Ankit and fielding ratings?

[statfreak]

I recently began a 1983 solo league using the Ankit Database... several players imported with fielding ranges of less than 20 (on a scale of 1-100).

None of these made sense to me. I think Tom Brookens was a 4 at 3B...

Has anyone else noticed extremely low fielding ranges using the Ankit database and more importantly, has the database been "fixed" to get more realistic ranges?!?

I just exported the general ratings and add to 20 to everyone with a fielding range under 30, but that only fixed the primary position.

As far as the league... I took the Brewers and only made 1 trade, sending Bob McClure and Don Money to San Diego for Sixto Lezcano. The computer AI originally offered Lezcano for McClure, but I added Money to try to make it fairer. Anyway, Milwaukee won 101 games and the East, then defeated California 3 games to 2 in the ALCS. Milwaukee is currenly up 2 games to 1 over San Francisco in the World Series...

[Eugene Church]

While both the Lahman and Ankit databases are really fine...I always had to do a lot of editing players ratings to get them somewhat accurate.

Range, arm, speed ratings had to be edited on the defensive side...pitchers groundball percentages seemed to be random...and pitchers batting ratings really were way off...if strikeouts while batting were not kept by the league...the databases give that player a "10" rating...because in the database it said the player had "0" strikeouts.

I prefer Ankit's Career Average Database...I have to do fewer edits...and also you get a player imported in his first productive season and his correct position.

[Malleus Dei]

Originally Posted by statfreak

I recently began a 1983 solo league using the Ankit Database... several players imported with fielding ranges of less than 20 (on a scale of 1-100).

None of these made sense to me. I think Tom Brookens was a 4 at 3B...

Has anyone else noticed extremely low fielding ranges using the Ankit database and more importantly, has the database been "fixed" to get more realistic ranges?!?

Yes, it's a problem and I've noticed it, too. No, I don't know of a fix.

Why not ask Ankit about it?

[statfreak]

Originally Posted by Malleus Dei

Why not ask Ankit about it?

1) I asked in a public forum Ankit has been active in. It is not unreasonable to think Ankit is capable of reading this thread and responding to it.

2) If another person, besides Ankit, has or knows of a fix, it would certainly behoove me to ask the public at large.

3) If other people have noticed the problem and there is a fix, asking publicly helps the community at large.

[Henry]

Originally Posted by statfreak

1) I asked in a public forum Ankit has been active in. It is not unreasonable to think Ankit is capable of reading this thread and responding to it.

2) If another person, besides Ankit, has or knows of a fix, it would certainly behoove me to ask the public at large.

3) If other people have noticed the problem and there is a fix, asking publicly helps the community at large.

I think what Mal meant was to send Ankit an email or PM. It would get to him faster than waiting for him to chance across it here.

[Carlton]

Originally Posted by statfreak

1) I asked in a public forum Ankit has been active in. It is not unreasonable to think Ankit is capable of reading this thread and responding to it.

2) If another person, besides Ankit, has or knows of a fix, it would certainly behoove me to ask the public at large.

3) If other people have noticed the problem and there is a fix, asking publicly helps the community at large.

Actually Ankit has become disenchanted with the OOTP franchise, so I do not think he will be updating.

As for a fix I had a thread with my Ankit changes in the dynasty forum, but ONCE AGAIN...OOTP crapped out on me, and/or the OOTP6 bugs were driving me crazy that I stopped yet another historic replay.

[Cowboys3356]

I think the problem with the defensive ratings/running ratings importing badly has more to do with OOTP than with the database. Maybe OOTP7 will take into consideration these issues and the database imports will work better.

[Ankit]

Save the PMs and the e-mails... Fielding ratings for range and arm are given out randomly by OOTP, independent of stats from the DB. I asked Markus if he could do something about ot or help me improve the DB so that historical leagues are better but he never sent me the things he said he was going to, even after repeated pleas.

Carlton, yes I have become disenchanted with OOTP, "6" has made historical leagues too hard and less fun to play. Historicals is what I like to play so for me it was money down the drain.

[statfreak]

Originally Posted by Henry

I think what Mal meant was to send Ankit an email or PM. It would get to him faster than waiting for him to chance across it here.

understood, but my question wasn't not aimed particularly at Ankit, but the community at large.

My thread title should have included the word database after Ankit, I did not intend to direct the thread to Ankit.

[jdew]

Originally Posted by statfreak

Has anyone else noticed extremely low fielding ranges using the Ankit database and more importantly, has the database been "fixed" to get more realistic ranges?!?

I've done historical leagues since OOTP3 and yes, I've noticed this (among other things in an import... both from Lahman & Ankit DB's). The problem is not the DB, but the way OOTP imports. OOTP randomly assigns values for many player attributes on import ("seems to" in some cases, "definately does" in other cases).

As Ankit mentioned, fielding and arm are random, or at least sometimes way off. Also random are the hitter type, leader, clutch and consistency ratings. All players import with injury rating of normal and all catchers import with a PB rating of 4. Running speed, stealing ability, running instincts, bunt for hit are suspect or at least sometimes wacky. The sacrifice bunt rating may be random (sometimes it looks right, other times it doesn't), though if stats aren't available everyone gets 3 for bunt and 1 for bunt for hit (on the 1-5 scale). Pitch type is random but isn't of much importance unless the pitcher was a knuckleball pitcher then you'll need to make sure that the pitcher has that pitch (Marcus stated someplace that knuckleball pitchers get a slight adjustment.) While I haven't yet examined it in any great detail, the pitcher's Hit Batsmen, Wild Pitch and Balk ratings may be random. The gound ball % and hold runners ratings also appear to be random. Also, I'm not sure how Velocity imports, but that sometimes looks a little wacky too.

Another, new, problem related to historical leagues is how OOTP now handles pitchers. Many relief pitchers import with good Stuff, Control, Velocity, etc and with Endurance ratings of 6 or greater. The OOTP AI will convert those pitchers to Starters since it evaluates them as better than the "real" starters. So I've also been editing relief pitchers endurance ratings to get them to stay in the bullpen for at least a season or two (until OOTP "develops" their endurance back up to starter range.) This probably isn't too noticable for more modern era imports, but its really noticable for imports in the 1920's, 30's, and 40's.

What this means is that to get the best accuracy there's a lot of research then editing of the initial import and then again for each rookie class imported.
This isn't necessarily a bad thing, since the research (call it "scouting") will give you an awfully good idea of each player's potential.

[Malleus Dei]

Originally Posted by Henry

I think what Mal meant was to send Ankit an email or PM. It would get to him faster than waiting for him to chance across it here.

Exactly, but it was somehow taken the wrong way.

[Malleus Dei]

Originally Posted by Ankit

Save the PMs and the e-mails... Fielding ratings for range and arm are given out randomly by OOTP, independent of stats from the DB. I asked Markus if he could do something about ot or help me improve the DB so that historical leagues are better but he never sent me the things he said he was going to, even after repeated pleas.

Carlton, yes I have become disenchanted with OOTP, "6" has made historical leagues too hard and less fun to play. Historicals is what I like to play so for me it was money down the drain.

Ankit, I hope Markus reads what you wrote.

[Ankit]

Running speed, stealing ability and running instincts are assigned via calculations that OOTP does. Markus was going to send me info on this a few months ago but hasn't yet...

[jbmagic]

you guys have any tips on getting a player range rating and arm rating in ootp 6


i know at http://www.baseballreference.com i can get there fielding % for any player, but how you firgure out what rating to give for Fielding range and arm range?



do i need to do any editing still with ankit database?

thanks

[Eugene Church]

Arm is very hard to determine...Range can be done fairly well. Even using Baseball-Reference, you can only get some idea of these...and sometimes these stats are misleading.

As you might guess it is very time-consuming and tedious to do this.
I have done it 3 times for my historical leagues...once at the public library using Total Baseball and the Encyclopedia of Baseball...then I discovered Baseball-Reference.com...which made it much easier to do the research.

If an OF has a great arm, teams don't run on him, therefore his assists will be lower than someone who has a poor arm and everybody runs on him, therefore he will look good statistically, but in reality, be poor. That is, the poor arm will have many more chances for assists, but a very poor percentage. This also applies to catcher arms.

Sometimes in the Baseball Era you have fewer strikeouts, therefore more balls in play and more assists and putouts, so it is hard to compare players of different eras...or teams run more on the bases, therefore more outfield assists...teams bunted more, therefore catchers, 3b and 1b had more assists.

You really need to compare players against their contemporaries to be somewhat accurate.

I use Putouts to determine the Range of an OF...this seems to be a pretty good determinant. I divide the games scheduled by the number of games played and multiply this by the total Putouts. If a player plays 100 games with 100 PO in a 154 game season ...it would be 154 (games scheduled) divided by 100 (games played) , then multiplied by 100 (putouts)...his rating would be 154 putouts per season.

an OF with 100 games played with 200 putouts would be rated at 308 putouts and be considered much better than the first player.

an OF with 140 games played with 375 putouts would be rated at 412 putouts and be considered much better than the previous two players.

This formula allows you to compare a player who play 50 games to those that play 100 or 154 or 162 games. It's not perfect, but it give a pretty good indication.

Then you have to set up a range to assign ratings in OOTPB. Basically I used this range to assign OF range ratings: over 350=A or 5 on a 1-5 scale...325-375=B or 4...290-325=C or 3...260-290=D or 2...below 260=E or 1.

I check assists to get Range for infielders...although as stated previously...assists may vary by the era...I used the same formula used for outfielders...Games Scheduled divided by Games Played, then multiplied by total assists.
To assign OOTPB ratings for 2b and SS, this scale was used: over 475 assists per season=A or 5 (on 1-5 scale)...450-475=B or 4...425-450=C or 3...400-425=D or 2...under 400=E or 1.

3b scale: over 325=A or 5...300-325=B or 4...275-300=C or 3...250-300=D or 4...under 250=E or 1.

1b scale: over 125=A or 5...110-125=B or 4...90-110=C or 3...75-90=D or 2...under 70=E or 1.

C scale: this is the hardest position to do...for range I use the player's Rfg rating in Baseball Reference and compare it to the LGrfg (league average for the postion)...also look at assists, double plays and passed balls per season.
It would be nice if "runners thrown out" stats were kept like they currently are...but this is only recently.

For catcher's arm, I look at assists and double plays per season...although this can be misleading as a poor arm might have more TOTAL assists because teams run on him so often and give him more chances to increase his totals.

Pitchers: very hard to compare eras...use assists and double plays to as some indication...mainly go by fielding percentage.

This may help you some...but be aware it is a lot of work.

[jbmagic]

wow thanks so much