don't understand Win Shares then I don't have to call you anything. You are what you are.
If you haven't read it then I stand 100% by what I said.
I've never been on his website so I don't know what you have and haven't seen. And maybe things have changed since Win Shares was published, but it makes no difference. Nobody is in any position to debate the system's merits or deficits if they don't understand what they are. What is it, specifically, about the computation of Win Shares for shortstops (or anyone else) with which you take exception? I'm happy to discuss or debate it. But, as I said, I have no time to debate Win Shares with someone who has willfully chosen to remain ignorant about what they are. First, read the book. Second, tell me what you read in it that you think is wrong. Third, we debate Win Shares. I'm not starting at 3, it would be pointless and a waste of everyone's time.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
I have read his books. I’ve also built contexts with his handbook updates which show his gradual acceptance of his error.
What do I mean by “read his books”. I’ll be honest. I thought the opening section where he goes through the rules and the building of the statistic to be peculiar and monumentally tedious. Big assumptions on one page become the bedrock of ground truth on the next. I found his analysis of players to be the most valuable. I don’t disagree on his analysis on players but his system is just contrived by him to match his own labels. It’s one of those books where you see the problems and you want to skip ahead because you’re mired in the mind of a mad scientist.
His system is based on what actually happened rather than what might have happened. Over the years he has adapted to reason by adopting what WAR already does. Anyways you keep saying books and I don’t know if you’re talking about his 2002 books or all of the updates through the years.
Win shares is his original and then he adapted it slowly towards WAR by adding loss shares but that was released in a series of updates online which are not in his book. I just don’t want to start with a dissection of win shares to find out we’re talking about “WAR with different weights” - win/loss shares.
As he improved his system his correlation with WAR increased. That’s telling. As his accuracy improved he approached the accepted baseline.
I understand why you don’t want to explain win shares from his book yourself. It sounds like a witches brew.
When he rolled out loss shares around 2008 he (Bill) explained that his original allocations were the result of a “bookkeeping mistake” - he had taken a pitchers positive runs that a pitcher had created and awarded them as runs the pitcher had prevented. This constructed his entire 52-48 allocation for pitching vs defense which he corrected to 50-50 once he realized his error. Online. Not in the book. You haven’t read all his corrections to the work you hold as word of god and that makes you the bozo here. Again. What an absolute joke you are. His book is interesting if tedious and error rich but hey. It’s a great door stop.
I think Bill has done us a great service in his analysis but it’s still less accurate than WAR. His abstract is more interesting as even a coarse stat can shine a light on interesting comparisons. You should really update yourself with what Bill has said about his original work because you’re promoting something he left behind and acknowledged errors within.
Before you tell the people here they should read his books maybe also explain what they will be reading. His main book is something most people would need to be forced to peruse to gain an understanding. These are manuals with brutal verbosity - and that should say a lot coming from me.
It sounds like what you want is a stat that takes into account how many strikeouts the pitching staff had, in addition to the basic stats like A, PO, DP, and E. And for it to be more accurate it would also take into account how many base runners were already on base. And for it to be really accurate, it would take into account how many innings were being pitched by LHP and RHP, since more innings pitched by LHP means more RH hitters and more balls hit to the left side. And once you'd done all your calculations for one SS, you'd want to compare that to how other SS were doing on the same basis. Problem is, you'd need a massive database to do something like that, and a whole lot of time. I've never done it and I'm never going to. You've never done it, and you're never going to. So where does that leave us?
I won't speak for you, but where it leaves me is relying heavily on the one person who did assemble that database, and did do all of those calculations for every SS in history and explained it all in painstaking detail in a series of books that I have read and understood. And there may be a nit to pick here and there in some of his defensive evaluations, but there's literally nothing with which anyone could take serious exception. And he rates Ozzie Smith an A+ SS - as does every other person on the planet besides you who ever saw him play. He rates Bowa a C shortstop, which is pretty much where I thought everyone rated him until this thread.
I actually don't want any of that. In order to know which shortstop is getting to more balls, and thus more range, I want to know how many routine ground balls were hit within two steps of their positioning either to the left or to the right, 10 feet to their left, 10 feet to their right. 15 feet to their left, 15 feet to their right. How many they converted and how many they didn't.
Then you have to go further into knowing the velocity of the balls and the types of hops they were taking. Usually you can see this, and Ozzie didn't get to any more balls under what I am looking for than Bowa. Ozzie simply had more of them hit his way, so yes, he technically got to more of them, but he had more opportunities.
James tried to make a point with Bowa, since Bowa was more sure handed than Ozzie Smith, more sure handed than Belanger, James was trying to show that fielding percentage wasn't a good measure, and his model showed Bowa didn' have range and he used him as an example to highlight his premise. He created a statistical model that did not account for the most important things that mattered in determining range(what I asked for above), instead he relied on a calculation that assumed things would be equaled out over time based on team strikeouts, hits, etc.....although he didn't even know the amount of ground balls or fly balls.
In the end, his model was missing the most important parts, and he called Bowa a statue, of which Bowa was anything but.
The vast majority of ground balls are of the MLB routine variety(not Joe Schmoe) routine. Give Bowa the same amount of chances as Belanger or Ozzie and he fields 99% of those balls too....because Bowa moved just as well, and his hands and arms were better.
James needed glasses.
PS what were Bowa and Ozzie's final total career defensive win shares from james?
OK, and what I want is a billion dollars. I was thinking you wanted something from among the theoretically available options. Since what you want isn't theoretically possible, then that's it. The beginning, middle, and end of your argument is that we have absolutely no way of knowing how good a fielder anyone is: we can't believe the stats and we can't believe our own eyes. It's a logically tenable position, but one which is also obviously immune to debate or discussion. It makes your OP that much more bewildering, but I'll have to let that go since neither you nor I has even a shred of the evidence you think is needed to make or refute such a claim.
And as with bgr, I just don't see the point of debating the merits of Win Shares with someone who hasn't read the book. James didn't call Bowa a statue so I have no idea why you said that he did. You said "in the end, his model was missing the most important parts" when you don't know what parts it has. Yes, it is "missing" data that doesn't exist - the data you say you want. But you are also missing that data so you just make it up in your head and draw conclusions from it.
James tried to estimate ground balls and fly balls, you just make stuff up. You also said James "used Bowa as an example", and I don't know what you mean by that. Where did James use Bowa as an example, and as an example of what?
Anyway, I'm done. If you or anyone else wants to discuss/debate something from James' book I'm your huckleberry. But debating things you or bgr have imagined might be in James' book is just frustrating. Read the damn book before you opine on it. Is that really too much to ask? And if I'm way off base here and you have read it, then quote something from it and tell me why it's wrong or could be improved. Such examples no doubt exist, but James not accounting for the velocity of the grounder that Jerry Grote hit off Dick Ruthven in the bottom of the 7th is not one of them, because that data doesn't exist. If your argument is more than "my preconceived notions are better than James's analysis of all of the available data", then I've missed.it.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
now it sounds like you're talking about the BIS data when you refer to James' work with ground balls vs. fly balls. the fielding bible. the real game info. The BIS charting wasn't really fully cooked when James released his books, but James then pivoted away from legacy, and error prone, traditional record and range factor with his (somewhat arbitrary) positional coefficients. Win Shares was based on this legacy data and Bill says he wanted to be consistent over accurate and that's why when data got better, he kept the input the same - consistency. so no. the original Win Shares does not include ground ball vs. fly ball or batted-ball location. The batted ball charting data is pretty awesome and I have digitized it... from Dewan's fielding bible however. Are you messing with us?
@dallasactuary said:
OK, and what I want is a billion dollars. I was thinking you wanted something from among the theoretically available options. Since what you want isn't theoretically possible, then that's it. The beginning, middle, and end of your argument is that we have absolutely no way of knowing how good a fielder anyone is: we can't believe the stats and we can't believe our own eyes. It's a logically tenable position, but one which is also obviously immune to debate or discussion. It makes your OP that much more bewildering, but I'll have to let that go since neither you nor I has even a shred of the evidence you think is needed to make or refute such a claim.
And as with bgr, I just don't see the point of debating the merits of Win Shares with someone who hasn't read the book. James didn't call Bowa a statue so I have no idea why you said that he did. You said "in the end, his model was missing the most important parts" when you don't know what parts it has. Yes, it is "missing" data that doesn't exist - the data you say you want. But you are also missing that data so you just make it up in your head and draw conclusions from it.
James tried to estimate ground balls and fly balls, you just make stuff up. You also said James "used Bowa as an example", and I don't know what you mean by that. Where did James use Bowa as an example, and as an example of what?
Anyway, I'm done. If you or anyone else wants to discuss/debate something from James' book I'm your huckleberry. But debating things you or bgr have imagined might be in James' book is just frustrating. Read the damn book before you opine on it. Is that really too much to ask? And if I'm way off base here and you have read it, then quote something from it and tell me why it's wrong or could be improved. Such examples no doubt exist, but James not accounting for the velocity of the grounder that Jerry Grote hit off Dick Ruthven in the bottom of the 7th is not one of them, because that data doesn't exist. If your argument is more than "my preconceived notions are better than James's analysis of all of the available data", then I've missed.it.
For example, here is an old man Bowa on the Cubs ranging up the middle, of which James said he could not do via his statistical model. Starting at the 5:48 mark you can see the nimbleness of Bowa up the middle(f of which James said he couldn't do...and he shows his arm off on a couple plays.
Bowa could move. He was fast, quick, and nimble...nothing Luzinski like about him.
It was obvious to any observer that he was nimble and had range...and had the best hands of his era. He could make nearly any moving play as good as Ozzie or Belanger. He made less errors. He had a better arm....yet these defensive measures are showing ridiculous statistical difference between their runs saved, and you already admitted that the valdiity of the defense measurements is clearly not good, so not sure why it is a surprise that you are so adamant against my point, other than being an Ozzie Smith fan.
The one thing Ozzie probably had over Bowa was the extreme diving play. Ozzie was good at those, but those were few and far between....but Bowa made up for that with a better hands and a stronger arm.
PS...start the video around 5:30 to see a vintage Old Style commercial
Phillies video where Harry Calas calls him all time great fielder later in the video. At the 1:29 mark he shows off his arm. Callas is back half of video.
3:23 Calas called him a glove man with no equal. So I'm not the only one. At the time he had highest fielding percentage of all time.
When I ask you, @dallasactuary, to explain it to me I am not doing so because I don't understand how Bill constructed Win Shares. I want you to explain it because I think that would make it clear to everyone what it is doing under the hood. There are 100-150 pages which walk through the construction of Win Shares and why this and that was selected before it gets into the application by story. Perhaps that's a flaw in itself... that Win Shares can't really be explained without digging that hole yourself. To pretend like it couldn't be understood by most-anyone is ridiculous. I don't recall Bill being a trained statistician, data-scientist, or mathematician.... He was a huge baseball fan who was quite brilliant and incredibly diligent.
What I was after was the next level down which discusses the linear extrapolation once a player has been normalized vs. his team's baseline. The big brush strokes could be covered in a handful of sentences - such as the ratio of real wins to win shares and the distribution of merit for a win - the shares and marginal shares parts. You don't spend any time describing the insightful parts of Win Shares - perhaps because that introduces the misleading parts as well. The only reason I am holding you accountable here is because you've spent a lot of time and effort dissing dWAR while Bill, himself, eventually succumb to it's recipe. You have ignored his further effort past the release of his books in 2002... probably have committed much of the almost 2000 pages of Win Shares and his Abstract to memory.. but failed to keep reading as Bill's system evolved over the years.
I really respect Bill James - the guy was a pioneer and there's probably no sabermetrics like we have today without Bill. His Baseball Abstracts, which I read long after they were published formulate a large part of how I see the game of baseball today. He created the questions and built the groundwork for the answers. But data-driven process for statistical player evaluation was happening already and elsewhere. James synthesized a lot of other work and took it to the next level of rigor... with a healthy dose of self-evangelism.
Win Shares has been replaced by WAR, Statcast has obsoleted his batted-ball tracking, and wOBA has supplanted his runs created and linear weights.
I believe James said about Win Shares that it was designed to ask questions - not answer them - I paraphrase that from what I remember in his book... Win Shares. Since you never were an online member you probably never used his online ask and answer where you could discuss his ideas directly with him - you missed out.
Shoulders of Giants. I learned a few things from Bill James over the years...
Earl Weaver was a really fascinating dude... He was using a form of sabermetrics during his tenure as Baltimore's manager starting in the 1970s. He had his index cards on every hitter - how they performed against every pitcher and he was one of the first to create some of the first platoons - based on data vs. gut feeling. He was using batter vs pitcher and batter vs handedness data decades before most. Outs are precious, OBP and power drive offense, and bunting and base-stealing are usually counterproductive. Weaver has his own book - Weaver on Strategy on these subjects.
Branch Rickey doesn't get enough credit... for many things.. Being possibly the inventor of sabermetrics might be one of them. After... you know... breaking the color barrier and inventing the farm system he figured more was needed. Rickey believed in the 40s that batting average, RBIs, and pitcher wins were misleading and he hired a data analyst to identify the right correlations... and guess what. OBP matters most, batting average misleads, and offense and defense decompose into measurable run contributions... gee. I wonder where I have heard this before? Was it... Win Shares? Have you read it? It's a fun read on the beach. On Branch Rickey there is a 1954 Life which has an article on it. https://invention.si.edu/invention-stories/branch-rickey-baseball-innovator
Another very insightful book that came long before Win Shares was The Hidden Game of Baseball. Palmer was an actual statistician and he gave us the first real actionable method for run estimation. His linear weights are still used today - refined certainly, but they form the backbone of modern run-estimation thinking.
@Darin said:
Dang….. I was just about to explain the linear normalization and the pool hierarchy. ☹️
I guess I’ll just save it for another time and place.
I appreciate your protective instincts. You guys are buds so protect your homie. Dallas is off base here and he's been slingin' mud - he's a mud slinger. I'll certainly stop sending it back when you guys grow up. I feel the embarrassment for him and I know he's gonna die on this hill because he can't walk it back or take the bridge when it's there. He just sees red.
Linear
Normalize
Pool
Hierarchy
Linear Normalization - this is the step in Win Shares which forces every player's individual contribution to sum exactly to the team's total.
Pool Hierarchy - the Win Shares pools cascade through several nested layers, splitting shares at each stage until being distributed to the individual players.
Team Pool - a 50 win team has 150 win shares.
Offense vs. Defense Pool - Team Pool divides into offensive and defensive categories. Dallas likes the 2002 52 vs. 48, but Bill James prefers his 2008 50 vs. 50. Deeper Dive - Marginal Runs.
Defense Pool branches into Pitching vs. Fielding - This is where James encountered one of his errors.
Now prepare yourself for about 40 pages discussing how the apportionment happens...
The team-level pitching/fielding split is one of the more contentious issues from Bill's work and this relates to his "defensive efficiency". Doing this at the team level is a coarseness that may have been necessary at some point, but I think it introduces much unnecessary error.
Point being; I was asking him to explain these two foundational parts of Win Shares to show that I know where the skeletons are and that I am familiar with the material from the book. To which he just pesters that I'm too stupid and haven't read his books so he doesn't need to say anything other than I'm stupid and he's right. Hitch your wagon to him all you want - I at least support my opinion and I'm not denigrating his.
Someone saying dWAR is terrible and Win Shares is great isn't making a whole lot of sense. I don't like or dislike any of you guys and I'm certainly not going to embark of a path of self-delusion just because I think someone's a total clown. Support it... don't. you'll look like a fool to people with sense.
@Darin said:
Dang….. I was just about to explain the linear normalization and the pool hierarchy. ☹️
I guess I’ll just save it for another time and place.
I appreciate your protective instincts. You guys are buds so protect your homie. Dallas is off base here and he's been slingin' mud - he's a mud slinger. I'll certainly stop sending it back when you guys grow up. I feel the embarrassment for him and I know he's gonna die on this hill because he can't walk it back or take the bridge when it's there. He just sees red.
Linear
Normalize
Pool
Hierarchy
Linear Normalization - this is the step in Win Shares which forces every player's individual contribution to sum exactly to the team's total.
Pool Hierarchy - the Win Shares pools cascade through several nested layers, splitting shares at each stage until being distributed to the individual players.
Team Pool - a 50 win team has 150 win shares.
Offense vs. Defense Pool - Team Pool divides into offensive and defensive categories. Dallas likes the 2002 52 vs. 48, but Bill James prefers his 2008 50 vs. 50. Deeper Dive - Marginal Runs.
Defense Pool branches into Pitching vs. Fielding - This is where James encountered one of his errors.
Now prepare yourself for about 40 pages discussing how the apportionment happens...
The team-level pitching/fielding split is one of the more contentious issues from Bill's work and this relates to his "defensive efficiency". Doing this at the team level is a coarseness that may have been necessary at some point, but I think it introduces much unnecessary error.
Point being; I was asking him to explain these two foundational parts of Win Shares to show that I know where the skeletons are and that I am familiar with the material from the book. To which he just pesters that I'm too stupid and haven't read his books so he doesn't need to say anything other than I'm stupid and he's right. Hitch your wagon to him all you want - I at least support my opinion and I'm not denigrating his.
Someone saying dWAR is terrible and Win Shares is great isn't making a whole lot of sense. I don't like or dislike any of you guys and I'm certainly not going to embark of a path of self-delusion just because I think someone's a total clown. Support it... don't. you'll look like a fool to people with sense.
Hey.. I’m not buds with Dallas. I like it better when it’s tax season and he’s busy hiding under his desk from his clients and not posting here. Lucky us, tax season is over. 😩
bgr….. I like your posts but I’m just not into the mathematics of the complicated stats like win shares etc.
I look at threads like this more for the entertainment value. Personally I don’t know why Dallas doesn’t acknowledge that Bowa was a great defensive SS. The numbers seem to say he was.
I think we both love baseball Darin... maybe not you so much this year but... in general. I'm not as much of a statistics guy as it might seem - I love baseball for the art and nuance while appreciating the baseline of statistics. It's the randomness in baseball that wins a World Series or throws a Perfect game. The moments. I think the statistics are really cool, but I'm hardly dominated by them. So I like to see the approach of it all while appreciating the game for what it is. The concept of best player or best shortstop is something I find incomprehensible - I can't even begin to construct this while some people can just rank order a list from 1 to 100. Understanding the statistical methods is something I enjoy doing. I feel like I could learn something from dallas, but I don't think I have ever seen him actually discuss something without belittling or berating the people he's debating. Jousting with him is pretty fun too though so it's not all bad news. This place is hella-fun. Some days it's a zoo and some days I'm just reading cool posts while I watch a Brewers game. This place needs more pepper though. I'm from Minnesota and it's not spicy enough for me.
@Darin said:
Dang….. I was just about to explain the linear normalization and the pool hierarchy. ☹️
I guess I’ll just save it for another time and place.
I appreciate your protective instincts. You guys are buds so protect your homie. Dallas is off base here and he's been slingin' mud - he's a mud slinger. I'll certainly stop sending it back when you guys grow up. I feel the embarrassment for him and I know he's gonna die on this hill because he can't walk it back or take the bridge when it's there. He just sees red.
Linear
Normalize
Pool
Hierarchy
Linear Normalization - this is the step in Win Shares which forces every player's individual contribution to sum exactly to the team's total.
Pool Hierarchy - the Win Shares pools cascade through several nested layers, splitting shares at each stage until being distributed to the individual players.
Team Pool - a 50 win team has 150 win shares.
Offense vs. Defense Pool - Team Pool divides into offensive and defensive categories. Dallas likes the 2002 52 vs. 48, but Bill James prefers his 2008 50 vs. 50. Deeper Dive - Marginal Runs.
Defense Pool branches into Pitching vs. Fielding - This is where James encountered one of his errors.
Now prepare yourself for about 40 pages discussing how the apportionment happens...
The team-level pitching/fielding split is one of the more contentious issues from Bill's work and this relates to his "defensive efficiency". Doing this at the team level is a coarseness that may have been necessary at some point, but I think it introduces much unnecessary error.
Point being; I was asking him to explain these two foundational parts of Win Shares to show that I know where the skeletons are and that I am familiar with the material from the book. To which he just pesters that I'm too stupid and haven't read his books so he doesn't need to say anything other than I'm stupid and he's right. Hitch your wagon to him all you want - I at least support my opinion and I'm not denigrating his.
Someone saying dWAR is terrible and Win Shares is great isn't making a whole lot of sense. I don't like or dislike any of you guys and I'm certainly not going to embark of a path of self-delusion just because I think someone's a total clown. Support it... don't. you'll look like a fool to people with sense.
That's a good explanation.
Per the part where Bill wouldn't explain, don't fret, he can't explain because Bill needs a pair of glasses.
For example, here is an old man Bowa on the Cubs ranging up the middle, of which James said he could not do via his statistical model. Starting at the 5:48 mark you can see the nimbleness of Bowa up the middle(f of which James said he couldn't do...and he shows his arm off on a couple plays.
I read this pargraph, rolled my eyes, shook my head, and let out a heavy sigh. I lack the strength to read more.
Show me where "James said" Bowa couldn't do anything. I'd say I'll wait, but we both know you pulled that statement out of your nether regions because James never said any such thing.
As my last post in this thread, and last visit to this thread, I will repeat: read the damn book before you opine on the book. I know you aren't embarrassed by your performance in this thread, but I am embarrassed for you. You are better than this. You are smarter than this. But your willful ignorance regarding the things James actually did say is impenetrable, and I give up, Let me know when you've read the book and I will discuss defensive Win Shares with you. Until then, bye.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
For example, here is an old man Bowa on the Cubs ranging up the middle, of which James said he could not do via his statistical model. Starting at the 5:48 mark you can see the nimbleness of Bowa up the middle(f of which James said he couldn't do...and he shows his arm off on a couple plays.
I read this pargraph, rolled my eyes, shook my head, and let out a heavy sigh. I lack the strength to read more.
Show me where "James said" Bowa couldn't do anything. I'd say I'll wait, but we both know you pulled that statement out of your nether regions because James never said any such thing.
As my last post in this thread, and last visit to this thread, I will repeat: read the damn book before you opine on the book. I know you aren't embarrassed by your performance in this thread, but I am embarrassed for you. You are better than this. You are smarter than this. But your willful ignorance regarding the things James actually did say is impenetrable, and I give up, Let me know when you've read the book and I will discuss defensive Win Shares with you. Until then, bye.
From May, 1981 Sports Illustrated:
"Frequently, our visual sense of the great fielders is surprisingly accurate: Brooks Robinson , Graig Nettles and Schmidt all have had high Range Factors. But the eye can deceive, too. The balloon James most enjoys bursting is that of Shortstop Larry Bowa, who has, he says, "the range of the Birdman of Alcatraz ." During the past few seasons, according to James' figures, Bowa has gotten to and fielded cleanly substantially fewer batted balls than fellow National League shortstops Garry Templeton , Ozzie Smith , Ivan DeJesus , Dave Concepcion and Tim Foli. Bowa is a fielder "who looks good," James says, "on the balls he reaches." "
Like I said, two simple plays in that video, along with countless others, proved James needs glasses.
James uses 'his formula' to come to HIS conclusion...the very, very flawed formula as pointed out. Bowa was quick, nimble, fast, and had a plus arm. Also the best hands in the league while at the time of that 1981 writing, Bowa had the all time best fielding percentage for a SS.
So even James admits that 'the eye' is saying Bowa is a great fielder among those other greats listed...but since his VERY FLAWED formula said otherwise, he reversed what his eyes saw, and made a bad conclusion...just like you did for listening to him like he is a God.
So maybe I was wrong and he didn't need glasses because he admitted Bowa looked just as elite as Brooks and Schmidt...but he had his flawed formula, so he had to highlight something to 'shock' what we all saw...and what common sense says; that Bowa was a faster human than Mark Belanger, Bowa had a great glove, and it is silly to think Bowa couldn't range 20 feet while Belanger could. Just plain silly. Silly because I saw him do it.
I'll say this much... this discussion did prompt me to go back and read some of Bill's assessment of players in his archives... some I find did turn out to be incorrect over time, but the guy does really present valuable insight into the game. I don't recommend anyone studies the machination of Win Shares, but the companion abstract is worth the time.
Comments
That's the kind of stuff I was looking for in this discussion. There simply is no way Ozzie Smith was saving that many more balls than Larry Bowa(who also had range), made every play a shortstop can make, and was as sure handed as they come.
Bowa was certainly not a statue like Luzinksi, and while I understand why that comparison was made, it does not apply to Bowa in any way. Bowa had more lateral quickness range than someone like Cal Ripken who is viewed as an excellent defender. Bowa was not slow or limited in range in any way.
It is all the other factors I mentioned that made those 'metrics' view Ozzie as being that much better, when the reality was that Ozzie was simply getting more playable balls to field than Bowa got...and Bowa made less errors to make up for the five spectacular plays a season that Ozzie 'may' have made that Bowa 'might' not have made.
@dallasactuary
There simply is no way Ozzie Smith was saving that many more balls than Larry Bowa(who also had range), made every play a shortstop can make, had a plus arm, and was as sure handed as they come.
Bowa was certainly not a statue like Luzinksi, and while I understand why that comparison was made, it does not apply to Bowa in any way. Bowa had more lateral quickness range than someone like Cal Ripken who is viewed as an excellent defender. Bowa was not slow or limited in range in any way.
Bowa had speed, quickness, and agility. Bowa stole 318 bases in his career, so he could most certainly run. For comparison Belanger only had 167. While that isn't a pure determination of speed, we all saw that Bowa was not slow in any way. He could run and was quick.
So this notion that Bowa did not have physical foot speed range in the field is absurd. Bill James only came to that conlcusion based on the fielding statistic he created, the same one that simply does not account for the actual number of playable balls hit their way, and DOES INDEED also rely on assist and putout totals to manufacture his number...all of which are skewed and not equal among players.
I don't know if Bill James needed glasses or depth perception goggles or not because if he watched Bowa physically move, he would not come to the conclusion that Bowa was physically limited in his range at SS...and certainly not a statue.
It is all the factors I mentioned that made those 'metrics' view Ozzie as being that much better, when the reality was that Ozzie was simply getting more playable balls to field than Bowa got...and Bowa made less errors to make up for the five spectacular plays a season that Ozzie 'may' have made that Bowa 'might' not have made.
@1948_Swell_Robinson
It sounds like what you want is a stat that takes into account how many strikeouts the pitching staff had, in addition to the basic stats like A, PO, DP, and E. And for it to be more accurate it would also take into account how many base runners were already on base. And for it to be really accurate, it would take into account how many innings were being pitched by LHP and RHP, since more innings pitched by LHP means more RH hitters and more balls hit to the left side. And once you'd done all your calculations for one SS, you'd want to compare that to how other SS were doing on the same basis. Problem is, you'd need a massive database to do something like that, and a whole lot of time. I've never done it and I'm never going to. You've never done it, and you're never going to. So where does that leave us?
I won't speak for you, but where it leaves me is relying heavily on the one person who did assemble that database, and did do all of those calculations for every SS in history and explained it all in painstaking detail in a series of books that I have read and understood. And there may be a nit to pick here and there in some of his defensive evaluations, but there's literally nothing with which anyone could take serious exception. And he rates Ozzie Smith an A+ SS - as does every other person on the planet besides you who ever saw him play. He rates Bowa a C shortstop, which is pretty much where I thought everyone rated him until this thread.
Again, I don't want to speak for you but my suspicion is that the variable you are missing is innings pitched by LHP. I didn't review their entire careers but I looked at 1972 (Bowa's best year) and 1982 (one of several of Ozzie's phenomenal years). And the Phillies pitched 1,400 innings with 701.2 by LHP (50%) while the Cardinals pitched 1,465.1 innings with 293.2 by LHP (20%). The Phillies pitchers did strike out 200+ more hitters, which Bill James absolutely accounts for, but with 1,000 or so non-K outs recorded with a LHP on the mound, there were a whole lot more outs for the Phillies SS and 3B to make, too. If you've read Win Shares, I'd be happy to pick nits on his methods with you. If you haven't read it, then you should. Everyone should.
And for the record, I was always a fan of Larry Bowa. Being an average major league SS for 2,000+ games probably makes him one of the 30 or 40 greatest SS of all time. If he could have hit a lick he'd probably be in the HOF. If I could take out Jack Morris and put Bowa in I'd do it in a heartbeat. But there is nothing in the universe that would make me believe he was as good a SS as Ozzie Smith or that Vic Wertz was as good a hitter as Mantle. They are equally preposterous statements, and until this thread I's have laughed if someone told me there was even a single person who agreed with either statement.
Regarding Templeton from someone else's comments, I remember thinking "THIS (Ozzie Smith) is all we could get for such a great player??". But my next thought was "but then, what is the market value of an a$$hole?". Templeton had to go, on that point there was no debate. But the whole league knew that, and knew we were beggars, not choosers. Templeton did grow up after being traded for a .220 hitter, which I assume bruised his ego tremendously, but the Cardinals still came out miles ahead on that trade which was a huge and pleasant surprise.
I think win shares is deliciously arcane. I would love to have someone explain it. Past the headlines. Like explain the team and game context data and how it was developed! Don’t explain the linear normalization or the pool hierarchy. Explain the stories that make the spreadsheet work.
Explain it. No one ever will because it’s so arbitrary in spots. Trust and faith.
Dang….. I was just about to explain the linear normalization and the pool hierarchy. ☹️
I guess I’ll just save it for another time and place.
I think we need to take a longer look at advanced pre participle nano distribution theory in a contiguous multi transference hypothesis relative to a flowing continuum to better measure a players defensive value.
bgr…. do you sometimes mistake this board for a NASA engineers’ board?
I don’t but I have to use the right terms to describe what he’s talking about. I know another book people take at face value.
This is an odd comment. I mean, there's a book called, appropriately, "Win Shares", that explains what Win Shares are in excruciating detail. If you haven't read it then, well, read it. Bill James holds the readers hand and walks him through how Win Shares are calculated from start to finish, and at every position for fielding. And if you have read it and still find them arcane then nobody else is going to be able to explain them better than the man that invented them.
As I said, they are hardly Gospel but only one person in history has attempted to do what Bill James did, and while he surely made some mistakes and more surely made some assumptions that are different than other people might have made, he is 100% transparent about what he did so we are all free to debate or nitpick what he did from a level playing field. I can't imagine anything more ignorant than dismissing what he did. Well, the pinnacle of ignorance would be dismissing what he did without even reading his book. That's willful ignorance, and I've got no time for debating the willfully ignorant.
But that’s the point. I understand his work also and I subscribed online years ago. I haven’t checked his site in many years but it hasn’t been updated for many years last I had checked.
You want people to believe because he compiled a lot of information that it’s meaningful. Because you’ve reviewed it. You understand his basic math and variable selection and normalization method.
It’s telling that he realized that win shares required negative win shares. More so that an acolyte won’t defend their masters hallucinations - I liked how you used that and it has the feels.
He hasn’t updated his site since I was last a member it appears. No new data to siphon.
It’s tough to defend a system which has been obsoleted. I just wanted you to frame how Bill constructs his… value. I guess people just have to trust you. End of the day it was his own rigid stubbornness which launched WAR into general acceptance.
Books. And handbooks? Because a lot changed since 2002 in how he approached the problem. Are you talking about before he basically did what WAR does or after?
I love how you just assert you’re correct and that I’m too stupid to understand it. Basic.
If you've read his book - and you never said you did - and still don't understand Win Shares then I don't have to call you anything. You are what you are.
If you haven't read it then I stand 100% by what I said.
I've never been on his website so I don't know what you have and haven't seen. And maybe things have changed since Win Shares was published, but it makes no difference. Nobody is in any position to debate the system's merits or deficits if they don't understand what they are. What is it, specifically, about the computation of Win Shares for shortstops (or anyone else) with which you take exception? I'm happy to discuss or debate it. But, as I said, I have no time to debate Win Shares with someone who has willfully chosen to remain ignorant about what they are. First, read the book. Second, tell me what you read in it that you think is wrong. Third, we debate Win Shares. I'm not starting at 3, it would be pointless and a waste of everyone's time.
I have read his books. I’ve also built contexts with his handbook updates which show his gradual acceptance of his error.
What do I mean by “read his books”. I’ll be honest. I thought the opening section where he goes through the rules and the building of the statistic to be peculiar and monumentally tedious. Big assumptions on one page become the bedrock of ground truth on the next. I found his analysis of players to be the most valuable. I don’t disagree on his analysis on players but his system is just contrived by him to match his own labels. It’s one of those books where you see the problems and you want to skip ahead because you’re mired in the mind of a mad scientist.
His system is based on what actually happened rather than what might have happened. Over the years he has adapted to reason by adopting what WAR already does. Anyways you keep saying books and I don’t know if you’re talking about his 2002 books or all of the updates through the years.
Win shares is his original and then he adapted it slowly towards WAR by adding loss shares but that was released in a series of updates online which are not in his book. I just don’t want to start with a dissection of win shares to find out we’re talking about “WAR with different weights” - win/loss shares.
As he improved his system his correlation with WAR increased. That’s telling. As his accuracy improved he approached the accepted baseline.
I understand why you don’t want to explain win shares from his book yourself. It sounds like a witches brew.
When he rolled out loss shares around 2008 he (Bill) explained that his original allocations were the result of a “bookkeeping mistake” - he had taken a pitchers positive runs that a pitcher had created and awarded them as runs the pitcher had prevented. This constructed his entire 52-48 allocation for pitching vs defense which he corrected to 50-50 once he realized his error. Online. Not in the book. You haven’t read all his corrections to the work you hold as word of god and that makes you the bozo here. Again. What an absolute joke you are. His book is interesting if tedious and error rich but hey. It’s a great door stop.
I think Bill has done us a great service in his analysis but it’s still less accurate than WAR. His abstract is more interesting as even a coarse stat can shine a light on interesting comparisons. You should really update yourself with what Bill has said about his original work because you’re promoting something he left behind and acknowledged errors within.
Before you tell the people here they should read his books maybe also explain what they will be reading. His main book is something most people would need to be forced to peruse to gain an understanding. These are manuals with brutal verbosity - and that should say a lot coming from me.
I actually don't want any of that. In order to know which shortstop is getting to more balls, and thus more range, I want to know how many routine ground balls were hit within two steps of their positioning either to the left or to the right, 10 feet to their left, 10 feet to their right. 15 feet to their left, 15 feet to their right. How many they converted and how many they didn't.
Then you have to go further into knowing the velocity of the balls and the types of hops they were taking. Usually you can see this, and Ozzie didn't get to any more balls under what I am looking for than Bowa. Ozzie simply had more of them hit his way, so yes, he technically got to more of them, but he had more opportunities.
James tried to make a point with Bowa, since Bowa was more sure handed than Ozzie Smith, more sure handed than Belanger, James was trying to show that fielding percentage wasn't a good measure, and his model showed Bowa didn' have range and he used him as an example to highlight his premise. He created a statistical model that did not account for the most important things that mattered in determining range(what I asked for above), instead he relied on a calculation that assumed things would be equaled out over time based on team strikeouts, hits, etc.....although he didn't even know the amount of ground balls or fly balls.
In the end, his model was missing the most important parts, and he called Bowa a statue, of which Bowa was anything but.
The vast majority of ground balls are of the MLB routine variety(not Joe Schmoe) routine. Give Bowa the same amount of chances as Belanger or Ozzie and he fields 99% of those balls too....because Bowa moved just as well, and his hands and arms were better.
James needed glasses.
PS what were Bowa and Ozzie's final total career defensive win shares from james?
OK, and what I want is a billion dollars. I was thinking you wanted something from among the theoretically available options. Since what you want isn't theoretically possible, then that's it. The beginning, middle, and end of your argument is that we have absolutely no way of knowing how good a fielder anyone is: we can't believe the stats and we can't believe our own eyes. It's a logically tenable position, but one which is also obviously immune to debate or discussion. It makes your OP that much more bewildering, but I'll have to let that go since neither you nor I has even a shred of the evidence you think is needed to make or refute such a claim.
And as with bgr, I just don't see the point of debating the merits of Win Shares with someone who hasn't read the book. James didn't call Bowa a statue so I have no idea why you said that he did. You said "in the end, his model was missing the most important parts" when you don't know what parts it has. Yes, it is "missing" data that doesn't exist - the data you say you want. But you are also missing that data so you just make it up in your head and draw conclusions from it.
James tried to estimate ground balls and fly balls, you just make stuff up. You also said James "used Bowa as an example", and I don't know what you mean by that. Where did James use Bowa as an example, and as an example of what?
Anyway, I'm done. If you or anyone else wants to discuss/debate something from James' book I'm your huckleberry. But debating things you or bgr have imagined might be in James' book is just frustrating. Read the damn book before you opine on it. Is that really too much to ask? And if I'm way off base here and you have read it, then quote something from it and tell me why it's wrong or could be improved. Such examples no doubt exist, but James not accounting for the velocity of the grounder that Jerry Grote hit off Dick Ruthven in the bottom of the 7th is not one of them, because that data doesn't exist. If your argument is more than "my preconceived notions are better than James's analysis of all of the available data", then I've missed.it.
now it sounds like you're talking about the BIS data when you refer to James' work with ground balls vs. fly balls. the fielding bible. the real game info. The BIS charting wasn't really fully cooked when James released his books, but James then pivoted away from legacy, and error prone, traditional record and range factor with his (somewhat arbitrary) positional coefficients. Win Shares was based on this legacy data and Bill says he wanted to be consistent over accurate and that's why when data got better, he kept the input the same - consistency. so no. the original Win Shares does not include ground ball vs. fly ball or batted-ball location. The batted ball charting data is pretty awesome and I have digitized it... from Dewan's fielding bible however. Are you messing with us?
https://mediaburn.org/videos/cubs-1983-highlights/
For example, here is an old man Bowa on the Cubs ranging up the middle, of which James said he could not do via his statistical model. Starting at the 5:48 mark you can see the nimbleness of Bowa up the middle(f of which James said he couldn't do...and he shows his arm off on a couple plays.
Bowa could move. He was fast, quick, and nimble...nothing Luzinski like about him.
It was obvious to any observer that he was nimble and had range...and had the best hands of his era. He could make nearly any moving play as good as Ozzie or Belanger. He made less errors. He had a better arm....yet these defensive measures are showing ridiculous statistical difference between their runs saved, and you already admitted that the valdiity of the defense measurements is clearly not good, so not sure why it is a surprise that you are so adamant against my point, other than being an Ozzie Smith fan.
The one thing Ozzie probably had over Bowa was the extreme diving play. Ozzie was good at those, but those were few and far between....but Bowa made up for that with a better hands and a stronger arm.
PS...start the video around 5:30 to see a vintage Old Style commercial
https://www.facebook.com/PhilliesHighlights/videos/larry-bowas-greatest-moments/648715039120413/
Phillies video where Harry Calas calls him all time great fielder later in the video. At the 1:29 mark he shows off his arm. Callas is back half of video.
3:23 Calas called him a glove man with no equal. So I'm not the only one. At the time he had highest fielding percentage of all time.
When I ask you, @dallasactuary, to explain it to me I am not doing so because I don't understand how Bill constructed Win Shares. I want you to explain it because I think that would make it clear to everyone what it is doing under the hood. There are 100-150 pages which walk through the construction of Win Shares and why this and that was selected before it gets into the application by story. Perhaps that's a flaw in itself... that Win Shares can't really be explained without digging that hole yourself. To pretend like it couldn't be understood by most-anyone is ridiculous. I don't recall Bill being a trained statistician, data-scientist, or mathematician.... He was a huge baseball fan who was quite brilliant and incredibly diligent.
What I was after was the next level down which discusses the linear extrapolation once a player has been normalized vs. his team's baseline. The big brush strokes could be covered in a handful of sentences - such as the ratio of real wins to win shares and the distribution of merit for a win - the shares and marginal shares parts. You don't spend any time describing the insightful parts of Win Shares - perhaps because that introduces the misleading parts as well. The only reason I am holding you accountable here is because you've spent a lot of time and effort dissing dWAR while Bill, himself, eventually succumb to it's recipe. You have ignored his further effort past the release of his books in 2002... probably have committed much of the almost 2000 pages of Win Shares and his Abstract to memory.. but failed to keep reading as Bill's system evolved over the years.
I really respect Bill James - the guy was a pioneer and there's probably no sabermetrics like we have today without Bill. His Baseball Abstracts, which I read long after they were published formulate a large part of how I see the game of baseball today. He created the questions and built the groundwork for the answers. But data-driven process for statistical player evaluation was happening already and elsewhere. James synthesized a lot of other work and took it to the next level of rigor... with a healthy dose of self-evangelism.
Win Shares has been replaced by WAR, Statcast has obsoleted his batted-ball tracking, and wOBA has supplanted his runs created and linear weights.
I believe James said about Win Shares that it was designed to ask questions - not answer them - I paraphrase that from what I remember in his book... Win Shares. Since you never were an online member you probably never used his online ask and answer where you could discuss his ideas directly with him - you missed out.
Shoulders of Giants. I learned a few things from Bill James over the years...
Earl Weaver was a really fascinating dude... He was using a form of sabermetrics during his tenure as Baltimore's manager starting in the 1970s. He had his index cards on every hitter - how they performed against every pitcher and he was one of the first to create some of the first platoons - based on data vs. gut feeling. He was using batter vs pitcher and batter vs handedness data decades before most. Outs are precious, OBP and power drive offense, and bunting and base-stealing are usually counterproductive. Weaver has his own book - Weaver on Strategy on these subjects.
Branch Rickey doesn't get enough credit... for many things.. Being possibly the inventor of sabermetrics might be one of them. After... you know... breaking the color barrier and inventing the farm system he figured more was needed. Rickey believed in the 40s that batting average, RBIs, and pitcher wins were misleading and he hired a data analyst to identify the right correlations... and guess what. OBP matters most, batting average misleads, and offense and defense decompose into measurable run contributions... gee. I wonder where I have heard this before? Was it... Win Shares? Have you read it? It's a fun read on the beach. On Branch Rickey there is a 1954 Life which has an article on it. https://invention.si.edu/invention-stories/branch-rickey-baseball-innovator
Another very insightful book that came long before Win Shares was The Hidden Game of Baseball. Palmer was an actual statistician and he gave us the first real actionable method for run estimation. His linear weights are still used today - refined certainly, but they form the backbone of modern run-estimation thinking.
I appreciate your protective instincts. You guys are buds so protect your homie. Dallas is off base here and he's been slingin' mud - he's a mud slinger. I'll certainly stop sending it back when you guys grow up. I feel the embarrassment for him and I know he's gonna die on this hill because he can't walk it back or take the bridge when it's there. He just sees red.
Linear
Normalize
Pool
Hierarchy
Linear Normalization - this is the step in Win Shares which forces every player's individual contribution to sum exactly to the team's total.
Pool Hierarchy - the Win Shares pools cascade through several nested layers, splitting shares at each stage until being distributed to the individual players.
Team Pool - a 50 win team has 150 win shares.
Offense vs. Defense Pool - Team Pool divides into offensive and defensive categories. Dallas likes the 2002 52 vs. 48, but Bill James prefers his 2008 50 vs. 50. Deeper Dive - Marginal Runs.
Defense Pool branches into Pitching vs. Fielding - This is where James encountered one of his errors.
Now prepare yourself for about 40 pages discussing how the apportionment happens...
The team-level pitching/fielding split is one of the more contentious issues from Bill's work and this relates to his "defensive efficiency". Doing this at the team level is a coarseness that may have been necessary at some point, but I think it introduces much unnecessary error.
Point being; I was asking him to explain these two foundational parts of Win Shares to show that I know where the skeletons are and that I am familiar with the material from the book. To which he just pesters that I'm too stupid and haven't read his books so he doesn't need to say anything other than I'm stupid and he's right. Hitch your wagon to him all you want - I at least support my opinion and I'm not denigrating his.
Someone saying dWAR is terrible and Win Shares is great isn't making a whole lot of sense. I don't like or dislike any of you guys and I'm certainly not going to embark of a path of self-delusion just because I think someone's a total clown. Support it... don't. you'll look like a fool to people with sense.
Hey.. I’m not buds with Dallas. I like it better when it’s tax season and he’s busy hiding under his desk from his clients and not posting here. Lucky us, tax season is over. 😩
bgr….. I like your posts but I’m just not into the mathematics of the complicated stats like win shares etc.
I look at threads like this more for the entertainment value. Personally I don’t know why Dallas doesn’t acknowledge that Bowa was a great defensive SS. The numbers seem to say he was.
I think we both love baseball Darin... maybe not you so much this year but... in general. I'm not as much of a statistics guy as it might seem - I love baseball for the art and nuance while appreciating the baseline of statistics. It's the randomness in baseball that wins a World Series or throws a Perfect game. The moments. I think the statistics are really cool, but I'm hardly dominated by them. So I like to see the approach of it all while appreciating the game for what it is. The concept of best player or best shortstop is something I find incomprehensible - I can't even begin to construct this while some people can just rank order a list from 1 to 100. Understanding the statistical methods is something I enjoy doing. I feel like I could learn something from dallas, but I don't think I have ever seen him actually discuss something without belittling or berating the people he's debating. Jousting with him is pretty fun too though so it's not all bad news. This place is hella-fun. Some days it's a zoo and some days I'm just reading cool posts while I watch a Brewers game. This place needs more pepper though. I'm from Minnesota and it's not spicy enough for me.
That's a good explanation.
Per the part where Bill wouldn't explain, don't fret, he can't explain because Bill needs a pair of glasses.
I read this pargraph, rolled my eyes, shook my head, and let out a heavy sigh. I lack the strength to read more.
Show me where "James said" Bowa couldn't do anything. I'd say I'll wait, but we both know you pulled that statement out of your nether regions because James never said any such thing.
As my last post in this thread, and last visit to this thread, I will repeat: read the damn book before you opine on the book. I know you aren't embarrassed by your performance in this thread, but I am embarrassed for you. You are better than this. You are smarter than this. But your willful ignorance regarding the things James actually did say is impenetrable, and I give up, Let me know when you've read the book and I will discuss defensive Win Shares with you. Until then, bye.
From May, 1981 Sports Illustrated:
"Frequently, our visual sense of the great fielders is surprisingly accurate: Brooks Robinson , Graig Nettles and Schmidt all have had high Range Factors. But the eye can deceive, too. The balloon James most enjoys bursting is that of Shortstop Larry Bowa, who has, he says, "the range of the Birdman of Alcatraz ." During the past few seasons, according to James' figures, Bowa has gotten to and fielded cleanly substantially fewer batted balls than fellow National League shortstops Garry Templeton , Ozzie Smith , Ivan DeJesus , Dave Concepcion and Tim Foli. Bowa is a fielder "who looks good," James says, "on the balls he reaches." "
Like I said, two simple plays in that video, along with countless others, proved James needs glasses.
James uses 'his formula' to come to HIS conclusion...the very, very flawed formula as pointed out. Bowa was quick, nimble, fast, and had a plus arm. Also the best hands in the league while at the time of that 1981 writing, Bowa had the all time best fielding percentage for a SS.
So even James admits that 'the eye' is saying Bowa is a great fielder among those other greats listed...but since his VERY FLAWED formula said otherwise, he reversed what his eyes saw, and made a bad conclusion...just like you did for listening to him like he is a God.
So maybe I was wrong and he didn't need glasses because he admitted Bowa looked just as elite as Brooks and Schmidt...but he had his flawed formula, so he had to highlight something to 'shock' what we all saw...and what common sense says; that Bowa was a faster human than Mark Belanger, Bowa had a great glove, and it is silly to think Bowa couldn't range 20 feet while Belanger could. Just plain silly. Silly because I saw him do it.
Check mate. Have a good night.
I think Bill James was closer to the answer than anyone before him and he made a substantial contribution to the field of sabermetrics.
I'll say this much... this discussion did prompt me to go back and read some of Bill's assessment of players in his archives... some I find did turn out to be incorrect over time, but the guy does really present valuable insight into the game. I don't recommend anyone studies the machination of Win Shares, but the companion abstract is worth the time.