Normalization: A Humble Request to Strat-O-Matic

For quote, I just hit the quote button on your reply.

AHA. I always used the "quote" button above the input box (across from the smiley faces ).
I've seen it a million times but didn't realize there was a RED quote button on the right.
My pattern recognition really sucks.

......

...

This topic has been debated endlessly over the years. I do remember a thread a few years back that said that the game wasn't normalized, but the cards were to adjust for the league conditions for the season on the player card. Also, apparently there was a difference in cards if there was a season set, or a card created for ATG exclusively. ATG has introduced some crazy super-charged cards in the past (Arky Vaughan, Eddie Matthews) and these were removed after an uproar (if anyone remembers those, please refresh my memory on the issues) so there has been some tinkering by Strat in past that altered the results. That Arky card was insane! My take is, the same people seem to win all the time, so if there was a normalization feature, they would not win as often as their team construction would be compromised in some way given reasonable competitive leagues and reasonable random variations.

I remember this guy, and I remember this same discussion going on and on and on over ten years ago. Once a conspiracy theorist, always a conspiracy theorist, I guess.

It's really this simple: It's a bell curve. "Normalization" is something that happens when you get the thousands and thousands of die rolls that you get over the course of a 162 game season. There's only so much standard deviation that you'll see after a sample size that big, so when Eddie Mathews hits 40 home runs through the first half of the season, that pace isn't going to keep up through the end of the season. That's not normalization. That's not a conspiracy theory. That's a return to the statistical mean, and it happens every time. It happens in Strat. It happens in craps. It happens in Yahtzee. Every. Time.

I remember this guy, and I remember this same discussion going on and on and on over ten years ago. Once a conspiracy theorist, always a conspiracy theorist, I guess.

It's really this simple: It's a bell curve. "Normalization" is something that happens when you get the thousands and thousands of die rolls that you get over the course of a 162 game season. There's only so much standard deviation that you'll see after a sample size that big, so when Eddie Mathews hits 40 home runs through the first half of the season, that pace isn't going to keep up through the end of the season. That's not normalization. That's not a conspiracy theory. That's a return to the statistical mean, and it happens every time. It happens in Strat. It happens in craps. It happens in Yahtzee. Every. Time.

Not that I think there is normalization, but if Eddie hits 40 home runs in a first half of a season, as often as that happens in the first half of the season, there should be a bell curve of results for the second half of the season, along with a possiblity of 40 HR's in the second half of the season. If EVERY time that pace does not continue then that is the very definition of normalization.

I remember this guy, and I remember this same discussion going on and on and on over ten years ago. Once a conspiracy theorist, always a conspiracy theorist, I guess.

It's really this simple: It's a bell curve. "Normalization" is something that happens when you get the thousands and thousands of die rolls that you get over the course of a 162 game season. There's only so much standard deviation that you'll see after a sample size that big, so when Eddie Mathews hits 40 home runs through the first half of the season, that pace isn't going to keep up through the end of the season. That's not normalization. That's not a conspiracy theory. That's a return to the statistical mean, and it happens every time. It happens in Strat. It happens in craps. It happens in Yahtzee. Every. Time.

I mean, I could say "Once a sucker, always a sucker, and apply that logic to you." Don't forget that you have absolutely zero proof of the legitimacy of the code. You're also displaying a lack of knowledge of probability with your example. It is quite possible for a card to get "lucky" and hit 40 home runs and then get even better results during the second half of the season. You're also speaking as though a strat season is an infinite sample where everything should balance out, as shown by your repeated usage of the phrase "Every time." If that were the case we would never see any anomalies over the course of a season. Did you know that after 1,000,000 coin flips it is not a guarantee to be even close to a 50/50 split? This has been tested and in some cases it is a 60/40 split or worse.

Lastly, you speak as though being a conspiracy theorist (in a single instance) is an inherently bad thing. It's called critical thinking. The ideal customer is someone that spends readily and doesn't ask questions. The company says "This is the truth." and you say "Great. Now I know the truth."

Did you know that after 1,000,000 coin flips it is not a guarantee to be even close to a 50/50 split? This has been tested and in some cases it is a 60/40 split or worse.

For 1,000,000 flips of a fair coin, the probability of 60% or more heads is virtually zero (I can't get a binomial calculator to give me a non-zero answer). The binom.dist.range function in Excel tells me the chance to get 51% or more heads is 2.77 x 10E-89

I'm pretty sure this whole thing would get wrapped up pretty quick if folks' idea of "critical thinking" included taking a statistics class.

I'm pretty sure this whole thing would get wrapped up pretty quick if folks' idea of "critical thinking" included taking a statistics class.

Cannot argue that.

Anyway, the reason the chance will never go to zero for that coin flip is because the function is asymptotic...approaches but never quite reaches it.

This is in fact the reason we use quantum mechanics to understand modern physics of particle systems.

One of the tricks we used to have to derive in statistical physics applications when first introduced to quantum mechanics is the equation for a tennis ball to go through a wall. That is, classical physics will show it never can happen. Quantum mechanics approach yields a non-zero but tiny chance that the ball will slip right through. Of course, we know we would never be able to get it to happen in a nanillion lifetimes.

Why?
One way to think of it is because classical physics is quantum mechanics as you progress the functional limit of the independent variable(s) toward infinity. Or--that it is a limit of our math systems such that using quantum mechanics to solve a bulk problem that is sufficiently at equilibrium is not correct.

So the stats will never show zero--but in actuality--the chance of being 60-40 after that many flips is essentially zero because we will never witness it happen.

Wow you guys are gettin down and dirty!