What makes for good discussion here? Options

[E5 Yaz]

The problem is - and you are one of the posters who is frequently poor with this - there is data to support points out there. Lots of it. When your fundamental hypothesis is "well they know what they are doing!" it makes the post nonsense. It is the lack of data used to support bold statements of fact that hurts that forum, the fact that an overwhelming number of posters truly believe that data isn't even needed to support points that kills it.

I would say that while I will rarely cite statistical chapter and verse for my opinions, it's not that I don't look at the stats. I just use stats as one of many factors as to why I form an opinion about a move -- and include salary level, significant experience, injury history, what I can gather about a player's intrinsic value to a team, do they step up when it matters most, disciplinary track record, etc.

The problem becomes when posters who don't use stats get labeled as being simply those with blind faith of "they know what they're doing." The Pats leadership knows a helluva lot more than anyone on this board about picking up players, so there is a sense of established trust that goes along with following the team. I know what you're getting at here, and I appreciate the sentiment. But as poor as a lack of support data might be in the midst of a discussion, so is the branding with broad swipes all those who don't use that data in their posts.

I spent the first 15 years of my career in sports as a writer and editor, and during that time covered a fair share of Patriots games. When I post in a fan forum, I'm doing so as a fan, and feel little compulsion to use the sort of supporting evidence that I would on a game story or team analysis. I come here to enjoy myself; not to do what I have had to do at work.

[Rick Burlesons Y...]

I would say that while I will rarely cite statistical chapter and verse for my opinions, it's not that I don't look at the stats. I just use stats as one of many factors as to why I form an opinion about a move -- and include salary level, significant experience, injury history, what I can gather about a player's intrinsic value to a team, do they step up when it matters most, disciplinary track record, etc.

The problem becomes when posters who don't use stats get labeled as being simply those with blind faith of "they know what they're doing." The Pats leadership knows a helluva lot more than anyone on this board about picking up players, so there is a sense of established trust that goes along with following the team. I know what you're getting at here, and I appreciate the sentiment. But as poor as a lack of support data might be in the midst of a discussion, so is the branding with broad swipes all those who don't use that data in their posts.

I spent the first 15 years of my career in sports as a writer and editor, and during that time covered a fair share of Patriots games. When I post in a fan forum, I'm doing so as a fan, and feel little compulsion to use the sort of supporting evidence that I would on a game story or team analysis. I come here to enjoy myself; not to do what I have had to do at work.

This post makes my point. There is really not much I can add.

To clarify a little bit for you:

a) If you were a writer and editor who covered the Pats - I will take it at face value, sure - wouldn't finding a little data to support your points be incredibly easy? I mean, like.......the kind of thing you would have at your fingertips?

b) When one makes the decision to never support their points with data, they should expect to be attacked. A lot. In all forums. At least show that you understand the data. Show us that you follow the data and understand it and don't just think player X is super awesome because that's what the guys calling in to EEI said yesterday. Then you can get a little looser with what you post. Myt1, Saints Rest, NickEsasky.....these guys have established that they get the numbers, they get the nuances, they can post stuff without data. Others need to establish credibility.

[Jnai]

b) When one makes the decision to never support their points with data, they should expect to be attacked. A lot. In all forums. At least show that you understand the data. Show us that you follow the data and understand it and don't just think player X is super awesome because that's what the guys calling in to EEI said yesterday. Then you can get a little looser with what you post. Myt1, Saints Rest, NickEsasky.....these guys have established that they get the numbers, they get the nuances, they can post stuff without data. Others need to establish credibility.

I think this is one of the biggest things that causes problems on SoSH, especially in Member/Lurker interactions. It's not a bad thing, it's just a natural result of having a conversation with some people you know (other members) and people you don't (mostly lurkers).

It's like, if you were sitting at a bar with a bunch of friends, discussing who has the best cheeseburger, and your friend Mark Black* offers up "McDonalds". Everyone laughs and takes another sip of beer. Why? Because a) you know your friend has a thing for quality food, and he's not really offering it up as a serious suggestion for best burger of all time, and also b) you know that last time you hung out, Mark was drunk as hell and ate four McDonalds cheeseburgers.

Some guy across the bar hears "McDonalds", and immediately he gets pissed. "McDonalds?!", he says. He's thinking to himself: McDonalds barely qualifies as food, it tastes like shit, and it's clearly not the best hamburger ever. So he offers up some bar next to his house that he thinks serves the best Burger ever. He yells across the bar, "Mickey's has the best Hamburgers ever, you are an idiot for thinking it's McDonalds".

So, you and your friend look at each other, and ask "Mickey's? Why is that the best burger ever?"

And THIS is the moment where shit goes sour.

Instead of giving a litany of reasons why Mickey's kicks the shit out of McDonalds, this guy across the bar starts yelling at you. "Your friend over there offered up McDonalds, and you didn't question HIM! Why do you question ME? I've been eating Hamburgers for 37 years, and let me tell you, McDonalds sucks. They barely qualify as food. How could you say McDonalds?" Etc, etc, etc.

Some subset of people get defensive really fast, because they don't understand that your friend doesn't have to qualify everything he says, whereas if you randomly interject into a conversation, people want to know a) who the hell you are and b) why you have the opinions you do.

And then the conversation/thread turns to shit, because instead of talking about who serves the best hamburger, now you're arguing about how rude this other guy is who just butted in across the room.

The moral of the story is, if you have good opinions, it's not really that hard to find facts for them, and it makes it a whole lot less awkward when you jump into an ongoing conversation. And it certainly doesn't help to get defensive right away. Usually, people just want to know who you are and why you think you're right, they're not trying to hold you to some higher standard or try to nitpick the shit out of you.

*Names changed to protect innocent.

[Rdsxmbnt]

That is awesome. The wiki specifically was made to give folks a place to be a bit more descriptive about stathead stuff. Perhaps our error there was dilluting the wiki with a bunch of other stuff.

If you type [wiki]BRARP[/wiki] (for example), it will autolink. Those stat-leaning folks can go out there and exapnd on the pages or (hopefully) someone who doesn't understand the stat would go out there and do some research, adding it to the wiki as a resource for others once he or she has collected it.

(wiki)term(/wiki) but use [s instead of (s

I really like this idea, but I don't feel like people will bother with the brackets. Is there anyway to devise a list of terms or pages that will autolink to the wiki when they appear in a thread?

[E5 Yaz]

To clarify a little bit for you:

a) If you were a writer and editor who covered the Pats - I will take it at face value, sure - wouldn't finding a little data to support your points be incredibly easy? I mean, like.......the kind of thing you would have at your fingertips?

Exactly, and I did it on a daily basis for 15 years. But that's not why I come HERE. I simply don't agree that establishing credibility as a fan is solely dependent on posting numbers that anyone (regardless of experience) can find on a Web site

b) When one makes the decision to never support their points with data, they should expect to be attacked. A lot. In all forums. At least show that you understand the data. Show us that you follow the data and understand it and don't just think player X is super awesome because that's what the guys calling in to EEI said yesterday. Then you can get a little looser with what you post. Myt1, Saints Rest, NickEsasky.....these guys have established that they get the numbers, they get the nuances, they can post stuff without data. Others need to establish credibility.

I live in Oregon. I haven't listened to EEI, or any sports talk radio for that matter, in about 10 years. I'm not quite sure which player(s) you believe I've labelled as "awesome" without credentials, but as I said in the earlier post, the evaluation of a player or teams should not be a stats-first exercise. Data is part of the equation, but not the whole equation.

The concept that posters must show fluency with statistical analysis in order to be accepted is, frankly, sad. There are plenty of posters here who can speak sabr, but not all of them can put numbers into perspective. Some simply cut it off at stats and don't accept other factors. The mocking of intangibles by many on this site, for example, is shortsighted.

Each year there's a bash. I've seen the photos and read the accounts of what happened. I sincerely doubt that a significant number of attendees sit in the stands and break down OPS in the middle of the game. The folks are there to have a great time. The thought that posting on this site should put enjoyment on the side while establishing bona fides doesn't make sense.

I didn't bring up my career to puff out my chest. I did so to explain that many people who work in sports media want to have a place to shoot the bull with others about players and teams without having to do so in a fashion they've been paid to do for years.

[Jnai]

The mocking of intangibles by many on this site, for example, is shortsighted.

Not to drag this thread down a different path, but could you expand on this point?

Perhaps I read "intangibles" too literally, but if something is immeasurable or unquantifiable by definition, why would you think it holds any explanatory power? What do "intangibles" buy us? What are they?

Pavlov was fond of the saying:
Can I see it?
Can I measure it?
Can I repeat my results?

If something fails in any of these criteria, it seems either a) unobservable, b) untestable, or c) unsubstantiated. I fail to see why theories about baseball should be held to any less standard than a theory about physiology or a theory about neuroscience.

[E5 Yaz]

I fail to see why theories about baseball should be held to any less standard than a theory about physiology or a theory about neuroscience.

That's actually it in a nutshell. You consider those elements a "less standard." I would argue that intangible assets are not a lesser measure.

I hate to bring this up, but one of the core debates on this site is Tek's value to a pitching staff. Those who break it into measurable statistics can make the claim that Tek's "value" is minimal. We all agree that the Red Sox have the resources to acquire a higher quality of pitcher; we should expect their performance level to justify that ability.
Yet, pitcher after pitcher talks about the work Tek puts into helping the pitching staff. How can that be measured? If, when in doubt, a Lowe or Schilling or Lester gets through a difficult inning by putting themselves in Tek's hands? That doesn't end up on a spreadsheet. Some will call that overrated, and brush off the pitchers' remarks as teammates backing each other up.

Yet stats themselves are faulty. What is the likelihood of Player A getting a hit off Pitcher B in at-bat C? Well, what inning is it? How long has the pitcher been in the game? Is either of them nursing a nagging injury? Is this just a pitcher/hitter who "owns" the other guy for some unknown reason? The outcome doesn't just rely on a pitcher's BABIP or a player's OPS.

All the circumstances of that moment come into play. Joba gets attacked by gnats. ARod yells "I got it" or slaps a ball out of Arroyo's hand. An umpire makes a bad call. Joe West makes the call of his life and the Red Sox win four in a row against the Yankees. All of those things are out of the hands or probability. All of those things matter as much as the calcuable data that is held in higher esteem.

Again, the notion of not being allowed to be looser with commentary until you prove you known formulations is a dictum that seems in direct opposition to being "dedicated to all things Red Sox."

[Ed Hillel]

I agree with Yaz.

Not only are stats in football less indicative of an individual's talents, but it's clearly not the only way to contribute to a football conversation. For instance, if someone wants to talk about Marvin Harrison's route-running what stat should they cite? If someone is talking about Tom Brady's vision and pocket presence, must they cite to his career stats every post? Football isn't as simple as baseball, becaues of the complex patterns and defensive schemes, so it's not always clear what the stats represent. In these circumstances, observation is the best tool. It's why teams have scouts and why coaches and players spend night and day reviewing the film, because it's far more valuable than stats, unlike baseball.

This doesn't mean that people can't make irrational posts and that chaos should be allowed. It means that holding use of stats as a threshhold for acceptable posting is nonsensical. People can make insightful and have a positive influence on the forum, and in many other sports-related areas of the board, by simple observation. Now, if people are in there just inciting anger and derailing threads or not putting any effort into posting and it goes on for a sustained period of time, I agree there isn't much point to keeping them around. It is a Patriots forum, though, so you have to expect some bias and I don't understand why lack of complete objectivity surprises or upsets as many people as it does.

If something fails in any of these criteria, it seems either a) unobservable, b) untestable, or c) unsubstantiated. I fail to see why theories about baseball should be held to any less standard than a theory about physiology or a theory about neuroscience.

Because people exhibit a wide range of often unpredictable emotions that affects both themselves and others around them. For instance, Bill Belichick is considered a master of motivating his players. There's no real objective way to measure that. Just because it's not testable doesn't mean that it's something that shouldn't be mentioned in discussion on this board.

This post has been edited by Ed Hillel: Mar 14 2009, 10:12 PM

[jayhoz]

That's actually it in a nutshell. You consider those elements a "less standard." I would argue that intangible assets are not a lesser measure.

I hate to bring this up, but one of the core debates on this site is Tek's value to a pitching staff. Those who break it into measurable statistics can make the claim that Tek's "value" is minimal. We all agree that the Red Sox have the resources to acquire a higher quality of pitcher; we should expect their performance level to justify that ability.
Yet, pitcher after pitcher talks about the work Tek puts into helping the pitching staff.

snip

When asked by a member of this site which stat was the best measure of a player's performance, Mike Lowell answered RBI's. I'm not sure players are the best judge of other player's "worth/value" with respect to wins.

This post has been edited by jayhoz: Mar 14 2009, 10:09 PM

[E5 Yaz]

When asked by a member of this site which stat was the best measure of a player's performance, Mike Lowell answered RBI's. I'm not sure players are the best judge of other player's "worth/value" with respect to wins.

And I'm not sure those who delve deeply into stats are the best judge.

This post makes me hate sports. And people.

This would appear to be an intangible conclusion.

This post has been edited by E5 Yaz: Mar 14 2009, 10:15 PM

[jayhoz]

And I'm not sure those who delve deeply into stats are the best judge.

Nice straw man.

Your answer reinforces the point. What evidence other than your opinion backs up the above claim?

[Jnai]

That's actually it in a nutshell. You consider those elements a "less standard." I would argue that intangible assets are not a lesser measure.

I hate to bring this up, but one of the core debates on this site is Tek's value to a pitching staff. Those who break it into measurable statistics can make the claim that Tek's "value" is minimal. We all agree that the Red Sox have the resources to acquire a higher quality of pitcher; we should expect their performance level to justify that ability.

Yet, pitcher after pitcher talks about the work Tek puts into helping the pitching staff. How can that be measured? If, when in doubt, a Lowe or Schilling or Lester gets through a difficult inning by putting themselves in Tek's hands? That doesn't end up on a spreadsheet. Some will call that overrated, and brush off the pitchers' remarks as teammates backing each other up.

Yet stats themselves are faulty. What is the likelihood of Player A getting a hit off Pitcher B in at-bat C? Well, what inning is it? How long has the pitcher been in the game? Is either of them nursing a nagging injury? Is this just a pitcher/hitter who "owns" the other guy for some unknown reason? The outcome doesn't just rely on a pitcher's BABIP or a player's OPS.

All the circumstances of that moment come into play. Joba gets attacked by gnats. ARod yells "I got it" or slaps a ball out of Arroyo's hand. An umpire makes a bad call. Joe West makes the call of his life and the Red Sox win four in a row against the Yankees. All of those things are out of the hands or probability. All of those things matter as much as the calcuable data that is held in higher esteem.

Again, the notion of not being allowed to be looser with commentary until you prove you known formulations is a dictum that seems in direct opposition to being "dedicated to all things Red Sox."

In fact, you and I are not so far apart.

You think that there is a model for player performance. It includes many things.

We could write it as something like:

Likelihood of Getting a Hit = A + B + C + D + E + F + ... V

where A=, say, a Player's skill at Hitting, B=The Pitcher's Talent, etc., etc., down to V, which equals "random variation".

We both believe that a player's skill at hitting is, in some way, measurable. For example, we believe that Manny Ramirez is a better hitter than Julio Lugo and this appears in a number of metrics of performance, such as OPS, Home Runs, whatever.

We both believe that a pitcher's skill at pitching is, in some way, measurable. For example, we believe that Josh Beckett is a better pitcher than Javier Lopez, and this appears in a number of metrics of performance, such as Strikeouts, WHIP, etc.

We also both believe that there is random variability, "V". Because of this random variability, the best we can generate is an estimate of performance. But, across many situations, this random variability should be measurably different from the factors that actually influence performance. That is why we say some pitchers are better than others, even though they do not ALWAYS perform better in all situations.

I assume we're both on the same page. Let's now turn to Tek's ability as a catcher.

If Tek's ability (or any catcher's ability) should get a parameter in the model, then it must be clearly separable from the influence of random variability. To do so, by definition, it must not be "intangible". We should be able to conduct some analysis (even a hypothetical analysis) which will parse situations in which Tek's ability as a catcher had an impact (presumably: when he is catching) from those in which Tek's ability as a catcher did not have such an influence. Do you agree? If not, then Tek's ability as a catcher cannot be isolated and therefore has no effect.

It is not up to me to do so. It is up to you to do so. If you think that Tek's ability as a catcher has an influence, then demonstrate it. Quoting players is not sufficient demonstration, because your assertion is that Tek's ability has an influence on performance. Demonstrate that.

"Intangible" means either unobservable, untestable, or unsubstantiated. These attributes are necessarily held at a lower standard because logic dictates that we must first looks to those features of the situation that have an observable, tested, and substantiated effect on the outcome. If we see a car suddenly roll to a stop, we generate a system of predictions about why that might be. "Did it run out of gas?" (Such a scenario would cause a car to stop). "Is the engine damaged?" "Are the tires punctured?"

We don't ask things like, "What is today's Zodiac sign?" We don't trust the driver when he told us that his car stopped because earlier he broke a mirror and he has been condemned to seven years of bad luck. We look to things that have observable, testable, and substantiated effects on the car.

So, it's not that I hold things like Tek's skill as a defender to be less valid than Josh Beckett's ability to throw a fastball. It's just, I can conceive of a hypothetical analysis in which I could measure the latter, and I can conduct it, and I can find an effect on performance. Can you do the same with Tek's skill as a defender? If not, have you bothered to ask yourself why you think it has any impact on the situation worth discussing if it cannot be isolated from random variability?

This post has been edited by Jnai: Mar 14 2009, 10:40 PM

[Jnai]

I agree with Yaz.

Not only are stats in football less indicative of an individual's talents, but it's clearly not the only way to contribute to a football conversation. For instance, if someone wants to talk about Marvin Harrison's route-running what stat should they cite? If someone is talking about Tom Brady's vision and pocket presence, must they cite to his career stats every post? Football isn't as simple as baseball, becaues of the complex patterns and defensive schemes, so it's not always clear what the stats represent. In these circumstances, observation is the best tool. It's why teams have scouts and why coaches and players spend night and day reviewing the film, because it's far more valuable than stats, unlike baseball.

Obviously, not all post need include statistics. Football is a game of complex strategy and high variability. But, in those cases where you do discuss something, it should be either documentable or previously documented.

Compare the following statements:
1) Tom Brady is a good quarterback because he has good vision in the pocket.
2) Tom Brady is a good quarterback because he likes cheetos.

The second statement is far more verifiable than the first (During a Press conference, someone can simply ask Tom, "Do you like Cheetos?"). But the first is far more likely to have an impact on his behavior as a quarterback.

When coaches are reviewing film, they are absolutely using statistics. Coaches don't review film for fun, they review it so that they can learn the opponent's tendencies, learn how to best "match up" against their opponent, and generally look at all sorts of MEASUREABLE and TESTABLE events. In fact, I think you could argue that a football game is one of the purest examples of experimentation in sports. Coaches start with a strategy that they have hypothesized will work before the game. If it works, they stick with it. If not, they refine what they are doing based on what they see during the game. Football is absolutely determined by measurable and observable events.

I promise you that coaches do not watch film so that they can see how motivated players seem or how much cheetos they have eaten. They are absolutely looking for measurable events that will have a meaningful impact on a game.

Because people exhibit a wide range of often unpredictable emotions that affects both themselves and others around them. For instance, Bill Belichick is considered a master of motivating his players. There's no real objective way to measure that. Just because it's not testable doesn't mean that it's something that shouldn't be mentioned in discussion on this board.

You can talk about it all you want. But, what are you using it to explain? But if there's no way to measure it, how can it offer justification for why things are the way they are? If the variable has no measurable impact, then what good does it do to explain things?

This post has been edited by Jnai: Mar 14 2009, 10:40 PM

[E5 Yaz]

Nice straw man.

Your answer reinforces the point. What evidence other than your opinion backs up the above claim?

And this is yet another problem, the automatic "straw man" response. Players who actually have to play the game, and have succeeded as much as Mike Lowell, believe in RBI. Who am I to say they're wrong, whether I have done deep statistical analysis or not?

[ScubaSteveAvery]

So essentially, people on this board have to approach baseball, football, hockey, whatever with a complete scientific approach in order for them to gain credibility?

[Jnai]

So essentially, people on this board have to approach baseball, football, hockey, whatever with a complete scientific approach in order for them to gain credibility?

Framed that way, it sounds awfully scary. But, people claiming to know why certain events happen should have measurable and testable reasons for thinking so.

That's not anything special about science.

[Jnai]

But not ever event can be attributed fo measurable and testable reasons. Al Weis in the 1969 World Series comes to mind.

Right. Read my post above, and be on the lookout for "V". The difference is that we both assume that that event was caused by random variation.

You are claiming that there are other "intangibles" that cannot be isolated from V but yet still play a role in determining whether or not someone succeeds. And I am saying, if they cannot be isolated from V, then there is no point in discussing them, because they have no measurable impact on the game.

[jayhoz]

And this is yet another problem, the automatic "straw man" response. Players who actually have to play the game, and have succeeded as much as Mike Lowell, believe in RBI. Who am I to say they're wrong, whether I have done deep statistical analysis or not?

The straw man was your assertion that I thought "those who delve deeply into stats are the best judge". Something I never said.

[fletcherpost]

If Tek's ability (or any catcher's ability) should get a parameter in the model, then it must be clearly separable from the influence of random variability. To do so, by definition, it must not be "intangible". We should be able to conduct some analysis (even a hypothetical analysis) which will parse situations in which Tek's ability as a catcher had an impact (presumably: when he is catching) from those in which Tek's ability as a catcher did not have such an influence. Do you agree? If not, then Tek's ability as a catcher cannot be isolated and therefore has no effect.

Dude, I understand everything in your post except I'm a bit...I'm unfamiliar with some of the terms as you use them. Could you - I dunno if translate is the word but - I 'd like to understand what you are saying. I know you're saying some good shit...cos i understand the words in the other paragraphs but this one above is getting in my way. If i can understand the bit prior to the bit in bold...i might be on my way. I'm not sure I agree with you 100% see...but that's just my intuition...which i don't trust.

EDIT I have now read the entire post three times, each more slowly than the last and I'm understanding less and less. I think i know what you are trying to say...but I do not know what you are actually saying. By this I mean the a-b-c of your post seems to jar with me...even though I know what you are trying to say. (That's me partly trusting my intuition that you know what your talking about, but I'm losing in translation and I rally want to understand your position.)

PM me if you like. But let me know if you're gonna fob me off.

This post has been edited by fletcherpost: Mar 14 2009, 11:20 PM

[ScubaSteveAvery]

Framed that way, it sounds awfully scary. But, people claiming to know why certain events happen should have measurable and testable reasons for thinking so.

That's not anything special about science.

I think what E5 and others are trying to say is that what makes you so sure that the equations and stats you use are any better at explaining what happened than observation or basic instinctual thoughts? Baseball statisticians are constantly updating and improving stats because the prior stats were inadequate at properly explaining the situation. So in citing current stats, people might actually be wrong. But according to your posts in this thread, just the presence or act of people using the scientific method makes them more legitimate (as long as its not dumbass stuff that wouldn't fly in a basic stats class).

Part of the fun of baseball, and life, is that things just happen sometime and there isn't some statistical explanation for it. This may be offensive to you, but it would be nice if some of the boards best statisticians (I'm using this term loosely) would acknowledge that there is another way of looking at the game. It may be less formal and it may sound crazy, but using stats doesn't necessarily precede intelligent conversation.

In no way does this make pitchfx, for example, illegitimate in many people's eyes. But for me, and I'm guessing a lot of other people, it is a cool tool that has some great explanatory power. I've tried to understand it but cannot completely wrap my head around it. I appreciate when you or sprowl can describe the charts in layman's terms. But baseball isn't as fun to me when I have to understand that stuff in order to talk about it here. I know I'm not alone in my opinion (and if I am, maybe this isn't the place for me). I thought the point of this board was to have fun talking about the Red Sox.

However, some people may have intelligent conversations about baseball and pitching and may not use pitchfx or ERA+ and still have a great conversation. I'm not saying that people should be allowed to unquestionably say "Josh Beckett was great today because I ate a tri-tip sandwich." There is no logical way to explain that. But, if somebody says "Manny's antics near the trade deadline last year really affected how the team played during those two weeks," why should have to start providing a bunch of stats to back it up. It's certainly plausible and I think logical in some ways. It was observable for sure. It may even be testable by looking at the teams win/loss record before and after the trade, batting averages of players before and after the trade. But Manny's antics cannot be put into an equation because how do you quantify them? And of course, there can be other factors that can substantiate the claim, some of them psychological effects that are unquantifiable. I don't think "v" in this case can be isolated. But why does that make it not worth talking about? And why would a post like that be less legitimate?

This post has been edited by ScubaSteveAvery: Mar 14 2009, 11:18 PM