2005 Crane Pool Forum Rankings Thread

[metsmarathon]

ok, so , it wouldn't be a ranking project without me trying to come up with a novel, overly complicated (and likely misguided) way of doing things.

and so, i present to you, the metsmarathon player ranking system.

here's what i've done...

basically, for offensive players, i take the players OPS+ and multiply it by time played. to get this number, i take the players' plate appearances and divide by 700.

it looks like this:

OPS+ x (PA / 700)

for david wright, we get the following:

138 x (657 / 700) = 129.5

is this a problem? well, see for yourself. douggie mientkiewicz gets himself up to the 8th best position player, which clearly seems to disagree with his win shares.



as soon as i figure out how i can convey my ridiculous pitcher rankings, i'll throw that up, and then try to meld the two. comments for now?

Edited December 30, 2005 by Guest

[Edgy MD]

Another problem with OPS+ and ERA+ is that he figures get distorted in the extremes. ERA+, in fact, moves toward infinity. A player with an ERA+ of 200+ might have gotten a 2.00 ERA in a league with a 4.00 ERA. A player with an OPS+ of 200 may have gotten a 1.500 OPS in a league with a .750 OPS. Assuming player A was a starting pitcher and player B a full-time batter, player B is a much much more productive guy.

[metsmarathon]

so how do i apply this to the pitchers?

here's where it gets more difficult, complicated, and wrongheaded!

if i look at just innings pitched, i overvalue the starters at the expense of the relievers.

if i look at games pitched, then i do the opposite. and of course, how do i account for a guy like heilmann, who both started and relieved?

well, i did this.

i created three new provisional stats with which to label my columns:

%IP, % GS, and %GR

%IP is (IP / 250); %GS is (GS / 34); %GR is (Games Relieved / 82)

where did the numbers come from? my head, mostly. a full season for a starter would be 250 innings, and/or 34 games started. a full season by a reliever would be 82 games relieved.

by averaging %IP with %GS for a starter, i get a better valuation of his overall playing time than by looking at just one or the other. ditto for a reliever. for heilmann, i add %GS with % GR, and then average that with %IP. i hope that all makes sense.

mathematically, it looks like:

ERA+ x [iP% + (GS% + GR%)] / 2

for pedro, roberto, and heilmann, respectively, we see the following:

148 x [217/250 + (33/34 + 0/82)] / 2 = 148 x [0.87 + (0.97 + 0)] / 2 = 131.7
162 x [69.7/250 + (0/34 + 67/82)] / 2 = 162 x [0.28 + (0 + 0.82)] / 2 = 88.8
132 x [108/250 + (7/34 + 46/82)] / 2 = 132 x [0.43 + (0.21 + 0.56)] / 2 = 79.1

and, to account for their small part of batting, i add in their offensive contribution, calculated the same way as david wright.

and so, this gives me the following mess:

Edited December 30, 2005 by Guest

[metsmarathon]

sure, before i can try to tackle the problem of mixing the lists, edgy points out the biggest inherent problem.

darn you!

[Guest Johnny Dickshot]

Finally I have company in my lofty Hernandez positioning.

[Edgy MD]

One complicated way to compare apples and oranges is to calculate a pitchers' OpS+ yielded, and use that, instead of ERA+. That not only allows you to better compare pitchers to batters, but clears up the records of pitchers (usually relievers) who have misleading ERAs by runs they're largely responsible for yielding counting against the ERA record of another, or runs they're largely responsible for saving embellishing the ERA record of another.

[metsmarathon]

complicated? that's prolly the least complicated, and most sensible way of doing things out there!

it really would make sense if pitchers were rated based on OPSa instead of ERA...

maybe i'll toss that into my next figuring.

edit: oh, i've been thinnking up a new way of doing OPS+ and ERA+ that should yield more linear, potentially more sensible (tho slightly more squashed) results.

basically, for OPS+, its always bugged me that the formula is:

OPS+ = 100 * (OPB/lgOBP + SLG/lgSLG - 1)

why not actually compare OPS to the league OPS? wouldnt that make more sense? (also, two players on the same team with the same OPS can end up with different OPS+)

and so, my OPS* is then, quite simply:

OPS* = 100 * (OBP + SLG) / (lgOBP + lgSLG)

for ERA+, i've made it more linear, by doing the following...

ERA* = 100 * [1 - (lgERA - ERA) / lgERA]

ERA* no longer approaches infinity for the extremely good pitcher, tho it now would approach infinity for the extremely bad pitcher. but linearly with ERA.

how do these look? look at the following to find out:



which, i think, undervalues poor pitching significantly. a 6.00 ERA is the same, here, as a .500 OPS!

tweaking is needed, of course. and i'm thinking too hard, too.

anyways, using (these two) formulation(s), i get the following ranking(s).



i'm more inclined to go with the score+ than score* rankings, as it utilizes fewer of my own machinations. also, score* gives pedro top billing, and i just can't abide by that. maybe if i fiddle with my OPS* and ERA* some more...

Edited December 30, 2005 by Guest

[Edgy MD]

Well, it's more complicated because nobody I know lists an OPSA+, so you have to calculate it standardized to the league with park factors by youself.

Believe me, I try to do this. Then, if you measure batters from an OPS+ of 60 (presumably the replacement level), do you measure pitchers in the opposite direction from an OPSA+ of 140?

[Frayed Knot]

Well some kind of consensus is emerging from the 12 different opinions floating around at this point: the list below is an averaging of the ones submitted so far plus both a list of Win Shares & our PotG standings but not KC's & Marathon's lists, both of which seem to be incomplete.

And while I don't want to stop anyone from submitting their opinions if they haven't already, what might be interesting at this point is to hear from those who are differing from the masses on certain players and have them either defend their positions in an attempt to cajole the rest of us into seeing things their way, or own up to their temporary insanity and explain how they've "seen the light" due to arguments from the others.

Remember, these are ranked in 30 (best) to 1 (worst) order so your list is still in the correct order even if you printed yours in a 1-30 format. I've just converted the data into what we've been using all along.
Our consensus to date:

30 - Wright: it was unanimous
29 - Martinez
28 - Floyd
there were a few which had these two flipped but virtually all had them at one spot or the other
27 - Glavine: a few had him as low as 25 points (6th place) but nothing major

26 - Reyes
25 - Beltran
This was a farily close race and it would be interesting to hear from those on either side of the argument as to the reasoning, particularly those who had Reyes several spots above Beltran: Zvon (Reyes - 27, Beltran - 23); YSG (27/25); Ole-Owned (27/25) - or from those who had it the other way around: Nymr (26 Beltran / 23 Reyes)

24 - Piazza: Dickshot has him down at 20
23 - Cameron: Seawolf at 18
22 - Benson: Agbodan (18), Nymr (17) and Dickshot (19) dragged his overall ranking down
21 - Seo: As high as 24 (Agbodan, Nymr) and as low as 17 (Dickshot)
20 - Hernandez: Yancy at 17, Dickshot at 26
19 - Heilman: Valadius is low man (15), Seawolf (22) & Dickshot (23) are high
18 - Diaz: Dickshot (21), Zvon (15)
17 - Zambrano: has a variety of opinions (SURPRISE!) from Zvon at 11 to Valadius at 21 ... also PotG had him at 22
16 - Castro: Nymr (13)
15 - Jacobs: Nymr (19), Frayed Knot & Dickshot (12)
14 - Anderson: Zvon (18), Seawolf (20) vs FK (9) and Nymr (11)
13 - Looper: pretty even - but maybe only because orig jb hasn't checked in yet
12 - Mientkiewicz : All voters are within two slots here but it's PotG (17) and Win Shares (8) who disagree the most
11 - Matsui: Again PotG is considerably higher (16) while Nymr (7) is low and Yancy (16) high
10 - Woodward: Dickshot and Seawolf on the high end (14) and FK on the low side (7)
9 - Padilla: Nymr real low (4), while Zvon gives him the most love (13)
8 - Cairo: Nymr 1 !! vs Zvon (12) ... that'll translate into a boatload of difference in total points (8 vs 1,195)
7 - Ishii: Almost as big a spread here: Agbodan & Nymr (2) vs FK (11)
6 - Trachsel: Nymr rewards Trax's 6 starts with 12 pts, while Zvon leaves him off the list entirely!!
5 - Koo: Nymr puts him at 10, FK only to 3
3.5 (tie for 4th) - Offerman: 7 & 8 from Nymr & Agbodan, 'No List for You!' from Yancy & Seawolf
3.5 - Bell: Not on 'Olerud Owned's list
2 - Williams
1 - Graves - 5 from Nymr

DeJean: 6 from Nymr ... and Ring: 4 from Abogdan
although neither makes the top 30 or got more than 1 or 2 pts from anyone else

Edited December 30, 2005 by Guest

[nymr83]

i have no idea how Cairo, regardless of your opinion of mediocre full-timers vs. good part timers, gets anywhere on anyone's list. he wasn't mediocre, he was TERRIBLE, one of the worst hitters in the entire major leagues and his contribution to the team was a negative one.

[Guest rpackrat]

Here's my list:

1. Wright
2. Floyd
3. Beltran
4. Pedro
5. Reyes
6. Glavine
7. Cameron
8. Piazza
9. Seo
10. Hernandez
11. Heilman
12. Benson
13. Castro
14. Diaz
15. Zambrano
16. Padilla
17. Jacobs
18. Matsui
19. Caro
20. Woodward
21. Looper
22. Anderson
23. Mienbtkiewicz
24. Koo
25. Takatsu
26. offerman
27. Williams
28. Ishii
29.Santiago
30. Trachsel

[Guest Bret Sabermetric]

Could the final cumulative POTG postings be appended to this thread?

[Elster88]

Jacobs is cumulatively above Looper and Matsui.

[Willets Point]

Bret Sabermetric wrote: Could the final cumulative POTG postings be appended to this thread?

Here they are in a link from our archives.

[metsmarathon]

there'd be two ways of doing OPA+, to my thinking.

one is more like OPS+, the other more like ERA+

OPA+ = 100 x [ 3 - ( SLA / lgSLA + OBA / lgOBA )]
or
OPA+ = 100 x [ lgSLA / SLA + lgOBA / lgOBA -1]

i like the first one better. it gives pedro a 147 OPA+ for 2005 (as opposed to 162 using the other method)

so, if i then apply my same methodology for playing time corrections for pitchers, and mix the list with hitters, i get my final ranking for the 2005 mets.



edit: bah, stupid traffic limiter on geocities. thats what i get for not paying attention to file sizes... (not like there's anybody sitting at their computers thinking, darnit, that guy had some good ideas, and now i cannot see them!) smaller pics should reappear in an hour, i think

anyways, below is the newest ranking!

30 Wright
29 Martinez
28 Floyd
27 Glavine
26 Benson
25 Beltran
24 Reyes
23 Heilman
22 Zambrano
21 Hernandez
20 Piazza
19 Cameron
18 Looper
17 Seo
16 Diaz
15 Mientkiewicz
14 Ishii
13 Padilla
12 Bell
11 Castro
10 Anderson
9 Matsui
8 Jacobs
7 Koo
6 Woodward
5 Cairo
4 Trachsel
3 Ring
2 Aybar
1 Graves
0 Offerman
-1 Takatsu
-2 Heredia
-3 Valent
-4 Daubach
-5 DeJean
-6 Williams
-7 Hernandez
-8 Matthews
-9 Santiago
-10 DiFelice
-11 Hamulack

[Frayed Knot]

Nymr83 wrote: i have no idea how Cairo, regardless of your opinion of mediocre full-timers vs. good part timers, gets anywhere on anyone's list. he wasn't mediocre, he was TERRIBLE, one of the worst hitters in the entire major leagues and his contribution to the team was a negative one.

Not arguing that should be anywhere near the top, only that simply the
fact that he held a starting position for a good portion of the season puts
him ahead of those who contributed both briefly and not very well either.
I'm talking about those like Graves and his 5+ ERA in 20 IPs or DeJean
and his 2.1+ WHiP/near 6.00 ERA in 25 IPs.
Not to mention that it's tough to explain why Matsui's: 255/300/352
season rates so high above Cairo's 251/296/324 ... not a whole lot of
difference there.

[nymr83]

i put matsui ahead of cairo because matsui hurt the team for a shorter amount of time and slightly less per AB, cairo was last to me but #25-29 on my list are not things i care to argue with much passion, if you want to flip them around i wouldnt care really

[Frayed Knot]

[Zvon]

Frayed Knot wrote: 6 - Trachsel: Nymr rewards Trax's 6 starts with 12 pts, while Zvon leaves him off the list entirely!!

This was purely intentional.
I was scoring on positives and negatives.
Trax went negatively off the chart,lol.

[Frayed Knot]

So Trachsel's 37 IPs @ 1.32 WHiP & 4.14 ERA somehow ring up more
negatives than:

- Graves' 21 IPs, 1.82 WHiP, 5.57 ERA
- Aybar's 25 IPs, 1.50, 6.04
- DeJean's 26 IPs, 2.10, 6.31

??

[Guest Johnny Dickshot]

This was purely intentional. I was scoring on positives and negatives. Trax went negatively off the chart,lol.

If we're going to get ignorant responses like this, just throw out the whole vote.

[Guest Bret Sabermetric]

Johnny Dickshot wrote: This was purely intentional. I was scoring on positives and negatives. Trax went negatively off the chart,lol. If we're going to get ignorant responses like this, just throw out the whole vote.

If we ask the death penalty for ignorant responses, the undertakers are going to be mighty busy around here.

[Guest Johnny Dickshot]

C'mon. This project till now has had earnest participation and compromise. If some are now unwilling to engage in such, f 'em.

[Guest Bret Sabermetric]

I was just funnin' ya. My serious response, offered when this project began, is that we just exclude the one or two most extreme outliers from the final tabulation, thus eliminating misinformed views and, much more important, discouraging people from posting rankings without thinking them through sensibly. If we do that, then such posts are just amusing to read (as well as to write).

[Guest abogdan]

So is this thing done? Have the rankings been added to the master list?

[Frayed Knot]

Nothing's been added to the master list yet but,since no one else seems particularly interested in adding their two cents or defending/explining their choices if they have made some,then I guess this year's order is pretty well set.

Taking the lists submitted and coming to a consensus by taking an average after throwing out the high & low extremes, leaves us with:

30 - Wright
29 - Martinez
28 - Floyd
27 - Glavine
26 - Reyes
25 - Beltran
24 - Piazza
23 - Cameron
22 - Benson
21 - Hernandez
20 - Seo
19 - Heilman
18 - Diaz
17 - Zambrano
16 - Castro
15 - Jacobs
14 - Looper
13 - Anderson
12 - Meintkiewicz
11 - Matsui
10 - Woodward
9 - Padilla
8 - Cairo
7 - Ishii
6 - Trachsel
5 - Koo
4 - Bell
3 - Offerman
2 - Williams
1 - Aybar

Got some votes but remain on the outside looking in:
Graves, Takatsu, DeJean, Ring, Daubach, Santiago

[Edgy MD]

I'm working on something.

[Zvon]

Frayed Knot wrote: So Trachsel's 37 IPs @ 1.32 WHiP & 4.14 ERA somehow ring up more negatives than: - Graves' 21 IPs, 1.82 WHiP, 5.57 ERA - Aybar's 25 IPs, 1.50, 6.04 - DeJean's 26 IPs, 2.10, 6.31 ??

37 innings from a starter though.
And 4 losses out of 6 games started.
I seriously consider Trachsels performance not worthy.

Those three you mention does make that a tuff call.
But they were relievers.
And I think Graves threw more innings than that.

[Frayed Knot]

Edgy DC wrote: I'm working on something.

I'm sure you are.

[Edgy MD]

Don't go bein' a punk.