 |
View Old Forum Thread
Old Forum Index » Baseball Beta Testing » Beta News » More Errors
Admin
Notably, I have added back in the code where a 1B can get an error on a throw from another player, as 1B were getting far too few errors (about 76 per season per league, whereas MLB would be at about 315 per 16 teams per season).
Let's take a look at error totals during the season and at the end.
Other things to look at:
- What problems were created by moving the 2 teams to Expansion cities?
- Is Abe tracking Active and Injured days correctly?
- Is Abe giving out Batter ICs in the correct amount based on the new formulas?
Tyson
More Errors
October 02, 2008 at 08:30PM View BBCode
Based on some research done by celamantia, we have far too few errors in Sim Dynasty. I have made some adjustments to the error logic which should at least double the number of errors in some cases.Notably, I have added back in the code where a 1B can get an error on a throw from another player, as 1B were getting far too few errors (about 76 per season per league, whereas MLB would be at about 315 per 16 teams per season).
Let's take a look at error totals during the season and at the end.
Other things to look at:
- What problems were created by moving the 2 teams to Expansion cities?
- Is Abe tracking Active and Injured days correctly?
- Is Abe giving out Batter ICs in the correct amount based on the new formulas?
Tyson
celamantia
157 per 16 teams per season. The data I sent you is for two seasons.
October 02, 2008 at 09:51PM View BBCode
Originally posted by Admin
whereas MLB would be at about 315 per 16 teams per season).
157 per 16 teams per season. The data I sent you is for two seasons.
Admin
Tyson
October 02, 2008 at 10:21PM View BBCode
Oops, let me work on that a bit before I turn on the next season.Tyson
Closer
October 02, 2008 at 11:42PM View BBCode
Is the increased errors going to just be figured to increase for every team proportionately to fit the 315 number or will it be tweaked to give more errors to the less capable firstbaseman in a year?
Admin
1B will have a slight chance (about 2 in 1000 chance) of making an error on fielding a throw from another infielder.
Catchers will be more prone to errors (about twice as many errors), and more prone to making a bad throw during a steal attempt as these numbers were too low.
SS and 3B will be generally more prone to make errors (about twice as many).
1B will be about 1.5 times more prone to errors.
I have left P, 2B, and OF alone as these seemed to be about right. There is a discrepancy with pitchers, but it is due to us not having pickoff throws and errors on those pickoff throws.
Tyson
October 03, 2008 at 03:34PM View BBCode
Ok, I have made a few adjustments. Errors are always based on the defensive competence of the players involved.1B will have a slight chance (about 2 in 1000 chance) of making an error on fielding a throw from another infielder.
Catchers will be more prone to errors (about twice as many errors), and more prone to making a bad throw during a steal attempt as these numbers were too low.
SS and 3B will be generally more prone to make errors (about twice as many).
1B will be about 1.5 times more prone to errors.
I have left P, 2B, and OF alone as these seemed to be about right. There is a discrepancy with pitchers, but it is due to us not having pickoff throws and errors on those pickoff throws.
Tyson
ScooterPie
Something just occurred to me, though: in your MLB data, were you able to look at how many bases the errors give up? I seem to remember that Chris broke it down by fielding/throwing, but I've noticed that in Sim Dynasty, every infield throwing error is two bases, as is every outfield error. All infield fielding errors are one base. Of course this isn't the case in MLB, but maybe the differences amount to a wash.
I think your increasing the number of errors will get us closer to reality no matter what, because it was off by so much. Just wondered whether you can check the number of bases to make sure it's that much better.
scooter
October 04, 2008 at 11:57AM View BBCode
I'm really glad to see this change. Thanks.Something just occurred to me, though: in your MLB data, were you able to look at how many bases the errors give up? I seem to remember that Chris broke it down by fielding/throwing, but I've noticed that in Sim Dynasty, every infield throwing error is two bases, as is every outfield error. All infield fielding errors are one base. Of course this isn't the case in MLB, but maybe the differences amount to a wash.
I think your increasing the number of errors will get us closer to reality no matter what, because it was off by so much. Just wondered whether you can check the number of bases to make sure it's that much better.
scooter
Admin
Tyson
October 04, 2008 at 02:18PM View BBCode
Hmmm, I will try to look at this next week. I know one thing I wanted to do is that if an OF makes an error with a guy on 1st or 2nd, it should be only a 1 base error.Tyson
tworoosters
Through 15 games none of my positional players are on track to earn more than 40 off season improves. I have only one platoon so this seems very low to me.
Glancing at other teams it appears to be the same everywhere.
Improvement Chances
October 04, 2008 at 04:54PM View BBCode
Originally posted by Admin
Other things to look at:
Is Abe giving out Batter ICs in the correct amount based on the new formulas?
Through 15 games none of my positional players are on track to earn more than 40 off season improves. I have only one platoon so this seems very low to me.
Glancing at other teams it appears to be the same everywhere.
Admin
I should weigh this back in during the season so players end up where I intended them too. Let me add a couple percentage points to the bad health bonus.
Tyson
October 06, 2008 at 04:09PM View BBCode
For the first 10 games of the season, all players are treated as A+ health for the improvement chance calculation. Starting with Game 11, players will get the health bonus. I think this is why there is a discrepancy.I should weigh this back in during the season so players end up where I intended them too. Let me add a couple percentage points to the bad health bonus.
Tyson
Admin
Tyson
October 06, 2008 at 04:12PM View BBCode
Correction, this applies for only the first 5 games.Tyson
Admin
http://beta.simdynasty.com/player.jsp?player=nobody&mode=stats&id=18604
Let me do some more looking around.
Tyson
October 06, 2008 at 04:23PM View BBCode
Nevermind, something looks off. This guy should have closer to 17 ICs, he has played every game:http://beta.simdynasty.com/player.jsp?player=nobody&mode=stats&id=18604
Let me do some more looking around.
Tyson
Admin
Tyson
October 06, 2008 at 04:35PM View BBCode
Alright, the issue is that no players got ICs for games 1-5. I have changed this around so that the first 5 games of the season are treated the same as the rest.Tyson
tworoosters
The error totals are way up, but now they appear too high and inconsistent. At SS for example [url=http://beta.simdynasty.com/player.jsp?playername=nobody&mode=stats&id=16803]Ratzer[/url] who is 88/97 has committed 22 errors and is +13 whereas [url=http://beta.simdynasty.com/player.jsp?id=15269&statsorimps=fielding]Pierce[/url] (84/70) has committed only 12 errors and is also +13.
I can understand Ratzer's increased range meaning a few more errors, but not almost double, and that range and better arm should create more + plays than a pedestrian fielder like Pierce.
SS range factors are also skewed higher than MLB for the most part, in MLB 4.6+ is Gold Glove material while right now in Beta 12 SS who played 1,000+ innings are over 4.5.
[Edited on 10-14-2008 by tworoosters]
October 14, 2008 at 04:26PM View BBCode
Where are we on the errors ?The error totals are way up, but now they appear too high and inconsistent. At SS for example [url=http://beta.simdynasty.com/player.jsp?playername=nobody&mode=stats&id=16803]Ratzer[/url] who is 88/97 has committed 22 errors and is +13 whereas [url=http://beta.simdynasty.com/player.jsp?id=15269&statsorimps=fielding]Pierce[/url] (84/70) has committed only 12 errors and is also +13.
I can understand Ratzer's increased range meaning a few more errors, but not almost double, and that range and better arm should create more + plays than a pedestrian fielder like Pierce.
SS range factors are also skewed higher than MLB for the most part, in MLB 4.6+ is Gold Glove material while right now in Beta 12 SS who played 1,000+ innings are over 4.5.
[Edited on 10-14-2008 by tworoosters]
hobos
Just a quick look at SS's in general:
Pisoni 71/68, 5.5 E/y -> 27
Rader 76/66, 5.57 E/y -> 25
Aven 53/63, 7.22 E/y -> 25
Ratzer 88/97, 2.43 E/y -> 23
Hill 87/53, 9 E/y -> 21
Carrasquel 92/71, 3.67 E/y -> 19
Grott 96/77, 4.22 E/y -> 19
Obviously this is a pretty worthless analysis, but 'about twice as many' seems to be off. 3-4x seems to be more the standard.
October 15, 2008 at 11:20PM View BBCode
Ratzer ended at 23 errors, which is 6 more than he made in the preceding 7 years combined! However his +/- was right in line with his career average (+21). Obviously guys like him need to see more errors than just ~2 a season, but it seems to be far more than even a doubling effect.Just a quick look at SS's in general:
Pisoni 71/68, 5.5 E/y -> 27
Rader 76/66, 5.57 E/y -> 25
Aven 53/63, 7.22 E/y -> 25
Ratzer 88/97, 2.43 E/y -> 23
Hill 87/53, 9 E/y -> 21
Carrasquel 92/71, 3.67 E/y -> 19
Grott 96/77, 4.22 E/y -> 19
Obviously this is a pretty worthless analysis, but 'about twice as many' seems to be off. 3-4x seems to be more the standard.
Admin
Tyson
October 16, 2008 at 01:09PM View BBCode
I didn't make any changes to the +/- logic, only to errors. I'll take a look at this data today or tomorrow and see how we did so far.Tyson
celamantia
October 16, 2008 at 01:31PM View BBCode
Note we are looking at an extremely small sample size. I think we need a minimum of three seasons worth of data before we can make comparisons. And if we move players from position to position, it'll be difficult to tell anything.
tworoosters
Note all these guys played the position at 100%. And yes, I realize we're dealing with small samples.
1B - [url=http://beta.simdynasty.com/player.jsp?id=16296&statsorimps=fielding]Craig[/url] - 66 arm/86 range, current year 10 errors, previous 7 seasons 8 errors in total.
I'm okay with increased errors but 10 is on the high side when compared to MLB (Miguel Cabrera only made 9) and Craig's 86 range should, I think put him in the upper range of defensive 1B.
2B - [url=http://beta.simdynasty.com/player.jsp?id=17069&statsorimps=fielding]Hickerson[/url] - 53 arm/75 range, current year 4 errors, previous 3 years 17 errors in total.
Way too low, Hickerson is an average defender at best and should be making 10-12 errors a year if we're to be in line with MLB.
SS - [url=http://beta.simdynasty.com/player.jsp?id=15910&statsorimps=fielding]Phil Grott[/url] - 96 arm/77 range, current year 19 errors, previous 7 seasons 26 errors in total.
I've always felt SS error totals in SIM were ridiculously low so this change puts things back in line with MLB.
As mentioned above however the SS in SIM simply makes too many plays. Grott has average, or slightly better than average range for the position and has a career range factor considerably higher than Ozzie Smith.
3B - [url=http://beta.simdynasty.com/player.jsp?id=18641&statsorimps=fielding]Looper[/url] - 62 arm/56 range, current year 31 errors, previous year 25 errors.
Looper's not a very good defensive player so I think the 31 errors are within reason, maybe a little low.
October 16, 2008 at 03:34PM View BBCode
Looking at the starting infield for Detroit:Note all these guys played the position at 100%. And yes, I realize we're dealing with small samples.
1B - [url=http://beta.simdynasty.com/player.jsp?id=16296&statsorimps=fielding]Craig[/url] - 66 arm/86 range, current year 10 errors, previous 7 seasons 8 errors in total.
I'm okay with increased errors but 10 is on the high side when compared to MLB (Miguel Cabrera only made 9) and Craig's 86 range should, I think put him in the upper range of defensive 1B.
2B - [url=http://beta.simdynasty.com/player.jsp?id=17069&statsorimps=fielding]Hickerson[/url] - 53 arm/75 range, current year 4 errors, previous 3 years 17 errors in total.
Way too low, Hickerson is an average defender at best and should be making 10-12 errors a year if we're to be in line with MLB.
SS - [url=http://beta.simdynasty.com/player.jsp?id=15910&statsorimps=fielding]Phil Grott[/url] - 96 arm/77 range, current year 19 errors, previous 7 seasons 26 errors in total.
I've always felt SS error totals in SIM were ridiculously low so this change puts things back in line with MLB.
As mentioned above however the SS in SIM simply makes too many plays. Grott has average, or slightly better than average range for the position and has a career range factor considerably higher than Ozzie Smith.
3B - [url=http://beta.simdynasty.com/player.jsp?id=18641&statsorimps=fielding]Looper[/url] - 62 arm/56 range, current year 31 errors, previous year 25 errors.
Looper's not a very good defensive player so I think the 31 errors are within reason, maybe a little low.
Admin
Here were the errors in 2019 by position played:
P: 149
C: 96
1B: 160
2B: 193
3B: 349
SS: 412
LF: 71
CF: 93
RF: 99
Let's compare that to 2018:
P: 128
C: 67
1B: 56
2B: 211
3B: 213
SS: 161
LF: 92
CF: 84
RF: 82
Here are the numbers we were striving for:
P: 112
C: 163 (of which almost 100 are throwing errors on steal attempts)
1B: 157 (40 of which are drops)
2B: 209
3B: 324
SS: 338
LF: 84
CF: 82
RF: 101
Everything looks pretty close, with the exception of Catchers being a bit low, and SS being a bit high.
I'm going to run beta up to 81 games to get half a season's data. Be back in a few minutes.
Tyson
October 23, 2008 at 06:20PM View BBCode
Ok, this was put in at the beginning of last season. So let's look at league-wide totals as a starting point.Here were the errors in 2019 by position played:
P: 149
C: 96
1B: 160
2B: 193
3B: 349
SS: 412
LF: 71
CF: 93
RF: 99
Let's compare that to 2018:
P: 128
C: 67
1B: 56
2B: 211
3B: 213
SS: 161
LF: 92
CF: 84
RF: 82
Here are the numbers we were striving for:
P: 112
C: 163 (of which almost 100 are throwing errors on steal attempts)
1B: 157 (40 of which are drops)
2B: 209
3B: 324
SS: 338
LF: 84
CF: 82
RF: 101
Everything looks pretty close, with the exception of Catchers being a bit low, and SS being a bit high.
I'm going to run beta up to 81 games to get half a season's data. Be back in a few minutes.
Tyson
Admin
P: 134
C: 124
1B: 164
2B: 202
3B: 334
SS: 444
LF: 92
CF: 98
RF: 76
Shortstops look a bit on the high side still. I'm going to make an ajustment for the 2nd half of this season for Shortstops.
Tyson
October 23, 2008 at 06:29PM View BBCode
Here's where we are through 81 games this year ( I multiplied all the totals by 2 for easier comparison)P: 134
C: 124
1B: 164
2B: 202
3B: 334
SS: 444
LF: 92
CF: 98
RF: 76
Shortstops look a bit on the high side still. I'm going to make an ajustment for the 2nd half of this season for Shortstops.
Tyson
celamantia
October 23, 2008 at 06:31PM View BBCode
Before you do that... what is the average Range at each of these positions now? Is it possible that the overall fielding talent of SS'es right now is on the low side?
Admin
Tyson
October 23, 2008 at 06:46PM View BBCode
The 15 teams that have a SS vs RHP have an average range of 71.9 and average arm of 76.8. That is actually a bit above average. Last time I checked, the average arm + range of SSs used in pay leagues was 139.Tyson
tworoosters
As I mentioned earlier in this thread the range factors of the SS currently in Beta are all high when compared to MLB.
October 23, 2008 at 07:23PM View BBCode
I think one of the problems with SS errors may have to do with the fact that they just make too many plays in SIM.As I mentioned earlier in this thread the range factors of the SS currently in Beta are all high when compared to MLB.
celamantia
October 23, 2008 at 07:28PM View BBCode
I don't know if there is an easy fix for that. Is there any positions that are obviously making too few plays?
ballplayer3
With a cursory look at the RF of other positions, after sorting by Innings, they seem to be about right. I did not go to actual MLB data to verify though - just going on my experience.
Mike
October 24, 2008 at 12:10AM View BBCode
As far as too few plays goes, I looked at assists from 1st Basemen and noticed that those are extremely low. That probably means the pitcher putouts must be way too low, but I can't view pitcher fielding stats.With a cursory look at the RF of other positions, after sorting by Innings, they seem to be about right. I did not go to actual MLB data to verify though - just going on my experience.
Mike
tworoosters
October 24, 2008 at 05:10PM View BBCode
Mike there are no pitcher putouts in SIM the 1st baseman fields all balls unassisted.Pages: 1 2