« Pete Walker | Main | Time For Yankees Fans To Join The Crowd? »

May 15, 2007

The Long Odds

Baseball Prospectus predicts that the Yankees now have a 3.87% chance of winning the A.L. East.

The folks at CoolStandings.com predict that the Yankees now have a 8.2% chance of winning the A.L. East.

Further, Prospectus says the Yankees have a 23.34% of making the post-season, in general, whereas Cool Standings has them at 36.6%.

Wow, and it's only May 15th.

A 4-8% chance of winning the East and a 23-37% chance of making the playoffs, period. Those are some very long odds, huh?

Posted by Steve Lombardi at May 15, 2007 02:46 PM

Comments

Wait till next year... :-). Just kidding. I like our chances. Till I see, "The Yankees are eliminated", then I won't pay attention to odds.

Go Yanks!!!!

Posted by: Garcia [TypeKey Profile Page] at May 15, 2007 03:25 PM

Aren't these the same people who predicted 95-97 wins?

Posted by: dave24s [TypeKey Profile Page] at May 15, 2007 03:57 PM

Tradesports has Yankees to win the AL East at 27-30%. If you really believe Prospectus or CoolStandings, you should go short Yankees (as painful as that would be) and also check out the other mispricings. Or maybe the market has good reason to keep the faith, even as we all begin to lose it.

Posted by: Dave [TypeKey Profile Page] at May 15, 2007 03:58 PM

I'm with you, Garcia.

The bats can't stay dormant forever--now that the pitching seems to be working itself out, it'll come together.

Posted by: snowball003 [TypeKey Profile Page] at May 15, 2007 04:31 PM

Those are fatally flawed odds.

Those odds presume that the teams will continue at their current levels (in terms of each criteria the system throws together).

In order for the system to have any credibility, it would have to be the case that, if the Yankees gain a bunch of ground, then the odds would have had to remain where they are right now. The system would just have to say, sure it's happening, when there is a 3.87% of something happening (just to use the first percentage that you've cited), you're just seeing one of the 3.87% times. If the odds were correct in the 1st place, they should remain constant and, if the Yankees win, well the system would say of course that happens sometimes. If it didn't happen sometimes, then instead of the odds beings 3.87%, the odds would be 0%.

Whenever you see a system that, as a season progress, one day's worth of odds will look unrecognizable not that long afterwards, you know it's just an ad hoc throw it together as you're going along, screw accuracy kind of system.

BP claims to be able to predict things with such accuracy that they can tell us the percentage down to the 4th digit (a 3.87% chance is .0387). So, if they can claim to make such a pinpoint scientific statement like that, let's see the proof. Let's see them proof it. Let's see them show us that teams in the Yankees current position did go on to win the division 3.87% of the time. And then let's see them repeat the proving and justify the rest of their numbers.

Otherwise, their numbers are no better than a random number generator. My random number generator gives me a 60.71% of the Yankees winning the division title. My method was grabbing the closest book to me, opening up to a random page, take the 1st 2 numbers that appeared in the text that I had opened to, put them in front of the decimal place, take the next 2 and put them afterwards.

And, regardless of whether the Yankees win or lose the division, I can later claim that my method produced the correct odds. If the Yankees win, I will tell you that, when something says there is a 60.71% chance of something happening, then sometimes it happens and see, this was one of the sometimes in which it happened. If they lose, I will tell you that, when there is a 60.71% chance of something happening, sometimes it doesn't happen and this was one of those times.

And, just for the record, my method also gives us a 12.41% chance of the Red Sox winning the division title. But, it's a good thing that my system didn't say just take the 1st number you see and put that in front of the decimal place. If that was the case, then since I had turned to page 124, it would have meant the Red Sox had a 124% chance of winning the division title.

Posted by: Lee Sinins [TypeKey Profile Page] at May 15, 2007 05:38 PM