Fielding

[Reed]

Attached is Brooke's fielding stats from my OOTP game (not his real stats). I used 3 year recalc.. Number of errors pretty much on par with his real number of errors.

[Brad K]

Originally Posted by Rain King

So, if development is on then the players are potentially going to regress during the season...particularly if they are older.

The Recalc will reset their ratings each year, but we know that the game will regress older players defensively quite often (understandably).

A lower TCR won't curb aging effects (there is a setting for that also though).

Why would development always regress players? If it always does that it needs to be called something else.

It reminds me of the discussions on the adjust/weaken settings where it's said they make better players stronger. The assumption is the routine always makes players subjected to it worse. They're even called "make bad" settings by some. But why aren't some of the low AB players being made better?

[Rain King]

Originally Posted by Brad K

Why would development always regress players? If it always does that it needs to be called something else.

It doesn't. I never said that.

The same development engine is used in historical as is used in fictional. Fictional players obviously develop positively. However, development is also tied to Potential ratings. So, for a player to develop positively they need to be both younger (i.e. in the age most development happens) and have room to develop. Historical leagues are going to have fewer of those players than they have of players at risk of aging. Also, those players that develop make the older players relatively worse whether they age or not.

It reminds me of the discussions on the adjust/weaken settings where it's said they make better players stronger. The assumption is the routine always makes players subjected to it worse. They're even called "make bad" settings by some. But why aren't some of the low AB players being made better?

That is not an accurate assumption. Some of the low AB players ARE being made better.

A player that was 0-10 doesn't come in with ratings that would give him a .000 batting average. Players that didn't hit a home run in a small amount of at-bats will have ratings with some potential to hit home runs, etc.

They are being pulled partially to around league average* (if they fall into the adjust pool) and partially to replacement* level (if they fall into the weaken).

*I don't know if it is exactly average/replacement. Just that it is somewhere around those levels based on past discussions on this topic.

[Brad K]

Originally Posted by Rain King

That is not an accurate assumption. Some of the low AB players ARE being made better.

A player that was 0-10 doesn't come in with ratings that would give him a .000 batting average. Players that didn't hit a home run in a small amount of at-bats will have ratings with some potential to hit home runs, etc.

They are being pulled partially to around league average* (if they fall into the adjust pool) and partially to replacement* level (if they fall into the weaken).

*I don't know if it is exactly average/replacement. Just that it is somewhere around those levels based on past discussions on this topic.

It has been stated on the board multiple times by devs and by Garlon that the "make bad" adjustments combine with LTMs to change the performance of top players. That requires that there is more removed from the ratings of the adjusted players than is added. The acceptance of this explanation by the public is shown by the adoption of the phrase "make bad".

While many clearly accept it as fact, I call it an assumption. In my view it could easily be an incorrect explanation resulting from erroneous analysis. Purely accidental. Nothing has ever been concealed or misrepresented.

[Rain King]

- The Adjust/Weaken process changes the ratings of low AB/IP players.

- Since an MLB season has a specific statistical output and the Adjust/Weaken process changes the "expected" results of those low AB/IP players it also has an impact on the "expected" results of players who are not adjusted.

There is not a perfect way to handle sample size issues for historical play, but this game does it better than anything on the market.


I would love for you to actually provide a link for any of the posts that you are referencing as I can guarantee you have taken them out of context. You have been trolling this topic for years now.

[Brad K]

Originally Posted by Rain King

- The Adjust/Weaken process changes the ratings of low AB/IP players.

- Since an MLB season has a specific statistical output and the Adjust/Weaken process changes the "expected" results of those low AB/IP players it also has an impact on the "expected" results of players who are not adjusted.

There is not a perfect way to handle sample size issues for historical play, but this game does it better than anything on the market.

The way to handle the adjustment is to keep the additions and deductions within the group of players being adjusted so the total adjustment for the group is zero.

I would love for you to actually provide a link for any of the posts that you are referencing as I can guarantee you have taken them out of context. You have been trolling this topic for years now.

If you know I've been posting about this for years then you witnessed the discussions so no link is necessary. A troll is a post you don't like but can't refute. And so in frustration you try to degrade the motives and morals of the poster.

[Rain King]

Originally Posted by Brad K

The way to handle the adjustment is to keep the additions and deductions within the group of players being adjusted so the total adjustment for the group is zero.

I encourage you to take any historical season, figure out the math, apply it to the group of players you think it should be applied to and show us the work.

[Garlon]

The was a significant change to the way the game is handling Adjust/Weaken settings for OOTP25. In previous edition of OOTP the weaken (aka made bad) was basically a ratings cliff where the players were basically unusable if they fell into that category. That ratings cliff is no longer there with OOTP 25. The coefficients used for adjust and weaken are far more conservative in OOTP 25.

A player can import above the adjust limit, within the adjust limit or below the adjust limit, but if they are below the adjust limit that means that they are being partially adjusted and partially weakened.

If you are adjusting at 250 AB and weakening at 175 AB and a given player has a total of 150 AB, they will receive 25 weakened AB and another 75 Adjusted AB to bring them up to 250 AB. These additional 100 AB get added to their original 150 AB to produce their ratings.

The player ratings and results have been fantastic with OOTP 25.

[Garlon]

Brooks Robinson Real FLD at 3B: .957

OOTP 25 in a saved game I have from Aug 5th, Brooks had .966 at 3B, was the career leader in ZR at 3B (game went from 1871-2023), and won 14 Gold Gloves.

Brooks Real OPS .723 vs OOTP .731

[Rain King]

Garlon, can you tell us more about what the Adjusted/Weakened levels mean compared to league average?


One thing to look at regarding this change based on some info in another thread is that pitchers hitting Power ratings may be getting propped up to much. They are, in a couple examples I tested, hitting 2-3 times more home runs than they should across the league.

[sprague]

I notice sometimes very odd fielding ratings right out of the box.
Below are from a new started standard 1983 game
Tom Foley is rated best in the league, better than Ozzie Smith at SS
While Dave Parker has an arm that might barely thrown though paper.
Just in this one season I noticed several "fielding" oddities such as this that I would have to go in manually and course correct.
Example the best throwing arms in the outfield are Barfield and Henderson each with a 237 value, so better than Clemente in his prime.

[Reed]

Adjust/Weaken. Like Rain King says, If a player comes in with low AB and terrible stats then their ratings will be adjusted upward.
The ratings only change for players below that number of ABs (or IP). Doing so may help player’s resulting stats that had more ABs than the adjust number but it doesn’t change their ratings.

IMO it may help all the players that had more ABs than the adjusted number, including the “top” players but the main purpose (again IMO) is to prevent someone that played limited ABs or IP from dominating your league. Without it a player that had 4 hits in 7 ABs would be used full time by the AI and would hit over .500.
It also helps make sure several several limited AB players from becoming starters making IRL starters becoming role players.

What would you recommend to prevent limited used players from dominating the league?

[David Watts]

Originally Posted by Garlon

Brooks Robinson Real FLD at 3B: .957

OOTP 25 in a saved game I have from Aug 5th, Brooks had .966 at 3B, was the career leader in ZR at 3B (game went from 1871-2023), and won 14 Gold Gloves.

Brooks Real OPS .723 vs OOTP .731

That's kind of what I was hoping for. Instead I get a Brooks Robinson that's putting up negative zone ratings 3 years in a row. He does rebound in 69 when his fielding ratings get a huge boost, but his fielding stats from 63-69 are rather blah. There's a screen shot earlier in this thread.

[David Watts]

Originally Posted by LansdowneSt

Were '62, 'and '68 in-game gold gloves and 2 inherited from before the sim? It's one-year recalc but the fielding is different, was that on same year or three-year?

His TZ was down in '65, and especially '66 per FanGraphs (Rtot in BBR).

Yes down, but not down so much that he was below zero. I don't know where to find such a thing, but where did Robinson rank among other 3B in 65 and 66?

[sprague]

Originally Posted by David Watts

That's kind of what I was hoping for. Instead I get a Brooks Robinson that's putting up negative zone ratings 3 years in a row. He does rebound in 69 when his fielding ratings get a huge boost, but his fielding stats from 63-69 are rather blah. There's a screen shot earlier in this thread.

I have found Robinson to have rather average fielding stats at third base in most of the sims I have played. He tends to win a gold glove or two, but nothing like the fielding dominance he seemed to employ during the 1960s.

[LansdowneSt]

Originally Posted by David Watts

Yes down, but not down so much that he was below zero. I don't know where to find such a thing, but where did Robinson rank among other 3B in 65 and 66?

Stathead on BBR can probably sort by individual but for a glance at him relative to league, look to Range Factor in BBR fielding and then far right is the league number for same. He’s a touch above average to league in that “down” year. I also don’t know if in-game or in BBR, it matters he is in the NL in the sim so might be measured against a different data set than before. I just don’t know.

[LansdowneSt]

Originally Posted by sprague

I have found Robinson to have rather average fielding stats at third base in most of the sims I have played. He tends to win a gold glove or two, but nothing like the fielding dominance he seemed to employ during the 1960s.

Garlon would know better than I but on the top end (and bottom) there may be backstop caps of how good or bad one can be, esp on one year calculations.

[Rain King]

I ran a sim from 1962-1972.

Default Mode, 1-Year Recalc, Injuries Off. 1-100 scale (100% accurate)

Here are the Ratings and resulting Zone Rating from each season.

The ratings definitely dip in '64-'66, probably a bit more than I would expect looking at Fangraphs TZ (just the best quick/dirty place to look IMO). But also these are just seasons that the single season defensive stats aren't great for Robinson.

My takeaway is that it is going to be really tough to get good year-to-year defensive ratings from a 1-year recalc as the metrics available to turn stats into ratings just aren't stable over that sample size.

I'd be curious to hear from Garlon on the nitty gritty of those seasons and how the ratings end up where they do, but would also generally recommend 3-year defensive recalc at all times to anyone playing historical. In fact, I'll run a new sim with that option and report back.

[David Watts]

Originally Posted by Rain King

I ran a sim from 1962-1972.

Default Mode, 1-Year Recalc, Injuries Off. 1-100 scale (100% accurate)

Here are the Ratings and resulting Zone Rating from each season.

The ratings definitely dip in '64-'66, probably a bit more than I would expect looking at Fangraphs TZ (just the best quick/dirty place to look IMO). But also these are just seasons that the single season defensive stats aren't great for Robinson.

My takeaway is that it is going to be really tough to get good year-to-year defensive ratings from a 1-year recalc as the metrics available to turn stats into ratings just aren't stable over that sample size.

I'd be curious to hear from Garlon on the nitty gritty of those seasons and how the ratings end up where they do, but would also generally recommend 3-year defensive recalc at all times to anyone playing historical. In fact, I'll run a new sim with that option and report back.

Will be interesting to see your results. Thanks

[Rain King]

I forgot to turn off Injuries, but here is 3-Year Fielding recalc (1-Year everything else).

Robinson lost most of his 1963 season, but was healthy aside from that.