Brock,

Thank you for your explanation! I am pretty clear on the concept of the database, as you surmised.

I am completely ignorant about the jargon, however. Although it may be "just" jargon, knowing it is essential for following this conversation, and I am unable to follow much of it. Much of it I probably don't have to follow, but I don't know when I am missing something that I ought to know.

I have also never set up my spreadsheets so that they can be used in a database. I know that there must be principles and conventions, but I don't know what they are and I have never given them any thought. It also appears that there are existing databases of major-league baseball data that the HoM database will need to work with, and that it has conventions. I don't know what those are.

I have used spreadsheets mainly to handle complex computations that I need to do for many players, not for data management. Any time I have ever needed to pull data out of a spreadsheet for presentation, I have always done it by hand, so the organization of data within the spreadsheet has never mattered much, as long as all of the formulas were correct. _I_ know where the numbers are that I need, and I'm the one who uses them, so I have never bothered to set them up so that another person could easily use them, much less an automated information retrieval system.

As far as I can judge, arranging this data for use in a database will require a totally different setup than anything I have in my spreadsheets now.

So some of the things I need to know are the following:

1) Will I need to put my data for MLEs into a spreadsheet of the kind you describe, with one row per player season, and all player seasons included on one sheet? Are there multiple alternatives? If so, what would be preferable?

2) Will data items need to be arranged in a standard order, e.g. column 1 player name, column 2 team name, column 3 year, column 4 league, column 5 plate appearances, column 6 at bats, column 7 hits, and so on? Or will the MLE data be separate, such that I could arrange that data in any order, as long as the joins are properly indicated?

3) If I am to make a spreadsheet that fits established standards and conventions, how do I find out what those are? If there are not pre-existing standards to hook into, how can I best establish the standards that I will need to know in order to make my spreadsheets link up with everything properly, e.g. the format for names, teams, leagues, seasons, types of stats, etc.?

4) If there is some flexibility in the arrangement of the data, how should I arrange it, and what should I include? Given that these are MLEs, should my spreadsheets include the raw NeL stats as well as the translated stats? How much information about the translation formulas should be accessible through the database? What data do I have the right to include in a publicly accessible database? How far broken down should the translated stats be (For my purposes, I have generally just aimed for PA, hits, walks, stolen bases, total bases, and the percentage stats that can be derived from those. Because Dan R has wanted distribution if extra base hits for baserunning value estimates, I have started learning how to turn total bases into reasonable XBH distribution projections, for career, but not for seasons. Is getting seasonal XBH projections something I should work on getting into the data?

5) How much of the calculations are to be built into the DB? Dan R's spreadsheets show the seasonal totals from his WAR, but not how they are calculated? Do I need to be concerned with putting together spreadsheets of results, or do I need to be concerned with operationalizing the calculations my spreadsheets do to generate MLEs?

6) How much of the stuff that I am talking about here is going to be handled by the database designers, and how much am I to do prior to sending my data to whomever will link it into the database?

7) Who will decide what the answers to these questions will be, and when?

In all honesty I am not too sure what the discussion is about right now. However, I think I can answer chris's questions under the assumption we are talking about the database for the website and using his spreadsheets to load data into it...

1) You should be able to arrange it in virtually any format you please, as long as it makes sense. Whomever loads the data into the database can probably do the transformations "on the fly."

2) Any order. The script that loads the data will determine which column goes into which column in the db.

3) We have to decide, I would think.

4/5) Any columns that can be derived from other columns need not be included. For example, if you have TB and AB you don't need to include SLG. And if you have 1B,2B,3B,HR you don't need to include TB.

6) All you need to be concerned about is the data. How it's formatted, etc. really can be handled by the db designers. If you have all the data in a readable format, your work is done.

7) Me, right now ;-) Unless of course I have no idea what we are talking about which is completely possible. The scope of this discussion seems to have changed but I'm not sure how.

Shock, thanks for responding so quickly.

Let me first be more specific about the situation.

My particular batch of data that could be included in the HoM database is major-league equivalencies for many of the Negro-League HoMers and serious candidates. Right now, that data is structured, more or less, as one spreadsheet per player. Each spreadsheet includes the raw NeL data, various elements of intermediary data necessary for the MLE calculations, including estimates that have to be done by educated guesswork rather than a formula, and the MLE results. I have a hard time believing that these spreadsheets can just be loaded into the database in their current form, unless the uploaders do a lot more work than I expect they ought to have to do, or the scripts are much more sophisticated that I can imagine. What I am would like to know is what I need to do to get them into a form that the uploaders and the scripts can work with and that will make it possible for end users to access, interpret, and make use of the data easily.

Now to more on each point.

1) You should be able to arrange it in virtually any format you please, as long as it makes sense.

What determines whether a format "makes sense"?

2) Any order. The script that loads the data will determine which column goes into which column in the db.

Then the labels for the columns will need to phrased in ways that will be understood by the script?

I wrote:

3) If I am to make a spreadsheet that fits established standards and conventions, how do I find out what those are? If there are not pre-existing standards to hook into, how can I best establish the standards that I will need to know in order to make my spreadsheets link up with everything properly, e.g. the format for names, teams, leagues, seasons, types of stats, etc.?

Shock replied:

3) We have to decide, I would think.

How, then, should we work on making those decisions? Should I give an overview of the data I have? E-mail someone one spreadsheet as an example? Ask questions about how other parts of the database are fomatted? Is this something that should be discussed now, or should I wait until the designers get many of the more fundamental issues hashed out?

4/5) Any columns that can be derived from other columns need not be included. For example, if you have TB and AB you don't need to include SLG. And if you have 1B,2B,3B,HR you don't need to include TB.

What about derivations that are particular to these MLEs, rather than usual? For instance, one step in the process of calculating the MLEs is the regression of the first estimate to a five-season baseline. Would that kind of derivation be built into the database, or does the database have a set of "standard baseball derivations, like slugging from TB and AB that it will perform, but nothing else? To put it another way, will the database want the results of MLE calculations, or will it want the data from which the results were derived and the formulas to do the deriving?

6. If you have all the data in a readable format, your work is done.

What are the criteria for a readable format?

7) Me, right now ;-) Unless of course I have no idea what we are talking about which is completely possible. The scope of this discussion seems to have changed but I'm not sure how.

I think you have addressed the questions that I was asking, though, as you can see, I am still looking for further clarification. Since I haven't been able to follow most of the conversation so far, I am not the right person to address whether the scope of the discussion has changed :-). I am just butting in with my own concerns about what I should be doing with my little bit of data. Now may not be the right time for anyone to bother about my concerns, and, if that is the case, I am content to wait until whatever the right time would be.

1.

Kevin Johnson is the single person closest to having a list of all the blackball & Cubaball & white minors & white independent team-season affiliations that appear in all the playing records that have earned Hall of Merit MLE attention. He is also the proximate source of many raw playing records for players and teams/ballparks, and many estimates of park factors --and one of the few people most likely to know that more data or better estimates have become available since the generation of some MLE (beside the main point of the website project but the likely source of some valuable byproduct). He also has some experience handling the playername, teamname, leaguename issues.

2.

Tom Tango #32 to Chris Cobb, quoted by Paul Wendt #90, cited by Chris Cobb #99:

CC, This is the first half only, which I interpret to be TT's only instruction to you. The second half, not re-quoted here, is TT's aside to the Chorus and to Shock.

Chris, the only requirement is that for players that are not in the BDB database, then we need to make sure they exist somewhere to link them. So, it would seem that would need to not only link to the BDB MASTER table, but also include other master tables, so that we have a "super duper" master table that the Players table will link to.

So, you should create your own master table with:
- player_id
- player_name
and everything else that you want

I understand that Kevin Johnson has just volunteered to satisfy the underline need with my assist, which leaves the last four lines as TT's instruction to you.

I understand that TT sees this almost entirely open to you (and Kevin or whoever is working with you on the MLE database). I suppose that is because he imagines the MLE database to be essentially separate from the HOM and non-critical.

3.

Shock's replies suggest to me that he and TT are "on different pages". It's clear that he imagines greater intimacy for the HOM and MLE databases and he sees greater immediacy in answering some design questions.

bump
(perhaps I have lost my reply, composed between #102 and 103. The index does say 104 comment.)

Yes, it's really tangotiger who can give much better answers to these questions. I am simply trying to talk in a vague point of view about how, technologically, a spreadsheet gets put into a database.

I don't know anything about MLE or NeL or anything like that. The actual data that you guys want to include is 100% up to you. I am just saying that, in terms of formatting in the spreadsheet, it doesn't matter much (from a tech point of view) how it's actually organized. Eg, what order the columns are in, or whether two columns should actually be one column, etc. That can all be automated.

Anyway, I'll just shut up and let tangotiger talk. :-)

"Then maybe you could maintain a blog on the site every week where you show us a cool list and talk about it for a bit. ;-)"

I'm thinking "Mondays with Howie" - seems original, and yet derivative, too

:)

I am just butting in with my own concerns about what I should be doing with my little bit of data. Now may not be the right time for anyone to bother about my concerns, and, if that is the case, I am content to wait until whatever the right time would be.

Chris - Not really trying to sidestep your very valid questions and concerns, but the easiest route for the project would probably be for you to just email me your spreadsheets, then I can consolidate the relevant data from them into 3 or 4 tables/spreadsheets which will match the format I'm using for Eric Chaleck's MLE's.

Chris - Not really trying to sidestep your very valid questions and concerns, but the easiest route for the project would probably be for you to just email me your spreadsheets, then I can consolidate the relevant data from them into 3 or 4 tables/spreadsheets which will match the format I'm using for Eric Chaleck's MLE's.

KJOK: If sidestepping my concerns means less work for me and getting the MLEs into the database, by all means sidestep! Is the e-mail address that I should use available through the BBTF? If so, I'll just look it up. I'm under a couple of deadlines that I have to meet before Christmas, so I may not be able to send them until next week, but if I can get it done tomorrow night, I will.

Just want to check in - what is the next thing anyone needs from me regarding this? As I said earlier, I will be somewhat AWOL from from 12/24-1/5, just want to tie up loose ends before then.

Post deleted, wrong thread!

Here's what the database will look like. If there are any issues (and I have two noted below), now is the time to bring it up. We'll finalize shortly.

The "K" refers to the Primary Key of the table. And F refers to this field being a primary key of ANOTHER table (foreign key).

VOTERS
K voter_id
- voter_name

VOTER_ALIASES
K voter_id
- screen_name
- effective_date (or current_indicator)

Justication for this table: if we are going to be receiving ballots where the voter's screen name is listed or various other aliases, then we need to be able to map to it.

WEBPAGES
K webpage_id
- webpage_type (discussion, players, data, plaque, etc)
- webpage_date
- url

ELECTIONS
K year_id
- election_date (can be removed if implied by the WEBPAGES record; please advise)
F electee_count
F discussion_webpage_id
F results_webpage_id

POINTS
K electee_count
- rank
- points

PLAYERS
K player_id
- player_name
- player_bio (large text field)
F bio_webpage_id
F discussion_webpage_id
F plaque_webpage_id
F primary_position_id
F secondary_position_id

This table should be fed from either Lahman or Retrosheet, and appended with other player ids.

EVALUATIONS_DANR
K player_id
K year_id
- and a
- whole bunch
- of fields

EVALUATIONS_JOED
K player_id
K year_id
- and a
- whole bunch
- of fields

And anyone else that wants to contribute a table.

BALLOTS
K voter_id
K year_id
K player_id
K ballot_rank
- voter_comment
- selection_rejected_switch

(Please advise as to how the rejection of ballots works.)

Tom

Tango we have had very few rejected ballots . . . basically if a voter does not post any comments, or has a criteria that doesn't apply consistent logic, or flatly refuses to consider a certain group of players, it could be challenged.

If the voter doesn't respond, or responds and isn't willing to change, or provide an explanation that is acceptable, it could be excluded. I can think of only a few times where this has happened, and the specifics are in a part of my brain that's inaccessible without a trigger (I'm sure someone remembers some specifics).

Does that help?

And is the rejection at the ballot level, or player/selection level?

That design looks solid, Tango.

I am not too sure I understand the purpose of the points table. Can points not be derived from Ballots?

Is the e-mail address that I should use available through the BBTF? If so, I'll just look it up.

I'm not sure it's visible thru BBTF - it just allows you to send a message to a hidden email address. So, you can send me a mesage via BBTF, and I can respond from my email account.

Or, even easier:
KJOKBASEBALL
AT
YAHOO.'the usual extension'.

Tango rejection would be at the ballot level.

Ok, then we need to amend it as:
BALLOTS
K voter_id
K year_id
- ballot_id (unique)
- ballot_rejected_switch

BALLOT_SELECTIONS
K ballot_id
K ballot_rank
- player_id (unique)
- voter_comment

This sounds fantastic, I really look forward to it!

Chris -

In the interest of getting you up to speed on database jargon, do you know what a "key" is? If you don't, please post up and I'll explain (or someone will). The need for and presence of keys is one of the very biggest differences between a database system and a spreadsheet system. It's also a lot of why the db people here aren't obsessing so much over exactly what you send them in the spreadsheet. They'll have to fit it into the table system that their keys dictate anyway. - Brock

Who is actually going to be setting up the database at the post? I can do it right now or very shortly using Tango's design if Joe wants to give me the keys. If someone else is doing it that's fine, too. Then I can start loading in some data and putting up some test pages so we can talk about layout, etc.

Tango we have had very few rejected ballots . . . basically if a voter does not post any comments,

This is of course not accurate. Not even close to accurate. What Joe meant to say is that if Sunnyday submits a ballot without comments it is rejected. If anybody else does, it is accepted. Approx. 25 out of 26 were accepted over the history of the project.

Yeaahhh, let's not drag that into this thread please.

Yes, I'm not up on the politics, nor do I want to be. This is a post purely for facts, which is what I need to create a database.

***

I will create the database, and I will provide the SQL for it, so that everyone can create one also if they wish.

***

The next step is for people to be assigned responsibility for each table. These are the tables, and the responsible person has to provide the necessary data. It doesn't matter how the data is provided. I'll get it in. That'll be my job.

Once that's done, then Shock (I imagine) will take the database, and turn it into a website.

VOTERS
VOTER_ALIASES
WEBPAGES
ELECTIONS
POINTS (Tango)
PLAYERS
EVALUATIONS_DANR (Dan?)
EVALUATIONS_JOED (Joe?)
BALLOTS (*)
BALLOT_SELECTIONS (*)

(*) It would be lovely if each voter actually supplied his ballot for each year.

If you want to know what each table means, see Post 112 and 118.

So, from this point onward, I'd like to see people grab one of the lines above and post as follows:
WEBPAGES (John Doe)

And so on.

That person will then be charged with getting the necessary data. This is the point where this project succeeds or fails. So, by the time the HOF announces in mid-Jan that Tim Raines is not elected, we'll know if we've got a viable project here.

Go...

It might be best simply not to get into rejected ballots, for the following reason: since I've started counting, the most common reason for rejecting a ballot has been technical defect. A case in point: the ballot submitted by stax for the 2009 election had 16 names including a doubled number. His ballot was also strongly criticized, and the criticism may have influenced his decision not to resubmit - but it wasn't rejected because of the criticism. It was rejected because it was technically defective and the ballot counters don't want to play mind reader to determine his intent from what he said. A technically defective ballot cannot even be properly entered into the database, precisely because of the defects that caused it to be rejected in the first place.

The second most common category of rejected ballot: late ballots that missed the deadline.

I do think that the file I sent Tango included some of those. Speaking of which: Tango - was that file useful?

OCF:

Yes, if the file you are referring to is called: HoM_vote_file.xls
then this is the best data I have for 1917-2009 ballots.

So, where we are lacking for those who want to contribute is these tables for the 1916- and earlier data:
BALLOTS (*)
BALLOT_SELECTIONS (*)

So, if you have your individual ballots for these years, that would be great.

Well, since I've already gathered most of the discussion/ballot URLs ...

WEBSITES (Shock)

That should, obviously, read WEBPAGES.

VOTERS
VOTER_ALIASES
WEBPAGES (Shock)
ELECTIONS
POINTS (Tango)
PLAYERS
EVALUATIONS_DANR (Dan?)
EVALUATIONS_JOED (Joe?)
BALLOTS (OCF *)
BALLOT_SELECTIONS (OCF *)

* starting in 1917

Is anybody else contributing?

I would guess that this week will be the real indicator. The holidays are officially over and most people will be returning to work tomorrow.

If Friday comes and we don't have any more volunteers then I guess we can stick a fork in it.

Who was responsible for official ballot tallying in the early years? And does whoever it is have records?

Well, the pre-1917 won't hold up the project. We can just start it with 1917.

Other than the PLAYERS table, the other tables (VOTERS, ELECTIONS) are pretty simple, but they are a bit important. It would be nice to get a couple of volunteers.

And for the pre-1917, it would be great for each voter to have kept a record of what they did, and submit those.

I would be willing to help, but I'm a bit confused by the format.

If something is listed as an "F", does that mean you don't need that included in the data?

And I just don't understand what would be required for the PLAYERS data.

I will e-mail my pre-1917 ballots to Tango today.

VOTERS
VOTER_ALIASES
WEBPAGES (Shock)
ELECTIONS
POINTS (Tango)
PLAYERS
EVALUATIONS_DANR (Dan?)
EVALUATIONS_JOED (Joe?)
BALLOTS (OCF *)
BALLOT_SELECTIONS (OCF *)

* starting in 1917

Is anybody else contributing?

Post 58 (from me): "I think I can 'own' the players table. I can help with the others, especially the 'ballots' one as needed."

I'm out sick today . . . but working my way back into the flow . . . nasty cold/flu thing going around.

Who is actually going to be setting up the database at the post? I can do it right now or very shortly using Tango's design if Joe wants to give me the keys. If someone else is doing it that's fine, too. Then I can start loading in some data and putting up some test pages so we can talk about layout, etc.

As far as setting it up at the post . . . I can give the keys, but I have to find them first :-) . . . let me check into exactly what we are going to do - I need to talk to Jim F. as well (in terms of hosting - I've got the domain name taken care of), I'm planning to call him this week. We tried to talk before the holidays, but didn't get it done in time.

What should I be doing this week, besides talking to Jim?

VOTERS
VOTER_ALIASES
WEBPAGES (Shock)
ELECTIONS
POINTS (Tango)
PLAYERS (JoeD)
EVALUATIONS_DANR (Dan?)
EVALUATIONS_JOED (Joe?)
BALLOTS (OCF *)
BALLOT_SELECTIONS (OCF *)

* starting in 1917

"I would be willing to help, but I'm a bit confused by the format."

Ideally, you would just present something like this for the ELECTIONS table:
year_id,election_date,electee_count,discussion_webpage_id,results_webpage_id
2008,2008-06-07,3,138,169

And in the WEBPAGES table, you'd need to add:
138,http://www.bbtf.org/hom/whatever/link/you/need/
169,http://www.bbtf.org/hom/whatever/otherlink/you/need/

Don't worry about K and F and whatnot.

OK, then, I'll take ELECTIONS. Just let me know what the IDs will be.

Oh, duh, electee count, not voter count. That makes it even easier than I thought. In that case, I'll also check the pre-17 ballot threads to see what ballots are still there.

Ideally, Shock would generate a list of all the webpages, and then assign an id for them. He'd publish it, and then you would pick the id from that list.

The list of webpages would be easy enough to get, if he has DB access to the underlying database in expressionengine.

Shock wrote, "The discussions about individual players is my absolute favorite part of the HOM. I can't tell you guys how much I've learned about players from before my time from here!"

I too find this really valuable, fun, interesting, etc. Hypothetically, with a lot of editing and condensing it'd make the meat of a great HoM book. A much less difficult degree of editing and condensing (but still work) would make a great online resource. Rough, but way less rough (rambling, hard for the uninitiated to hunt thru, signal to noise problems) than it is now. Right now, it's kinda daunting to get into, compared to someone's experience who was part of the discussion as it happened. (I was an involved voter myself for a little while sorta early on.)

I believe (although I'm not sure) that all the ballot/discussion and maybe results threads have the same format:

http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/2009_ballot/

http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/2009_ballot_discussion/

http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/2009_results/

The player discussion threads are a little more problematic.

I have text files for the ballots from 1898-1917 that I prepared for my own purposes years ago. The format is like:

[voter name]
1 [player last name]
2 [player last name]
...
15 [player last name]

repeat

I can zip them up and email them if they would be useful.

OCF says that his data is complete 1920-2009.
My text ballot files cover 1898-1917.
I can enter 1918 and 1919 in the same format if that is useful.
If a different format is preferable, I can attempt that instead.

Let me know (here).

I have started the webpages one. I just have results and discussion for now:

http://spreadsheets.google.com/pub?key=pV7hxOzO9ThiKvItIKmXkZQ

VOTERS
VOTER_ALIASES
WEBPAGES (Shock)
ELECTIONS (Devin)
POINTS (Tango)
PLAYERS (JoeD)
EVALUATIONS_DANR (Dan?)
EVALUATIONS_JOED (Joe?)
BALLOTS (OCF *)
BALLOT_SELECTIONS (OCF *)

Just need one more taker for the VOTERS table, then we can begin.

***

* Devin supplied his own for pre-1917

***

jimd: yes, email it to me at tom~tangotiger~net (replacing ~ as appropriate).

I sent Tango my consensus score history file; one thing that could be used for is a list of voters from sometime in the early 1920's (I left my flash drive at home today so that's not in front of me.) That means that the primary need in identifying voters would be for the first three decades of the project.

One complication: the early years of voting were before site registration.

Did John Murphy have a list of voters? I remember him thanking a long list of people in the 2008 Results thread, but I don't know if they were all voters or not.

Here are the voters (with obvious typos). The person who comes forward to be in charge will need to resolve the typos, as well as using their Primer internal numeric IDs as their voter Id:

Adam Schafer
AJM
Al Peterson
andew siegel
Andrew M
Andrew M.
Andrew Siegel
Ardo
B. Wiliams
B. Williams
B. Williams doubled to catcher
bjhanke
Bleacher
Bleed the Freak
Brad G
Brad G.
Brad Harris
BradG
Brent
Brian H
Bryce B (Tanketra)
BryceB (Tanketra)
Buddha
Carl G
Carl Goetz
Casey Elston
caspian88
Chris Cobb
Chris Fluit
Chris J
Chris J.
Clint
Craig B
Craig K
dan b
Dan G
Dan R
Dan Rosenheck
danb
DanG
daryn
Daryn (George Sisler's #1 Fan)
David C Jones
David C. Jones
David Foss
DavidFoss
Dayrn
Devin McCullen
Devin Mccullough
DL from MN
Dolf Lucky
Don F
Don F (Pedro Feliz Navidad)
DonF
Dr. Chaleeko
dzop
ed
Eric C
Eric Chalek
Eric Enders
EricC
essteban rivera
Estaban Rivera
Esteban Rivera
favre
flaxseed
fra paolo
Gadfly
Got Melky
Got Melky?
Guapo
Howie Menckel
James Newburg
Jeff M
Jeff M.
jhwinfrey
Jim Sp
Jim Spencer
jimd
Joe Dimino
John Murphy
John Murphy
jscchmeagol
jschmeagol
jshmeagol
Juan V
kalmagnus
karlmagnus
Kelly from SD
Kelly in SD
Ken Fischer
kenn
KJOK
Lennox HC
Marc
Mark Donelson
Mark McKinnis
Mark McKinniss
Mark McKinniss (Dolf Lucky)
Mark S
Mark S. (jsch)
Mark Shirk
Martin
Martin (mdb1mdb1)
Matt B
MattB
Max Parkinson
mbd1mbd1
mdb1mdb1
Michael Bass
Michael D
MichaelBass
MichaelD
Mike Webber
mulder & scully
OCF
Patrck W
Patrick W
Patrick W.
Pedro Feliz Navidad
Philip
PhillyBooster
rawagman
rdfc
redsox2004
Rick A
Rick A.
RickA
rico vanian
RMc
Rob Wood
Robbie Cano (dzop)
Robby Cano
Robby Cano ('zop)
RobC
Ron Wargo
ronw
RonWargo
Rusty Priske
Sean (questionable)
Sean Gilman
Sean M
Sean M.
Seaver 1969
Seaver1969
stephen
sunnyday2
SWW
Tanketra
Thane of Bagarth
TheGoodSamaritan
Tiboreau
Tom D
Tom D.
Tom H
TomH
Trevor
Trevor P
Trevor P.
Vaux
Webber
YardApe
yest
zapatero

Here's a first crack at the list Tango just posted, sorted into clusters. I took out most of the "singles" - handles that appear only once on that list. As far as I know, the names appearing between each pair of lines below all belong to the same individual.

It looks like some of Tango's source for this was a file I sent him, but he had at least on other source besides me. However, I'm also pretty sure there were 1898-1920 voters not listed here at all. And some of those may have no BTF ID# since that was before registration.

Am I correct that Marc = sunnyday2? Or am I wrong about that? And what was PhillyBooster's other handle?

Also, I don't know who Sean (questionable) was - that doesn't look like my notation.

-------------------
andew siegel
Andrew Siegel
-------
Andrew M
Andrew M.
--------
B. Wiliams
B. Williams
B. Williams doubled to catcher
----------
Brad G
Brad G.
BradG
--------
Bryce B (Tanketra)
BryceB (Tanketra)
--------
Carl G
Carl Goetz
----------
Chris J
Chris J. (so he never voted as Dag Nabbit?)
----------
Craig B
-----------
Craig K
-----------
dan b
danb
-----------
Dan G
DanG
-----------
Dan R
Dan Rosenheck
---------
daryn
Daryn (George Sisler's #1 Fan)
Dayrn
-----------
David C Jones
David C. Jones
------------
David Foss
DavidFoss
-----------
Devin McCullen
Devin Mccullough
-------------
Don F
Don F (Pedro Feliz Navidad)
DonF
Pedro Feliz Navidad
----------
dzop
Robbie Cano (dzop)
Robby Cano
Robby Cano ('zop)
-----------
Eric C
EricC
-------------
Dr. Chaleeko
Eric Chalek
------------
Eric Enders
------------
essteban rivera
Estaban Rivera
Esteban Rivera
----------
Got Melky
Got Melky?
-----------
Jeff M
Jeff M.
-----------
Jim Sp
Jim Spencer
----------
John Murphy
John Murphy
---------------
kalmagnus
karlmagnus
-------------
Kelly from SD
Kelly in SD
-------------
Marc
sunnyday2
----------------
Mark Donelson
------------------------
Dolf Lucky
Mark McKinnis
Mark McKinniss
Mark McKinniss (Dolf Lucky)
---------------------------------
Mark S
Mark S. (jsch)
Mark Shirk
jscchmeagol
jschmeagol
jshmeagol
------------------
Martin
Martin (mdb1mdb1)
mbd1mbd1
mdb1mdb1
-----------------
Matt B
MattB
-----------------
Michael Bass
MichaelBass
--------------
Michael D
MichaelD
--------------
Mike Webber
Webber
--------------
Patrck W
Patrick W
Patrick W.
----------------
Rick A
Rick A.
RickA
--------------
Rob Wood
--------------
RobC
--------------------
Ron Wargo
ronw
RonWargo
-------------------
Sean (questionable)
----------------------
Sean Gilman
--------------------
Sean M
Sean M.
-------------------
Seaver 1969
Seaver1969
---------------
Tom D
Tom D.
--------------
Tom H
TomH
-------------
Trevor (?? I think this is the same but I'm not sure)
Trevor P
Trevor P.
------------

I'll add that I mostly didn't use long handles in my vote recording. John Murphy has always been "John Murphy" on my lists, with no mention of Grandma; Dan Rosenheck has always been "Dan R" on my lists, not "David Concepcion ... " As for the obvious typos above: those would be the typos of the person recording the vote (either me or someone else) and are not the responsibility of the voter in question.

IIRC, Matt B and Phillybooster are the same (before and after registration).

Also flaxseed and Chris J? (not so sure about this one)

No, flaxseed (the short version of a much longer user name) was most definitely not Chris J. These days, Chris J. uses "Chris Jaffe" for his articles on THT and posts on BTF as "Dag Nabbit." And his most recent HoM election ballot was in 1963.

If anyone is doing little articles about particular voters and posters who drove the discussion, Gadfly (who voted from 1946 through 1972) would deserve a section. We may not all have agreed with him (his consensus score average is about the same as karlmagnus's) but he got our attention.

And Paul Wendt is long-time member of the HoM voters group even though he doesn't vote.

Hi, just to let everyone know, I am also Kelly in / from SD

Also flaxseed and Chris J? (not so sure about this one)

I apologize for the faulty memory.

It was James Newburg that voted as "flaxseed".

Addendum to #149, summarizing the subsequent posts:
(I posted this yesterday without the appropriate computer file on had; I have it now and can confirm the flaxseed-James Newburg connection.)

-------------
Kelly from SD
Kelly in SD
mulder & scully
-------------
James Newburg
flaxseed
-------------
Matt B
MattB
PhillyBooster
-------------

OK, the ELECTIONS database is done. I cheated and used the dates that Shock put in WEBPAGES for the results threads, but they should be the same, right?

Tango, do you want it posted on GoogleDocs, or e-mailed to you?

Devin: email please.

Chris J. (so he never voted as Dag Nabbit?)

Not that I'm aware of.

Is there anyone willing to do the VOTERS tables?

At a minimum, take the list in post 148, and assign a voterid (using the Primer ID... e.g., Dag Nabbit / Chris J is 192), one record per line.

Once someone does that, then I can begin.

I'll do it.

I could use some help tracking down the people who are going by a completely different sceenname (e.g. B williams).

Okay. I entered what I could figure out. I'm posting it on GoogleDocs again so y'all can look at it and make corrections or fill in some blanks:

http://spreadsheets.google.com/pub?key=pV7hxOzO9ThjrVlD28PQRPg

VOTERS (Shock)
VOTER_ALIASES (Shock)
WEBPAGES (Shock)
ELECTIONS (Devin)
POINTS (Tango)
PLAYERS (JoeD)
EVALUATIONS_DANR (Dan?)
EVALUATIONS_JOED (Joe?)
BALLOTS (OCF *)
BALLOT_SELECTIONS (OCF *)

***

* Devin supplied his own for pre-1917. Others should do the same.

***

Ok, we're all in. Here's what everyone should be providing me via email, once they certify their data in csv, or xls format.

***

SHOCK:
voter_id, voter_name (one record per id)

voter_id, screen_name, effective_date or current_indicator (many records per id)

webpage_type (discussion, players, data, plaque, etc), webpage_date, url

You can make the date optional

***

DEVIN:

You will have to wait until Shock gives me his data, and I generate the relevant IDs.

***

JOE D:

You can start compiling this:
player_id, player_name, player_bio (large text field)

The player_id is either from BDB or Retrosheet. Either works. The player_name will come from either source, so it is optional. But, is useful as a double-check.

The player_bio is whatever you happen to have on the bio page, or the plaque page, or whatever you want. Basically, you decide what text you want to store in the database, so that when Shock sets up the database, he will simply plop out whatever it is that is already stored in the database.

Joe: this particular data is already in the BDB. If you guys have a simple rule as to who you have under consideration, then I can generate the list.

However, you also have Negro Leaguers, etc. So, you should supply the specific data for whatever I can't get out of the BDB, and the algorithm for anyone that I can get out of the BDB.

I can't can put that together Tango . . . I would say offhand anyone who has received a vote, and probably anyone with "X" number of games or innings would be a good start.

I will put something together in the next week or so.

As an FYI, my time is going to be somewhat limited until 2/22. I'm taking the CISSP that day, a 250 question, 6 hour test; it's the main certification for Information Security professionals, and a pretty tough test, so I'm going to be doing a lot of studying over the next 5 1/2 weeks. But it shouldn't be a problem for me to find some time to do this also, and (this goes for everyone) please don't hesitate to email me if something comes up here that needs my attention, I probably won't be checking in as often as I should.

Argh - that first line should say I CAN put that together, LOL.

"I would say offhand anyone who has received a vote, and probably anyone with "X" number of games or innings would be a good start."

Which reminds me... I'm going to give you the list of players that I have that received votes. I'll email it to you. Your job therefore would be to go through each player, and assign his BDB (or Retro) Player ID. So, give me a few minutes, and I'll generate that for you.

I would also say that we should include Hall of Famers that didn't ever receive a vote, there are a few.

Joe: for the HOF who never received a HOM vote, I can add those programmatically, so no need for you to worry about those.

Shock, please send me an email, as I want to send you a file: tom~tangotiger~net .

I need to track down Primer ID's for the following:

- B. Williams / B. Williams doubled to catcher
- Casey Elston
- ed
- Gadfly
- Philip
- redsox2004
- Sean (questionable)

A little help?

(I was AWOL longer than expected. I could say that

Re: HOM database

I presume that the database will somehow include the special ranking elections of 2008 (Groups 1 to 4 and eight "positionals") plus the pitcher ranking elections of 2009. "Somehow" means at least links to those 12-15 discussion, ballot, and results pages.

If I am right, that is enough to imply that the PLAYERS table should include fielding position --the official fielding position for every member. On my own desktop I would use '2B', 'LF', etc, as the key values; and '1', '2', '3', '4' for the Groups. (Official fielding position is not identical to primary_position_id by Tom T, nor can it replace that derived fieldpos unless the Hall of Merit assigns everyone in the database to a fielding position.)

I do not suggest that special elections data beyond internet addresses should be in the database, hence not that any special elections VOTERS (eg, yours truly) or VOTER_ALIASES be in there.

--
Re MLE database:

Chris Cobb,

Did you ever compile the MLE debut and finale years with text explanations at least for all departures and clarifications of Riley yearspans? ("clarification" where provides multiple spans or includes non-player roles inconsistently)

If there is no compilation by Chris, I am willing to help compile specific internet addresses for what is scattered in the Negro League player pages. Here is one re John Lloyd, for example,
http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/john_henry_lloyd#836947

There Chris says simply,
"I will modify the i9s projections as follows:

1) drop 1906 and 1931 from Lloyd’s MLE career"

(I/we would skim the remainder of the thread for more info.)

Shock: I can answer two of your questions:

Gadfly = 2321
Philip = 679

I'm going to speculate that the rest of those names never had a Primer ID.

The only one of B. Williams, Casey Elston, ed, or redsox2004 who is even in my 1921-present consensus score database is Casey Elston. I have a record of his 1921 vote, but when I went to the 1921 ballot thread here, I noticed two things: first, there are people in that thread referring to posts by post numbers and the numbers don't match so clearly some posts have disappeared entirely, and second, I can't find any post by Casey Elston, certainly not where it should be. Would posts by those who never got Primer ID's have been scrubbed from the record? That seems likely.

B Williams was a relatively recent voter. I don't remember what his name was before that or what it may be now, but he was definitely voting in the last couple years.

He wasn't dzop, was he?

And I have no memory of RedSox2004, but based on his name he must have voted post-registration as well.

Yes, B Williams doubled to catcher was 'zop.

redsox2004 is one of our early voters. He voted as redsox12 from 1902-1914, changing to redsox2004 for 1915-1917.

ed was one of our charter voters, voting 1898-1934, missing 1932. He voted as "ed phatyou", ed, and TheGoodSamaritan.

Casey Elston voted once, in 1921.

"special ranking elections"

In that case, the ELECTIONS table would need to be modified to include a "type" attribute.

Can someone list all the different kinds of elections and votings that you have had.

And of course, if you have a different point system for each type of election, and if there is an "elect-3" or not for these special elections, or it's simply an ordered list?

On "Sean (questionable)". Not sure what the "(questionable)" is about, though it could be 1917 when his ballot was not counted (originally no player comments, amended ballot filed later but never counted).

There is little doubt that Sean and Sean M were the same voter, though the question was never asked of him. Sean voted 1915-1918, his 1917 vote not counted. Sean M voted 1919, 1921-23, 1926, and 1934. The 1918 and 1919 ballots are nearly identical. Flick was 6th in 1918 and elected; in 1919, positions 7-15 moved up, and Jimmy Sheckard moved in at 15th.

The special elections were a passtime conducted in 2008 after the regular HOM elections "caught up". There were two kinds. One kind ranked HOM members by position (SS, 1B, etc.; pitcher ranking has not yet been conducted). The other kind ranked HOM-not-HOF members within their current HOF election groups (e.g. post-1943 VC). Both kinds ranked all N members of the particular election group and used a simple point system from N to 1, where N is the size of the election group. Example: say there were 20 HOM catchers, 1st got 20 points, 2nd got 19 points, ... 20th got 1 point.

Ok, for the moment then, let's ignore these "past-time" elections.

Now that the rankings are done, it would be interesting for the HOM to go back to 1939 and then vote out every player elected to the HOF (each year is a different election), with 75% required by HOM voters to bounce an unworthy HOF inductee.

Tango: Do you want me to update the CSV with information from 175-178 or are you cool?

Shock, it's best if we always have one system of record, so that would be your csv file, of which I'd get a copy.

Hey guys.

So I guess Joe is in the process of switching registrars so we can get the domain name hooked up. Until then, I have set up shop at one of my other domains:

http://www.steeko.com

Most of the plumbing isn't hooked up yet, but feel free to check out what I have and give some feedback/suggestions.

PEACE

Confirm: I am B. Williams doubled to catcher

I guess I should show my face here at least once. :-)

First of all, thanks to Shock and Tango for volunteering to do this for us. While the present Plaque Room has been fine for our purposes, since I can only do so much with BBTF's ExpressionEngine, it was limited as to what we wanted. What has been proposed here will bring us into the 21st century and will be a substantial improvement. To say that I'm excited would be an understatement.

If there's anything I can do, please don't hesitate to e-mail me.

I know John Murphy went through there and rebuilt many previously destroyed posts, but I don't know if he wasn't able to find them all, or didn't have the time to find them all. If it's the later, I assume we should be OK.

I actually had corrected them all, but another BBTF periodic makeover destroyed a lot of them again. I'll see if I can do anything with them again.

I know Chris Jaffe had saved many of the early ones, so he definitely should be contacted about them.

Most of the plumbing isn't hooked up yet, but feel free to check out what I have and give some feedback/suggestions.

Extremely cool so far, Shock! I'm even more psyched now. :-)

DanG's advice regarding Internet Archive is try, try again. Access to all, most, or any of some old sites including this one (BaseballPrimer.com) comes and goes. We find the dating inscrutable, not to say indecipherable. Navigating under a particular date for the site, there may be some links that go forward in time to later dates for some pages.

DanG's advice regarding Internet Archive is try, try again. Access to all, most, or any of some old sites including this one (BaseballPrimer.com) comes and goes. We find the dating inscrutable, not to say indecipherable. Navigating under a particular date for the site, there may be some links that go forward in time to later dates for some pages.

Yeah, I know what to do, Paul, but it's still a daunting task. :-(

jimd also helped me out years ago with the corrupted threads, BTW.

I just realized that the site renders poorly under IE.

So my apologies to those of you who opened it in IE and saw it all garbled, I'll try to fix it later.

Pub first. Website later. ;-)

Great work, by all, says a technological Luddite

:)

Regarding the old posts, I have gone through Internet Archive and copied and pasted the elections from 1898 though 1915 (I think, I'm at work) into Word files. If that helps anyone, email me by clicking on my name.

Why is the Constitution repeated? I've noticed this on this site, and it's also repeated on Shock's site. Is that intentional?

Thanks for pointing that out, HGM. That must have happened as a side effect during the last BBTF makeover. Surprised nobody mentioned it in all that time, though.

Anything happening with this lately?

Hey jimd - thanks for the reminder . . . I owe Tango a table. I had completely forgotten. I'm out of town until Sunday, but hopefully I can work on it Monday and get it to him next week.

I finally have enough free time that I can do some work on this again. I forget what was going on or whose court the ball was in, but I'm eager to make something cool.

probably something has changed at steeko.com (#184)

What I recall is a color or color and font combination that was difficult to read. I supposed that that was just what others called cool.

Sorry, I'm using steeko.com for something different now. The (fairly empty) site is actually at its proper domain -- hallofmerit.com

I have color blindness issues do if the colors suck just tell me what to use instead.