- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Season Record Prediction Using Mathematical Analysis of Point Spreads 2015
Posted on 8/21/15 at 12:31 pm to moneyg
Posted on 8/21/15 at 12:31 pm to moneyg
quote:I tend to agree with you. The discussion about the underlying reason for the correlation is somewhat superfluous.
That's a whole lot of discussion that ignores the only thing that matters...is your chart based on actual data or not?
However, as I'm sure you know, The Rant is famous for attacking the minutest of minutia. I explained the underlying reason for the correlation because, as minor as it was, the underlying reason for the correlation was attacked by a poster.
Posted on 8/21/15 at 12:33 pm to Salviati
quote:
Dear Guava Jelly,
The analysis in this thread appears to be over your head.
You aren't even handling the probabilities correctly, and the analysis is over other people's heads?
Posted on 8/21/15 at 12:34 pm to Guava Jelly
He never said the games weren't independent events. Do you NOT think that having a 99 percent chance to win a game versus a 51 percent chance to win a game matters in terms of most likely win total? You're falling victim to the fallacy he described with the 52% coin where you assign each flip more likely to be heads as a head, so you predict all heads. That's not how expected values work.
You're correct to say 'There are ten games that LSU is probably going to win this year' but that doesn't mean LSU is probably going to win 10 games this year. Depending on how close to 50% those win percentages are, the expected value could be a few games lower. Now the variance around that mean value would change with the probabilities in the series, with a bunch of 99s and 1s virtually guaranteeing a certain record while a series of probabilities with several close to 50% would have a wider range that is reasonably likely. But the expected value, or most likely win total, is not simply counting the number of games in which we are favored, as illustrated in the coin example.
You're correct to say 'There are ten games that LSU is probably going to win this year' but that doesn't mean LSU is probably going to win 10 games this year. Depending on how close to 50% those win percentages are, the expected value could be a few games lower. Now the variance around that mean value would change with the probabilities in the series, with a bunch of 99s and 1s virtually guaranteeing a certain record while a series of probabilities with several close to 50% would have a wider range that is reasonably likely. But the expected value, or most likely win total, is not simply counting the number of games in which we are favored, as illustrated in the coin example.
This post was edited on 8/21/15 at 12:35 pm
Posted on 8/21/15 at 12:35 pm to Salviati
quote:
If you take a bent coin that has a 52% chance of landing on heads and flip it 100 times, your analysis would count all flips as a predictive "win" individually. Thus, your analysis would predict the coin landing on heads 100 times.
But I'm only predicting one flip. Comparatively, each additional coin toss would require a different coin with varying degrees of bend.
Posted on 8/21/15 at 12:35 pm to Guava Jelly
quote:
I'll say it again. You can't view probability as an aggregate.
8.6/12 = a 71.67% probability that we win an individual contest.
You don't even need to divide by 12.
P(AorB) = P(A) + P(B)
That is for mutually exclusive events. These aren't mutually exclusive.
Posted on 8/21/15 at 12:37 pm to jonboy
quote:
Comparatively, each additional coin toss would require a different coin with varying degrees of bend.
And those coins with varying degrees of bend are truly independent and mutually exclusive.
Posted on 8/21/15 at 12:38 pm to Wayne Campbell
quote:A preseason prediction based on betting lines is useful/useless to the EXACT same extent as any preseason prediction.quote:Doesn't that basically render a pre-season prediction based on betting lines useless? LSU will be 9-3 based off of preseason betting lines, but the lines will shift, sometimes drastically, after each game.
Thus, a prediction based upon betting lines would also change over time viewed from the day of the prediction.
Betting lines for subsequent games will shift, sometimes drastically, after each game.
However, in EXACTLY the same way, EVERYONE's predictions for the remainder of the season will change with the caveat that some people will change their predictions more intelligently than others.
EXAMPLE: If LSU wins its first three games convincingly, the results of those three games will cause betting lines to shift, possible drastically; however, the results of those three games will cause a similar shift in the predictions of all people, possible drastically.
Posted on 8/21/15 at 12:42 pm to Salviati
Ok, buddy. You're right. Even though as it sits (according to your odds) LSU has a greater probability of winning 10 of the games (when viewed individually), it's more likely that they'll win 8.
I'm out.
I'm out.
Posted on 8/21/15 at 12:47 pm to Salviati
I'm no statistician, but the way you did it just defies common sense, IMO.
By this rationale, a team who is favored 90/10 in each game they play should post a 11-1 record... which just seems odd to me, similar to the example I used earlier with FLA, SCAR, CUSE, and ARKY.
Even if statistically true, it feels wrong.
By this rationale, a team who is favored 90/10 in each game they play should post a 11-1 record... which just seems odd to me, similar to the example I used earlier with FLA, SCAR, CUSE, and ARKY.
Even if statistically true, it feels wrong.
Posted on 8/21/15 at 12:47 pm to Guava Jelly
quote::smh:
Saying "smh" to strong arguments is not a refutation.
PLEASE provide YOUR proposed analysis.
quote:You are adding an additional layer of complexity which is wholly unnecessary. I see games as independent events when viewed from the time that the prediction is made. Each future football game is EXACTLY like a coin flip at the present time. Each game is an independent event. When Vegas establishes betting lines, Vegas views each game as an independent event.
You keep using the "bent penny" analogy, but earlier you tried to say that you don't see games as individual events... which is what the coin flip analogy is meant to illustrate.
quote:I know what an average is. However, I don't add all of the odds together and then divide. You really need to look at what I did and the con flip analogy more closely.
However, when you add all of the odds of success together, then divide (read, "find the average") your likelihood of success is pulled down by an outlier.
quote::smh:
By totaling the odds, you are the one who is trying to view the season as a whole. I am viewing each individual game, as it well should be.
Your refusal to examine this issue more closely is causing me to :smh:.
I AM providing a prediction for "the season as a whole"; however, I AM viewing each game as an individual game.
Again, PLEASE provide YOUR proposed analysis.
Posted on 8/21/15 at 12:51 pm to Salviati
quote:
. I see games as independent events when viewed from the time that the prediction is made
Even assuming this is true... It isn't. Your conclusion is still wrong and shows a fundamental misunderstanding of what you calculated with those probabilities.
Posted on 8/21/15 at 12:53 pm to CptBengal
quote::smh:
The games are NOT mutually exclusive.
You need to argue with Guava Jelly.
The games ARE independent events. The odds were established by viewing the games as independent events.
quote:Yes, I did.
So basically, you summed the games.
quote:Yes, I did.
Which you then conclude means we will win 8.6 games.
quote:No. I did NOT. You need to take up this argument with Guava Jelly.
However that is not true, what your analysis did say is that there is a probability of .86 that we win any single game in your series.
quote:Yes, they are. You need to take up this argument with Guava Jelly.
That isn't true, because the events are NOT mutually exclusive.
quote:I suggest you STOP bullshitting in this thread.
I suggest you look up non-mutually exclusive and conditional probability and try again.
Show EVERYONE how SMART you are, and how DUMB I am.
PLEASE provide YOUR proposed analysis.
Posted on 8/21/15 at 12:56 pm to Salviati
Listen, moron.
This is what you did :
P(A) + P(B)...
The solution to that is the P(A or B)
Which means that the events are mutually exclusive. They aren't. Your math is wrong.
This is what you did :
P(A) + P(B)...
The solution to that is the P(A or B)
Which means that the events are mutually exclusive. They aren't. Your math is wrong.
Posted on 8/21/15 at 12:59 pm to Salviati
Since we've come to an impasse, thos is my last post in this thread.
I provided my proposed analysis. I look at the probability of winning vs one team. Make a judgment based on that probability. Then move on to the next game.
This method yeilds 10 wins since LSU's odds of winning in this scenario are above 50% in 10 games individually.
Given that it's a pointless hypothetical and given that the season changes week to week, it merits no further examination.
In this example, the odds of season success are tied to the odds of individual game success. Since the odds of winning vs McNeese have no bearing on the odds of winning vs Bama, each of these games should be viewed as independent events, not part of a single whole.
I provided my proposed analysis. I look at the probability of winning vs one team. Make a judgment based on that probability. Then move on to the next game.
This method yeilds 10 wins since LSU's odds of winning in this scenario are above 50% in 10 games individually.
Given that it's a pointless hypothetical and given that the season changes week to week, it merits no further examination.
In this example, the odds of season success are tied to the odds of individual game success. Since the odds of winning vs McNeese have no bearing on the odds of winning vs Bama, each of these games should be viewed as independent events, not part of a single whole.
Posted on 8/21/15 at 12:59 pm to Salviati
My summary:
"go to hell Ole Miss." LSU wins that one, they are in the playoff along with 2 seed Bama, THE Ohio State and the TCU horny toads.
LSU beats THE Ohio State and faces Bama a 2nd time, coming away with a win, avenging the 21-0 debacle.
"go to hell Ole Miss." LSU wins that one, they are in the playoff along with 2 seed Bama, THE Ohio State and the TCU horny toads.
LSU beats THE Ohio State and faces Bama a 2nd time, coming away with a win, avenging the 21-0 debacle.
Posted on 8/21/15 at 1:00 pm to Guava Jelly
quote:PLEASE READ THIS POST: MY ANALYSIS TREATS THE GAMES AS INDEPENDENT EVENTS.
His problem is that he's trying to say he both believes games are not independent events (which in this example I believe they should be) while also using the "coin flip" analogy.
quote:Yes, that is correct. The odds demonstrate that LSU will likely win 8.6 out of 12 games. Of course, LSU cannot win .6 of a game, so the numbers must be rounded. Therefore, LSU the odds suggest that LSU is likely to win 9 out of 12 games.
In this example, he's saying that we win 8.6/12 games.
quote:There you go again. YOU are AVERAGING the odds. I am NOT averaging the odds.
Which means that there is a 71.6% chance we win an individual game.
quote:Okay, so you started posting in this thread with ignorance.
This is patently false because his own numbers bear out that our probability of success vs Bama is far lower, and vs McNeese far higher.
Thus you can't view the odds as an aggregate.
Then you figured out you wrong.
Now you're just spewing silliness in hopes that people with think you are trying to troll me.
Posted on 8/21/15 at 1:05 pm to Salviati
What a waste of time this was.
Posted on 8/21/15 at 1:06 pm to Salviati
I've not changed what I said from the beginning. I'm not trolling anyone.
I don't agree with your methodology. And I don't think you understand what it means to treat the games as independent events. If you did, you wouldn't have tried to add all the percentages up to divine a win total to begin with.
Call me "ignorant" if you like, pal.
Enjoy your hokey pseudo math. I'm done here.
This post was edited on 8/21/15 at 1:09 pm
Popular
Back to top
Follow TigerDroppings for LSU Football News