People Will Never Accept Advanced Statistics (Or: You Can't Win If You Don't Buy A Ticket)

by Jonathan Willis on Apr 2, 2009 11:22 AM MDT in Statistical Analysis

People as a whole will never accept advanced statistics because they are bad at math. The MSN website has an article up on it's main page right now talking about the odds of winning the lottery; here's the relevant portion:

Lotteries are often used as examples of innumeracy or financial illiteracy because it's common to misunderstand the probabilities on which the games are based...

We can digest the big numbers involved in the probabilities of winning the Lotto 6-49 in the very same way. The nice folks at the lottery assure us that the odds that any particular number will be drawn is one in 13,983,816. To state that same statistic in another way, any number may be expected to win 0.000007 per cent of the time.
So the next time you buy a 6-49 lottery ticket, you can be 99.999993 per cent confident that you will not win the big jackpot. Now, go ahead and compare that to your conviction about, well, pretty much anything else. (For instance, based on the most recent Transport Canada stats, on average you're more than four times more likely than that to die in a motor vehicle...today.)
Talk about buying with confidence! There may not be another product about which you can be so certain. You are virtually guaranteed not to win. In fact, you may as well just tear up your lottery ticket right in front of the clerk who sold it to you, and cavalierly toss the scraps in the trash.

And yet, people continue to buy lottery tickets.

Now, I understand that there are real and identifiable problems with advanced statistics right now. But the simple fact of the matter is that people as a whole will never accept them, even if they are proven to work 90% or 95% of the time. It's something that I need to remember, something that anyone interested in pushing this viewpoint needs to remember.

Keep in mind that I'm not attacking people who don't accept these numbers now - because they're a long ways from being proven. My point is that no matter how well grounded the argument eventually becomes, large portions of society will not accept it, because some people simply won't have the necessary faculties to process the information. It reflects more on them than on the argument itself.

I don't really care how arrogant that sounds because it is true. In any society where lotteries such as 6-49 can make tremendous profits, there are just too many people who aren't capable of comprehending basic mathematics.

Tweet Comment 17 comments | 0 recs |

Do you like this story?

Comments

Agreed. I assume you don't bother with lottery tickets then? ;)

On the flip side, every time I bug my uncle about throwing his money away on the 6-49 he replies, "Yes I know. All I'm doing is buying a little piece of a dream."

And he's right. Some people buy a $5 mochachiino-frappe-latte somethinghorother to make going to their crappy job more tolerable. Others buy a $2 dream.

by R-Gib on Apr 2, 2009 1:54 PM MDT reply actions

When it come to gambling I'm one step above all out degenerate but I won't buy lotto tickets because I can't control it at all. At least a sports game, blackjack, or poker has some level of control for the player but the lotto is completely random.

Lotteries like 6/49 are the as VLTs, they're a tax on stupidity. You can't tax stupid people directly so you give them a way to allow them to give you their money.

by ryanbatty on Apr 2, 2009 2:53 PM MDT reply actions

I have absolutely no problem with:

1) admitting advanced stats are over my head... I have no idea what some of the abbreviations you guys toss around even reference and therefore often can't follow your discussions.

2) acknowledging that advance stats should play a role in the scouting departments of every NHL team. Teams should have their traditional scouts and then one math wizard who can crack the numbers as just another point of reference on a player's projection.

3) that as someone who watches a lot of junior hockey (but is not a paid scout) I fall more in LT's "seen him good" category.

I like the intangibles that don't show up in stats like: attitude when I talk to him, what a guy does in the pre-game skate, how he plays away from the puck, does he skate off the ice hard at the end of his shift, does he whine and complain to the refs or teammates, does he back check, etc.

4) I don't buy lottery tickets.

by Guy Flaming on Apr 2, 2009 3:24 PM MDT reply actions

Guy: Sounds good ;)

I'm not saying that everyone who won't accept advanced stats does so because they don't get the math; only that people who don't get the math won't accept the stats and there are a lot of them out there.

by Jonathan Willis on Apr 2, 2009 3:29 PM MDT reply actions

I doubt the majority of people who buy 6/49 tickets are scoffing at statisticians and discounting the odds of winning. Sure, they may not realize the ridiculous odds AGAINST winning, but there are other motivations:

- extremely low risk vs. extremely high reward scenario (albeit unlikely)
- as R-Gib said, spending $2 to hope against hope about a "better existence" than the person already has
- "you can't score if you don't shoot, and what's $2 anyways?"

That said, I do agree that most people out there won't readily embrace some of the stats thrown around on these blogs. Like you said, not everyone is a math whiz and it's a helluva lot easier grasping a "goal" and an "assist" and even "+/-" than trying to figure out the nebulous meanings of QC and Corsi etc.

Part of it is communication. Not everyone has a math/stats background, and being a science guy by training myself, we're often the worst at making ourselves understood. So if these stats could be explained or communicated in more "everyday language", there would be broader acceptance.

And otherwise, you're right - people will always gravitate more toward a hard/counting number than stats that measure trends. The end outcome is always easier to see, easier to measure. But the trends are extremely useful at helping to understand what drives the end outcomes, so hopefully you all keep up the good work and don't get discouraged by others who don't value the numbers.

A fan who doesn't care about the stats isn't any less of a fan than one who does (though personally I'd much rather see the rational explanation than the "saw him good" explanation). If you're actually a GM in the NHL, on the other hand, I can't see why you wouldn't want to use these kinds of measures...

(and pls don't confuse me with the other "William" who teed off on you in the comments of ON)

by William on Apr 2, 2009 3:48 PM MDT reply actions

I don't really get math, but I try. (there is a reason I study political philosophy not political statistics).

I don't know if the goal should be to get everyone to accept advanced stats, but instead try to develop them into something, those that choose to use them, can play with. No matter what, you can always shrug and complain in your head that others are stupid/ignorant.

I also don't buy lotto tickets: not out of ideological reasons, but mainly out of poverty (oh, plus I smoke).

by B.C.B. on Apr 2, 2009 3:57 PM MDT reply actions

Tag: arrogance

Ha.

But in my humble opinion there is more to the psychology behind probability and understanding of it. There's actually some interesting studies specifically on how people understand probability in situations of reward.

Other than that, I think the issue is that most fans just can't be bothered to look at underlying statistics which are convey mean on-ice performance.

by jon k on Apr 2, 2009 6:05 PM MDT reply actions

Please excuse my lack of proof-reading.

The point I was trying to make is that most fans watch the game, buy the jerseys, but just can't be bothered to think about the team beyond the boxcars. Stupidity may or may not enter into the equation but I don't think it's the critical factor.

by jon k on Apr 2, 2009 6:07 PM MDT reply actions

I don't buy lottery tickets, I'll gladly sell them to seniors. It's the only way to keep them from yelling at me that there's a single piece of toilet paper on the bathroom floor of the store.

I've taken two statistics courses in University over the past four years. This shit is easy when compared to the stuff they're teaching these days. I still for the life of me don't know what a multiple linear regression is, and how it's different than a single linear regression. Nor do I care now that I never have to take another math class in my life.

This stuff isn't advanced statistics. This is the ability to manipulate an Excel spreadsheet. And that is something I simply cannot do.

by doritogrande on Apr 2, 2009 6:07 PM MDT reply actions

Granted, the lottery is a bad bet. However, once the expected payout reaches a certain point, the story changes. Given the odds of winning are 1 out of every 13,983,916 and it costs two dollars to buy a ticket, whenever the jackpot is greater than $27,967,630 ($13,983,916 * 2) the expected payout of a lottery ticket becomes positive.

Specifically:
($27,967,630*(1/13,983,816)) + (-2*(1-(1/13,983,816)) = 0

Although, this presumes that you are the only individual with the winning numbers. While I don't know for sure, I would guess that as the jackpot increases, more people play the lottery, which decreases the chances that you hold the only winning ticket.

by Zachary on Apr 2, 2009 6:26 PM MDT reply actions

Agreed on lottery tickets.

Here's a point, though: I don't really like the implication that fans who discount "advanced" (by what measure?) stats are probably idiots. I don't think I'm particularly stupid, and I read and understand (and generally agree with) the more stat-heavy blogs - this, MC79, IOF, and so forth.

But here's the thing: watching hockey is a leisure activity to me; manipulating spreadsheets is tiresome drudgery. If someone else is willing to run the numbers, I'm interested, but not to the point that I'll spend hours analyzing things, as it feels too much like work. I suspect that you, Vic, etc. enjoy the stat end of things enough that it's a key part of your enjoyment of hockey. For me? Nah, not worth it the trouble, but that certainly doesn't mean that I can't understand the simple math involved, or that I think Schremp will ever have an NHL career.

If GMs don't pay attention to this stuff, though, they probably are idiots.

by Peeeete. on Apr 2, 2009 6:49 PM MDT reply actions

I hate the fact that I can't edit typos on blogger. Irritating.

One other point: many smart hockey fans don't post on the Oilogsphere; most smart hockey fans don't post on the fansites. Doesn't mean they're not out there.

by Peeeete. on Apr 2, 2009 6:57 PM MDT reply actions

And that I guess is another point; for a lot of fans, there's no real desire to go beyond watching the game, and that's fine.

For the people who make a living at the game though, it's something that should at least be considered.

by Jonathan Willis on Apr 2, 2009 6:57 PM MDT reply actions

Yeah, exactly. I don't think teams should be scouting off spreadsheets or anything, but if these numbers aren't at least part of the discussion there's a problem. It's not 1955. "Saw him good" is fine and fun, for me, the casual fan, but I'm not getting paid $100K to help build a multi-million dollar business.

I would assume all 30 teams employ someone who can crunch numbers. Maybe not?

by Peeeete. on Apr 2, 2009 7:05 PM MDT reply actions

I don't think teams should be scouting off spreadsheets or anything, but if these numbers aren't at least part of the discussion there's a problem.

Absolutely, and I think we can safely say that there are a bunch of teams who don't use the numbers at all, based on their moves.

I'll never advocate moving away from physical scouting, and character references and whatnot, but these numbers deserve a place in the discussion.

by Jonathan Willis on Apr 2, 2009 7:27 PM MDT reply actions

Zachary has it right. At some point, any progressive lottery becomes a +EV bet - when the jackpot grows large enough. That's when I play.

As for stats, I look at it as a parallel of the science versus faith argument. Just keep proving it and eventually you'll get all but the most hard-headed. Let them enjoy variance and commit emotional suicide over regression to the mean. In the long run, if the forumlas work, the proof is in the pudding.

by Coach pb9617 on Apr 2, 2009 11:23 PM MDT reply actions

The other thing is that's an economic question. The marginal value of a $1 to me is pretty low. The marginal value of $10,000,000.00 is pretty high.

by mc79hockey on Apr 3, 2009 11:35 AM MDT reply actions

Comments For This Post Are Closed

Northwest Standings

	GP	W	L	OTL	PT
Vancouver Nucks Misconduct	52	32	15	5	69
Minnesota Hockey Wilderness	52	25	19	8	58
Calgary Matchsticks and Gasoline	53	24	22	7	55
Colorado Mile High Hockey	54	26	25	3	55
Edmonton The Copper & Blue	53	21	27	5	47

(updated 2.7.2012 at 7:26 AM MST)

21 - 27 - 5

Streak: Lost 1

Oilers Schedule

Next 5 Games

@ Detroit Winging It In Motown	Wed 02/08	5:30 PM MST
@ Ottawa Silver Seven	Sat 02/11	12:00 PM MST
Toronto Pension Plan Puppets	Wed 02/15	8:00 PM MST
Colorado Mile High Hockey	Fri 02/17	7:30 PM MST
Vancouver Nucks Misconduct	Sun 02/19	7:00 PM MST

Last 5 Games

@ Toronto Pension Plan Puppets	Mon 02/06	L 3 - 6
Detroit Winging It In Motown	Sat 02/04	SO - W 5 - 4
Chicago Second City Hockey	Thu 02/02	W 8 - 4
Colorado Mile High Hockey	Tue 01/31	W 3 - 2
@ Vancouver Nucks Misconduct	Tue 01/24	SO - L 2 - 3