Generating Historical Interactive Odds

BlackAndBlue · 10-04-2002, 05:24 PM

Machiavelli,

I've been playing interactives for the past couple of years on and off by using my gut.
I'm a small winner over that time. Reading yours posts got me thinking about
doing some data analysis and computing hisorical win percentages based on
inning, starting line, and current score differential (I'm starting with baseball
simply because the innings seem to be easier than time in a FB game).

As a first shot I ignored the line and just looked at the value of average teams
at different points. Now I'm trying to look at incorporating the line into the
calculation and have run into some difficulty and am looking for some advise.

What I've done is tabulate win% by half inning, money line and current score
differential. My goal was to do some regression on these figures to be able
to account for the "skill difference" between the two teams (using the line as
a proxy for skill difference). The problem I've run into is that as the data gets
that scattered (eg there are only a handful of games where a -150 home favorite
was down by one run at the end of the bottom of the third).

This scattered data is leading me to believe that any regression will be unreliable
at best. Any suggestions? Am I essentially on the right track?

Machiavelli · 10-04-2002, 07:33 PM

You are very much on the right track.

I also feel that a regression is unreliable way to introduce team strength.

I am trying to think what else I can say without giving away my methods for doing this, which I don't want to do just yet.

This might end up being my last season trading, as I plan to take a long vacation/retirement starting in early December. I have not yet decided what to do with my knowledge regarding interactives. Do I give it to no one? Do I give it to a select group of people? Do I give it to WSEX (if they want it), whose treatment of me has been extraordinarily fair despite my tremendous success? Do I just post it in a public forum?

As to further hints on how to model this, I would point out (and you probably already know this) that the problem with regression is the big standard error that is introdcued when you try to model using such a small sample size. For example, trying to model how a -180 favorite will fair after four innings in a tied game can produce screwy results unless you have many thousands of samples. Simple logic says that the answer should be much higher than 50% and at least a little bit less than the approximately 64% chance they had at the outset. That is, unless there is a serious imbalance between one team's starting pitching strength vs. their bullpen's strength. For example, if the Braves are starting a AAA pitcher vs. Milwaukee's Ace that price might actually rise in a tie game as the game progresses.

Perhaps an easier thing to model would be each team's potential for future scoring....

KingOfTheSquares · 10-04-2002, 09:49 PM

Geeze Mach...you and I need to have a long serious talk...

Giving stuff to the Book? They already get 11/10 (or more) per wager!!!!

Come on down to my condo by the sea here in Tijuana Norte and I'm sure I'll get you to see things in a different light...

KingOfTheSquares · 10-04-2002, 09:51 PM

BTW Black...regression won't hack it, gotta build a sim. Whole problem is pretty non-linear and we all know where non-linear regression goes....

deepsouth · 10-09-2002, 11:43 PM

<< This might end up being my last season trading, as I plan to take a long vacation/retirement starting in early December. I have not yet decided what to do with my knowledge regarding interactives. Do I give it to no one? Do I give it to a select group of people? Do I give it to WSEX (if they want it), whose treatment of me has been extraordinarily fair despite my tremendous success? Do I just post it in a public forum?
>>

I'm surprised this hasn't gotten more attention. All of these choices sound interesting except the one about giving to a select group of other people (though perhaps you would be able to sell it to a select group of other people, and thus cash out with a lump sum payment ... not that this hasn't occurred to you).

It'd be interesting to see what would happen if you posted on a public forum, or gave to WSEX. I don't think it would be shocking if the markets became unbeatable or very close to it shortly thereafter. Your methods probably aren't completely formulaic, and besides, people could probably come up with minor improvements ... but still, it's likely that your methods becoming public knowledge or WSEX knowledge would make the market prices close enough to the "true prices" that profitable opportunities would become very rare. Or maybe I'm overestimating how good you are at this [img]i/expressions/face-icon-small-smile.gif[/img].

Anyways, I'll be interested to hear what you decide, and I'm sure I won't be the only one.

Machiavelli · 10-10-2002, 12:25 AM

<< I'm surprised this hasn't gotten more attention. >>

I don't believe many people even view this forum. 88 views, and each person can multiple view's count multiple times (so 5 to 10 of them are probably mine).

<< It'd be interesting to see what would happen if you posted on a public forum, or gave to WSEX. >>

I am leaning towards just posting them here. I would have no control over the information if I gave it to a select group, and I wouldn't want to upset anyone for excluding them from the group.

<< WSEX knowledge would make the market prices close enough to the "true prices" that profitable opportunities would become very rare. >>

Yes and no. I believe the everyday value that I was able to reap over the course of the baseball season might disappear unless someone enhances my method, but in certain events it is my sense that the prices are highly influenced by square action. Monday night football, any game involving the Rams (maybe not anymore), postseason baseball, and the NBA playoffs are all examples. At WSEX, I am only inferring this through my 75 share orders being filled for more than the standard 50 shares at WSEX. At tradesports, I can see other people's limit orders and they confirm my suspicions.

<< Your methods probably aren't completely formulaic >>

They definitely aren't. In football, I insist that any sim or formula would have to be highly advanced if it is going to be used, there are too many subtleties.
That is the other issue with release of my baseball methods. There is a pretty major "leap of logic" in there, which many people might have objections with. I believe I mentioned it to you at the MW convention (reminder hint: geometric series).

AussieVamp2 · 10-10-2002, 02:13 AM

Similar sorts of baseball type stuff is out there, I think. Of course you still have to do the work.

AussieVamp2 · 10-10-2002, 02:15 AM

The other thing I guess it sometimes leaps of logic are needed, and you certainly do end up finding things 'this doesn't make sense' or this is very counter-intuitive, but it still works.

Machiavelli · 10-10-2002, 02:23 AM

<< Similar sorts of baseball type stuff is out there, I think. >>

For in-progress?

AussieVamp2 · 10-10-2002, 04:27 AM

well, modelling games at certain stages and states, yeah

Pancho Sanza · 10-10-2002, 02:20 PM

baseball is easy to simulate, all you need is inning by inning scoring for a decent amount of games, need to know scoring patterns in extra innings and your set.

The Actuary · 10-10-2002, 06:53 PM

Hi Mach,

I hope you've considered other alternatives besides the one's you've outlined in your post, a few come to mind, but even something as simple as offering it to a video game manufacturer to enhance their games could be lucrative I would imagine.

Either way, the "new" era of interactives will never return, and that's a shame.[img]i/expressions/face-icon-small-smile.gif[/img]

TA

AussieVamp2 · 10-10-2002, 11:53 PM

<< baseball is easy to simulate, all you need is inning by inning scoring for a decent amount of games, need to know scoring patterns in extra innings and your set. >>

well, not quite that simple, really

Pancho Sanza · 10-11-2002, 01:23 AM

uhhh, Mr Aussie, I beg to differ, but for what I use it for, it is that simple.

KingOfTheSquares · 10-11-2002, 09:01 PM

Vamp, I agree with Pancho. Compaired with say, NFL football, Bases are simple. Why don't you tell us why you think it isn't?

Pancho Sanza · 10-11-2002, 09:34 PM

Because he never goes over 10 words per post.[img]i/expressions/face-icon-small-smile.gif[/img]

AussieVamp2 · 10-11-2002, 11:46 PM

Using an generic aggregated sample to model a specific situation? Or using a generic aggregated sample to model a generic situation?

KingOfTheSquares · 10-12-2002, 02:30 AM

well, he got 21 in on that one...