How is the accumulation of position experience done in the game?

Drstrangelove · 06-17-2018, 05:33 AM

I looked at player stats in the editor from my Spring Training games showing how many games of experience were "earned" over the course of eight spring training games.

Players not only earn different amounts of points (some earn zero while others earn 4, 6, 8, 10, 12 or 16), but they aren't always consistent even for the same player at the same position.

E.g., one player in CF earned 0 for one game, then, 8, 0, 2, 0. When I moved him to LF, he earned 0, 8 and 16. Another player earned 10 every game at 3B for 5 games, then 10 every game at 1B for 3 games. A third player earned 0, 0, 8, 4 and 8 in LF, then 4, 4, and 4 in CF.

OutS|der · 06-17-2018, 06:09 AM

From the Manual

Learning New Positions
Players learn new positions through practice. The best way to accomplish this is by having them play in a new position. Players learn more quickly when they play regularly in the minor leagues, or in spring training. The more a player plays at a certain position, the more he improves. After he has acquired a certain amount of experience at a position, he will be given a position rating that will display on his profile.

OOTP follows the concept of the "defensive spectrum." The defensive spectrum is as follows: DH - 1B - LF - RF - 3B - CF - 2B - SS. Generally speaking, the further to the right, the harder the position is to play and the harder it will be to convert a player to that position. Position players can learn to play catcher, but it often takes a very long time, and they typically don't make very good ones.

Note: A player cannot 'lose' positions. Once he has learned a position, he will always have a rating at that position.

I think adaptability also plays a role in how quick they learn

Drstrangelove · 06-17-2018, 09:33 AM

Thanks for the reply. I hadn't considered either of those. These are some good ideas, so I went to look at the data:

1) Adaptability - Games Earned (scores are below. E.g., Player 1 had a rating of 47 and earned 72 "game" XP in 8 spring training games.

47 - 72
44 - 32
13 - 34
7 - 84
2 - 66

These are only a few data points, but these numbers are not correlated (r = -.27.)

2) Using the spectrum, (DH - 1B - LF - RF - 3B - CF - 2B - SS):

Player 1 earned 10 points in 5 games in CF, but a higher rate, 24 in 3 games in LF, which agrees with the spectrum.

Player 2 skipped as he played the same position in all 8 games.

Player 3 earned 20 points in 5 games in LF, but 12 in 3 games in CF, so, the same rate, which does not agree with the spectrum.

Player 4 earned 36 points in 5 games at 1B, but 30 in 3 games at SS, a higher rate, which does not agree with the spectrum.

And player 5 earned 10 points per game in every game whether 3B or 1B, which does not agree with the spectrum.

So, I'd say, in this small sample, the spectrum isn't supported, at least, as single variable.

When I examined this in OOTP 18, it was a flat 11 per game for everyone at every position, with no modifiers. (I.e., play 10 spring training games, and you will get 110 games added to XP up to the cap of 200.)

I also considered the following:

3) Work Ethic and Intelligence. Work Ethic was r=.41, intelligence was r=.66. Again, only a handful of data points, but not clearly explaining this.

4) I also looked at whether game XP was being earned on some type of diminishing returns, so the early XP comes faster than the last XP. (Not sure that'd be done if skill sets are designed along bell curves, but I looked anyway.) One player learning a new position earned 10 points per game, while a second player was only getting 4 per game. So, that would suggest that it's not diminishing returns.

5) Random versus step-wise skill attainment. Is it random? (In 6 times out of 40 game-scores, a player earned zero, which suggests it could be random.) Moreover, I had two cases of back to back games where a player earned zero XP despite playing a full game. (This would I think exclude partial credits that rolled over into the next game.)

Bluenoser · 06-17-2018, 11:48 AM

One thing not mentioned in the manual - like real life, some players just won't learn a position, no matter how long you play them there.

NoOne · 06-17-2018, 01:08 PM

in cases of zero points earned -- did they field any balls in that game?

heh, i've always thought intelligence had a stronger correlation to developement than work ethic... a glimmer that i might be right in the differing correlation coefficients for learning a position.

Drstrangelove · 06-17-2018, 08:17 PM

Quote:

Originally Posted by Bluenoser

One thing not mentioned in the manual - like real life, some players just won't learn a position, no matter how long you play them there.

I've seen this, but only when a player lacks a unique sub skill. E.g., a player who has a very low Turn DP skill, might not show up as "able" to play 3B. They can have a great infield arm and great range, but the profile won't show 3B as a learned position.

But that's not the problem I'm discussing. My players do "show" the positions as learned. They simply aren't getting game XP (when they aren't capped at the position) even when they play it, so their position skill doesn't change despite playing the position.

Drstrangelove · 06-17-2018, 08:35 PM

Quote:

Originally Posted by NoOne

in cases of zero points earned -- did they field any balls in that game?

This is a great question. In one game, both my CF and LF earned zero points. Sure, it's possible they both had no chances, but it's not the way the game worked in OOTP 18, iirc. Players "earned" a flat amount of XP every game in ST. And it wasn't random or close to zero.

Quote:

Originally Posted by NoOne

heh, i've always thought intelligence had a stronger correlation to development than work ethic... a glimmer that i might be right in the differing correlation coefficients for learning a position.

Certainly plausible. Sure. But, intelligence alone, even taking into account the position, isn't predicting this. In the stats above, a 110 INT guy who is already a great CF is on pace to need 50 games to learn how to play LF, while a 117 INT guy who is already a good 2B, will learn 3B in just 20 games. This would work better if the CF was a moron or the 2B a genius, but both are almost identical INT wise, while CF to LF is clearly easier than 2B to 3B.

This would seem to be a huge change to me since OOTP 18. If that's how it works now, that's fine, too, but it's difficult to reconcile to the idea that in ST:

1) the AI moves players all around on it's own volition
2) the AI pinch hits and subs as it pleases
3) the AI doesn't give credit if there are no plays

So, in a theoretical manner, it's possible for a player to earn no or virtually no credit at all in ST because the AI removes him in the game before he gets a play. This might be true as to how the game engine works, but imo it's not a good model for how players really learn the position (in those other 6-8 hours we don't see them playing.)

damientheomen3 · 06-18-2018, 09:33 AM

Quote:

Originally Posted by Drstrangelove

I've seen this, but only when a player lacks a unique sub skill. E.g., a player who has a very low Turn DP skill, might not show up as "able" to play 3B. They can have a great infield arm and great range, but the profile won't show 3B as a learned position.

But that's not the problem I'm discussing. My players do "show" the positions as learned. They simply aren't getting game XP (when they aren't capped at the position) even when they play it, so their position skill doesn't change despite playing the position.

IIRC this was changed as of the recent patch. The thresholds for fielding attributes determining whether or not a player can have a rating at a position were lowered or removed according to the patch notes.

zappa1 · 06-18-2018, 10:11 AM

I've been watching this closely as I have Scott Kingery on the team. He had no rating at all at shortstop. I started playing him there almost regularly. He started off with a 1 on defense. Little by little as he's played more games at short, he's now a 5 after about 45 games into the season. This is on the 1-20 scale. I like the way this is working out. In past versions, it seemed to take forever for a player to learn a position.

06-17-2018, 05:33 AM	#1
Drstrangelove Major Leagues Join Date: Jun 2014 Posts: 346	How is the accumulation of position experience done in the game? I looked at player stats in the editor from my Spring Training games showing how many games of experience were "earned" over the course of eight spring training games. Players not only earn different amounts of points (some earn zero while others earn 4, 6, 8, 10, 12 or 16), but they aren't always consistent even for the same player at the same position. E.g., one player in CF earned 0 for one game, then, 8, 0, 2, 0. When I moved him to LF, he earned 0, 8 and 16. Another player earned 10 every game at 3B for 5 games, then 10 every game at 1B for 3 games. A third player earned 0, 0, 8, 4 and 8 in LF, then 4, 4, and 4 in CF.

06-17-2018, 06:09 AM	#2
OutS\|der Hall Of Famer Join Date: Apr 2010 Location: In A Van Down By The River Posts: 2,711 Infractions: 0/1 (1)	From the Manual Learning New Positions Players learn new positions through practice. The best way to accomplish this is by having them play in a new position. Players learn more quickly when they play regularly in the minor leagues, or in spring training. The more a player plays at a certain position, the more he improves. After he has acquired a certain amount of experience at a position, he will be given a position rating that will display on his profile. OOTP follows the concept of the "defensive spectrum." The defensive spectrum is as follows: DH - 1B - LF - RF - 3B - CF - 2B - SS. Generally speaking, the further to the right, the harder the position is to play and the harder it will be to convert a player to that position. Position players can learn to play catcher, but it often takes a very long time, and they typically don't make very good ones. Note: A player cannot 'lose' positions. Once he has learned a position, he will always have a rating at that position. I think adaptability also plays a role in how quick they learn Last edited by OutS\|der; 06-17-2018 at 06:10 AM.

06-17-2018, 09:33 AM	#3
Drstrangelove Major Leagues Join Date: Jun 2014 Posts: 346	Thanks for the reply. I hadn't considered either of those. These are some good ideas, so I went to look at the data: 1) Adaptability - Games Earned (scores are below. E.g., Player 1 had a rating of 47 and earned 72 "game" XP in 8 spring training games. 47 - 72 44 - 32 13 - 34 7 - 84 2 - 66 These are only a few data points, but these numbers are not correlated (r = -.27.) 2) Using the spectrum, (DH - 1B - LF - RF - 3B - CF - 2B - SS): Player 1 earned 10 points in 5 games in CF, but a higher rate, 24 in 3 games in LF, which agrees with the spectrum. Player 2 skipped as he played the same position in all 8 games. Player 3 earned 20 points in 5 games in LF, but 12 in 3 games in CF, so, the same rate, which does not agree with the spectrum. Player 4 earned 36 points in 5 games at 1B, but 30 in 3 games at SS, a higher rate, which does not agree with the spectrum. And player 5 earned 10 points per game in every game whether 3B or 1B, which does not agree with the spectrum. So, I'd say, in this small sample, the spectrum isn't supported, at least, as single variable. When I examined this in OOTP 18, it was a flat 11 per game for everyone at every position, with no modifiers. (I.e., play 10 spring training games, and you will get 110 games added to XP up to the cap of 200.) I also considered the following: 3) Work Ethic and Intelligence. Work Ethic was r=.41, intelligence was r=.66. Again, only a handful of data points, but not clearly explaining this. 4) I also looked at whether game XP was being earned on some type of diminishing returns, so the early XP comes faster than the last XP. (Not sure that'd be done if skill sets are designed along bell curves, but I looked anyway.) One player learning a new position earned 10 points per game, while a second player was only getting 4 per game. So, that would suggest that it's not diminishing returns. 5) Random versus step-wise skill attainment. Is it random? (In 6 times out of 40 game-scores, a player earned zero, which suggests it could be random.) Moreover, I had two cases of back to back games where a player earned zero XP despite playing a full game. (This would I think exclude partial credits that rolled over into the next game.) Last edited by Drstrangelove; 06-17-2018 at 10:04 AM.

06-17-2018, 11:48 AM	#4
Bluenoser Hall Of Famer Join Date: Mar 2002 Location: In The Moment Posts: 14,211	One thing not mentioned in the manual - like real life, some players just won't learn a position, no matter how long you play them there.

06-17-2018, 01:08 PM	#5
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 0/1 (3)	in cases of zero points earned -- did they field any balls in that game? heh, i've always thought intelligence had a stronger correlation to developement than work ethic... a glimmer that i might be right in the differing correlation coefficients for learning a position.

06-18-2018, 10:11 AM	#9
zappa1 Hall Of Famer Join Date: Apr 2012 Location: Pennsylvania Posts: 2,134	I've been watching this closely as I have Scott Kingery on the team. He had no rating at all at shortstop. I started playing him there almost regularly. He started off with a 1 on defense. Little by little as he's played more games at short, he's now a 5 after about 45 games into the season. This is on the 1-20 scale. I like the way this is working out. In past versions, it seemed to take forever for a player to learn a position.