01-08-2010, 08:45 PM | #151 | |
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
You obviously have no concept of how current advanced statistical analysis judges players and are throwing out a bunch of uneducated crap to try to defend your position. |
|
01-08-2010, 08:46 PM | #152 | |
Pro Starter
Join Date: Feb 2006
|
Quote:
I stick more with the statistical analysis part of things. The problem with OPS+ is that when people go around quoting it, there is no weight as to how many PAs this player had. An OPS+ of 130 in 5 full seasons is great but I'll gladly take an OPS+ of 120 in 10 full seasons. |
|
01-08-2010, 08:55 PM | #153 | |
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
|
|
01-08-2010, 09:20 PM | #154 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
Actually, I peeled that right off your precious statistical lists. Obviously, you painted yourself into a corner and now you can't throw out anything beyond insults. Because when the numbers start to work against you, you're just running for the hills. Simply put, you can analyze anything you want and make the numbers break your way. There's WAR, win shares, WAA (wins above average), objective functions, modified objective functions and a thousand more ways to dissect the numbers. You can almost always find a list to prove your point. You know what? Statistics are great. They're very useful. But baseball isn't just a science. It's just as much an art and there's no quantifying that part of the game. Now maybe you believe that or maybe you don't, but baseball isn't played by machines. It's played by guys like Jackson and Puckett, Rice and Molitor, Yaz and Yount. Statistics don't take into account what happened after Bill Buckner let a grounder through his legs, or what happened when Kirby Puckett hit a HR in the 11th off Leibrandt or when Willie Mays caught Wertz' drive. According to statistics, those were just an error, a home run and an out. Anyone who watches baseball knows those events were much, much more (no wonder the Red Sox lost, the Twins won and the Giants swept). That's why statistical analysis only takes you so far. Hey, if you want to tell me that Edgar Martinez was a better slugger than Harmon Killebrew (and that's what the stats say), go right ahead. Me, I'll take the guy who played in a semi-dead ball era, who played outside, had 260 more HRs and 300 more RBIs over the guy who played in a juiced-ball era, played in a dome and hit over 30 hrs only once in his life. Last edited by Blackadar : 01-08-2010 at 09:23 PM. |
|
01-08-2010, 09:31 PM | #155 | |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
Quote:
Was there a particular number that went against him that I missed because I cant recall this? Everything he has pointed out has been fact. Everything you have made mention of is opinon. The reason people depend on stats is because there is no possible way for people to watch 15 games a day 6 months out of the year. You are crazy if you think you can paint the picture by watching a player a few times a year. As a Twins fan I thought Kirby Pucket was the best player in the game at the time. Once I decided to quit being stubborn I realized he wasnt even close. His biggest asset was his character. Last edited by jbergey22 : 01-08-2010 at 09:38 PM. |
|
01-08-2010, 09:40 PM | #156 | ||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
Then you decided to use OPS+ as a way to judge overall value of players, and as I noted it's not. You didn't paint me into a corner at all - you revealed your ignorance of modern statistical analysis. Quote:
You use the correct tool for the job. Just because some don't know which stats to use for what analysis doesn't mean that the stats used correctly are wrong - it just means the person using the stats incorrectly is wrong. Quote:
If you want to say you prefer player X to player Y because of something like how graceful he was as a fielder, or how he came up big in a situation that really mattered to you, that's your prerogative. But if you want to have a discussion about the value of players, that is something that is most accurately judged by using the appropriate statistical analysis. Quote:
Quote:
Quote:
Martinez had a higher slugging percentage over his career than Killebrew, yes. But he had a lower isolated slugging percentage, and by quite a bit. And Killebrew had a higher peak slugging percentage than Martinez. And to address your specific characterizations, Martinez played 5 1/2 of his 13 full season outdoors in a park that kills right handed hitters; doubles have value too, and Martinez hit far more (and far more frequently) than Killebrew; I don't care so much about RBI totals - if you care about RBI's, then what you should be looking at is how frequently a player drove in runs as compared to the opportunities to do so, i.e. who maximized their opportunities. |
||||||
01-08-2010, 10:03 PM | #157 |
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Dola - I also want to make it clear that I take no pleasure in ripping your arguments Blackadar, but your arguments in this thread are so piss-poor that I can't let them slide.
|
01-08-2010, 10:08 PM | #158 |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
And that leads me to another point - there are so many things that modern statistics don't measure.
They don't measure the talent dilution as compared to the MLB of 50 years ago. Especially since baseball was once much more popular. If there are more mediocre/bad guys now, then the good guys should seperate themselves more in things like adjusted OPS. If Stephen Hawking is in a room full of dunces rather than more average guys, then accordingly he's going to be further ahead of the class. They don't measure how things like balls and bats are far more consistent than they were even 25 years ago. They test the balls to ensure that they're not juiced. Supposedly, they're not - I'm not sure I always believe that, but let's take MLB's word for it. They did measure a box and found them to be all within limits. But they all were consistently on the highest end of the scale. Do you think they had that consistency back in the 60s or do you think Harmon Killebrew had to hit more than a few duds in his day? Let's see your stats calculate that. They average out the stadiums to reduce the impact of hitters' parks, but you don't see the cavernous stadiums of yesteryear. So averaging it out only goes so far because if all the stadiums are smaller, the coefficient is still 1. How do you think Fenway went from being a hitters' park in the 70s to being an average stadium today? It's not like the dimensions have changed all that much. Yet it's not a hitters park today (there have been a few changes). It's that many more teams play in a bandbox and very few play in the cavernous stadiums of the past. Much of the statistical analysis doesn't even take into account how what's important changes over time. If the game didn't emphasize bases on balls 40 years ago, then a stat like OBP is going to punish older players. 7 out of the top 14 guys in all-time intentional walks have played in the last 10 years. All but 2 of the top 15 seasons have taken place in the last 20 years. Do you think that's made a difference? Are we to believe that guys today are much better at getting on base? Or that they're just directed to get on base? The top 7 strikeout seasons of all time are by active players. Is it factored in that guys then would be told to do anything but strike out and guys now can swing freely? Do you think Jim Rice would still have the all-time high season for double plays if he knew he could strike out 160 times a season instead of being told early in his career to cut down the 120ks he was averaging? Let's take a stat we deem crucial - on base percentage - which was largely neglected for 60 years. From 1960-1990 - 30 years of baseball - only 3 performances hit the top 100. There are 16 such performances in the last 18 years. How is it that OPS+, which is supposed to smooth things out, records only 2 seasons out of the top 50 that occurs from 1960-1990, yet records 10 such seasons in the last 20 years? Are we to think that we didn't produce any great hitters for 30 years? Do you find it reasonable that 6 of the top 18 seasons in runs created have occurred in the last 10 years? The problem is that statistical analysis often doesn't take factors like these into account. Don't get me wrong, it tries. But ultimately, some of these factors aren't quantifiable. They're real, but many of the statistical models (here are some examples for those who don't know: http://members.cox.net/~harlowk22/atgwsobj.html) don't factor these things in or don't do it fully. It's ok to use statistics. But I fully reject the notion - as does every other GM in the league, including Theo Epstein - that they tell the whole story or even most of it. That's why you watch the game. |
01-08-2010, 10:38 PM | #159 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
I'm headed to bed, but I'll address this. Win Probability measured absolutely nothing about the context of those events in the subsequent games that followed. Each game, each event is not isolated. That's what I kind of enjoy about baseball. Frankly, I think you're taking all the fun out of it. By the way, your continued use of "strawman" is incorrect. It's not a strawman to say that you believe that statistical analysis is pretty much the end of the discussion, whereas I believe that it's only the start of it. It's not a strawman to say that the statistics you believe in suggest the Martinez was a better slugger than Killebrew. And your other use of strawman in the above post is a failure in your reading comprehension. Plus, the term is unnecessary when we're having a friendly discussion, especially when the main gist of it comes down to a difference between the art and science of baseball. So save it for something important. I don't think we're ever going to see eye to eye and I doubt you'll ever convince me of your views. As you said yourself, "much of what can't be quantified though doesn't matter in terms of objective analysis of the value of a player." That's a notion I fully reject. I think the statistics that some of the Bill James followers cling to are based on today's evaluations of what is important. That's why some of the stats I mention above seem very skewed towards players of a certain era. Normal distributions don't look like that. As such, I think some of these evaluations are made on some very incomplete models for factors that aren't easily quantified. It's easy to say that those things that can't be quantified don't matter. It's also very lazy. You think you're making decisions based on great statistical analysis. I think you're basing your decisions on very flawed, incomplete models and without any of the human factor that makes baseball a game and not a scientific theorem. That's where we are. I'm not going to change your mind. I don't think you're going to change mine. Personally, I'm going to balance the stats with things like peer evaluation and gut intuition. I think there's a balance to be had that's the best of both worlds. That's why it'll be damn difficult to try to convince me of something like Edgar Martinez being a more feared hitter than Jim Rice. You can trot out a ton of statistics to suggest otherwise (and I can throw numbers back at ya). But what it comes down to on something like that is simply I know better. And I'm not the only one. Those who watched the games know better. 76% of the voters knew that Rice belonged in the HoF last year, while only 36% think that Martinez belonged this year. It's not quantifiable. It simply boils down to what you saw with your own two eyes. What your gut tells you. What you know. The stats on Martinez won't change. If that's all voters looked at, then he'll get about 36% of the vote next year. What will change is the perception of him. Of what a voter's gut tell him/her. Of what they know. Whether that increases or decreases his vote totals I don't know. He's a good guy and I don't have a problem with him making it. I just wouldn't vote for him at this juncture. Why? Because I believe he's not a HoFer. It's the same reason I wouldn't vote for Blyleven. And since this has gone unanswered: Can anyone reasonably say that at any point in his career (say for a 3 or 4 year stretch), Blyleven was one of the top 5 pitchers in the game? That's what I want to hear...when was Blyleven actually great? Last edited by Blackadar : 01-08-2010 at 10:53 PM. |
|
01-08-2010, 10:48 PM | #160 | |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
Quote:
73-77 he was pretty darn good. Great is an opinion based question. ERA+ of 158,(lead league) 142, 129, 125, 151. You'd have a hard time today finding a pitcher that was "great" for 3-4 straight years. Pitching is a very fickle position when it comes to stats. Comparing him to Roy Halladays best 5 year stretch 158, 145, 115,(21 starts) 184,(19 starts) 143 Last edited by jbergey22 : 01-08-2010 at 11:01 PM. |
|
01-08-2010, 11:12 PM | #161 | |||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
Quote:
Quote:
Quote:
And that's exactly what you did when you said I must think Brett was a better player than Martinez (etc.) because of the OPS+ measure, and it's what you did when you said I must think Martinez is a better slugger than Killebrew by focusing on career SLG marks rather than looking at ISO and discounting peak value as a means of judgment. If you want to have a healthy debate, stick to what I say and stop using faulty assumptions you attribute to me to argue against. Quote:
Quote:
Quote:
|
|||||||
01-08-2010, 11:35 PM | #162 | |||||||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
Quote:
We can measure how players perform relative to their contemporaries, and how much above or below their contemporaries they rank. Quote:
Quote:
Quote:
Quote:
The top 7 strikeout seasons of all time are by active players. Is it factored in that guys then would be told to do anything but strike out and guys now can swing freely? Do you think Jim Rice would still have the all-time high season for double plays if he knew he could strike out 160 times a season instead of being told early in his career to cut down the 120ks he was averaging? Quote:
Quote:
Quote:
Quote:
Quote:
And to be clear, there's far less difference what statistical analysis says these days and what scouts say than in the past. Much of what was lacking in statistical analysis in the past was an appropriate appreciation for, and good tools for measuring fielding. |
|||||||||||
01-08-2010, 11:38 PM | #163 |
Dark Cloud
Join Date: Apr 2001
|
Blackadar comes back and shows people how to FOFC.
|
01-08-2010, 11:39 PM | #164 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
Damnit, was just about in bed. That's why I asked for 3-4 years and not 5. But I just don't see it. In 1973, his ERA+ was 2nd. In 1974, he was 5th. In 1975, he didn't make the top 10. In 1976, he didn't make the top 10. In 1977 he was 5th again. In that same span, he only is in the top 10 in wins once (and for those who discount wins...well, that's why you play the game!). He's only 8 games above .500 - not appreciably different in W/L percentage from the teams he plays for. He is 2nd, 2nd, 6th, 7th and 8th in Ks. For those 5 years, he's an average of 16-14, with an ERA of 2.75, a WIP of 1.132, an ERA+ of 140 and 1,141 strikeouts. That's pretty strong That's comparable in most ways to Carlton's best 5. Carlton averaged 19-9, ERA of 2.77, WIP 1.135, ERA+ of 137 and 1,037 strikeouts. That's our closest match. It's also comparable to Sutton. 18-11, 2.63, 1.036, 128 and 955. It pales in comparison to Seaver's best 5. Seaver averaged 21-10, with an ERA of 2.35, a WIP of 1.029, ERA+ of 154 and 1,280 strikeouts. Ryan was a different animal. 19-16, 3.03, 1.309, 117 and 1,604. Jim Palmer was 19-11, 2.40, 1.113, 143 and 778. How about some guys not in HoF? Ron Guidry went 17.7, 2.68, 1.118, 143 and 895 strikeouts in that same era. Vida Blue: 14-8, 2.72, 1.104, 123, 779. Luis Tiant: 19-12, 3.12, 1.170, 125, 778. It gives me something to think about. What I did notice about the other HOFers is that they had a whole other set of years that were close. Even the guys not listed - Gaylord Perry, Niekro, etc. - had continued excellence. Blyleven wasn't quite as consistent it seems. Also, those guys all put up much better win totals. I wonder why Blyleven wasn't appreciably better than his team (on average)? |
|
01-08-2010, 11:55 PM | #165 | ||||||||
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
Ok, have it your way. I tried to play nice. Unsubstantiated opinion. Quote:
It doesn't matter if you say it. The stats which you so highly tout do say that Martinez is a better slugger than Killebrew in their respective careers. Now you want to run away from that fact. Which is it? As such, it's not a strawman. It's you trying to weasel out of what you've already established. Strawman. Your failure. Live with it. Quote:
When you try to constantly change the definition of what you believe, I have to question your sincerity. Quote:
Yet you keep throwing out selective statistics. You're the one who casually tossed out OPS+ statistics as the end-all-be-all of a debate, then you ran like hell when I started tossing those stats back at you. Quote:
Then stop trying to throw out selective statistics based on incorrect models, especially across eras. Quote:
Earlier you said that these factors aren't quantified. Now you're saying they're not statistically significant. How do you know that they're not statistically significant if you didn't quantify them in the first place? That's an amazing hole in your logic. Quote:
He's in. Martinez is not. Everything else is conjecture. Quote:
And I think you'd find that decisions made sole with the use of incomplete statistical models don't fare much better. That's why I've already said it's not the sole basis for making a decision. Thank you for continuing to comprehend only half the posts before responding. Last edited by Blackadar : 01-09-2010 at 12:23 AM. |
||||||||
01-09-2010, 12:17 AM | #166 | |||||||||
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
Fair enough. Decent point. Quote:
But then you try to use those rankings against modern players. That's where it doesn't work very well in a lot of instances. Quote:
Ah, so the models aren't perfect. We finally have an admission. Oh joy, oh joy. We have a breakthrough. And how do we measure contemporaries? On statistics that were relevant at the time. And I believe it's appropriate to also see where they were ranked by their contemporaries while their career was ongoing. It's a very simple question - how good did the people of the day find that particular player? Obviously, his peers didn't find Blyleven to be all that great. Quote:
Strawman. The lists don't show that, nor does it address the fundamental issue of using statistics deemed important today that were not considered in the same light 40 years ago to judge players across eras. Quote:
Again, way to miss the point. The issue is the emphasis on modern techniques applied to past eras where those techniques were not emphasized to the same degree and then trying to evaluate the players across eras. If someone comes up with a model that shows that walks are better than hits, so everyone starts trying to draw walks, then we can evaluate players based on walks. We can even look at how past players dealt with walks as compared to their peers. But we cannot use that metric to evaluate the players across eras because walks were not as important in one era as they were in another. One player was told to draw walks. The other was not. Therefore, comparing walks is an illogical comparison. It's really that simple. Or, to break it down in case that's too complex: Player in era 1: Do A. Don't do B. Player in era 2. Do B. Don't do A. I can see how well B is done as compared to one's peers, but it's absolutely fucking useless to compare Era 1 to Era 2. The same emphasis wasn't on each player. Quote:
If you don't know if it's true, go look it up. Perhaps I miscounted. I am a human being and not a calculator. Unsubstantiated, much like your models. Quote:
Even discounting Bonds, the numbers are skewed. (x3, rather than feel some pathological need to quote it again) Quote:
I read that. No, I don't remember where, so go ahead and put "unsubstantiated" as your retort. I don't really care. But the gist was that he uses both sides of the equation to make player evaluations. Quote:
At last, something we can agree on. Last edited by Blackadar : 01-09-2010 at 12:21 AM. |
|||||||||
01-09-2010, 12:19 AM | #167 | ||||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
First off, define what you mean by slugger. My view (and it would by shared by most in the sabermetric community) is that ISO is the best metric for judging "slugger", as it measures only extra base hits. Second, you have to decide what you are valuing - entire career or peak value? Killebrew was superior in ISO in both measures. Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
And thank you for continuing to show you have very little understanding of current statistical analysis and poor debate skills. |
||||||||
01-09-2010, 12:33 AM | #168 |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
This one might go all night.
If I may add something. I really think you need a better understanding of the metrics being used Blackadar in order to try and tear them apart. I also was a non believer at one time. I tried every possible way to shread DIPS with no success. Bill James is so abvanced in his theories at this stage it is really hard to find an argument that is not covered. I dont believe dawgfan is trying to tell you to change your opinion because he is right. I believe he is just trying to tell you that their is evidence to refute a lot of things you are saying. If you want to ignore these stats that is your option but it doesnt mean they are saying anything directly incorrect. The numbers he is showing you are numbers that have been broken down to take out as many variables as can be done at this time. They are very accurate. They have been studied, debated, re-done, changed and re-done over again to get them where they are now. Last edited by jbergey22 : 01-09-2010 at 12:38 AM. |
01-09-2010, 12:37 AM | #169 | ||||||||||
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
When you have to ask for definitions on common terms, you've lost the debate. It's akin to the dreaded godwin. Quote:
Yet you were throwing out OPS+ earlier to talk about feared hitters. Seems your position is a bit incongruent. Quote:
Thank you. I'm having fun while doing it too. Quote:
Ah, I get it. Selective statistics! Didn't I mention this about 10 posts ago? About how stats can be manipulated in multiple ways to come to different conclusions? I'm not gullible enough to believe that if you create analysis from a dataset to get "A" and create another analysis from the same exact dataset to get "B", that I should take your word for it when you want to base your argument on A and then bring out B when it's in your favor. You know, when it comes to baseball, I like to use statistics that the players actually earned without a lot of manipulation. Like batting average. On base percentage. Home runs. ERA. Wins. I'm a meat and potatoes kind of guy. Quote:
Again, selective statistics in action. Quote:
Again, selective statistics in action. Didn't I already talk about this? Quote:
Too late. Quote:
Yet it's the current situation. Isn't that interesting? Yes, I'm purposely trolling on this one. Quote:
I don't dispute it. Do you actually think you're going to convince me (or anyone else) that statistical analysis is virtually absolute after finally admitting the models themselves aren't perfect? Quote:
Ad hominem. |
||||||||||
01-09-2010, 12:41 AM | #170 | |||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
Quote:
The lists that I see show that the yearly leaders in OBP and walks drawn from year to year don't show much difference between eras until you way back a century. The lists I see show that league average OBP doesn't change very much between eras. There's some variation, sure, but not a trend upward as you seem to be contending. In fact, in the era in which Babe Ruth played, league average OBP was quite a bit higher (.353) than it was in the era in which Edgar Martinez played (.337). Quote:
Quote:
Quote:
You made the contention, thus the burden of proof is on you. Especially considering that I have presented data that disputes your assertion. Quote:
Quote:
|
|||||||
01-09-2010, 12:42 AM | #171 | |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
Quote:
Slugger? It can be defined many ways. What definition to you want? Are you looking for the pure home run slugger? Do you want the slugger that hits a lot of doubles and home runs? Do you want a slugger that hits for a low average strikes out a lot with a lot of home runs? That is what he means by ISO(Isolated Power). Do you want to compare a Dave Kingman like hitter OR a Jim Rice type? That is what he is asking you. Last edited by jbergey22 : 01-09-2010 at 12:44 AM. |
|
01-09-2010, 12:47 AM | #172 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
Gee, I'm the only one who has actually linked any models here, yet you presume to tell me I don't understand them? That's seems rather presumptuous. And yes, I've already poked holes in them above - stuff Bill James does not and can not quantify. James' statistical analysis was/is good for the game because before there was none. But treating it like it's some sort of Holy Grail is a major mistake. It's not. That pendulum between gut feeling and statistical analysis will always be swinging. And neither is fully correct. I prefer it as Nuke LaLoosh said in Bull Durham, “This is a very simple game. You throw the ball, you catch the ball, you hit the ball. Sometimes you win, sometimes you lose, sometimes it rains." |
|
01-09-2010, 12:53 AM | #173 | |
Hall Of Famer
Join Date: Jan 2001
Location: Decatur, GA
|
Quote:
Who the fuck cares? Yes, I know... statistics are the devil, I know what I saw bullshit. So a player should be punished more if he's in an era with a lot of great pitchers, even though normalized stats would show that he was a great pitcher? And btw, yes I can. From 1973-75, Blyleven was 1st, 2nd, and 5th in ERA+. Those years he was also 1st, 2nd, and 4th in K/BB (and 2nd in K's all three years), and 2nd, 4th, and 3rd in WHIP.
__________________
"A prayer for the wild at heart, kept in cages" -Tennessee Williams |
|
01-09-2010, 12:53 AM | #174 | |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
Quote:
Well using your "gut" is not going to get you anywhere. Your sample size is going to be too small to get any accuracy in your results. You could have watched Arod in the playoffs 04-08 and thought he was an average player. You could see Mariano Rivera blow a save 6 out of 7 times you seen him pitch and think he is the worst closer of all time. I dont mean to presume that you dont understand these metrics however you are making it hard not to think that. The game is a dream for stat geeks and this is one game that the stats actually do tell the story. Last edited by jbergey22 : 01-09-2010 at 12:56 AM. |
|
01-09-2010, 12:54 AM | #175 | |||||||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
Yes, what a horrible thing to do in a debate - ask for clarification on the definition of something that is being argued. Better to leave that vague. Quote:
Quote:
Quote:
Quote:
Oh lord. Yeah, why use a more precise measurement when a less precise one is available. Quote:
Again, ignorance of what the different advanced metrics measure, and thus what they are best used for. Quote:
Quote:
Quote:
Quote:
I didn't say it was "absolute" - I've explained multiple times that there is definite room for improvement in fielding metrics (but much better data is on the way soon) and some room for improvement in pitching metrics, and not much room for improvement in hitting metrics. Quote:
|
|||||||||||
01-09-2010, 01:04 AM | #176 | |
Hall Of Famer
Join Date: Jan 2001
Location: Decatur, GA
|
Quote:
The problem is that Blackadar is a statistical Neanderthal and like those who deny that the earth goes around the sun, it may just be fruitful to ignore his idiotic arguments. And yes, I tried to say that in the nicest way I possible could.
__________________
"A prayer for the wild at heart, kept in cages" -Tennessee Williams |
|
01-09-2010, 01:09 AM | #177 |
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Yeah, I'm pretty much done with this argument - don't see much point in continuing. Blackadar either doesn't get what my points are or just doesn't care to admit when he's wrong.
|
01-09-2010, 01:10 AM | #178 | ||||||||
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
No, but if OBP wasn't valued as much at some points as others, then it's not a valid comparison to compare players from an era where it was emphasized versus another. If SLG has been emphasized equally, then it's a more valid comparison. The more "creative" the statistical model gets, the more likely it is to over-emphasize some component of performance. Quote:
Didn't we already have this discussion? If you don't like what I say when I back you in a corner, you say "strawman" thinking it's some sort of Get Out of Jail Free card. What, you don't like the same treatment? I understand the term just fine. It's your use of it that's nonsensical. Quote:
Yes, because using the outliers in each year is a great way to establish a baseline. Don't you know anything about statistics? Aren't you supposed to be the statistics buff? Quote:
Yes, and it's not coincidental that many of the great OBP seasons happened during those two live ball eras. That variation isn't background noise, yet that's what you're treating it like. Quote:
I already did. You chose to ignore it. No, you didn't. You ignored it by trying to post a non-relevant statistic. It's a simple principal, really. Player in era 1: Do A. Don't do B. Player in era 2. Do B. Don't do A. I can see how well B is done as compared to one's peers, but it's absolutely useless to compare Era 1 to Era 2. The same emphasis wasn't on each player. Quote:
Sarcasm detection fail. Quote:
Thank you for proving my point. I knew if I waited long enough that you'd do it. The highest single season OBP results are during the live ball era. Notice that similar stats that are supposedly adjusted neatly coincide with those results. That suggests that those stats aren't being adjusted to the degree with which they should. The distribution tables for an adjusted stat look wrong and any Statistics 101 student would realize it pretty quickly. But then again, I don't really care about the models, do I? Quote:
No, it's not. You're splitting hairs. He uses old-fashioned player evaluations along with sabermetrics. If sabermetrics were as foolproof as you portray, they'd all be using it full time. What does it tell you when the best baseball people in the world don't use it as the end-all-be-all of a discussion? That perhaps you shouldn't either? And that's what this all about, isn't it? Last edited by Blackadar : 01-09-2010 at 01:11 AM. |
||||||||
01-09-2010, 01:13 AM | #179 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
Me and every GM in the MLB today don't believe that statistics are the end of the discussion. If you want to think you're smarter than they are, enjoy your little Walter Mitty trip. I'm sure it helps your ego. |
|
01-09-2010, 01:24 AM | #180 |
College Starter
Join Date: Oct 2000
|
my multi-quote literacy level is about 3 multi-quotes per post, so this page just reads like a bunch of scrambled goo.
__________________
... |
01-09-2010, 01:26 AM | #181 | |||||||
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
It's a pretty common baseball term. I'm sorry you don't know what it means. No, but when talking about two non-base stealers, it's pretty easy to figure out. Except for you. Quote:
Let me know when you actually start to debate. Quote:
Bullshit. I've posted links that shows the different results from the same data set based on the emphasis placed on the underlying statistics. If you couldn't be bothered to actually look them up, then no wonder this is going in circles. Quote:
Why use an existing one when you can make a new one up and weight it however you'd like? Quote:
Ad hominem Ad hominem Spiders? Quote:
Entirely relevant. You just don't get it. Quote:
Yet changing the hitting metric gives different results. We've seen that countless times in the last 4 pages. Changing the emphasis gives different results. Why should I compare strikeouts from today's pitchers, even as an adjusted statistic versus players from the 70s. Pitchers today are taught that it's easier to make the out than it is to go for the K. So using it as a metric across eras makes little sense. Yet that's what you're continually doing. It all goes back to the complete games example I posted some pages ago. Are we going to compare complete games by a pitchers in this era versus complete games 30 years ago? It mattered then. It doesn't matter today. So how idiotic is that comparison? It's stupid. It's illogical. Yet this is what you want to do to try to prove a point. Worse, you're trying to be selective about the model and statistics you use. That's not advanced statistic modeling. It's dishonesty. Ad hominem. |
|||||||
01-09-2010, 01:26 AM | #182 | |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
Quote:
What the hell is this suppose to mean? First of all I doubt you have any relationship with any of the current GMs so youd have no way to back this up. And what exactly does this comment prove? Of course they dont believe stats statistics are the end of discussion. They have to try and discover the unknown.(What the 22 year old hot shot rookie will do in the future) Its not even comparable to what this debate is about. I know the A's, Twins, and Red Sox use these metrics to try and predict future success in minor leaguers and seem to be ahead of other organizations in producing talent. |
|
01-09-2010, 01:29 AM | #183 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
I get your points. I just don't believe them. It's not a matter of wrong or right here, bucko...it's a matter of opinion. Your opinion is that models are sound enough to evaluate across eras. My opinion is that they're not. That's something I tried to convey about 10 posts ago, but you wanted to Don Quixote this thing. So let's ride! In the meanwhile, I'm making some good progress on my CM2010 career. Hasn't anyone ever heard of the phrase "Lies, damned lies, and statistics"? Last edited by Blackadar : 01-09-2010 at 01:32 AM. |
|
01-09-2010, 01:34 AM | #184 | |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
Quote:
Other than the fact that the stats he is "trying" to show you are completely relevant while CG's has no relevance at all. ERA+ and OPS+ take era's into the equation. What other stat are you so hellbent over thinking that he is twisting it around to make his point? Honestly, You have completely lost this argument because instead of backing up your own statements you are going after his in which all he is doing is pointing out FACTS. Last edited by jbergey22 : 01-09-2010 at 01:34 AM. |
|
01-09-2010, 01:44 AM | #185 | |||||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
Quote:
And you know this how? Here's a challenge - take a look at Runs Created and explain to us what it's overemphasizing, and show why that is true with something other than "because I said so". Quote:
Definition via Wiki: "To "attack a straw man" is to create the illusion of having refuted a proposition by substituting a superficially similar proposition (the "straw man"), and refuting it, without ever having actually refuted the original position." This is what you did when you tried to argue against me by implying claims or assertions I did not make. You backed me into no corners - all you did was assume arguments on my behalf that I never made. Quote:
Quote:
Another assumption - did I say they were background noise? Ballpark factors and perhaps juiced balls played a role in boosting offensive numbers in those eras. You are contending that it was instead an emphasis on the importance of OBP, and this for some reason declined after the end of the '30's. That seems a difficult theory to support - why would players stop emphasizing approaches that led to such great results as occurred in the earlier live ball era of the '20's & '30's? Doesn't seem logical. Quote:
Quote:
Quote:
Quote:
|
|||||||||
01-09-2010, 01:53 AM | #186 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
No, I've lost this argument in your eyes because you're predisposed - by your very first post - to believe that Sabermetrics is some sort of Holy Grail. Fine, I didn't change your mind. I don't really care. But you're not the arbiter of a win/loss on a debate. That's childish. If that's your MO, then I'll treat you accordingly. I've posted more links, more stats and more facts than dawg has in the last 4 pages. Frankly, I'm the only one who has even posted an examples of the very models we're talking about. Yes, OPS can be adjusted to compare across eras. But if getting on base - a highly valued statistic in OPS - wasn't as highly emphasized in one era as it was another (swing for the fences versus draw the walks has been a subject of some debate for long time), then it's not a valid comparison. Why? Because you don't know what that player could/would have done to emphasize a positive result as it relates to his peers. It's like establishing the criteria for an exam after the test. Posting that a group of players had a high OPS in a live ball era doesn't mean a thing beyond that someone doesn't understand how to use valid data. Of course players in a live ball era had a high OPS. What the hell else would you expect? But to then try to use that to prove that players were aware of their OPS is a logical fallacy. I've seen a lot of those tonight. Wade Boggs once said he could hit 30 homers in a year if he wanted to drop his batting average. And he probably could have. But he didn't. If my statistical model favors home runs versus hits, then Boggs is going to look worse against his contemporaries and also against players of other eras. That's the simple nature of the beast. So you compare his OPS+ to Joe Shmoe and say "well, he wasn't as good". If my model discounts home runs (as many of James' models do), Boggs is going to look better than his contemporaries and therefore against players of other eras. I don't believe it's that hard and fast. It's not that easy. It's never that easy. Players have different roles in different times, different coaches with different methodologies. Statistics in something like baseball are a significant evaluation criteria. But I'll never believe that they're anywhere close to perfect and that's why you have to look at the whole picture. Then you start to look at the whole picture and find out that Boggs went to the All-Star game 8 times while Joe Shmoe went once. You know, at that point, it's time to question the model. Last edited by Blackadar : 01-09-2010 at 01:58 AM. |
|
01-09-2010, 02:05 AM | #187 |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
OPS+ is On Base Percentage plus Slugging (without the variables) and I know without even looking that Boggs would do quite well. It does not favor HRs or any other category. It favors what he has backed up it takes to score runs. Getting on base via base hit or walk and getting the runners in (SLG) using basically it computes to 1 pt for a single 2 pt for a 2b 3 pt for a 3b and 4 pts for a HR. This concept doesnt make sense? Home Runs should be worth more than a triple correct?
What is your magic trick to scoring runs? If you have a better method please share. Your statement doesnt even make sense if you think about it. So on base percentages were high in 1920-1939 then the emphasis changed to swing away and not score as many runs before someone suddenly figured out that drawing walks was the magical trick again? Where are these links you talk about? A better argument would have been that the umpires were calling the game differently from 1940-1989 or the mounds were higher instead of this bogus BS you came up with(getting on base wasnt as valued from 1940-1989???). Last edited by jbergey22 : 01-09-2010 at 02:16 AM. |
01-09-2010, 02:09 AM | #188 | |||||||||
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
See my above post. Quote:
No, you're the one using it. Prove that it works absolutely. I'm the one not treating it like the Holy Grail, remember? Quote:
No, I've used your examples in own cited models to show that they appears to be faulty. You wanted to call those strawman arguments. They're not. Instead, why don't you defend the models? Quote:
You never use the high mark as a basis for a sample. Statistics 101. Quote:
See above post on why your 20s and 30s arguments in regards to OBP don't hold water. Quote:
I don't, any more than you have proof. It seems to me that if you're the one valuing the stat that it's incumbent on you to prove that it's valid. Quote:
The burden of proof is still on you to show that the models are as good as you say they are. You've not even attempted to do so. Quote:
I've given examples where the models are limited and why those discounted factors are relevant. I'm not paid to deconstruct them, you know. Though if you want to hire me to do so, it might be an interesting challenge for a year or three. Quote:
The bar keeps getting lower. First, these stats were used as an absolute. Then it was "most". Now, it's a "big part". At what point do we see "majority"? |
|||||||||
01-09-2010, 02:13 AM | #189 | |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Quote:
Because if the game was called differently for everyone, then this wouldn't matter in an adjusted statistic, would it? Sheesh...and you're supposed to be on dawg's side? |
|
01-09-2010, 02:17 AM | #190 |
Retired
Join Date: Oct 2000
Location: Fantasyland
|
Ah, it's been fun folks. But I have to catch some Zs. Dawg, thank you for the entertainment. I've much enjoyed it. It's been fun messing with you again.
I'm going to go watch a game tomorrow and I'll make sure that I properly evaluate the players with my own two (flawed) eyes to pick out the best player on the field. I'm sure that those with great performances will rise to the top and be recognized as such without the need of detailed statistical analysis on the merits of rushing yardage versus return yardage. Oh, and Rice was still more feared than that pussy Martinez. Last edited by Blackadar : 01-09-2010 at 02:18 AM. |
01-09-2010, 02:27 AM | #191 | |
Grizzled Veteran
Join Date: Nov 2006
Location: Minnesota
|
Quote:
This isnt my argument this is what you should have used as it would have made a lot more sense than the crap you came up with. 1919-"Well boys we arent scoring many runs, maybe we should take a few pitches and work the count and see how it goes. While you are at it start swinging for the fences" 1939-"Well boys I think we might do better if we stop taking so damn many pitches so lets go out there hacking away on the first good pitch you see" 1940-"It looks like our runs are way down but Im not giving up on this hack away concept so get up there and be aggressive" This concept continues until some genious decides 1990 "Well the runs have been down the past 50 years, lets try to take some more pitches so we can draw some walks like they did in the 20's" Is this how it went down Blackadar? |
|
01-09-2010, 02:31 AM | #192 |
Head Coach
Join Date: Oct 2002
Location: Georgia
|
American League Batting Encyclopedia - Baseball-Reference.com
National League Batting Encyclopedia - Baseball-Reference.com Looks like OBP rates aren't that drastically different. They were high in the 30s through the mid-50s, beginning to take a dip through the 60s until reaching the infamous 1968 season, then picking pack up to current levels in the 80s and 90s.
__________________
Top 10 Songs of the Year 1955-Present (1976 Added) Franchise Portfolio Draft Winner Fictional Character Draft Winner Television Family Draft Winner Build Your Own Hollywood Studio Draft Winner |
01-09-2010, 03:06 AM | #193 | |
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
|
|
01-09-2010, 03:14 AM | #194 | |
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Quote:
|
|
01-09-2010, 03:52 AM | #195 | |||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Your above post is not supported by facts. Walk rates have been pretty consistent from the late '50's onward. The primary driver in the difference in OBP rates over the various eras is the difference in BA. Are you suggesting that players in the '50's, '60's, '70's and '80's de-emphasized getting a hit, while maintaining basically the same walk rates?
Quote:
Runs created - Wikipedia, the free encyclopedia Check the 2002 version of the formula, and show me how it's biased and flawed. And again, keep in mind I'm not advocating any one stat as the be-all, end-all for everything. Runs Created is an excellent metric for evaluating the hitting contributions of an individual player on a per-season basis. Quote:
I never said I thought Edgar Martinez was a better player than George Brett. You made a faulty assumption that because I value OPS+ as a metric for judging hitting and comparing season hitting results across eras, that I must then think Martinez was the better player. I never said I thought Martinez was the superior "slugger" to Harmon Killebrew - you assumed that I think SLG is the best measure of "slugging". Etc. Quote:
Quote:
The strong correlation at work here was ballpark changes, and then after 1968 the additional effect of the reduction in the pitching mound height. Quote:
Quote:
Quote:
"But I fully reject the notion - as does every other GM in the league, including Theo Epstein - that they tell the whole story or even most of it." I asked you to back up that statement with some evidence. The best you could come up with was this: "I read that. No, I don't remember where, so go ahead and put "unsubstantiated" as your retort. I don't really care. But the gist was that he uses both sides of the equation to make player evaluations." Once again, "both sides of the equation" isn't incompatible with "most of the story". |
|||||||
01-09-2010, 04:33 AM | #196 | |||||||||
Grizzled Veteran
Join Date: Oct 2000
Location: Seattle
|
Don't know how I overlooked this gem - some real doozies.
Quote:
A "slugger" to me is a hitter who hits for a lot of power, i.e. extra base hits. And that is most clearly measured by ISO, which removes singles from the equation used in SLG. And it's pretty easy to figure out that ISO is a better measure for judging a "slugger" than SLG. Except for you apparently. Quote:
Quote:
What links (plural)? The only link you've posted in this thread is this one: http://members.cox.net/~harlowk22/atgwsobj.html Much of what you've used as a criticism of the above is not valid. You've repeatedly contended that there was a difference in emphasis on walks in different eras, yet the actual walk rates from the late '50's onward has been basically the same and stayed fairly consistent. Quote:
Quote:
Quote:
Quote:
Comparing strikeout totals for pitchers from different eras is flawed, so you'd instead look at how many more strikeouts a guy like Walter Johnson recorded per IP compared to the league average, and see how much that delta differs from the same figures for Randy Johnson. What we're looking for was how much better at this aspect of the game was this guy compared to his peers vs. a guy from a different era. Quote:
Quote:
|
|||||||||
01-09-2010, 06:18 AM | #197 |
Pro Starter
Join Date: Feb 2006
|
Man...sorry I went to bed...
|
01-09-2010, 06:53 AM | #198 | |
Pro Starter
Join Date: Jan 2001
Location: Burke, VA
|
Quote:
The stats do NOT say Martinez was a better slugger than Killebrew. The stats say Martinex had a better slugging percentage than Killebrew. Attributing the qualities of a skill (slugging, defense, etc.) based on a single statistic is incorrect. And that's something you just did. Last edited by Toddzilla : 01-09-2010 at 06:54 AM. |
|
01-09-2010, 03:13 PM | #199 |
High School Varsity
Join Date: Jun 2001
Location: Houston, TX
|
I think a lot of statheads can come on way too strong, and do a disservice to their cause by turning off potential statistical analysis converts. I certainly don't think that's what is happening here - dawgfan and jbergey have done an excellent job of presenting their arguments - but it does happen a lot. Baseball is ideal for statistical analysis - it is a series of individual discrete outcomes. Basketball, football and hockey are all so team-oriented that it's very difficult to isolate the true causal effect of statistical results.
To think that one person's observation is more accurate than 100+ years of data is absolutely absurd. Combine that attitude with Blackadar's brick wall debating, and this thread has turned from a fairly good discussion into a pull-your-hair-out maddening experience.
__________________
I failed Signature 101 class. Last edited by Hammer755 : 01-09-2010 at 03:16 PM. |
01-09-2010, 03:31 PM | #200 | |
Pro Rookie
Join Date: May 2002
Location: Prairie du Sac, WI
|
Quote:
+1 I'm very statistically inclined and in terms of the conversation in this thread can't disagree with a thing jbergey or dawgfan said at all. My bone of contention with some of the statistical community mostly comes in terms of projection instead of analysis of past performance. And actually any author of any projection usually strongly advises people reading these projections not to take them as gospel. Yet I see so many people on baseball forums (can't really say I see it here at all) throwing these projections around as if the next season has already been played. Projections are our best guess, and I love looking at them too. I love sitting on the toilet and reading Baseball Prospectus. But some people just take them way too literally. Again, I'm not pointing the finger at anybody here. I see some of the most reasonable baseball talk here on a football forum. |
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
|
|