2014 Off Topic Thread

[poopoo333]

What is 2+2?

[trotterz]

Thanks a ton. That actually helps a lot.

trotterz, you of course right, but it is more depth than I want to go into right now, that's why I asked to assume the same number of cases for each judge. The reality is slightly more complicated, but I can figure it out from here...

Thanks again guys, youre the best... I am really jealous of people that can do math. It makes me mad, because I feel like I should be able to do something like that and after you express the problem appropriately, the answer becomes clear to me. But I just can't order my thoughts in mathematically, even though I often imagine I am an expert logician (I even taught formal logic classes at a University). I realize how limited I am in this way when the problem is expressed correctly and see how obvious it should have been to me.

I actually do love it. I love playing with numbers and it is actualy my full tme job to do analysis and reports all day. and I am making out $82,000/year doign it so numbers are good to me

[Luke]

I do. Great series, that really took a crazy turn last episode.

I'm not sure if I'm up to date or not . Are you talking about the episode where they start suggesting McConaughey is the killer?

[Mr. IWS]

Anyonr watch true detective? Hell of a show. Ive been watching it for some time now. There are eight episodes total and its on hbo.

I do. Great series, that really took a crazy turn last episode.

[Svino]

I understand your rationale Svino but only if you are stricly looking to know if a lawyer is better than the other ones. I would still take into consideration the amount of Cases won in front of each judges .... simply for the reporting purposes. It is totally possible that the lawyer with the 54% rate won 95% of his cases in front of judge 1 but but only won 10% of his cases in front of judge #2.

Yeah, this could definitely make a difference. But controlling for another factor like that, while doable, is kinda tricky. Also, you'd have to make certain assumptions about how the two factors interact, like that lawyers and judges can be modeled with a single parameter for persuasiveness/leniency and they make a linear contribution to the probability of success. It could be super hard to model if there are quirks about specific lawyers matching up better or worse with different judges (like if one lawyer was primarily a jiu-jitsu guy, but one judge has really good take-down defense).

[Luke]

Nope. Wht name does he post under?

Shadow_preistX I think. And you can use my account if you want:

"Dear SPXlikesBalz,

You have requested to reset your password on Sherdog Mixed Martial Arts Forums because you have forgotten your password"

[edman5555]

Nope. Wht name does he post under?

[Luke]

So is spx ever coming back?

I told you to go to sherdog and bring him back. Don't you have a sherdog account?

[Luke]

Anyonr watch true detective? Hell of a show. Ive been watching it for some time now. There are eight episodes total and its on hbo.

I watch it . Very good show. Also just started watching House of Cards which I recommend

[edman5555]

So is spx ever coming back?

[edman5555]

Anyonr watch true detective? Hell of a show. Ive been watching it for some time now. There are eight episodes total and its on hbo.

[MMA_scientist]

It's hard to say. Here's the quick version:

Thanks a ton. That actually helps a lot.

trotterz, you of course right, but it is more depth than I want to go into right now, that's why I asked to assume the same number of cases for each judge. The reality is slightly more complicated, but I can figure it out from here...

Thanks again guys, youre the best... I am really jealous of people that can do math. It makes me mad, because I feel like I should be able to do something like that and after you express the problem appropriately, the answer becomes clear to me. But I just can't order my thoughts in mathematically, even though I often imagine I am an expert logician (I even taught formal logic classes at a University). I realize how limited I am in this way when the problem is expressed correctly and see how obvious it should have been to me.

[Mr. IWS]

Fun fact: DiCapprio has never won an Oscar.

That is crazy. I mention that shit everytime the Oscars are on.

[trotterz]

It's hard to say. Here's the quick version:

given that:

1) You asked asked about the lawyers and not the judges
2) The judges are fairly clustered; 45-55 is not a huge range
3) You said each lawyer had the same number of cases with each judge

We can basically ignore the judges part of the whole thing, and reduce the problem to the lawyer percentage alone, or something like:

We have three coins of unknown weighting; we flip each 200 times, and they end up heads 54, 56, and 61 percent of the time respectively. What can we say about the weighting of the coins?

The most relevant thing I can tell you is that the margin of error (with 95% probability) for a sample size of 200 is given by "0.95 / sqrt(200)", or about 7%, which just happens to be exactly the span between the two extremes.

(So we're 95% certain that the true weighting of the first coin is between 47 and 61 percent, and we're independently 95% certain that the true weighting of the last coin is between 54 and 68 percent.) So it's fairly likely that the last coin/lawyer is actually better than the first one, harder to have a strong opinion between the first two.

I understand your rationale Svino but only if you are stricly looking to know if a lawyer is better than the other ones. I would still take into consideration the amount of Cases won in front of each judges .... simply for the reporting purposes. It is totally possible that the lawyer with the 54% rate won 95% of his cases in front of judge 1 but but only won 10% of his cases in front of judge #2.
Doing so, it is possible that the lawyer with the lowest cases won overall might have better odds to win a case (as opposed to the other 2 lawyers with beter overall stats) if it's in front 1 of the 2 judge in particular.
It is possible that there is no correlation whatsowever in the stats mentionned earlier about the judge to judge stats, but it might be interested to take a look at it from my point of view.

[Svino]

svino, could I get some non-gambling related math help?

I don't know how to express the question in the right terms, sorry. Here is the issue: There are 5 disability judges, each grants benefits between 45% and 55% or the time. There are 600 cases. If three lawyers each handled 200 cases and had individual approval rates of 54%, 56%, and 61%-- would that indicate something that is normal variance (just chance) or is there some indication of an advantage (ie one lawyer is "better" than the others)?

I know I can't ask the question right. Sorry.

If it is a lot of work just tell me to fuck off.

It's hard to say. Here's the quick version:

given that:

1) You asked asked about the lawyers and not the judges
2) The judges are fairly clustered; 45-55 is not a huge range
3) You said each lawyer had the same number of cases with each judge

We can basically ignore the judges part of the whole thing, and reduce the problem to the lawyer percentage alone, or something like:

We have three coins of unknown weighting; we flip each 200 times, and they end up heads 54, 56, and 61 percent of the time respectively. What can we say about the weighting of the coins?

The most relevant thing I can tell you is that the margin of error (with 95% probability) for a sample size of 200 is given by "0.95 / sqrt(200)", or about 7%, which just happens to be exactly the span between the two extremes.

(So we're 95% certain that the true weighting of the first coin is between 47 and 61 percent, and we're independently 95% certain that the true weighting of the last coin is between 54 and 68 percent.) So it's fairly likely that the last coin/lawyer is actually better than the first one, harder to have a strong opinion between the first two.