How accurate are ratings in Ankit's Career Average Database?

[AngelinOF]

Originally posted by Eugene Church
Ankit may be too difficult for a computer illiterate like me.

If you can use the original Lahman then you would probably have very little trouble with either Ankit's database. All 3 are basically the same, the only difference is that they change a players' debut year and the career average db averages out the players career as opposed to the player having ratings based on his rookie season.

The only trouble would be duplicate players for the first couple of seasons and depending on how you are assigning players. If you are having an amateur draft each season then finding these duplicates is fairly simple, you would just need to display all players in the league and then sort the players based on name, looking for players that don't have a team listed (since those would be the imported players 95% of the time) and look to see if any players have the same name. If you are placing the players on their original teams then finding the duplicates is a little more time consuming. You could probably e-mail Ankit to see if he could provide you with a list a players that had their debut years changed.

[JimiP]

When i decided to bring the WTL into existence,
WTL OUTPUT

i chose Ankits Career Db for several reasons. It is a 'career based DB, which has the players appearing at somewhat different dates fromthe normal Db - this was good for me because it added to the distinctive flair of the league and it also allowed abit more uncertainty to the ammy drafts. have found it works very well in that respect.
Certain players who only played a few games in early years of their career are delayed until they played significant amounts of bball an dthen are introduced into league.
I have found the ratings of positions certainly more accurate because of this factor than the Lahman DB - as we all know certain players came up as cathers for a week and then they actually played their career as 2bman - or whatever - this 'career' DB has reduced considerably the amount of time editing or 'correcting' positional faults in imports.
As for hitters and pitchers - they are abit more advanced because of the age and experience factor that is inherently interwoven inthe Ankit Db - the players are older, more experienced and therefore their ratings and abilities reflect this.
I think that this DB is a great alternative to the standard one - but, as with all things, it has certain limitations and 'faults'. These I have mentioned briefly above and fi anyone is interested i could go in to more detail - tho not sure how much i could offfer.
Scott Vibert also did an edit of the Lahman Db making some adjustments and stuff - i have used it also and it is very nice - eliminating some of the the positional and other 'faults' in it.
I beleive that you must figure out what type of league you want - and weigh the pluses and minuses of all the available DBs and choose accordingly.
Be aware of the fact that if you are switchingover youmay get some duplication and other things you may not expect.

[birdsofprey02]

How about the whole reason this thread was made anyways... Is anyone coming along with a trade deadline roster? I was about to do all the transactions myself, but I don't have the time right now. I'm about to go back to school and I wanted the rosters on the laptop before I went back.

Thanks,
Mike

[GForce22]

Originally posted by adventureran
Henry,

I don't think this gets said often enough. It should be a part of the "packaging" of OOTP. It was not designed to accurately recreate history. Many, especially in historical leagues, complain that "my player should have better ratings" or "why is that bum playing so well? he didn't play that way IRL."

To get this point across, the booklet should include the link to whoever ran the dynasty thread with the offensive juggernaut that was Joe Orsulak. I remember somebody ran a historical league and Orsulak was like a triple crown winner. Funny stuff.

GH

[Eugene Church]

Would someone please answer my original question?

How accurate are the ratings in Ankit's Career Average Database?

That is, does it require much editing to get accurate career stats and ratings? Or does it come complete with all players ratings and with a high degree of accuracy (90% or better)?

In Ankit will I have to edit fielding ratings (range, arm, percentage), speed and stealing ratings and outfield positions?

Will all the various ratings be at least 90% accurate?

(I know once I start a league with career averages, that OOTPB will develop the players as it wants and I will not necessarily get historically accurate performances. I just want to begin with accurate career averages for my Alltime Allstar Leagues).

[JimiP]

EC:
Below is a quote from Ankits himself with regards to how his Career DB was constructed......i beleive it will assist in answering your question......

As far as fielding positions, I summed up all games player for each player at each position. Then if a player played less than 10% of his total games at a certain position, the data for that position is deleted and the career average for the remaining position is all this left. So, every player will be rated for all position at which they played at least 10% of the games there and their total data is averaged by the number of seasons played in the majors to get per season data.

as for my experience with it - i have had little to no work on redoing the positions and arms of the players which i have imported into my league. I think it is one fine database and allows us to import players/rookies with little or no editing on our part -

hope this is good enuf for ya ...

[cooper_gd]

Eugene - I've had my fist experience with Ankit's DB the last several months. On a whole, I am MUCH happier with it than I was the original Lahman. Players do debute a little later, but their rating are much more accurate - from what I have been able to tell.

I am in 1922 of a league that I started in 1903. So far, the names that are the "best" were the RL best. Stats are being reproduced pretty well. Ty Cobb has 3300 hits at 35yr old (just under .360 career average) - Walter Johnson has 330 wins at 34. The Babe, in 1921, hit .390 44 128. Christy Mathewson retired with 318 victories (I drafted in 1903 - that is darn close to his RL 374 (1900)).

I do not check every player, but I did check most of the familiar players... I stopped feeling that I had to do that after 10 sims tho' - I just never needed to change anyone's ratings.

Some others off the top of my head . . . Joe Jackson .361 career average, Eddie Collins .340, George Sisler .344 . . . Tris Speaker .344.

Walter Johnson 332 wins, 2.38 ERA at age 34. Hoyt, Pennock, Colveleski, Shocker and others are doing well as well.

Only one obvious player stands out otherwise... that is Sam Rice. In 1922, he has not even broken into anyone’s starting line-up. His ratings are ok 6/2/8, but not enough to make him more than 2.5-3 stars and be a platoon player.

Hope this was the info type you were looking for.

Oh, also, early pitchers almost always imports with MR and SP endurance - kinda a fix for the whole endurance issue that everyone has commented on.

Players are rated at all positions that they had significant time at during their careers.

Ratings are based on career, not rookie season... also, Ankit gives good instructions on how to start using multiple databases for the initial draft and then for every seasons import to give the most accurate results..

[Ankit]

Originally posted by Eugene Church
Would someone please answer my original question?

How accurate are the ratings in Ankit's Career Average Database?

That is, does it require much editing to get accurate career stats and ratings? Or does it come complete with all players ratings and with a high degree of accuracy (90% or better)?

I will let others answer this one because if I answer then there is an obvious bias.

In Ankit will I have to edit fielding ratings (range, arm, percentage), speed and stealing ratings and outfield positions?

Will all the various ratings be at least 90% accurate?

The fielding data should be and is at least 90% accurate. The only exceptions here would be if in RL, a player had a career fielding percentage of .970, then due to rounding errors, he might import with a %age of .964 or something like that, but that is completely a rounding issue, nothing else. The reason for the rounding error is this: players fielding data is average per season and OOTP calculates fielding percentage solely on one season's data, not the sum of all seasons. All other fielding data is as accurate as possible. The Arm ratings are randomly given by OOTP, there is nothing in the Lahman DB or AnkitDB that can help OOTP assign proper ratings. Also, there is a lot of subjectivity involved and stats can not tell you accurately how strong the arm of a player is. Stealing ratings and speed ratings should be accurate but again OOTP doesn't have a good system for assigning speed and stealing ratings. In general though, the speed and steal ratings are highly accurate. One thing I have noticed though, Ty Cobb will import always with a speed of A, but his stealing rating jumps around, usually A or B and in rare cases E. This is due to an error in the import process of OOTP and not a DB problem. This issue is present in both Lahman DB and AnkitDB.

(I know once I start a league with career averages, that OOTPB will develop the players as it wants and I will not necessarily get historically accurate performances. I just want to begin with accurate career averages for my Alltime Allstar Leagues).

Also, my readme file in AnkitDB 5.0 has all the details on how I created the DB, that should help answer some of your questions as well.

[Eugene Church]

cooper gd:

Thanks for the useful info. You have direct knowledge using it, so this will really help me.

Ankit: I really appreciate your reply...and I really do appreciate you providing this database.

[hamonrye]

I'd like to give an enthusiastic endorsement for Ankit's career average database. I'm in a historical replay that began in 1903 and is currently in 1972 and I've found that it is much more reliable as far as an accurate historical reprsentation than the original Lahman database. Of course, you won't get a straight historical replay. There will be players you've never heard of who, on occasion, have outstanding careers, but I've always felt that that was the beauty of OOTP. There are games you can buy if you want total accuracy, but OOTP gives you an alternate version and the surprise of what will happen with certain players is part of the enjoyment (for me, anyway).

Looking at my career leaderboard, Ty Cobb is the all-time hits leader and Henry Aaron is the all-time home run king, with the usual suspects all over the leaderboard in every category, so I think it's safe to say that Ankit's database has worked fairly well for me.

[Eugene Church]

Thanks to all of your replies. I appreciate it.

JimiP: Sounds like "it's good enuf" for an old, obsessive compulsive ratings perfectionist like me.

I just want to start the league with players ratings as accurate as possible, realizing the their performances may not mirror Real Life.

Does the Ankit go through 2002?

To use Ankit's Career Database, do I just download it like the Lahman to the C drive and not in the game folder?

Does it replace the Lahman or work in concert with it?

Will I go to the "free agents" screen and import players with the source path "c:\statsankit" or some similar path?

Currently I use "c:\stats".

[Ankit]

Eugene, the stats in my DB do go through 2002. Also, I recommend you use AnkitDB in concert with Career Avg DB. Create a historical using AnkitDB, and then import subsequent years using the Career Avg DB. The benefit of this is that you will get actual player stats for the years they played prior to your starting year and their fielding data is based on career numbers so you get the best of both worlds.

The way I use it, and this is the way I have given directions for in the readme file for the Career Avg DB:

I have AnkitDB in C:\stats2
I have Career Avg DB in C:\stats

[Eugene Church]

Ankit:

Thank you for the help and the Career Database.