PECOTA projections

[BohKnowsBmore]

LOL. I wasn't commenting on the post itself, but rather the play on words that came to mind.

Some people believe that this offseason has been a complete cluster luck.

Sent from my iPhone using Tapatalk

[TradeAngelos]

Nice. I like this team a lot. They're not perfect on paper, but I think we win the division.

I like our guys.

I won't mention any names, but some people claim their "model" has the Orioles with a 4% chance to win division, and that Gallardo and Fowler won't change that much.

I told him he was delusional.

Vegas will open them at 4-1 and I will take that all day long, especially if Fowler is in the mix.

[jsbearr]

I won't mention any names, but some people claim their "model" has the Orioles with a 4% chance to win division, and that Gallardo and Fowler won't change that much.

I told him he was delusional.

Vegas will open them at 4-1 and I will take that all day long, especially if Fowler is in the mix.

Was that me?

To be clear the percentages I reported are not models. They simply look at what teams were projected to do and where they wound up. So a 4% mark means teams with a certain projected win total in the past wound up in first or whatever. Very simple and straight forward.

Past work shows that since around 2010, projection models and betting markets screw up equally. I could not find any rhyme or reason why they miss out on some teams but not others. It was easier before 2010 when projection models were not as prevalent by professional gamblers as they are now.

But yeah...projections usually have a standard deviation. Of about 6 to 10 games, so if a team is projected as an 80 win team...16 percent of the time thereabouts they win 90 or more and get into the playoffs.

[Frobby]

Was that me?

To be clear the percentages I reported are not models. They simply look at what teams were projected to do and where they wound up. So a 4% mark means teams with a certain projected win total in the past wound up in first or whatever. Very simple and straight forward.

Past work shows that since around 2010, projection models and betting markets screw up equally. I could not find any rhyme or reason why they miss out on some teams but not others. It was easier before 2010 when projection models were not as prevalent by professional gamblers as they are now.

But yeah...projections usually have a standard deviation. Of about 6 to 10 games, so if a team is projected as an 80 win team...16 percent of the time thereabouts they win 90 or more and get into the playoffs.

That's a very high standard deviation, and it's the reason I don't worry too much about projections, though I have fun reading them.

I see the AL East this year as very wide open. I don't think there is any team that is guaranteed to finish over or under .500. I'm excited to watch it unfold.

[jsbearr]

That's a very high standard deviation, and it's the reason I don't worry too much about projections, though I have fun reading them.

I see the AL East this year as very wide open. I don't think there is any team that is guaranteed to finish over or under .500. I'm excited to watch it unfold.

Sure...it is accurate yet imprecise. Still...a 6 to 10 win standard deviation means something. Being projected to win 72 games is much different than 82 games.

[DrungoHazewood]

Sure...it is accurate yet imprecise. Still...a 6 to 10 win standard deviation means something. Being projected to win 72 games is much different than 82 games.

I like the way Tango framed projections on his site the other day. When you say a team is an 82-win team you're really saying there's a 33% chance of 81-83 wins, a 33% chance at less than 81, and a 33% chance at more than 83. Realistically six wins is a near-perfect standard deviation in projecting a MLB season.

[Number5]

I like the way Tango framed projections on his site the other day. When you say a team is an 82-win team you're really saying there's a 33% chance of 81-83 wins, a 33% chance at less than 81, and a 33% chance at more than 83. Realistically six wins is a near-perfect standard deviation in projecting a MLB season.

I think the odds of not finishing within one game of what's projected are a lot more than 67%.

[jsbearr]

I think the odds of not finishing within one game of what's projected are a lot more than 67%.

In years where standard deviation is 6 wins...33% of falling within on game is pretty close to what you would see. With a ten win standard deviation that is probably plus or minus 3.

[PeteU]

I recall PECOTA in 2012 and 2014....seriously, screw PECOTA.

No, really, screw PECOTA.

[jsbearr]

I recall PECOTA in 2012 and 2014....seriously, screw PECOTA.

No, really, screw PECOTA.

PECOTA is as precise and accurate as today's forecasted high temperature. I am not mad at the weather forecast.

[weams]

Sure...it is accurate yet imprecise. Still...a 6 to 10 win standard deviation means something. Being projected to win 72 games is much different than 82 games.

I question why it is 72 games.

[weams]

PECOTA is as precise and accurate as today's forecasted high temperature. I am not mad at the weather forecast.

I'm not mad at Pecota, but it has been wrong on the Orioles the last four years. And it is often wrong on the Red Sox.

[DrungoHazewood]

PECOTA is as precise and accurate as today's forecasted high temperature. I am not mad at the weather forecast.

A lot of people are mad at the weatherman. A lot a people are just mad.

[Redskins Rick]

I think maybe, they could have got better forecast from PETCO!

But, I am glad, give this team a little fire in their belly!

[weams]

I think maybe, they could have got better forecast from PETCO!

But, I am glad, give this team a little fire in their belly!

Late night and early morning low clouds and fog along the coast, clearing to hazy afternoon sunshine.

Highs in the mid to upper 70s, lows in the mid to upper 50s.