Bruce,

I've been playing around with ADP_USER and standard ADP. What I'm finding with my leagues is that the experts are going on value, but the not-so-experts are still sticking with Magazines, etc.

There was an earlier thread from Anon where it was pointed out that even if you import ADP_USER, if you have some owners picked out as non-experts, they will follow the appropriate ADP, where as the experts will follow the ADP_User.

However, when I import ADP user, then go back and look at the ADP.csv file, ADP_Expert hasn't changed.

Here's my question, if I import ADP_user, and then set owner ADP draft preferences randomly in the setup area, what settings take precedence for the calculated dropoffs? ADP_User or ADP.csv? Or does it take both into account?

I'm trying to figure out how to best incorporate most of my data yet take into account the randomness of some of my other owners. I think figured out how to manipulate all the ADP's in EXCEL but I need to know how the data is actually incorporated.

Thanks.

Save your adp.csv. Then do an export of ADP, I belive this will create a new adp.csv based on the currently loaded settings (from the players.dat I think).
I believe anytime you import adp.csv or adp_usr.csv it consolidates them in the players.dat file. leaving the original files untouched. It gives you the option of re-importing adp.csv to remove your adp_user.csv changes (assumming you didn't previously do an export).

Here's what I've found:

ADP appears to be tied to the league.

I opened up a brand new league (Myleague) and see the STD ADP settings. Open up my league with the Custom ADP and the custom ADP sticks with the league.

I do get a latest projections to reset the ADP. Copied the ADP to another file.

After I import ADP_USER and the export the ADP file the ADP file looks nothing like the original file.

I've got all owners drafting by magazine average except me.

Brady is 1.06 in the Magazines and I had him as 2.03 (first time a QB went off the board last year) - in the mocks he keeps going at the 2.03, not the 1.06 I'd like to see him go out. I also see that the dropoffs match the ADP_USER file, not the magazine file.

The other thing I noticed -- took me a while to figure out -- if Brady is #06 - but Barber is listed as #06 in ADP User, the DD skips them both and goes to 1.07 (Gore).

Can't figure this one out. I'd like to be able to utilize the combination of ADP_USER and Magazine or other ADP.

If I update projections and bring in a new ADP then Brady goes #06 again.

Also if I haven't been able to import the copied ADP.csv without errors. I'll continue to play with it, but I don't think I'm going to be able to get it to work like I'd like.

Bruce, any hints or ideas?

3 hours later - an answer.

What I did was update to bring in the latest ADPs and overwrite the ADP_USER file. I then brought in my PD file to align my projections up with the standard ADP's. Then I exported the ADP and renamed it for future use.

Then I imported my ADP_USER and the exported the ADP again and renamed it too. Here's what I found when comparing the files.

When ADP_USER is imported it doesn't just overlay the ADP_Expert from DD.

First it renames the first column, from adp_expert to adpExp. I'm going to assume that's the majority of the problem I was having.

The 2ND issue is that if you are importing from an ADP_USER, the odds are that you are not importing an entire 250 player list. I have a 12 team league, 14 round draft - 168 players in my ADP_User. But when I was cutting and pasting I was leaving off rankings for the bottom 82 players. Unfortunately since the sheet is in alphabetical order, this caused some mismatch problems.

While checking this out I noticed that at the very end of the file, all of the kickers and defenses were stuck to the end -- in a semi-alphabetical order:

Seahawks Seattle Def SEA/4
Stover Matt PK BAL/10
Suisham Shaun PK WAS/10
Buccaners Tampa Bay Def TB/10
Vinatieri Adam PK IND/4
Redskins Washington Def WAS/10

Finally, when I carefully compared files I found that the ADP export of the standard ADP and the ADP export of ADP_User were naming the DEF's differently. Again, when I tried to merge, this caused some problems.

From the ADP_USER Export:
lname fname pos team adpExp
San Diego DEF SD 53
Minnesota DEF MIN 60
Chicago DEF CHI 75
New England DEF NE 76
Dallas DEF DAL 83
Pittsburgh DEF PIT 90
New York Giants DEF NYG 98
Jacksonville DEF JAC 112
Green Bay DEF GB 113
Seattle DEF SEA 115
Indianapolis DEF IND 130
Baltimore DEF BAL 140
Tampa Bay DEF TB 144
Philadelphia DEF PHI 150

From the ADP Export:
lname fname pos team adp_expert
Bills Buffalo Def BUF/6 204
Panthers Carolina Def CAR/9 218
Bears Chicago Def CHI/8 97
Browns Cleveland Def CLE/5 232
Colts Indianapolis Def IND/4 147
Cardinals Arizona Def ARI/7 244
Cowboys Dallas Def DAL/10 111
Broncos Denver Def DEN/8 249
Packers Green Bay Def GB/8 139
Jaguars Jacksonville Def JAX/7 123
Vikings Minnesota Def MIN/8 84
Patriots New England Def NE/4 99
Giants New York Def NYG/4 121
Jets New York Def NYJ/5 242
Titans Tennessee Def TEN/6 180
Eagles Philadelphia Def PHI/7 179
Steelers Pittsburgh Def PIT/6 113
Raiders Oakland Def OAK/5 235
Ravens Baltimore Def BAL/10 149
Chargers San Diego Def SD/9 83
Seahawks Seattle Def SEA/4 143
Buccaneers Tampa Bay Def TB/10 159
Redskins Washington Def WAS/10 210

I found all of these issues but did not test the impact of each individually - I basically fixed them all at once.

To actually get an ADP file that used the ADP_USER along with the standard ADP's I did the following:
Added a # column at the end of the sheet and numbered it top to bottom 1-250. this was to keep everything in order so that I could put it back into DD as it came out. Especially with the weird alphabetical problem at the end of the list.

I took each file and sorted them by rank. Note that when I refer to the ADP_USER file - I'm refering to the ADP file I created by exporting the ADP_USER file from DD.

Took the players ranked from 166 to 250 from ADP and copy pasted them to the end of the ADP_User file to ensure all players were captured.

I took each file and alphabetized it.

I inserted 5 columns into the ADP file (after adp_expert) and pasted in the first 5 columns of the ADP_USER file.

I then went row by row to ensure all the alphabetizing matched up, with the ADP file taking precedence. Getting the DEF's right was a bit of a hassle, but I got them right.

I saved the file as a master in case I made a mistake (learned from my mistakes)

I deleted the column with labeled adp_expert along with any other unneeded columns I'd copied over.
Renamed the adpExp coulmn to adp_expert.
Resorted the the entire sheet based on the # column (from above) and saved the sheet. I then did a save as adp.csv and imported it back into DD.

The mocks ran perfect.

GoBears84 - Thank for the detailed explanation.

I took a look at mine more closely and noticed that it also stopped at 192 players(12tx16R) Also, a number of players had 9999 listed as the ranking for each one.

At first I couldn't get the ADP view with the Team and bye, but realized I had to reopen the league "Fresh" and update the rankings, then export. As soon as you click SAVE, the adp file is sorted by the adpExp (adp_expert) and removed from Alpha order.

The ADP list is running 251 players today, so I added the extra positions to my adp_user list just so they had the same number and exported.

I did it slightly differently, but appear to have gotten the same results.
1)Updated ADP and Saved league so it was formated as we normally see it.
lname fname pos team adpExp adpMFL adpMock adpCBS adpMag
2)Exported, Renamed and saved adp so I wouldn't lose it...
3)Imported current user_adp file with the additional 59 positions added.
4)Exported, Renamed and saved the user_adp so I could work with it.
5)I opened ADP and sorted the columns by last name and first name. Then copied the 5 user_adp columns (sorted) as you did to match up names. I found a few different names in my list that I pulled out keeping their user_adp number with them. Just inserted or deleted rows to keep them lining up. Most were at the tail-end of draft picks, after 238. One notable one missing from my adp_user list was Shaun Alexander as he's a FA so never was included I guess. I just copied the names from adp across into the empty rows in the adp_user list and assigned the ranking numbers from the ones that I pulled out above. Won't really matter but then everything matched up.
6)Copied my revised adp_user list column and pasted into the adp_exp column.
7)Deleted the additional adp_user columns I had inserted so it looked exactly the same as a standard ADP file.
8)Saved as ADP.csv AND renamed and saved so I would have backup.
9)Imported as ADP.csv file. (NOT the user_adp)
10) Same result as you, players are being picked accordinly to ADP list for Mocks.

Pretty much the same as you did, but worked with the saved ADP file so I wouldn't have to match up the odd sorts on DEF. Seems to work as long as it's standardized and you have an equal number of players to the ADP file.

Thanks again for the info on this. Gave me a reason to mess around for a couple more hours with fantasy football this evening instead of some work that I really didn't want to do... ;>)

I played with it during a staff meeting a found the same thing you did..I could use the two exported files and they matched up better -- though the ADP_USER export was still a few players short. I think the problem I encountered with DEF's was because I used the actual ADP file that gets updated with DD updates.

I just want to pimp this a bit...

Last couple of years I've been using ADP_USER exclusively. I noticed this year that my ADP User was significantly different than standard ADP.

I combined the two ADP's using the steps above, ran a mock with the new ADP with each owner set accordingly. Then I took the results from the mock and turned those into ADP_User and used it for my custom dropoff calculator.

I was pleasantly surprised to see that I was within 1 player at each position for the first few rounds. A run on TE's was not expected, but I was still able to grab Gonzalez in the 6th.

I haven't done the full analysis, but I think that this is the way to go in the future...