prediction formula

[slw87]

didn't really know where this question belongs but does anyone know where I can find some formulas for fantasy predictions such as the ones that petcota or others like it use? thanks for any help that u can provide in pointing me in the right direction.

[Matthias]

Short answer, no, I don't and no, you won't be able to find it.

PECOTA doesn't use a "formula"; it looks at a player's production at various ages and levels and compares it to players in the past who look similar. And all of the rest of them are mostly educated guesswork, I think. The only one that really has a "formula" is Marcels and even then, he doesn't say (at least on that page) exactly what the formula is. But he might elsewhere.

[slw87]

does anyone know anything more about marcels or any other projection systems. I've heard chones mentioned as well. and is there anywhere I can get a hold of one of these formulas to study and learn how it works.

[Amazinz]

The basis of the Marcel system is your standard (fairly accurate) three-year weighted average (5/4/3). Then he does a regression to the mean for each player on the three year league average and an age adjustment: http://www.tangotiger.net/archives/stud0346.shtml.

[Matthias]

You can try e-mailing the guy who does Chone; his blog is here: http://www.lanaheimangelfan.blogspot.com/

But I think you're really thinking about it in the wrong way; it's not that there is, "a formula"... it's that they have, "a system" that has many formulas and iterations and everything else. It's much more complex than you seem to be thinking about it.

[Bobbleheadrusty]

You can try e-mailing the guy who does Chone; his blog is here: http://www.lanaheimangelfan.blogspot.com/

But I think you're really thinking about it in the wrong way; it's not that there is, "a formula"... it's that they have, "a system" that has many formulas and iterations and everything else. It's much more complex than you seem to be thinking about it.

Exactly. If there was some magical formula for player predictions then there would be no point in playing out the season. You can look at some of the well known predictions, which are based on formulas that have been proven out, but rarely can you apply that to fantasy with much accuracy. For every player like Aviles, Quentin or Bruce that exceeds the predictions you have an Alex Gordon or Travis Hafner who fail to meet the minimum.

So if you want to be hardcore you can look at Pecota, Chone, Marcel ect. but then you need to use your own judgement on when to take each player when compared to the value. Just because your prediction said that Longoria was going to be the 5th best third baseman last year you would have been wrong to take him there, as he was available much later in most leagues.

[slw87]

I agree completely with u rusty. There is and never will be a full proof system. However fantasy baseball is a numbers game. It is based completely around statistics and probabilities and I am intrested in finding out more about how these systems work. I think they are a good place to start in evaluating talent and making fantasy decisions. However there will never be a formula that acounts for freak injuries such as line drives of pitchers or people like zack grienke who is suicidal and quits baseball and then two years later is on his way to being the best pitcher in baseball. Ur right there is no perfect formula but we do need to take these probabilites into account to better ourselves as owners but the final judgement needs to come from our hearts, because at least for me I do it because I love the game and the players and even the numbers that create them. Justwant to get better though. Thanks for all the input.