BP Sox projection.

[sullythered]

Originally Posted by kittle42

Then that is taking a *very* shortsighted look at a model that necessarily requires a review of all its predictions and numbers across all of baseball to evaluate. You're basically admitting to cherry-picking.

By the way, I'm not defending BP, just logic.

I'm talking about the Sox, particularly, because that is the forum we are in. If the discussion were elsewhere, then it would be a different discussion (I do find PECOTA flawed, as far as full-team projections go, league-wide).

[TheOldRoman]

Originally Posted by Nellie_Fox

Gee, I must have missed those posts.

When you can't make a point, it's best to resort to hyperbole and constructing strawmen. That's what "intellectuals" do.

[dickallen15]

2012 PECOTA had 12teams projected correctly within 5 games, had 12 teams projections off by at least 10 (including the Sox who were at 10). There were 2 right on, and 5 missed by at least 16. The average the W-L were off in the AL was 8.86, and in the NL 8.06. Overall average off by 8.43 games.

It sparks a lot of debate, but good or bad for the team you are pulling for, it's really no more accurate in the end than the posts in the preseason prediction threads. We all know which teams have the best players. If you don't have a roster full of players that have been consistently successful, PECOTA isn't going to rate you that high.

[doublem23]

Originally Posted by dickallen15

2012 PECOTA had 12teams projected correctly within 5 games, had 12 teams projections off by at least 10 (including the Sox who were at 10). There were 2 right on, and 5 missed by at least 16. The average the W-L were off in the AL was 8.86, and in the NL 8.06. Overall average off by 8.43 games.

It sparks a lot of debate, but good or bad for the team you are pulling for, it's really no more accurate in the end than the posts in the preseason prediction threads.

PECOTA does not project wins and losses. It projects runs scored and runs allowed. I cannot for the life of me figure out which of those two sentences you guys cannot comprehend.

[kittle42]

Originally Posted by doublem23

PECOTA does not project wins and losses. Its projects runs scored and runs allowed. I cannot for the life of me figure out which of those two sentences you guys cannot comprehend.

THE GAMES AREN'T PLAYED ON PAPER, DAMNIT!

There is no point to projections, as before the games are played, every team has a similar chance of winning the World Series. I don't know how people here can predict the Astros to lose 100 games until they do lose 100 games.

[dickallen15]

Originally Posted by doublem23

PECOTA does not project wins and losses. Its projects runs scored and runs allowed. I cannot for the life of me figure out which of those two sentences you guys cannot comprehend.

And then projects wins and losses. The debate was how off they have been about the White Sox. The fact is, their projection was a bit off on the White Sox last year, and there were 11 teams that they were at least as "off on their projection" as they were on the White Sox. So for the guys that want the Sox to suck, and there are a couple on here, it's really no reason to puff out their chests with pride just yet.

[doublem23]

Originally Posted by dickallen15

And then projects wins and losses.

No, you can easily turn RS/RA into W-L with the Pythagorean W-L formula but the actual data is in RS/RA

[dickallen15]

Originally Posted by doublem23

No, you can easily turn RS/RA into W-L with the Pythagorean W-L formula but the actual data is in RS/RA

Then they are converting it. This entire argument has been about how off they have been about the White Sox projected win totals.

[asindc]

Originally Posted by doublem23

PECOTA does not project wins and losses. It projects runs scored and runs allowed. I cannot for the life of me figure out which of those two sentences you guys cannot comprehend.

Which is, of course, why BP presents their projections in terms of wins and losses instead of just runs scored and runs allowed.

As has been continually noted by those, including me, who express ridicule of BP's consistently wrong guesses with regard to the Sox's record, that should not be taken as an indictment of BP as a whole. They seem to generally do exceptional work, but for whatever (not subjective, IMO) reasons, BP generally fails to properly evaluate the Sox before the season is played out, at least in its application of PECOTA.

[sullythered]

Originally Posted by doublem23

No, you can easily turn RS/RA into W-L with the Pythagorean W-L formula but the actual data is in RS/RA

Isn't that exactly the point we're arguing? That converting PECOTA is not a very accurate system when it comes to W-L? This thread is about the BP projection for the White Sox W-L record.

[doublem23]

Originally Posted by dickallen15

Then they are converting it. This entire argument has been about how off they have been about the White Sox projected win totals.

I would assume they do it as a courtesy to people since the Pythagorean W-L forumla is pretty standard knowlegde amongst the baseball community and it presents the data in an easily digestable format.

The best analogy I can think of off the top of my head are bakers. Most professional bakers measure ingredients in weight (100 grams of flour), not the volume that you're used to seeing in cookbooks designed for everyday chums like you and me, because it's just easier and more accessable for most people. The conversions are pretty commonplace, but if you're operating in the world of professional cooking, you're not using cups and tablespoons to create the recipes, only translating at the very end.

Originally Posted by asindc

Which is, of course, why BP presents their projections in terms of wins and losses instead of just runs scored and runs allowed.

As has been continually noted by those, including me, who express ridicule of BP's consistently wrong guesses with regard to the Sox's record, that should not be taken as an indictment of BP as a whole. They seem to generally do exceptional work, but for whatever (not subjective, IMO) reasons, BP generally fails to properly evaluate the Sox before the season is played out, at least in its application of PECOTA.

Originally Posted by sullythered

Isn't that exactly the point we're arguing? That converting PECOTA is not a very accurate system when it comes to W-L? This thread is about the BP projection for the White Sox W-L record.

It's an important distinction to make for no other reason than for someone to make an accurate critique of their success or failures based on the PECOTA system, it would be measured against actual RS/RA output since, again, that is what PECOTA actually projects. If you're judging the model against W-L, well, it kind of underlines the fact that you're out of your league.

[amsteel]

If people knew how weather was predicted they would lose their ****.

[asindc]

Originally Posted by doublem23

I would assume they do it as a courtesy to people since the Pythagorean W-L forumla is pretty standard knowlegde amongst the baseball community and it presents the data in an easily digestable format.





It's an important distinction to make for no other reason than for someone to make an accurate critique of their success or failures based on the PECOTA system, it would be measured against actual RS/RA output since, again, that is what PECOTA actually projects. If you're judging the model against W-L, well, it kind of underlines the fact that you're out of your league.

What fact does it underline for those who use PECOTA's RS/RA projections to publish actual W/L projections?

[sullythered]

Originally Posted by doublem23

I would assume they do it as a courtesy to people since the Pythagorean W-L forumla is pretty standard knowlegde amongst the baseball community and it presents the data in an easily digestable format.





It's an important distinction to make for no other reason than for someone to make an accurate critique of their success or failures based on the PECOTA system, it would be measured against actual RS/RA output since, again, that is what PECOTA actually projects. If you're judging the model against W-L, well, it kind of underlines the fact that you're out of your league.

Then BP should not project wins, at all. They should only publish projections on RS/RA, so as to avoid confusion.

[Nellie_Fox]

Originally Posted by doublem23

I would assume they do it as a courtesy to people since the Pythagorean W-L forumla is pretty standard knowlegde amongst the baseball community and it presents the data in an easily digestable format.

It's an important distinction to make for no other reason than for someone to make an accurate critique of their success or failures based on the PECOTA system, it would be measured against actual RS/RA output since, again, that is what PECOTA actually projects. If you're judging the model against W-L, well, it kind of underlines the fact that you're out of your league.

If BP doesn't think that using the PECOTA projected RS/RA to plug into Pythagorean W/L, then they should say so, and certainly not do it themselves. Since they do it, that seems to say they think it's a legitimate thing to do.