PECOTA PREDICTIONS

[Mephistopheles]

Alfonso Soriano

.290/.335/.545 - Meph

.278/.336/.544 - Actual

 

NAILED!

 

Ryan Theriot

.279/.331/.368 - Meph

.270/.330/.347 - Actual

 

Almost, just 20 pts of SLG.

 

Kosuke Fukudome

.301/.381/.467 - Meph

.289/.401/.504 - Actual

 

Yeah.... They had a blogpost about him i dont recall if they had his PECOTA listed there or not.

 

Matt Murton

.295/.368/.456 - Meph

.295/.359/.462 - Actual

 

NAILED!

 

Mark DeRosa

.286/.341/.431 - Meph

.286/.360/.439 - Actual

 

20 pts of OBP.

 

Ronny Cedeno

.278/.325/.432 - Meph

.278/.326/.422 - Actual

 

NAILED!

 

Geovany Soto

.285/.356/.478 - Meph

.273/.352/.470 - Actual

 

NAILED!

 

Felix Pie

.275/.326/.465 - Meph

.291/.344/.479 - Actual

 

Meh...they had him with 15 pts higher BA which followed through on OBP and SLG.

 

Derrek Lee

.301/.373/.504 - Meph

.303/.387/.527 - Actual

 

They didn't project a HR drop as much as I thought they would.

 

Aramis Ramirez

.299/.354/.551 - Meph

.301/.363/.536 - Actual

 

NAILED!

 

Zambrano

3.93 - Meph

3.81 - Actual

 

NAILED!

 

Rich Hill

4.12 - Meph

4.02 - Actual

 

NAILED!

 

Ted Lilly

4.21 - Meph

4.25 - Actual

 

NAILED!

 

Jon Lieber

4.87 - Meph

4.62 - Actual

 

Meh....

 

Jason Marquis

5.12 - Meph

5.04 - Actual

 

NAILED!

 

Sean Marshall

4.65 - Meph

4.74 - Actual

 

NAILED!

 

Sean Gallagher

4.76 - Meph

4.91 - Actual

 

Close enough....

 

PECOTA sucks for RPs....so I missed on Wood and Marmol

 

All in all I thought it was an impressive projection of projections myself. You should all bow down to me.

[Mephistopheles]

WARP thinks that the difference between an average 2B and a replacement level 2B in terms of fielding runs is about 28. Last year year for Roberts anyway. He get a 30 VORP for his projected .293 EQA and 3 full wins for being an average defender. It's really stupid. Nobody gets a cup of coffee that's -30 fielding and a .225 EQA hitter. WARP is [expletive].

 

And yes, it projects Roberts to get something like 170 more plate appearances than DeRosa.

 

Not entirely true. WARP = BRAR + PRAR + FRAR. BRAR is NOT position adjusted. It is batting average over a replacement level hitter, which is .230 EqA. The FRAR adjustment is value for playing a tougher position. Ie in theory it *should* work this way:

 

VORP = BRAR + FRAR baseline for that position.

 

It really does not work that way because VORP does not figure out RP level the same way nor does it figure out runs the same way.

 

If you want a good indicator, EqA RP level for each position

C .227

1B .248

2B .231

3B .235

SS .228

LF .250

CF .242

RF .258

 

If you set the highest as +10 over a DH, you get and give them about 150 games:

 

33 C (actual 39)

18 1B (actual 10)

30 2B (actual 29)

33 SS (actual 33)

27 3B (actual 22)

23 CF (actual 24)

13 LF (actual 14)

10 RF (actual 14)

 

So it's still a pretty damn good rough indicator. It looks at the same thing but puts the fielding value based on a position. The actual way they get their run values for a position is the average change in defensive +/- output if you make the switch. IE a +4 SS moving to 2B would be expected to be a +8 2B. My beef with WARP really isn't that.

 

I found those EqA RP levels using the spreadsheet's RP level for VORP.

[haltz]

I actually did know that VORP is a different replacement level, and WARP is based around BR. But it's what's on the PECOTA sheet and it's close enough. I still don't understand how you can have a separate replacement level for hitting and fielding. It would make it way too low. Maybe you just explained that and I'm not following you.

 

What's your beef with it then?

 

EDIT: Actually, I see what you are saying.

[Geech]

PECOTA likes Fukudome a lot. Not only is the projection the most optimistic I've seen, take a look at his list of comparables:

 

J.D. Drew, Gene Hermanski, Jim Edmonds, Fred Lynn

 

I wonder what PECOTA sees that other projections and translations don't.

[hankp101]

I can't remember if it was actually reported somewhere or if it was dreamed up on here that he was a possible target, but PECOTA predicts very good things for AJ Burnett: 3.64 ERA, 38 VORP, 5.1 WARP.

[Mephistopheles]

I actually did know that VORP is a different replacement level, and WARP is based around BR. But it's what's on the PECOTA sheet and it's close enough. I still don't understand how you can have a separate replacement level for hitting and fielding. It would make it way too low. Maybe you just explained that and I'm not following you.

 

What's your beef with it then?

 

EDIT: Actually, I see what you are saying.

 

people think that it's denominated in wins (it is but it's not). It's fairly good framework on hitters. it is embarrassingly dumb for pitchers

[Mephistopheles]

anyways predicted runs off pecota

 

Chicago Cubs 803.1

Milwaukee Brewers 793.2

New York Mets 786.9

 

that is all ive done so far

[SABR Gamer]

anyways predicted runs off pecota

 

Chicago Cubs 803.1

Milwaukee Brewers 793.2

New York Mets 786.9

 

that is all ive done so far

 

 

Wow, I sure hope so, because our pitching is better than those two teams. If that happens, we would be the best team in the NL.