The flaws of the various methods of measuring hitters...
Skinpinch
Posts: 1,531
in Sports Talk
OK. I will say right now before anybody discounts the better methods, that there is no hitting measurement that is 100% to the truth. The best it can come is somewhere in the high 90's percent range.
The odd thing about human beings and this, is that I find it very silly that people will totally discount a superb measurement such as situational Batter Runs, by citing some very minor items that it may not cover. AND THEN EMBRACE A MEASURMENT THAT ALSO DOES NOT COVER THESE SAME MINOR ELEMENTS. NOT TO MENTION THE FACT OF IT NOT COVERING MANY MAJOR ELEMENTS.
Here are some measurements and their biggest flaws....
BATTING AVERAGE: Does not give credit for walks. Treats a HR as the same value as a 1B. Simply put, treats all hits as the same value.
Total bases: Does not give credit for walks. Says 4 for 4 with four singles is the same value as 1 for 4 w/ a HR...because it does not account for the negative impact of outs made to achieve high Total Base totals. Also says a double is twice as good as a single, a triple three times as good, and HR four times. This is not true.
OB%: Treats a BB and alll other types of hits as the same value. It is batting average that accounts for BB. Is one leg of the hitting equation.
SLG%: Ignores base on balls. It does not account for the number of times a runner reaches base. Is one leg of the hitting equation.
OPS: This is the combination of OB% and SLG%...combining the single legs to forumlate a pretty good measurement. But it is adding two flawed measurements together, and it still can be improved on.
PARK ADJUSTED: All of those measurements are flawed if they do not include a park adjustment. They all can be park adjusted.
SITUATIONAL HITTING: None of those measurements account for what a batter did based on base or out situation. A HR with two men on is better than with nobody on. A single with nobody out is better than a single with two outs. You can find what each player did with men on for each stat though.
LINEUP PROTECTION: A big falsehood. A batter is not made better because of the goodness/badness of the hitter behind him. Read the old thread regarding that. Nonetheless, none of the above measurements take that into account.
FACING BETTER PITCHERS: If for some odd reason a player faced ONLY the best pitchers, then it should be accounted for. None of these measurements account for that.
THE BEST MEASUREMENTS: The best measurements account for all of the key items above, but they put them together. Unlike Batting Average, it doesn't treat a HR as a single. Unlike Total Bases, it doesn't ignore outs made or base on balls. It accounts for ALL THEIR RELEVANT offensive hitting evnts, and gives it the proper weight as to how many runs it typically leads to. Some take it a bit further by adding the baserunner situation and out situation for that to give a really accurate portrayal of what a hitter did offesnively.
The tiny stuff that it does not cover(or needs to be covered), IS ALSO NOT COVERED BY ANY OF THE FATALLY FLAWED methods above.
Margin of error: Of course there is a margine of error on ballpark measurement, but it is small. It doesn't make a guy like Jim Rice who is dwarfed by his contemporaries, all of a sudden to the top...as many of his fans wish.
The Best Measurement as a law of lists!: When I put the lists of Batter Runs up, I don't rank them, but rather group them by how close they were to others. If there is a margin of error, it is going to be small, and a guy wold most likely only change place from within the group he is listed. He is not all of a sudden going to vault from the bottom of the guys presented(rice territory), leap frog 20 other guys, adn then be grouped with the guys at the very top.
If someone wishes to use extremely flawed measurements, instead of superficially flawed measurements, then have at it. I'm not sure why they would if they are trying to answer the 'who's best' or HOF type questions, as they seem to seek an understanding of where someone ranks. So it is of interest to them. SO if it is of interest to someone, then why stick with obviously flawed measurements? If someone wishes to, then state that they are only a casual fan, and only want a cursory knowledge of where a player stands. Don't try and say that inferior methods are better suited than the best comprehensive methods.
I'm not proclaiming the BR as gospel, as there are other very good comprehensive measurments out there that have slightly different results. If one is a naysayer, then bring up a Bill James winshares or something to compare/contrast. Don't bring up a supremely flawed method as Total bases to refute a vastly superior measurement.
What one will find, is that the best measurements...all through slightly different methods, tend to say pretty much the same thing. Things like Jim Rice's or Harmon Killegrew's or Mike Schmid'ts or Bill Buckner's true spot in value.
IT CORRECTS THE FATAL FLAWS OF THE POOR MEASUREMENTS...AND ANY FLAW THAT WOULD STILL EXIST WOULD EXIST IN THE INFERIOR METHOD AS WELL...THOUGH IT WOULD BE MINOR ONES
The odd thing about human beings and this, is that I find it very silly that people will totally discount a superb measurement such as situational Batter Runs, by citing some very minor items that it may not cover. AND THEN EMBRACE A MEASURMENT THAT ALSO DOES NOT COVER THESE SAME MINOR ELEMENTS. NOT TO MENTION THE FACT OF IT NOT COVERING MANY MAJOR ELEMENTS.
Here are some measurements and their biggest flaws....
BATTING AVERAGE: Does not give credit for walks. Treats a HR as the same value as a 1B. Simply put, treats all hits as the same value.
Total bases: Does not give credit for walks. Says 4 for 4 with four singles is the same value as 1 for 4 w/ a HR...because it does not account for the negative impact of outs made to achieve high Total Base totals. Also says a double is twice as good as a single, a triple three times as good, and HR four times. This is not true.
OB%: Treats a BB and alll other types of hits as the same value. It is batting average that accounts for BB. Is one leg of the hitting equation.
SLG%: Ignores base on balls. It does not account for the number of times a runner reaches base. Is one leg of the hitting equation.
OPS: This is the combination of OB% and SLG%...combining the single legs to forumlate a pretty good measurement. But it is adding two flawed measurements together, and it still can be improved on.
PARK ADJUSTED: All of those measurements are flawed if they do not include a park adjustment. They all can be park adjusted.
SITUATIONAL HITTING: None of those measurements account for what a batter did based on base or out situation. A HR with two men on is better than with nobody on. A single with nobody out is better than a single with two outs. You can find what each player did with men on for each stat though.
LINEUP PROTECTION: A big falsehood. A batter is not made better because of the goodness/badness of the hitter behind him. Read the old thread regarding that. Nonetheless, none of the above measurements take that into account.
FACING BETTER PITCHERS: If for some odd reason a player faced ONLY the best pitchers, then it should be accounted for. None of these measurements account for that.
THE BEST MEASUREMENTS: The best measurements account for all of the key items above, but they put them together. Unlike Batting Average, it doesn't treat a HR as a single. Unlike Total Bases, it doesn't ignore outs made or base on balls. It accounts for ALL THEIR RELEVANT offensive hitting evnts, and gives it the proper weight as to how many runs it typically leads to. Some take it a bit further by adding the baserunner situation and out situation for that to give a really accurate portrayal of what a hitter did offesnively.
The tiny stuff that it does not cover(or needs to be covered), IS ALSO NOT COVERED BY ANY OF THE FATALLY FLAWED methods above.
Margin of error: Of course there is a margine of error on ballpark measurement, but it is small. It doesn't make a guy like Jim Rice who is dwarfed by his contemporaries, all of a sudden to the top...as many of his fans wish.
The Best Measurement as a law of lists!: When I put the lists of Batter Runs up, I don't rank them, but rather group them by how close they were to others. If there is a margin of error, it is going to be small, and a guy wold most likely only change place from within the group he is listed. He is not all of a sudden going to vault from the bottom of the guys presented(rice territory), leap frog 20 other guys, adn then be grouped with the guys at the very top.
If someone wishes to use extremely flawed measurements, instead of superficially flawed measurements, then have at it. I'm not sure why they would if they are trying to answer the 'who's best' or HOF type questions, as they seem to seek an understanding of where someone ranks. So it is of interest to them. SO if it is of interest to someone, then why stick with obviously flawed measurements? If someone wishes to, then state that they are only a casual fan, and only want a cursory knowledge of where a player stands. Don't try and say that inferior methods are better suited than the best comprehensive methods.
I'm not proclaiming the BR as gospel, as there are other very good comprehensive measurments out there that have slightly different results. If one is a naysayer, then bring up a Bill James winshares or something to compare/contrast. Don't bring up a supremely flawed method as Total bases to refute a vastly superior measurement.
What one will find, is that the best measurements...all through slightly different methods, tend to say pretty much the same thing. Things like Jim Rice's or Harmon Killegrew's or Mike Schmid'ts or Bill Buckner's true spot in value.
IT CORRECTS THE FATAL FLAWS OF THE POOR MEASUREMENTS...AND ANY FLAW THAT WOULD STILL EXIST WOULD EXIST IN THE INFERIOR METHOD AS WELL...THOUGH IT WOULD BE MINOR ONES
0
Comments
This is a partial explanation on the methods you choose to use. None of it is opinion, it is all based on leading to creating runs.
I'm not going to take all this time to respond to the jabs at reliable stuff, and reinvent the wheel, when these basic questions are already covered.
If someone insists on using a batting average or a total bases as the foundation of their stance, fine. If somebody wants to refute the information I presented based on those measurements? Fine, but my response from now on to those will be to research it better, and start with Total Baseball first.
If one chooses to ignore all that stuff and use faulty information to have 'their' players shine in a brighter light, then being the generous guy I am, I won't get in the way of one's happiness.
If anybody chooses to respond to stuff I post, great. If somebody questions the basic stuff, I would refer you to the books to save me some time. If somebody questions something that is relevant or isn't easy to find, then I will explain, much like Detcher's question on Mattingly. Relevant question that has meaning.
<< <i>My team won the World Series this year. >>
that would be the bottom line
ISO 1978 Topps Baseball in NM-MT High Grade Raw 3, 100, 103, 302, 347, 376, 416, 466, 481, 487, 509, 534, 540, 554, 579, 580, 622, 642, 673, 724__________________________________________________________________________________________________________________________________ISO 1978 O-Pee-Chee in NM-MT High Grade Raw12, 21, 29, 38, 49, 65, 69, 73, 74, 81, 95, 100, 104, 110, 115, 122, 132, 133, 135, 140, 142, 151, 153, 155, 160, 161, 167, 168, 172, 179, 181, 196, 200, 204, 210, 224, 231, 240
That paste I did of the article is quite long
The bottom line is that there is no point on using faulty or very incomplete methods when better ones are good and available.
I probably would not have added to this particular post, as it does make fairly good sense and you finally do admit that any stat, even ones you are enamored with, are not perfect, however since you decided to address me specifically , I shall respond.......
My chosen methods are just that, methods, plural. I feel that no one stat regardless of any false perception of including EVERY SINGLE OFFENSIVE EVENT IMAGINABLE is just not good enough for evaluating a hitter. It is probably good to remember we are talking about hitting, just one aspect of a player's worth, not every imaginable factor which contributes to a ballplayer's value, like stolen bases, fielding, inspiration, throwing ability, intangibles, and several other important facets of being a good player.
I believe the BR stat does not take into account the game situation via the score and inning. A guy who gets any kind of a hit in the 9th inning of a tie ball game might be considered more valuable than one who gets a double, driving in the 7th run in the middle innings of a 7-0 rout. While the ballpark is a factor, all hitters in any one game will hit in the same park, they will however face different pitchers trying to get them out. Even a simple-minded traditionalist like myself would see a run produced in the 5th inning of a one-sided game played in June, might not be as important as a run produced in September, while in the pennant race, during the 9th inning of a tie ballgme. The impact of a game winning RBI has an identical effect as the 5th inning example prior noted, when compling BR.
It is naive to think the effect of the following hitter has no importance. Ted Williams or Barry Bonds batting after a hitter will very much influence the hurler's choice of pitches for the current batter. I would venture a guess even Mr Bill James could see that.
You feel traditional methods are " extremely flawed", I feel thsy are extemely accurate in evaluating hitters. As I stated before , a combination of these stats is a better representation, but it seems any single one, on its own, provides a very good indication of a hitter's skill.
Career totals Triple Crown hitting stats;
Batting Ave. 1 Cobb, 2 Hornsby, 3 Joe Jackson, 6 Ted Williams, 7 Ruth
Home runs 1 Aaron, 2 Bonds ,3 Ruth, 4 Mays
RBIs 1 Aaron, 2 Ruth, 3 Gehrig, 4 Musial, 5 Cobb
Career totals other stats;
Total Bases 1 Aaron, 2 Musial, 3 Mays, 4 Cobb, 5 Ruth
OB % 1 Ted Williams, 2 Ruth, 3 Gehrig, 4 Bonds, 5 Hornsby
SLG % 1 Ruth, 2 Ted Williams, 3 Gehrig, 4 Pujols 5 Foxx
Of course Pujols is still going to have to keep it up for some additional time to continue in his place, which is a percentage figure, among these hitters. Is any one of them NOT worhy of being called a great hitter ??? Do these traditional methods confirm some of the great hitters in baseball history ??? Would any exotic stat prove otherwise ???
Are the above methods really poor measurements ? do they have FATAL flaws ? Probably not, as the results confirm. Perfect measurements ? Maybe close when combined with each other.
All of those measurements tell a story, I don't disagree with that at all. The better measurements simply tell a more COMPLETE and truthful story. It fixes the wrongs portrayed by all the traditional numbers of BA, HR, RBI. TO further measure them, you take the next step of OB%, and SLG%
Please note that OB% and SLG% are NOT traditional stats, and are valuable. Combined, they are quite valuable. The next step is the comprehensive methods. With each step, you get more truth. But in Rice's case you were simply ignoring the park affect that screws with those measurements.
The best hitters of the universe are going to dominate the leaderboards in most conceivable measurements...but there are some that are totally off. A comprehensive measurement is never totally off, and that is the main point. It is your well rounded evaluation you crave, only it is accurate and accountable.
Ted Williams is 19th in Total Bases, 64th in hits, 7th in batting average, 16th in runs, 13th in RBI. Not one of those are even close to telling how good he is. Not until you use good measurements of OB% and SLG% do you start to get the gravity of the hitter.
MICKEY MANTLE was 36th in total bases, 98th in hits, not even close to the top 100 in Batting Average, 27th in Runs, 45th in RBI, and 13th in HOME Runs. Those traditional methods make Mantle look much worse than he was. Even the better methods of OB% 19th, and SLG% 26th, still don't put him into proper light.
Looking at your way of just guessing or assuming where Mantle ranks based on YOUR all inclusive method of just guessing, his highest ranking in anything important is 13th, followed by 19th, 26th, 27th, 36th, 45th, 98th, 100something. Your overall guessing, using your methods, could not reasonably put him anywhere better than 35th or so.
It isn't until the context of era, the ballpark effect, and the proper value of a comprehensive method used, that Mantle is put where he actually is....... a top TEN hitter based on lifetime, and a TOP FIVE based on PEAK.
You can bite and pull to decide where in the top ten, or where in the top five, BUT the comprehensive method puts him in that company. The comprehensive method like this avoids putting Rice in the echelon of Brett, Murray, and Schmidt. It clarifies things much more, much like clarifying Mantle's place.
By the way, from the very beginning, I noted the biggest flaw in the comprehensive batter runs, and never made claim that it represents the whole ballplayer. It is simply a hitting tool. Baserunning is an easy add on. Defense is a tougher add on. Neither one of those aspects(baerunning or fielding) will make Rice shine any brighter than what his hitting ranks him. It will make a guy like Tim Raines, Ozzie Smith, or Ryne Sandberg shine brighter, BUT NOT RICE(it may even make hime shine worse).
You mentioned not measuring inning or score? That is not a mystery by any means. You can check every single at bat if you wish. IN Rice's case his OPS in LATE/CLOSE games was .823. His Lifetime OPS was .854. Simple deduction shows that he hit worse in those tough spots you seek. In Murray's case, he raised his OPS 51 points in the LATE/CLOSE situations.
No, do ya say?
Steve
As I have previously posted several times, being a league leader is a very good indicator of how well one compares to his peers.
Some traditional stats make up hitting's triple crown, both Williams and Mantle were excellent enough to earn a triple crown ! That combined with their other league leading performances, show enough to rate them as great hitters. It seems you did not notice that Williams did in fact rank well in half of the traditional career stats which were listed, No one admires Teddy Ballgame anymore than I do.
You never did admit the traditional career stats listed were not at fault in evaluating great hitters. Could you possibly claim that Bonds, or Ruth, or Cobb, or Musial or Aaron or any other player who excelled via those old-fashioned time-tested stats, was not a great hitter ???
It seems you feel you have discovered something special like gravity, or an additional planet in the solar system, however there is no way a single stat will account for EVERY SINGLE OFFENSIVE EVENT IMAGINABLE. It is not new to combine various stats, and weight them as one sees proper, and adjust for circumstances like the park or the pitcher faced. BR may be another good combination stat which does confirm the validity of many traditional ones.
By the way Jime Rice hit about 2 more HRs at home, per season, than on the road, appears NOT a real big home field favor. Rice did lead the league in offensive events ( traditional stats ) some 13 times during his career, Eddie Murray who had several more seasons in his career, led 3 times.
But as you back several rungs down the greatness ladder you find an awful lot of players; some of them are great players and some of them aren't. "Traditional" stats are of little or no use in discerning between them since they take no account of eras, ballparks or teammates. And leading ones league in something can mean that player was the best or that player was the most fortunate - looking only a traditional stats there's no way to know which. Tony Armas led the AL in HR and RBI in 1984 but he wasn't among the 20 best players in the league that year; he was a right handed fly ball hitter playing in Fenway Park with a plum spot in a strong batting order.
<< <i>Skin-ilk,
As I have previously posted several times, being a league leader is a very good indicator of how well one compares to his peers.
Some traditional stats make up hitting's triple crown, both Williams and Mantle were excellent enough to earn a triple crown ! That combined with their other league leading performances, show enough to rate them as great hitters. It seems you did not notice that Williams did in fact rank well in half of the traditional career stats which were listed, No one admires Teddy Ballgame anymore than I do.
You never did admit the traditional career stats listed were not at fault in evaluating great hitters. Could you possibly claim that Bonds, or Ruth, or Cobb, or Musial or Aaron or any other player who excelled via those old-fashioned time-tested stats, was not a great hitter ???
It seems you feel you have discovered something special like gravity, or an additional planet in the solar system, however there is no way a single stat will account for EVERY SINGLE OFFENSIVE EVENT IMAGINABLE. It is not new to combine various stats, and weight them as one sees proper, and adjust for circumstances like the park or the pitcher faced. BR may be another good combination stat which does confirm the validity of many traditional ones.
By the way Jime Rice hit about 2 more HRs at home, per season, than on the road, appears NOT a real big home field favor. Rice did lead the league in offensive events ( traditional stats ) some 13 times during his career, Eddie Murray who had several more seasons in his career, led 3 times.
First, being a league leader is not necessarily a good indicator of how one compares to their contemporaries. There is a huge difference between leading the league in HR's when, say, you hit 45 and the league average is 22, and when you hit 30 HR's and league average is 8.
Secondly, you should allow for the possibility that BR may be a good combination stat that further illustrates the LIMITATIONS of more traditional statistical indicators. The fact that there may be a correlation between having an outstanding BR ranking and having an exceptional ranking in traditional stats does NOT mean the stats are of equal value.
The good news is that the time lag is narrowing in terms of how long it takes for a sabermetrics measure to become valued by the mainstream baseball community as well. OBP and OPS now, Win Shares and Batter Runs somewhere down the road. Sabermetrics will always be ahead of the curve.
Great info.
What is the formula for situational Batter Runs?
<< <i>I feel your pain. >>