Kirby to kick the Rangers to the curb G.T. 8/12/22

User avatar
bpj
Posts: 12739
Joined: Mon Apr 29, 2019 12:55 am

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by bpj » Sun Aug 14, 2022 5:07 pm

Well color me surprised that the global warming alarmist and vaccine mandate believer thinks projections- which are wrong more like 70% of the time- are solid enough to buy into. Like "climate change" and, "Get your vax!"- which didn't work to begin with.

Figures.

The Mariners were "projected" for 70 wins last season. And here you are. Lol.

Projections aren't statistics bozo.

User avatar
bpj
Posts: 12739
Joined: Mon Apr 29, 2019 12:55 am

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by bpj » Mon Dec 12, 2022 3:32 am

bpj wrote:
Sun Aug 14, 2022 2:33 am
Wonder how accurate "the projections" have been so far.. at quick glance they got close on most of the obvious good players and pinned the tail on the donkey picking the rest of their numbers.

Projections are absolute rubbish. They rely on enough of their single player projections being wrong on the up or down side that their team projections average out as close to the actual record as possible then the closest group of dorks can claim victory even though it's all bogus.

For example:

If I project a total of 4.9 wins for three players (Steamer) France, Kelenic and Moore and my projections break down like the following:

3.0 WAR for France
1.6 for Kelenic
0.3 for Moore

and they finish with:

3.0 WAR France
-0.5 for Kelenic
2.0 for Moore

Well, my overall projection was pretty close when in reality I was a giant dickhead throwing darts at a board to get there.

That's how projections work. It's hogwash.
Fun thread.. :lol:

The projections for last season, oof-
Attachments
Screenshot_20221211-193216_DuckDuckGo.jpg

GL_Storm
Posts: 2896
Joined: Sat Jul 10, 2021 9:00 pm

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by GL_Storm » Mon Dec 12, 2022 4:06 am

Why is this old thread coming up again?

User avatar
bpj
Posts: 12739
Joined: Mon Apr 29, 2019 12:55 am

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by bpj » Mon Dec 12, 2022 5:20 am

harmony wrote:
Sun Aug 14, 2022 3:00 pm
The science denier lacks a rudimentary understanding of statistics.

If done well baseball projections are wrong 40 percent of the time.

It's the 60 percent right that provides the advantage
.

Note the noisy ding-ding-ding in the slot machine area of a casino. That casino is paying out ... in essence losing at that specific moment on those precise machines.

But is the casino losing money on slot machines? Of course not.

Don't be confused by ding-ding-ding of Sam Haggerty's hot streak. Haggerty has been fun to watch augmenting a Seattle outfield depleted by injuries to Mitch Haniger, Julio Rodriguez, Jesse Winker and Kyle Lewis. The 28-year-old Haggerty -- who entered this season with a career OPS+ of 64 in 59 games -- deserves our accolades. Love the (positive) outlier.

At the end of the season (if not before) the publicly available projections systems will compare each other's accuracy. In an MLB front office, non-player personnel decisions may hinge on the accuracy of that team's internal projections. These are not your grandfather's baseball operations.

Analytics run industries and baseball was a relative latecomer to the game.
It's coming up because I was called a "science denier" for calling BS on the covid vaccines, climate crisis, and baseball projections.

A claim was made last year (see above bold) that projections are right 60% of the time so I kept receipts.

I posted pictures of the projections the experts made for 2022, and now we have the seasons numbers that we can test that theory

That post here- viewtopic.php?f=3&t=7071&start=170#p195987

I'm not sure how you guys define being "right", but the WAR projections (ZIPS & STEAMER) which I posted for future reference were within 20% of actual WAR in just one case (I only took the time to chart semi-regulars/starters).

Most were off between 50-200%.

Now people are starting to cite projections again so I'm pointing out, again, that looking at projections for any sample smaller than an entire team is like throwing darts at a dart board. Projections mean nothing, and they're right much closer to 0% than 60%.

Using them in discussion for how a player might do next season is meaningless.
Attachments
Screenshot_20221211-233152_Gallery.jpg
Screenshot_20221211-233152_Gallery.jpg (213.63 KiB) Viewed 379 times

User avatar
bpj
Posts: 12739
Joined: Mon Apr 29, 2019 12:55 am

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by bpj » Mon Dec 12, 2022 7:02 am

The point was what I wrote here:

viewtopic.php?f=3&t=7071&start=170#p195987
For example:

If I project a total of 4.9 wins for three players (Steamer) France, Kelenic and Moore and my projections break down like the following:

3.0 WAR for France
1.6 for Kelenic
0.3 for Moore

and they finish with:

3.0 WAR France
-0.5 for Kelenic
2.0 for Moore

Well, my overall projection was pretty close when in reality I was a giant dickhead throwing darts at a board to get there.

That's how projections work. It's hogwash.
Now what do we see in the "Projections"?

They were horribly wrong individually.

But once you add up the Actual WAR and compare it to the projected WAR it looks like this:

Actual WAR: 21.7
STEAMER Proj WAR: 22.6
ZIPS Projected WAR: 27.6

So, despite the closest individual projection still being wrong by 20%, and most being WAY off, once you add it all up, the cumulative product looks like it was accurate thanks to the Law of Averages.

STEAMER was within 1 WAR cumulatively for this group of players despite only 3 of 13 being within even 50% of their projection.

Because it's all a bunch of BS where the larger the sample size the more they can just throw whatever number strikes their fancy because it all averages out and people who lack critical thinking skills think they can claim the projections are anything resembling reliable when evaluating individual players despite the fact that the projections are completely bogus individually.
Screenshot_20221211-230007_Excel.jpg
Screenshot_20221211-230007_Excel.jpg (202.69 KiB) Viewed 392 times
After I pointed out this is how projections work, this is what was said to me-
harmony wrote:
Sun Aug 14, 2022 3:00 pm
The science denier lacks a rudimentary understanding of statistics.

If done well baseball projections are wrong 40 percent of the time.

It's the 60 percent right that provides the advantage.
So, yeah, when projections start getting FWIW'd at us again, I may be a little quick on the trigger to point out that their worth is nothing individually.

User avatar
D-train
Posts: 67910
Joined: Mon Apr 29, 2019 1:33 am
Location: Quincy, MA

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by D-train » Mon Dec 12, 2022 10:44 am

Great stuff brother!
dt

User avatar
Coeurd’Alene J
Posts: 5113
Joined: Wed May 01, 2019 2:56 am

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by Coeurd’Alene J » Mon Dec 12, 2022 1:53 pm

bpj wrote:
Mon Dec 12, 2022 7:02 am
The point was what I wrote here:

viewtopic.php?f=3&t=7071&start=170#p195987
For example:

If I project a total of 4.9 wins for three players (Steamer) France, Kelenic and Moore and my projections break down like the following:

3.0 WAR for France
1.6 for Kelenic
0.3 for Moore

and they finish with:

3.0 WAR France
-0.5 for Kelenic
2.0 for Moore

Well, my overall projection was pretty close when in reality I was a giant dickhead throwing darts at a board to get there.

That's how projections work. It's hogwash.
Now what do we see in the "Projections"?

They were horribly wrong individually.

But once you add up the Actual WAR and compare it to the projected WAR it looks like this:

Actual WAR: 21.7
STEAMER Proj WAR: 22.6
ZIPS Projected WAR: 27.6

So, despite the closest individual projection still being wrong by 20%, and most being WAY off, once you add it all up, the cumulative product looks like it was accurate thanks to the Law of Averages.

STEAMER was within 1 WAR cumulatively for this group of players despite only 3 of 13 being within even 50% of their projection.

Because it's all a bunch of BS where the larger the sample size the more they can just throw whatever number strikes their fancy because it all averages out and people who lack critical thinking skills think they can claim the projections are anything resembling reliable when evaluating individual players despite the fact that the projections are completely bogus individually.
Screenshot_20221211-230007_Excel.jpg
After I pointed out this is how projections work, this is what was said to me-

[quote=harmony post_id=196023 time=<a href="tel:1660489225">1660489225</a> user_id=111]
The science denier lacks a rudimentary understanding of statistics.

If done well baseball projections are wrong 40 percent of the time.

It's the 60 percent right that provides the advantage.
So, yeah, when projections start getting FWIW'd at us again, I may be a little quick on the trigger to point out that their worth is nothing individually.
[/quote]

That quote didn’t age well. Lol

Guess I’m a clot shot free science denier……..I must be a Mariner denier as well

harmony
Posts: 1350
Joined: Thu May 02, 2019 4:59 pm
Location: Portland OR

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by harmony » Mon Dec 12, 2022 3:12 pm

Anecdotal evidence is often presented to counter quantitative evidence.

The analogy might be the person who wins on the slot machines on a given day. That person is happy that he or she beat odds that clearly favor the house. That does not make playing the slot machines a wise ongoing investment.

Any person who prefers Player A over Player B is implicitly projecting that Player A will outproduce Player B in the future. How did that person arrive at that projection? Did that person look beyond a player's 2022 OPS?

For years the widely cited publicly available MLB projections systems were measured against each other for accuracy. Those external comparisons are harder to track down these days. However, MLB front offices almost certainly scrutinize their internal projection systems. The broad-scope evaluation looks past any anecdotal shortcomings.

Jarred Kelenic is clearly a polarizing figure as evidenced by his disparate 2023 projections from FanGraphs, Steamer and Marcel. Nevertheless Kelenic remains a 23-year-old with six years of team control earning the league minimum salary while on the 26-man roster. Kelenic has one minor league option remaining so the Mariners should not trade low on Kelenic this offseason. Still Kelenic needs to earn his playing time.

Thank you for the discussion.
Last edited by harmony on Mon Dec 12, 2022 4:39 pm, edited 1 time in total.

User avatar
D-train
Posts: 67910
Joined: Mon Apr 29, 2019 1:33 am
Location: Quincy, MA

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by D-train » Mon Dec 12, 2022 3:26 pm

I am sure all of the investors that refused to sell Enron stock low for $10 a share felt they were being wise as well...

Perhaps trading JK this off season would be a sell low move...

Or not. :)
dt

User avatar
bpj
Posts: 12739
Joined: Mon Apr 29, 2019 12:55 am

Re: Kirby to kick the Rangers to the curb G.T. 8/12/22

Post by bpj » Mon Dec 12, 2022 4:52 pm

harmony wrote:
Mon Dec 12, 2022 3:12 pm
Anecdotal evidence is often presented to counter quantitative evidence.

The analogy might be the person who wins on the slot machines on a given day. That person is happy that he or she beat odds that clearly favor the house. That does not make playing the slot machines a wise ongoing investment.

Any person who prefers Player A over Player B is implicitly projecting that Player A will outproduce Player B in the future. How did that person arrive at that projection? Did that person look beyond a player's 2022 OPS?

For years the widely cited publicly available MLB projections systems were measured against each other for accuracy. Those external comparisons are harder to track down these days. However, MLB front offices almost certainly scrutinize their internal projection systems. The broad-scope evaluation looks past any anecdotal shortcomings.

Jarred Kelenic is clearly a polarizing figure as evidenced by his disparate 2023 projections from FanGraphs, Steamer and Marcel. Nevertheless Kelenic remains a 23-year-old earning the league minimum salary while on the 26-man roster. Kelenic has one minor league option remaining so the Mariners should not trade low on Kelenic this offseason. Still Kelenic needs to earn his playing time.

Thank you for the discussion.
What a pointless drivel post to deflect. I understand. It's hard for you to just say when you're wrong.

Preferring a player may you be "projecting" that you think that player is the better option.

That's a far cry from people discussing options for the Mariners and piping in with, "FWIW Player X is projected to hit .276/.325/.450 with 4.2 WAR."

It's literally worth nothing. Between 2 systems (ZIPS & Steamer) with a sample of the 13 Mariners hitters- so a total of 26 projections- their guess was closer than 20% ZERO times.

Bloviate about anecdotal and quantitative evidence and lottery winners all you want, but I put it in black and red for you. Projections are a terrible argument when discussing individual players.

Say you think they're a better player, nobody has a problem with that, but posting imaginary projections isn't quantitative evidence. It's a guess. And not even very good ones.

Post Reply