***VD 13 Spreadsheet***

[mjl]

yeah I just used Lahman because it was trivial to download instead of scraping all of bbref (I actually already have a script to do that but I feel like they'd get mad if I ran it). And they use the same player IDs, don't they? Anyway, it's okay to use it for development, and Firestore (the Google Cloud Storage solution I used) lets you write Java objects as db entries directly which is convenient but meant that I'd have to do some parsing regardless of what form I got the data in.

[Kevin Seitzer]

And they use the same player IDs, don't they?

99.99% yes

[Kevin Seitzer]

Latest version of spreadsheet posted.

1. Fixed the aptly named 'pos' worksheet.
2. Updated some stats for 1870's hitters.

[johnnya24]

Latest version of spreadsheet posted.

1. Fixed the aptly named 'pos' worksheet.
2. Updated some stats for 1870's hitters.

I updated my shared file with your latest file. Both links contain the same version

[Kevin Seitzer]

The basics of how the 'pos' worksheet functions, for future reference (Johnny):

1. The tables in columns O:BN pull from the identical tables on the individual team worksheets. (In addition there is a MI column that is a 1 if either SS or 2B is a 1.)
2. The overall table in C2:K16 pulls from the labels in the light-green-colored columns V, AE, AN, AW, BF, BO. These columns are manual entry and need to be updated as the draft goes along.
3. Column M contains warning flags that show up red if there are hitters who have been selected by a team but no corresponding position label has been entered in the light-green-colored position-entry column for that team's player.
4. The overall table will highlight in blue once that position's quota is filled.

[johnnya24]

I recall trying to figure this sheet out on many an occasion!

How do you account for multi position players? Especially things like OF/1B eligibility.

[Kevin Seitzer]

I recall trying to figure this sheet out on many an occasion!

How do you account for multi position players? Especially things like OF/1B eligibility.

That's why it's manual entry. I couldn't figure out a sensible way to automate that. Whatever you put into the light green columns determines how the overall table tabulates things. If you screw up the position labels, it won't know any better. It will highlight obvious errors like listing three catchers for a team, but for the more subtle stuff you have to choose for each team whether they are putting, for example, Ed Delahanty at OF or 1B. In this update, I put Delahanty at OF (of2), but if mjl needs to move him to 1B later in the draft, that would need to be manually changed in cell V41.

It's not elegant, and it's not fully automated, but it does work if you keep the position labels updated.

[Kevin Seitzer]

Ah, and I discovered another little fix I need to make to that sheet. We have two UT players this draft, correct? So I need to change the conditional formatting for the UT column (Col K) to highlight blue on two UT players instead of one UT player.

Edit: Fix is uploaded.

[johnnya24]

What I do is colour-tag the position column next to each team on the Snapshot - dark grey for a "certain" position, and Light Red for an "uncertain" position ... typically only CI/OF eligibility. C, 2B, SS, 3B = always grey even if a player has multiple positions); 1B/OF = both 1B and OF will be given a red tag.

It's actually really fast and simple to accurately label each team with this method.

[Ken]

I think I can automate it in excel, I'll give it a shot this weekend.

[mjl]

I don't know if there's some chance this is a more common problem, but Bobby Abreu's career OBP is 21 points lower than it is in the spreadsheet. He's credited with 1840 BB but that's actually his career strikeout total, it should be 1476.

spot checking, that's the only one I saw among the better careers ending in 2014-2016. This seems like something that would only happen with bad transcription so it seems plausible that it's a one-off.

[johnnya24]

I don't know if there's some chance this is a more common problem, but Bobby Abreu's career OBP is 21 points lower than it is in the spreadsheet. He's credited with 1840 BB but that's actually his career strikeout total, it should be 1476.

spot checking, that's the only one I saw among the better careers ending in 2014-2016. This seems like something that would only happen with bad transcription so it seems plausible that it's a one-off.

Good call

Abreu Updated .. just as well no-one picked him expecting that inflated OBP.

That said, if someone does pick him, can we all remind them that his stats have been changed for the worse.

[Ken]

I don't know if there's some chance this is a more common problem, but Bobby Abreu's career OBP is 21 points lower than it is in the spreadsheet. He's credited with 1840 BB but that's actually his career strikeout total, it should be 1476.

spot checking, that's the only one I saw among the better careers ending in 2014-2016. This seems like something that would only happen with bad transcription so it seems plausible that it's a one-off.

*** I want to note that I'm NOT complaining, but just noting so it can be fixed ***

I see quite a few errors in the spreadsheet. Here's what I found for players that have been picked through the first 81 picks (and where relevant the year that they were picked). Can someone spot check me on a few of these to make sure I'm not crazy or reading the spreadsheet wrong?

Babe Ruth had 168 rbi in 1921 not 171
Lou Gehrig had 173 rbi in 1930 not 174
Cap Anson had 48 rbi in 1872 not 50
Ty Cobb had 1944 RBI not 1933
Dan Brouthers had 1301 RBI not 1296
Rudy York had 1149 RBI not 1152
Hank Greenberg had 1274 RBI not 1276
Al Simmons had 1828 RBI not 1827
Eddie Collins had 1299 RBI not 1300
Babe Ruth had 2214 RBI not 2213
Ed Delahanty had 456 sb not 455
Dan Brouthers had 257 sb not 256
Dan Brouthers had 1529 runs not 1523
Hank Greenberg had 1046 runs not 1051
Hugh Duffy had 1554 runs not 1552
Mickey Mantle had 1676 runs not 1677
Ty Cobb had 2243 runs not 2244
Hugh Duffys Career OBP is .386 not .384
Kid Nichols had 362 W not 361
Cy Young had 18 Sv not 17
Kid Nichols had 16 Sv not 17
Dazzy Vance had 12 Sv not 11
Ed Walsh had 35 Sv not 34
Warren Spahn had 28 Sv not 29
Lefty Grove had 54 Sv not 55
Hoyt Wilhelm had 228 Sv not 227
Babe Adams had 16 Sv not 15
Christy Mathewson had a 0.827 WHIP in 1908 not 0.837 (Due to 281 H instead of 285)


There are obviously many more for players/years that have not been picked which may be in error. I wonder how many there will be by the end. Most of these are off by 1 or 2 which makes me think someone is going back and correcting stats in bbref since the time when the spreadsheet was created. As we've discussed before I presume that bbref stats win when there is a discrepancy (i.e. I know we have gone back and fixed Deacon White which had been updated)

Also, do I win a prize for most mistakes found?

[mjl]

Yeah, this is because bbref updates old players occasionally and we don't usually make a point of rechecking the numbers for old players. It wouldn't hurt if people got used to checking bbref and updating stats when a player they're taking has a discrepancy.

[mjl]

Another minor change to make to the spreadsheet:

The only 2010s BY for Lance Berkman in the spreadsheet is his terrible 2013. Replace it with 2011:

0	0	berkmla01_2011	berkmla01	Berkman	Lance	2011	488	147	31	94	90	2	92	3	4	587	0.412	2010