0904: "Sports"

This forum is for the individual discussion thread that goes with each new comic.

Moderators: Moderators General, Prelates, Magistrates

Chrisfs
Posts: 49
Joined: Fri Apr 10, 2009 7:54 pm UTC

Re: 904: "Sports"

Postby Chrisfs » Fri May 27, 2011 5:35 pm UTC

from canada wrote:Image


The big difference is that science has on a number of occasions been able to make sturdy predictive models, and has structures in place to go back and check themselves for accuracy, whereas sports commentators or people chatting idlely in sports bars have not.

Chrisfs
Posts: 49
Joined: Fri Apr 10, 2009 7:54 pm UTC

Re: 904: "Sports"

Postby Chrisfs » Fri May 27, 2011 5:40 pm UTC

furgle wrote:
wlao wrote:
CorruptUser wrote:Financial analysis is otherwise rarely based on random numbers.

You obviously don't know anything about accounting, do you?

Forum Sports Commentator 1 aka Bob: Wow!, did you see that play exchange now wlao has been around her for some time since 2009 but this is his 6th post and he comes in with a strong troll move.
Forum Sports Commentator 2 aka Steve: now hold on their bob, it could be a genuine piece we'll have to wait and see.
Bob: Now I can't wait to see CorruptUser's response he's been in the competition since 2009 too but has over 900 posts. I expect some class response
Steve: Ofcourse this could be a ploy and a way to get this whole discussion off topic, I can't wait to see how this pans out. We might even see some grammar nazi tactics.
Bob: I sure hope so!


After typing this i think i know why sports commentators exist. Self indulgence.


Bob: Furgle is displaying an excellent command of meta for someone who has been at this game for just over a year
Steve: I think we can expect to promising things from this guy in the upcoming season.

User avatar
mani.cavo
Posts: 5
Joined: Fri May 27, 2011 12:24 pm UTC

Re: 904: "Sports"

Postby mani.cavo » Fri May 27, 2011 7:41 pm UTC

suzi wrote:If you only watch casually you might think that passes/plays/shots/goals/fouls are "random," but I find that a really belittling and so incredibly tedious to find truth in.
I'll note that you didn't say you find it to be incorrect, just belittling.

The word "random" has too much baggage and I think that's what's tripping you up here. Saying that a system is a "weighted random number generator" is not saying that there is no impact of skill, no trends, no patterns, etc. Those are all the factors/results of the weighting. Unless you think players and referees can control the weather, and everything else that might impact their performance in or outside of the game, then you need to acknowledge the element of randomness.

Yes, you're right that there's an element of randomness everywhere in life. That should be how you know he's not belittling the sports themselves: because it's true of everything. But we don't necessarily try to derive flavorful dramatic narratives from everything all the time, as though these were predictive explanations.

To give you an example: When I filled in scantron tests as a kid, I would imagine that "A" and "B" answers were a villain, and "C" and "D" answers were heroes trying to fight it. Based on how the answers wound up, I had choreographed entire dramatic battles in my head, and even had a very strict and consistent set of rules for interpreting the answers.

But if you asked me how my test went, and I said "There were a lot of reversals, and the heroes held their ground initially and tried to break the villain apart with strong counters, but in the end he was just too much for them and they ultimately fell under the barrage before being eliminated by a supercharged coup de grace!" you would correctly note that this
A. Is, although based on the experience of taking the test, pretty divorced from reality as a conclusion about the test
B. Is not predictive at all
C. Has nothing to do with the actual test (or design, including potential beauty, of the test), and most importantly
D. Doesn't give you an idea for what actually resulted (i.e. my grade) at the end of the day, because instead of focusing on the actual weights I'm just spinning a pretty thread from the events as I see fit

Basically, I'm really sick of otherwise smart, analytical, and thoughtful people dismissing sports based on what little contact they have with them in the mainstream media.
Nothing in the comic is dismissing sports; it's dismissing sports commentary in the mainstream media, specifically.
Our world, with its rules of causality, has trained us to be miserly with forgiveness.

zAlbee
Posts: 95
Joined: Fri Jun 04, 2010 9:21 pm UTC

Re: 904: "Sports"

Postby zAlbee » Fri May 27, 2011 8:46 pm UTC

Excellent comic. Esp re: financial analysis.

clanders
Posts: 47
Joined: Thu Jan 14, 2010 12:07 am UTC

Re: 904: "Sports"

Postby clanders » Fri May 27, 2011 11:00 pm UTC

Randall once again "writes" a comic that grossly oversimplifies and misunderstands something. I thought scientists were all about being precise and exact?

User avatar
Schadenfreude
Posts: 26
Joined: Thu Jul 15, 2010 5:03 am UTC
Location: "I'm loyal to nothing, general...except the dream."
Contact:

Re: 904: "Sports"

Postby Schadenfreude » Sat May 28, 2011 12:48 am UTC

clanders wrote:Randall once again "writes" a comic that grossly oversimplifies and misunderstands something. I thought scientists were all about being precise and exact?


Forum bait, dude. The man answers to people who identify themselves as nerds, not to sports fans. Given that pretty much all humans tend to clump themselves into "us-vs-them" groups, it's understandably more marketable to pander to the fan base than to provide us with a deeper and more enriching understanding of "them" groups. You know, even if some people identify with sports fans.

Either that, or he's just copying the onion. You know, just like he did six comics ago.

Oh, hell. I'ma just say he's shooting for both.
"Shake my hand. Come on, boys, won't you shake a poor sinner's hand?"

User avatar
vodka.cobra
Posts: 371
Joined: Thu Mar 27, 2008 6:50 pm UTC
Location: Florida
Contact:

Re: 904: "Sports"

Postby vodka.cobra » Sat May 28, 2011 2:49 am UTC

clanders wrote:Randall once again "writes" a comic that grossly oversimplifies and misunderstands something. I thought scientists were all about being precise and exact?

Wow, and here I thought he was a humorist.
If the above comment has anything to do with hacking or cryptography, note that I work for a PHP security company and might know what I'm talking about.

Brooks Hatlen
Posts: 24
Joined: Tue Feb 15, 2011 5:47 am UTC

Re: 904: "Sports"

Postby Brooks Hatlen » Sat May 28, 2011 6:55 am UTC

Schadenfreude wrote:Either that, or he's just copying the onion.

I don't see a connection between this article and the comic other than they are both humorous and related to sports.

Humanist Geek
Posts: 6
Joined: Wed Jan 05, 2011 7:21 am UTC

Re: 904: "Sports"

Postby Humanist Geek » Sat May 28, 2011 7:59 am UTC

sotanaht wrote:The mode for multiple fair dice of any size is (range/2)+min, with odd ranges resulting in 2 modes (round up AND down). Specifically, 2d6 is 7 ((12-2)/2)+2, 3d6 is 10 AND 11 ((18-3)/2)+3, 2d20 is 21 ((40-2)/2)+2) and so on.

TBH i just worked that out for examples. If someone can show me a proof for it (by induction most likely) I would like to see it.


Actually, it's just...
((highest sum of dice rolls possible)+(lowest sum of dice rolls possible))/2 = (sum of average dice rolls)

When all dice involved have a "1" side (which is usually the case), the above equation can be simplified to
((highest sum of dice rolls possible)+(number of dice))/2 = (sum of average dice rolls)

Over the spread of multiple rolls, the average value generated by any fair die with a "1" side should be
((highest possible value)+1)/2 = (average result)

And if someone already posted this, oh well. I'm tired, I'm not going to read anymore to find out if this is redundant. And I'm not going to figure out how to use the math formatting tools that are available to me in the authorship of this forum post.

'Night.

EDIT: Wait, did I just restate what I quoted? I dunno...

Wolfram Alpha has a limited amount of functionality for probability distributions of dice rolls. Sure, it can figure out 4d56, but... wait, they've improved the functionality! Now one can do stuff like 2d6+1d4. Yay! Still can't do stuff liketwo six-sided dice plus one or [urlhttp://www.wolframalpha.com/input/?i=1d20+%2B+1d6+-1d2]1d20+1d6-1d2[/url]. </rambling>

<sleep>zzzzzzzzzzzzzzzz.......

User avatar
eran_rathan
Mostly Wrong
Posts: 1846
Joined: Fri Apr 09, 2010 2:36 pm UTC
Location: in your ceiling, judging you

Re: 904: "Sports"

Postby eran_rathan » Sat May 28, 2011 1:23 pm UTC

"Y aquí viene Thorgrim el enano, que tiene su hacha a cabo, es
teniendo un columpio, cerca de perder por Lorienanthelar, pase bien, sin embargo,
que está saqueando el elfo inmediatamente inferior, pero está arriba y balanceo de nuevo,
corriendo los malditos bastardos poco lo que puede ser rápido y GOOOOOOOOOAAAAAAAAALLLLLLLL!!!!!"


Mexican soccer announcers make everything better.



Also, I have no idea what this says in Spanish really, I just used Google Translate.
"Does this smell like chloroform to you?"
"Google tells me you are not unique. You are, however, wrong."
nɒʜƚɒɿ_nɒɿɘ

abhi2point0
Posts: 3
Joined: Thu May 26, 2011 6:41 am UTC

Re: 904: "Sports"

Postby abhi2point0 » Sat May 28, 2011 2:26 pm UTC

This strip should be dedicated to Nicholas Nassim Taleb of Black Swan fame. It's got all his pet peeves aka financial prediction, competitive sport, nerdiness and weighted random numbers becoming narratives.

User avatar
vodka.cobra
Posts: 371
Joined: Thu Mar 27, 2008 6:50 pm UTC
Location: Florida
Contact:

Re: 904: "Sports"

Postby vodka.cobra » Sat May 28, 2011 4:52 pm UTC

eran_rathan wrote:"Y aquí viene Thorgrim el enano, que tiene su hacha a cabo, es
teniendo un columpio, cerca de perder por Lorienanthelar, pase bien, sin embargo,
que está saqueando el elfo inmediatamente inferior, pero está arriba y balanceo de nuevo,
corriendo los malditos bastardos poco lo que puede ser rápido y GOOOOOOOOOAAAAAAAAALLLLLLLL!!!!!"


Mexican soccer announcers make everything better.



Also, I have no idea what this says in Spanish really, I just used Google Translate.

So you went out of your way to post something in a language other than the language of the forum and most of its users, and you don't even know the language in question?

"Mexican soccer announcers make everything better."

Yeah well maybe some of them make everything worse. Why are you generalizing and patronizing? Are you prejudiced?
If the above comment has anything to do with hacking or cryptography, note that I work for a PHP security company and might know what I'm talking about.

TheFreeloader
Posts: 2
Joined: Tue Feb 01, 2011 8:42 pm UTC

Re: 904: "Sports"

Postby TheFreeloader » Sat May 28, 2011 5:56 pm UTC

I think he meant technical analysis or quantitative finance when he said financial analysis.

Financial analysis actually does tell you something about how a company is doing fundamentally. Technical analysis and quantitative finance on the other hand are in mainly just mindless pattern recognition.

User avatar
mootown2
Posts: 4
Joined: Tue May 17, 2011 1:44 pm UTC
Location: Rural America (currently central Kentucky - previously West Virginia)

Re: 904: "Sports"

Postby mootown2 » Sat May 28, 2011 6:01 pm UTC

ok comic, I laughed more audibly at from canada's spoof ... thanks fixblor for the link unfortunately I won't be able to use that Vonnegut clip in the math classes I teach since he uses a little profanity ...

User avatar
vodka.cobra
Posts: 371
Joined: Thu Mar 27, 2008 6:50 pm UTC
Location: Florida
Contact:

Re: 904: "Sports"

Postby vodka.cobra » Sat May 28, 2011 6:07 pm UTC

mootown2 wrote:ok comic, I laughed more audibly at from canada's spoof ... thanks fixblor for the link unfortunately I won't be able to use that Vonnegut clip in the math classes I teach since he uses a little profanity ...

Damn you, anti-profanity laws.

I don't see why it's so taboo. Rude? Sometimes. But for fuck's sake everyone in my high school cussed why would it matter if a teacher did?
If the above comment has anything to do with hacking or cryptography, note that I work for a PHP security company and might know what I'm talking about.

User avatar
neoliminal
Posts: 626
Joined: Wed Feb 18, 2009 6:39 pm UTC

Re: 904: "Sports"

Postby neoliminal » Sat May 28, 2011 6:24 pm UTC

D&D makes commentary on the results of dice rolls in the same manner that sports commentators make them on seemingly random events in sports. The narrative in D&D is a story and the best sports commentators also weave stories because audiences prefer context while watching sports (and roleplayers like context while rolling dice...)

However, since my mother died in a sports commentator related accident, this comic isn't funny.
http://www.amazon.com/dp/B0073YYXRC
Read My Book. Cost less than coffee. Will probably keep you awake longer.
[hint, scary!]

User avatar
Ephemeron
Posts: 282
Joined: Sun Jan 31, 2010 12:39 pm UTC

Re: 904: "Sports"

Postby Ephemeron » Sun May 29, 2011 1:01 am UTC

Radical Pi wrote:Wait, did Randall just prove that D&D is equivalent to mainstream sports? Sweet!


No, no, that is wrong. The content of his non-argument is an abstraction to an analogy that describes both. I could use the same logic to prove that these forums and the Xkcd Sucks blog are equivalent.

Image

User avatar
vodka.cobra
Posts: 371
Joined: Thu Mar 27, 2008 6:50 pm UTC
Location: Florida
Contact:

Re: 904: "Sports"

Postby vodka.cobra » Sun May 29, 2011 2:10 am UTC

Ephemeron wrote:
Radical Pi wrote:Wait, did Randall just prove that D&D is equivalent to mainstream sports? Sweet!


No, no, that is wrong. The content of his non-argument is an abstraction to an analogy that describes both. I could use the same logic to prove that these forums and the Xkcd Sucks blog are equivalent.

Image

Secondary picture caption: Captain Obvious edits a webcomic.
If the above comment has anything to do with hacking or cryptography, note that I work for a PHP security company and might know what I'm talking about.

Hirg
Posts: 18
Joined: Mon May 16, 2011 10:04 pm UTC

Re: 904: "Sports"

Postby Hirg » Sun May 29, 2011 2:47 am UTC

cream wobbly wrote:
SpringLoaded12 wrote:
CorruptUser wrote:
monty30 wrote:
CorruptUser wrote:
Brooks Hatlen wrote:
glasnt wrote:Since when is D&D weighted?!

As in any game with multiple dice, there are numbers more likely to come up than others, namely half the maximum total roll.


Only when there is more than 1 die rolled. Otherwise it's (almost) a uniform distribution.


It isn't half of the maximum total roll, although that could be a rough approximation.
For example, the most common sum from rolling two dice is 7, and not (6+6)/2 = 6.

Edit: assuming six sided dice


I know it's not correct, but was correcting a different problem; single die rolls have no mode, assuming a fair die of more than 1 side. No die is truly perfect, but close enough that you can assume it to be.

PEOPLE.
PLEASE.
The alt-text describing D&D as based on weighted random number generators was a joke about the fact that you could call dice weighted random number generators, in that they are used to generate random numbers and are "weighted" in a sense of the term because they are physical objects that have some physical weight. See, it's a joke.


Oh good grief. While you are correct in your final sentence, the rest of what you wrote is complete nonsense, I assume based on tangential knowledge of the craft. See, the joke is that D&D has a referee who rolls dice secretly as a guide only: when the DM wants to move the action along or kill off a character because nobody likes a smartass (especially a DM, attributable to self-loathing), then he (always "he", even if technically female) "interprets", i.e. lies about the result. In D&D, this is a dirty little secret. In other RPGs, it's explicitly called out in the rules: the referee is a player too. Moreover his (the gender-ambiguous pronoun this time) role is that of storyteller, in addition to arbiter.


The entire conversation about D&D is somewhat ridiculous, because the simple truth is that D&D does function on weighted random number
generators. One does not merely role a die and use the value shown on the die; nearly every toss of the dice is modified by some value that reflects a character's abilities. For example, a stronger character will add a larger value to all die rolls involving feats of strength, and a weaker character will add a smaller value (or maybe even subtract a value) from die rolls involving strength.

Even the standard method of designing a character involves weighted random number generators: all ability scores are calculated using 3d6 (the total shown on three six-sided dice), but one rolls 4d6 (four six-sided dice) and takes the three highest numbers for each score. Furthermore, a player generally has an idea of what kind of character they want to play before rolling the dice, thus adding another weight: a player who wants to play a strong character will naturally end up with a higher ability score assigned to strength than a player who doesn't care about strength.

Just rolling dice doesn't let you play D&D. Weighting dice rolls lets you play D&D.

User avatar
addams
Posts: 10336
Joined: Sun Sep 12, 2010 4:44 am UTC
Location: Oregon Coast: 97444

Re: 904: "Sports"

Postby addams » Sun May 29, 2011 3:39 am UTC

from canada wrote:
WolfieMario wrote:
michaelsaunders77 wrote:Draw 2 well-placed line segments and the comic becomes two spheres balanced atop tetrahedra.

Done.

Image


i think you both need to learn what a tetrahedron is...

This is so cute. Minimalist.
If, this were submitted by the Master, then, I would expect criticisms.
Because, it was submitted by one of the crowd, we can enjoy it.

Minimalist art work allows the viewer to find meaning.

I love this minimalist work. People sometimes see patterns in one another.
Shoot. We always see patterns in one another.

The next frame could have hearts floating above the heads of the figures.
That would be sweet.

Or; It could be a tragedy. Hearts above one of the heads and a different figure above the other. Don't so it. There is too much heart ache in the world, now.
Let the stick figures find one another and wander off solving one interesting problem after the other.
Life is, just, an exchange of electrons; It is up to us to give it meaning.

We are all in The Gutter.
Some of us see The Gutter.
Some of us see The Stars.
by mr. Oscar Wilde.

Those that want to Know; Know.
Those that do not Know; Don't tell them.
They do terrible things to people that Tell Them.

Cal Engime
Posts: 38
Joined: Mon Apr 26, 2010 8:41 am UTC

Re: 904: "Sports"

Postby Cal Engime » Sun May 29, 2011 4:33 am UTC

Thread title needs a leading zero.

User avatar
vodka.cobra
Posts: 371
Joined: Thu Mar 27, 2008 6:50 pm UTC
Location: Florida
Contact:

Re: 904: "Sports"

Postby vodka.cobra » Sun May 29, 2011 5:20 am UTC

Cal Engime wrote:Thread title needs a leading zero.

This has been secretly pissing me off, which is probably why I keep getting drawn back to it.
If the above comment has anything to do with hacking or cryptography, note that I work for a PHP security company and might know what I'm talking about.

cptjeff
Posts: 44
Joined: Thu Sep 03, 2009 4:42 am UTC

Re: 904: "Sports"

Postby cptjeff » Sun May 29, 2011 6:23 am UTC

Chrisfs wrote:
CorruptUser wrote:
If you were to quiz actual market participants, they would be ignorant as to the supposed reasons for their actions. I often buy or sell, not based on broad economic themes (or whims), but rather on the outlook for a certain stock. Some people will sell not because of their lack of confidence in the market but simply because they need the funds.



They probably would be. It's the same reason that we can't predict tomorrows weather, but we can model the climate 100 years from now with a fair degree of consistency. The randomness cancels out over time, revealing broader trends. Enough samples will do that.

The, say, skepticism over the dollar may not drive every single person involved in the market to sell .01% of their portfolio, but it does tip a proportion of people in a general direction. There may also be subconscious processing of information going on that traders aren't consciously aware of. Read the book Blink. With enough training and repetition in something, you don't have to think about it on a surface level, which is a good thing. It turns out our subconscious is really, really good at taking in a shiatload of information and processing it really quickly. In athletics, it's known as muscle memory, and it's the reason A-Rod can determine where a pitch will be located nearly as soon as it leaves the pitcher's hand, and you and I would be swinging randomly. Randall seems unable to comprehend that people can actually react and make judgments on that sort of thing, which is what makes him think that things like batting averages are actually random rather than skill based. Which, along with this comic, is complete and utter bullshit. A career batting average is a pretty reliable indicator of how good a player is at hitting a baseball. Sports commentary based on a player being due, or hot or cold is silly if you assume a series of discrete, random events, but while the events are certainly discrete (at least in baseball), they sure as hell ain't random in any way.

Unless you're defining random number to include any number ever. But I suspect this is one of those cases of Randall trying to make fun of something that he thinks he knows much more about then he does, and looking like an idiot as a result.

cptjeff
Posts: 44
Joined: Thu Sep 03, 2009 4:42 am UTC

Re: 904: "Sports"

Postby cptjeff » Sun May 29, 2011 6:41 am UTC

mani.cavo wrote:
suzi wrote:If you only watch casually you might think that passes/plays/shots/goals/fouls are "random," but I find that a really belittling and so incredibly tedious to find truth in.
I'll note that you didn't say you find it to be incorrect, just belittling.

The word "random" has too much baggage and I think that's what's tripping you up here. Saying that a system is a "weighted random number generator" is not saying that there is no impact of skill, no trends, no patterns, etc. Those are all the factors/results of the weighting. Unless you think players and referees can control the weather, and everything else that might impact their performance in or outside of the game, then you need to acknowledge the element of randomness.

Yes, you're right that there's an element of randomness everywhere in life. That should be how you know he's not belittling the sports themselves: because it's true of everything. But we don't necessarily try to derive flavorful dramatic narratives from everything all the time, as though these were predictive explanations.

To give you an example: When I filled in scantron tests as a kid, I would imagine that "A" and "B" answers were a villain, and "C" and "D" answers were heroes trying to fight it. Based on how the answers wound up, I had choreographed entire dramatic battles in my head, and even had a very strict and consistent set of rules for interpreting the answers.

But if you asked me how my test went, and I said "There were a lot of reversals, and the heroes held their ground initially and tried to break the villain apart with strong counters, but in the end he was just too much for them and they ultimately fell under the barrage before being eliminated by a supercharged coup de grace!" you would correctly note that this
A. Is, although based on the experience of taking the test, pretty divorced from reality as a conclusion about the test
B. Is not predictive at all
C. Has nothing to do with the actual test (or design, including potential beauty, of the test), and most importantly
D. Doesn't give you an idea for what actually resulted (i.e. my grade) at the end of the day, because instead of focusing on the actual weights I'm just spinning a pretty thread from the events as I see fit

Basically, I'm really sick of otherwise smart, analytical, and thoughtful people dismissing sports based on what little contact they have with them in the mainstream media.
Nothing in the comic is dismissing sports; it's dismissing sports commentary in the mainstream media, specifically.


The random elements are not a significant element in the system of a baseball game. Yes, they're there, but would you describe driving your car as a random event? Hopefully, you don't just think that your job is to point it and hope for the best. No, you're supposed to use skill to drive it. If everyday random systems are affecting your performance, then I sure as hell don't want you on the road.

Same with a baseball game. Both the pitcher and hitter are highly skilled, and the random elements like weather might affect how they approach something, but it's still their skill that is the driving factor. A batting average is fairly predictive over a large sample. So is an on base percentage, becuase some hitters are better at fouling off balls and making at bats longer, thus increasing the chances a pitcher will screw up and walk them.

Yeah, there's still randomness. Happens in science too. In Science, nothing is 100%. But once something hits a certain point, you can disregard the butterfly flapping its wings in right field when calculating when that sinker is gonna break, because it's not gonna have a measurable effect on the outcome.

Badly Shaved Monkey
Posts: 4
Joined: Sat Oct 09, 2010 6:59 am UTC

Re: 904: "Sports"

Postby Badly Shaved Monkey » Sun May 29, 2011 9:03 am UTC

madock345 wrote:
CorruptUser wrote:I know it's not correct, but was correcting a different problem; single die rolls have no mode, assuming a fair die of more than 1 side.  No die is truly perfect, but close enough that you can assume it to be.

A die of more than one side? I don't think they make dice in the one sided variety, but that might just be me.


I've been trying to work out how you'd throw a Möbius Strip....

For me this strip scratches an itch I feel every time I watch sport. As has been pointed out there is an overwhelming desire for sports fans and commentators to draw general conclusions from insufficient data. This especially affects cricket, and particularly test cricket, where we are asked to consider whether a particular team is the "best" of the current world test sides where I cannot believe there are ever enough data to draw a valid conclusion. The teams meet each other too rarely and the performances lurch erratically. Furthermore, because these infrequent meetings are spread out over numbers of years that comprise a significant fraction of any player's career the idea of making a statement about the teams' rankings at a given instant is even more problematic. 

Contrast this, say, with football leagues where the trial of one team versus another is repeated several times in a season so the actual outcome at the end of a season is closer to a 'fair test' of a team's actual merits. This leads to another of my unscratchable itches. Football makes sense as a league game but no sense as a knockout tournament game. Ask yourself in what sense is it reasonable to conclude that Holland is the Best national football team in the world? 

Finally, reading this thread, several posters have focused on what we might call the in-game statistics, batting averages, yards run etc etc. What I am highlighting  is the contrast between these statistics where the reported result is the result of hundreds or thousands of trials versus the the big outcomes that really matter- national championships where the number of trials is too small to give a reliable result. There is an unresolveable contradiction here. 

Badly Shaved Monkey
Posts: 4
Joined: Sat Oct 09, 2010 6:59 am UTC

Re: 904: "Sports"

Postby Badly Shaved Monkey » Sun May 29, 2011 9:19 am UTC

To enlarge upon an already lengthy argument. What underlies this whole discussion is a contrast between two modes of thinking. 

Sports competitions are effectively posing a question, offering an hypothesis: which is the best team/player? This is testable in principle using the methods of data collection and statistical testing.  But, enjoyment of sport can occur independently of the proving the hypothesis. A well-played drop-shot or an exciting to-and-fro rally can be things of beauty to enjoy for their own sakes as aesthetic objects.

I lack much of that aesthetic feeling for sport and merely have a slight interest in knowing whether Federer is the best grass-court tennis player or Barcelona is the world's best club football team. I see the lack of a proper answer to those questions as simply irritating rather than intriguing. 

The proliferation of in-game statistics so noticeable in American sports seems to occupy a space between these two approaches. The numbers are used sometimes to prove (or pretend to prove) hypotheses and other times merely to decorate an essentially aesthetic argument about how a player or team is performing. 

User avatar
Jorpho
Posts: 6290
Joined: Wed Dec 12, 2007 5:31 am UTC
Location: Canada

Re: 904: "Sports"

Postby Jorpho » Sun May 29, 2011 1:07 pm UTC

cptjeff wrote:A career batting average is a pretty reliable indicator of how good a player is at hitting a baseball. Sports commentary based on a player being due, or hot or cold is silly if you assume a series of discrete, random events, but while the events are certainly discrete (at least in baseball), they sure as hell ain't random in any way.
You're contradicting yourself here. "Not random in any way"? So everyone should be able to predict from moment to moment with absolute certainty exactly what is going to happen next? And therefore those who claim to be able to do so are worthy of reverence?

teucer
Posts: 42
Joined: Wed Aug 08, 2007 4:28 pm UTC

Re: 904: "Sports"

Postby teucer » Sun May 29, 2011 2:37 pm UTC

Chrisfs wrote:
from canada wrote:Image


The big difference is that science has on a number of occasions been able to make sturdy predictive models, and has structures in place to go back and check themselves for accuracy, whereas sports commentators or people chatting idlely in sports bars have not.


I'm a soccer fan. There's a match-predicting game for the league I follow (NASL), in which players pick the final score of each game. You get one point for each team whose number of goals you get exactly right, two points if you get the match outcome right, and no partial credit on either of those. Each match is worth up to four points.

I have a spreadsheet that implements a model wherein I assume that when team A plays team B that team A's number of goals obeys a poisson distribution with lambda equal to ((A's goals scored + B's goals allowed) / (A's games played + B's games played)). The spreadsheet then calculates the probability of every possible result that includes teams scoring ten or fewer goals (because frankly even half that is pretty absurd in soccer) and figures out the expected number of points I would earn by picking that result. Then I pick the highest-scoring prediction, save that I will never pick my own team to lose or to merely draw at home (which is working pretty well for me since they're currently the best in the league).

Right now I am eight points behind the best player in this game, and that only because there wasn't enough data to make meaningful predictions for the very early part of the season. For the month of May, it's only three points' difference. I'd say that's a sturdyish predictive model, and as we get into June's games I expect it to only get better.

At the less flimsy end, the various baseball statisticians actually do have a pretty good track record at predicting the results over the course of a season. Not on a per-game level (it's too noisy), but they come pretty close on what a given team's overall record will be. Certainly they do better than my soccer spreadsheet.

But then you get things like the "hot hand" phenomenon that gets discussed in every statistics class ever: a good basketball player makes x% of his shots; what are the odds that he'll have an impressive streak like the one he's on? Turns out they're pretty good; the streak is merely indicative of the fact that he's good enough to have that kind of result by chance, rather than because he's especially "hot" at the moment. Sports commentary loves the latter interpretation, though, because streaks feel impressive even if the numbers say they're to be expected. That's making up a narrative based on the output of a weighted RNG, and it kind of sucks. I'd rather the commentators admit they know those things are bogus. The fact that Etienne Barbara's streak of scoring at least one goal in the first eight games of the season (to return to my soccer team) ended last night doesn't mean he isn't still objectively the best striker in the league - it just means that he isn't God, which we already knew. Streaks happen, streaks end, and it would be a mistake to look for a reason why it ended in this game in particular other than luck - but I guarantee I'll be seeing coverage this morning that asks what changed.

To me, it's much more interesting to look at the output of a weighted RNG, admit that's what you're seeing, and ask yourself, "How is it weighted?" How good is this team, this player, to produce these results? What makes them good - not why did they perform so well tonight (because that is almost entirely down to chance), but what makes the probability of it happening to them higher than to that guy? There's a lot of meat there to spin far truer narratives out of and to analyze, but so much sports commentary tries to keep the drama of a single game alive past its end rather than looking at the often grander drama of a full season or a full career.

User avatar
mani.cavo
Posts: 5
Joined: Fri May 27, 2011 12:24 pm UTC

Re: 904: "Sports"

Postby mani.cavo » Sun May 29, 2011 5:34 pm UTC

cptjeff wrote:The random elements are not a significant element in the system of a baseball game.
Arguably. Hence, "weighted."

Yes, they're there, but would you describe driving your car as a random event?
I wouldn't, but that's largely because of the connotative baggage that the term "random" has, which i mentioned before; not because it's inaccurate.

Hopefully, you don't just think that your job is to point it and hope for the best.
Sure it is. (I don't drive, but I'm running with your analogy regardless.) We tacitly accept that there are many factors we can't control that could affect anything we do, make decisions based on their likelihood, control what we can, and - tacitly if not consciously - "hope for the best." You look both ways when crossing on the assumption and hope that people don't come banking around the curve like maniacs and hit you after you looked; you don't avoid crossing the street altogether.

If everyday random systems are affecting your performance, then I sure as hell don't want you on the road.
Pretty sure the weather - like rain - affects anyone's driving performance. And I'm pretty sure it'd be more dangerous to drive exactly the same way in every weather condition than it would be to adjust your performance based on what weather conditions do arise. Again: "Weighted." I said right up front that calling something "random" doesn't necessarily mean that there is no room or place for things like "skill" - and even if it did, we've been talking about weighted randomness since the beginning.

Yeah, there's still randomness. Happens in science too. In Science, nothing is 100%. But once something hits a certain point, you can disregard the butterfly flapping its wings in right field when calculating when that sinker is gonna break, because it's not gonna have a measurable effect on the outcome.
Agreed 100%.

But in the example you described, one acknowledges and understands the degree of influence of various randomized components in the system, and based on that understanding of the probabilities involved makes an informed assessment of what is and isn't negligible in what context.

Which is not what the majority of sports commentators do; instead they focus on whatever suits their narrative and/or seems obvious or significant to them, and they run with it weaving the conjecture into a story. There's a reason why Randall wrote "narrative" instead of "trend"; the two are not synonymous.
Our world, with its rules of causality, has trained us to be miserly with forgiveness.

Inkstain
Posts: 11
Joined: Fri Dec 05, 2008 8:52 pm UTC

Re: 904: "Sports"

Postby Inkstain » Sun May 29, 2011 7:51 pm UTC

This comic really hits home because I am all of the following:

1) A professional sports reporter
2) A person with a layman's understanding of human cognitive biases
3) A person who enjoys arguing on the internet.

My fellow sports journalists are especially difficult to persuade. They are people whose very livelihood depends on the idea that they can watch sporting events and come up with meaningful stories. Trying to tell them that most of what they think they see is just variance is a very hard sell.

The most important cognitive bias at work, besides of course the tendency to see patterns in randomness, is: http://en.wikipedia.org/wiki/Fundamenta ... tion_error

We see what happens as other people to be a result of innate characteristics of that person, regardless of the influence of outside circumstances and variance.

The upshot of this is that when a really great basketball player happens to play with another really great basketball player and a few good ones, just at the same time as a few other key players in the league start to get worse from aging, that player is on six championship teams in eight years, the other two years being part of an ill-planned retirement.

Rather than saying "Wow, Michael Jordan is a great basketball player on some great teams that won a lot of championships," the narrative becomes "Michael Jordan is a winner," with the idea that he has innate qualities beyond his basketball skills that force the universe to align in such a way that his team's winning is the only possible result. As a corollary, players who are not as fortunate, but equally or nearly equally skilled, must not be Winners.

Inkstain
Posts: 11
Joined: Fri Dec 05, 2008 8:52 pm UTC

Re: 904: "Sports"

Postby Inkstain » Sun May 29, 2011 7:58 pm UTC

cptjeff wrote:Same with a baseball game. Both the pitcher and hitter are highly skilled, and the random elements like weather might affect how they approach something, but it's still their skill that is the driving factor. A batting average is fairly predictive over a large sample. So is an on base percentage, becuase some hitters are better at fouling off balls and making at bats longer, thus increasing the chances a pitcher will screw up and walk them.


1) I guess "fairly predictive" is a relative term, but there's a ton of variance in batting average. The standard deviation on individual batting average for an entire season is about 25 points. Considering the difference in perception between a .250 season and a .300 season, when both are equally likely from a .275 hitter, that's huge. But even more interesting to this point than batting average is Batting Average on Balls in Play, that is batting average excluding strikeouts and home runs. The standard deviation on BABIP is *huge* from season to season and is the engine that drives all the variance in batting average. It's very common for a batter's BABIP to vary wildly from season to season, quite simply because he's either lucky that the hits are falling in or unlikely that they aren't.

2) OBP is highly predictive, but you are misunderstanding the skill set involved. The correlation between frequency of foul balls and walks is actually negative. Batters who foul off a lot of pitches are less likely to be taking a lot of walks. The skill set that leads to walks is simply not swinging. Great offensive players are extremely discerning about which pitches to swing at (the ones they can hit hard) and which ones not to (the pitches that may be borderline strikes, but aren't pitchers pitches). The guys who are fouling off a lot of pitches are the ones who are swinging at lots of pitches, knowing that even if they don't hit it hard they might still get a hit. As an aggregate, these batters are generally less valuable than the more selective ones, but there are of course exceptions.

User avatar
Eternal Density
Posts: 5590
Joined: Thu Oct 02, 2008 12:37 am UTC
Contact:

Re: 904: "Sports"

Postby Eternal Density » Mon May 30, 2011 1:20 am UTC

vodka.cobra wrote:And I've never heard commentary on the events in an ongoing D&D campaign. :\
No, the D&D campaign itself is the narrative built from random numbers. Or so I heard.
Play the game of Time! castle.chirpingmustard.com Hotdog Vending Supplier But what is this?
In the Marvel vs. DC film-making war, we're all winners.

User avatar
Jorpho
Posts: 6290
Joined: Wed Dec 12, 2007 5:31 am UTC
Location: Canada

Re: 0904: "Sports"

Postby Jorpho » Mon May 30, 2011 1:00 pm UTC

Really now, a dungeon master who shows up with no material and expects random dice rolls alone to completely plan out a session for him is going to be a rather boring dungeon master indeed.

User avatar
AvatarIII
Posts: 2098
Joined: Fri Apr 08, 2011 12:28 pm UTC
Location: W.Sussex, UK

Re: 904: "Sports"

Postby AvatarIII » Mon May 30, 2011 1:35 pm UTC

monty30 wrote:
CorruptUser wrote:
Brooks Hatlen wrote:
glasnt wrote:Since when is D&D weighted?!

As in any game with multiple dice, there are numbers more likely to come up than others, namely half the maximum total roll.


Only when there is more than 1 die rolled. Otherwise it's (almost) a uniform distribution.


It isn't half of the maximum total roll, although that could be a rough approximation.
For example, the most common sum from rolling two dice is 7, and not (6+6)/2 = 6.

Edit: assuming six sided dice


to make it make sense for any size and number of dice, i guess it would always be
n = number of dice
s = sides of dice

(sn/n)+n-1 = statisically most common roll

right?

roydesoto
Posts: 3
Joined: Mon May 31, 2010 11:03 pm UTC

Re: 0904: "Sports"

Postby roydesoto » Mon May 30, 2011 1:42 pm UTC

Some of the other posters have used terms that suggest that they are aware of this important sports/statistics reference, but no on has made it explicit.

Stephen Jay Gould discussed the phenomenon of "hot hands" and randomness in an article:

http://www.nybooks.com/articles/archives/1988/aug/18/the-streak-of-streaks/.

The upshot is that Joe DiMaggio's 1941 hitting streak is indeed an outlier, and beyond that predicted by chance. Of course, the proof offered is elusive, and smells vaguely of "...but this margin is too small to write it all here."

Nonetheless, just wanted to make this element of the discussion explicit.

User avatar
Eternal Density
Posts: 5590
Joined: Thu Oct 02, 2008 12:37 am UTC
Contact:

Re: 0904: "Sports"

Postby Eternal Density » Tue May 31, 2011 5:00 am UTC

A very good example of randomly generated narratives is Dwarf Fortress.

Examples: http://dfstories.com/
Play the game of Time! castle.chirpingmustard.com Hotdog Vending Supplier But what is this?
In the Marvel vs. DC film-making war, we're all winners.

User avatar
vodka.cobra
Posts: 371
Joined: Thu Mar 27, 2008 6:50 pm UTC
Location: Florida
Contact:

Re: 904: "Sports"

Postby vodka.cobra » Tue May 31, 2011 5:55 am UTC

Eternal Density wrote:
vodka.cobra wrote:And I've never heard commentary on the events in an ongoing D&D campaign. :\
No, the D&D campaign itself is the narrative built from random numbers. Or so I heard.

I wish it worked that way. The people in charge of my D&D campaigns wrote the story beforehand and if anything went wrong they hated you for fucking up their story.
If the above comment has anything to do with hacking or cryptography, note that I work for a PHP security company and might know what I'm talking about.

User avatar
mani.cavo
Posts: 5
Joined: Fri May 27, 2011 12:24 pm UTC

Re: 904: "Sports"

Postby mani.cavo » Tue May 31, 2011 10:26 am UTC

vodka.cobra wrote:
Eternal Density wrote:
vodka.cobra wrote:And I've never heard commentary on the events in an ongoing D&D campaign. :\
No, the D&D campaign itself is the narrative built from random numbers. Or so I heard.

I wish it worked that way. The people in charge of my D&D campaigns wrote the story beforehand and if anything went wrong they hated you for fucking up their story.
Sounds like some DMs need to stop railroading their players. (Some systems are complex enough that it's harder to just wing something, though. I once GM'd a whole session of Don't Rest Your Head off improvisation.)
Our world, with its rules of causality, has trained us to be miserly with forgiveness.

Gamer_2k4
Posts: 34
Joined: Sat Feb 13, 2010 10:09 am UTC

Re: 0904: "Sports"

Postby Gamer_2k4 » Tue May 31, 2011 4:53 pm UTC

goomhr randall all of existence is just narration and interpretation of weighted random chance too

User avatar
vodka.cobra
Posts: 371
Joined: Thu Mar 27, 2008 6:50 pm UTC
Location: Florida
Contact:

Re: 904: "Sports"

Postby vodka.cobra » Tue May 31, 2011 6:30 pm UTC

mani.cavo wrote:
vodka.cobra wrote:
Eternal Density wrote:
vodka.cobra wrote:And I've never heard commentary on the events in an ongoing D&D campaign. :\
No, the D&D campaign itself is the narrative built from random numbers. Or so I heard.

I wish it worked that way. The people in charge of my D&D campaigns wrote the story beforehand and if anything went wrong they hated you for fucking up their story.
Sounds like some DMs need to stop railroading their players. (Some systems are complex enough that it's harder to just wing something, though. I once GM'd a whole session of Don't Rest Your Head off improvisation.)

Yeah it was kinda bullshit in retrospect. It also kinda ruined D&D for me, but I'm still making a text-based game built on D&D's systems... go figure.
If the above comment has anything to do with hacking or cryptography, note that I work for a PHP security company and might know what I'm talking about.


Return to “Individual XKCD Comic Threads”

Who is online

Users browsing this forum: AluisioASG, Majestic-12 [Bot] and 103 guests