A little OOTP 11 update... - Page 15

pstrickert · 02-10-2010, 01:57 PM

Quote:

Originally Posted by Qwerty75

Can't remember exactly how I came to be convinced this is how the engine works. I think it's a mixture of comments by Markus, talking to a beta tester, and my own observations. For example, I had a team where I had a first baseman learning to play left field and it brought the team BABIP up to a rate out of proportiion to what you would expect with simply the number of balls recorded as being played by the left fielder.

I'm not raggin' on you, qwerty. You seem to be a good guy. But it is amazing how easily misinformation is spread. Here, at least, Markus (or someone else) has a chance to correct it. On other boards, I've seen all kinds of examples of misinformation about OOTP. Which is careless -- though I often think those who post it don't care. They seem more than happy to denigrate OOTP, whether justified or not.

thbroman · 02-10-2010, 02:41 PM

Some Thoughts on New and Exotic Stats

I agree with MD -- this has turned out to be an interesting thread for all sorts of reasons!

But something posted by Eugene a couple of pages back has really stuck with me, and I wanted to put it up for discussion: what is it that new stats packages are supposed to do? Eugene claimed -- and I agree with him completely -- that he never looks at the new-fangled stats. I don't either, because in part I don't understand them (and don't care enough to learn how to use them) but also because I wonder about the role of such statistical breakdowns in the game.

So what are they supposed to do? When Bill James and GM's use them in real life, they're basically attempting to produce a new method for squeezing out the essence of superior play by an individual from the murky complexity of the game situation. The individual player is a real unit (you can see that for yourself), and since these unitary but mysterious players function at the same level on which the game of baseball is played, it makes sense to try to find effective ways of identifying which players are most valauble in producing winning baseball. At least, that seems to be what many stats are meant to do.

But what role do such stats play in OOTP? As much as we all love the game, it is -- ahem! -- only a model, and the players are, sad to say, mere lines in a database. So on the one hand, it's perfectly comprehensible why we would want to have stat outputs on something like L/R batting averages, because we can have faith that the game engine takes L/R batting into consideration when it does it cranking. Thus the L/R averages are "real" in the sense of representing something that the game itself factors into the equation, so to speak. But there are other stats in OOTP that appear to be purely cosmetic --BA in May versus August, for example. The numbers are there in the database, you can "do the math", as they say, but it is not so obvious (not to me, anyway) that the game engine actually considers month-to-month fluctuations in its calculation of results.

So here's a question -- when Markus decides to include something like WAR in a forthcoming version of the game, does its inclusion mean that:

a) he's just going to dig out some numbers and produce a new column in the already crowded stat output, or

b) he is going to add yet another modifier to the calculation of the basic outcome of any single play?

In other words, which stats approaches are important enough to warrant inclusion as factors in how the game produces outcomes?

just wondering . . . .

robc · 02-10-2010, 03:02 PM

Quote:

Originally Posted by pstrickert

Speaking of which . . .

There is an interesting thread on the Puresim forums in which you, MD, have caused quite a reaction. What's up with that?

Some people responded a little emotionally in my opinion. We are all people after all sometimes and we do that. After Shaun thought about it I think he appreciated MD's comments and realized there was some validity to them. I'm not 100% sure he wasn't trying to 'stir the pot' but based on a prior comment abut Shaun I think MD was trying to be helpful.

actionjackson · 02-10-2010, 03:17 PM

Quote:

Originally Posted by robc

Some people responded a little emotionally in my opinion. We are all people after all sometimes and we do that. After Shaun thought about it I think he appreciated MD's comments and realized there was some validity to them. I'm not 100% sure he wasn't trying to 'stir the pot' but based on a prior comment abut Shaun I think MD was trying to be helpful.

MD regularly stirs the pot and is helpful. The two aren't necessarily mutually exclusive. That's why they call him the surly bastage.

vyshka · 02-10-2010, 03:31 PM

Quote:

Originally Posted by robc

Some people responded a little emotionally in my opinion. We are all people after all sometimes and we do that. After Shaun thought about it I think he appreciated MD's comments and realized there was some validity to them. I'm not 100% sure he wasn't trying to 'stir the pot' but based on a prior comment abut Shaun I think MD was trying to be helpful.

It looked like the majority of posters just ignored everything he had to say and decided he was a troll because he is a fan of OOTP. If he wanted to stir the pot he could have asked Shaun what he meant when he was talking about developers and honesty with regards to that 'other' game.

robc · 02-10-2010, 03:32 PM

Quote:

Originally Posted by actionjackson

MD regularly stirs the pot and is helpful. The two aren't necessarily mutually exclusive. That's why they call him the surly bastage.

PhillieFever · 02-10-2010, 03:36 PM

Quote:

Originally Posted by OakDragon

Doesn't a ruling of "passed ball" or "wild pitch" require that the runner is safe?

You are absolutely correct, let me rephrase that. When the ball gets away from the catcher, players always advance safely. I would like to see the runners thrown out occasionally.

PhillieFever · 02-10-2010, 03:37 PM

Quote:

Originally Posted by OakDragon

Now, if the ball gets away from the catcher but the runner does not advance, that, I agree, would be nice to see. However, that would not be termed a wild pitch or passed ball, legalistically speaking, and I haven't played out enough games lately to recall whether that actually happens or not in OOTP.

I've never seen this either, and would like this added in as well.

robc · 02-10-2010, 03:40 PM

Quote:

Originally Posted by vyshka

It looked like the majority of posters just ignored everything he had to say and decided he was a troll because he is a fan of OOTP. If he wanted to stir the pot he could have asked Shaun what he meant when he was talking about developers and honesty with regards to that 'other' game.

It was unlike Shaun to take a bit of a shot like that. As I said some people were already worked up a bit before MD made his comment. Shaun has stated he has played OOTP and thinks it is a good game.

There are several people who have Puresim who are known players / fans of OOTP too. When I first started posting over there, I was mainly pointing out the things I like better in OOTP. After playing Puresim there are many things I like better in Puresim. I'm waiting until OOTP 11 before passing final judgement.

The Puresim forums are generally one of the more welcoming, non-hostile forums I have posted on. A lot less sniping goes on over there than over here. I guess each forum develops its own culture. Don't get me wrong though, there have been plenty helpful people over here too.

actionjackson · 02-10-2010, 05:27 PM

General question that anyone can jump in on, with a rather lengthy setup/introduction. I'm working on creating a custom Lahman database that will present more of a challenge for a historical dynasty by creating more of a "Fog of War" and level the playing field between me and the AI, editing only the master.csv file. My aim is to take a bunch of randomly generated names courtesy of the random name generator site I mentioned above and plug them into the nameFirst and nameLast columns. Before doing this I think I would blank out the birthState, birthCity (so that whatever Babe Ruth is called I won't be able to figure out that it's him by MD, Baltimore in those two columns), nameNote, nameGiven, and nameNick columns, so I won't be influenced by a George Herman (though I can't recall if given names show up anywhere in the game), here or a "The Bambino" there. I would leave everything else alone.

...Now (finally) for the question. Would writing over the nameFirst and nameLast columns and blanking out the other 5 in the master.csv file affect the functioning of the game? Or does it all just key off of the lahmanID and playerID columns and whatever's in these 7 columns doesn't matter much in terms of game performance? Marcus, Spritze and any other database geniuses, please feel free to jump in. Thanks in advance for your response(s).

Cryomaniac · 02-10-2010, 06:41 PM

Quote:

Originally Posted by actionjackson

Would writing over the nameFirst and nameLast columns and blanking out the other 5 in the master.csv file affect the functioning of the game? Or does it all just key off of the lahmanID and playerID columns and whatever's in these 7 columns doesn't matter much in terms of game performance? Marcus, Spritze and any other database geniuses, please feel free to jump in. Thanks in advance for your response(s).

As far as I'm aware that would work how you want it to, the disclaimer being that I know very little about the technical side of historical imports.

fhomess · 02-10-2010, 06:47 PM

The different files in the database key off of the lahmanID to match players to their stats. I'm pretty sure that the game loads them into its internal database and then doesn't look at the files again until it's time to load the next batch of players. In short, changing the names shouldn't affect anything other than the names they import with.

fhomess · 02-10-2010, 06:55 PM

On the issue of advanced/sabermetric stats, most of the basic ones are easily calculated from the traditional stats, so they don't require any additional database space to store. Things like basic wOBA and BP's new SIERA can be calculated on the fly. Other stats that require park and/or league adjustments are a bit harder to do as they require calculating something else first, so they might require a bit of database space being taken up.

As far as the addition of markUsZR to provide some better defensive statistics, I do hope that the underlying data will be available in the data dump somehow.

Speaking of the data dump, I've got numerous suggestions from my work with StatsLab that would be nice to have in there. Some examples:
1. RBOE - reached base on error isn't in there
2. Game Logs aren't reproducable. Would allow for things like WPA to be calculated for the more statistically inclined, but would also allow for evaluation of key game events/external game summaries.
3. Stint - there's currently no way to tell the order a player played for different teams during a single season.
4. Transactions - there's no record of transactions/trades (kind of goes with #3, but provides more)

Malleus Dei · 02-10-2010, 06:59 PM

Quote:

Originally Posted by pstrickert

Speaking of which . . .

There is an interesting thread on the Puresim forums in which you, MD, have caused quite a reaction. What's up with that?

No idea. PM me the link, would you?

Sweed · 02-10-2010, 07:13 PM

Quote:

Originally Posted by pstrickert

i'm not raggin' on you, qwerty. You seem to be a good guy. But it is amazing how easily misinformation is spread. Here, at least, markus (or someone else) has a chance to correct it. On other boards, i've seen all kinds of examples of misinformation about ootp. Which is careless -- though i often think those who post it don't care. They seem more than happy to denigrate ootp, whether justified or not.

qft.

Qwerty75 · 02-10-2010, 07:22 PM

Quote:

Originally Posted by pstrickert

I'm not raggin' on you, qwerty. You seem to be a good guy. But it is amazing how easily misinformation is spread. Here, at least, Markus (or someone else) has a chance to correct it. On other boards, I've seen all kinds of examples of misinformation about OOTP. Which is careless -- though I often think those who post it don't care. They seem more than happy to denigrate OOTP, whether justified or not.

I appreciate your concern for accuracy of information ... I know I feel frustrated myself when I read someone's misunderstanding about the game and such. I brought up my version of how the game works because I have had it as an assumption about the game for a long time, even though I can't remember precisely now how I made that conclusion in the first place. I would just like what I perceive as flaws in the game design to be known so that others don't go on thinking something is relevant to the game when it isn't, and possibly so that it can make improvements in the future.

Note that Markus didn't directly contradict what I said about the entire defensive makeup being factored in, but just addressed my concern about the double-counting of fielders. Maybe I don't have my facts completely straight, but, if I'm right in my suspicion, the system would seem to be illogical and might benefit from adjustments.

Shaun Sullivan · 02-10-2010, 07:44 PM

Quote:

Originally Posted by Malleus Dei

No idea. PM me the link, would you?

Not sure if you saw my apology, but I was wrong and you were right. It has been a rough week

Thanks for your very insightful comments.

Shaun

Eugene Church · 02-10-2010, 08:37 PM

Quote:

Originally Posted by PhillieFever

You are absolutely correct, let me rephrase that. When the ball gets away from the catcher, players always advance safely. I would like to see the runners thrown out occasionally.

Yesterday I sent Markus a new list of basic plays that "seem" to be missing from OOTPB. I will send him this missing play, too.

Thanks for pointing it out.

bubba22plus · 02-10-2010, 09:03 PM

Just read this after a long break from OOTP.

All i can say is:

Why can't i invite Free Agents to Spring Training yet?

Joking Aside...

Nice list, looking forward to it.

Killing Time · 02-10-2010, 11:44 PM

Quote:

Originally Posted by Shaun Sullivan

Not sure if you saw my apology, but I was wrong and you were right. It has been a rough week

Thanks for your very insightful comments.

Shaun

Whoa!

02-10-2010, 02:41 PM	#282
thbroman All Star Reserve Join Date: Jun 2002 Location: Madison, WI Posts: 867	Some Thoughts on New and Exotic Stats I agree with MD -- this has turned out to be an interesting thread for all sorts of reasons! But something posted by Eugene a couple of pages back has really stuck with me, and I wanted to put it up for discussion: what is it that new stats packages are supposed to do? Eugene claimed -- and I agree with him completely -- that he never looks at the new-fangled stats. I don't either, because in part I don't understand them (and don't care enough to learn how to use them) but also because I wonder about the role of such statistical breakdowns in the game. So what are they supposed to do? When Bill James and GM's use them in real life, they're basically attempting to produce a new method for squeezing out the essence of superior play by an individual from the murky complexity of the game situation. The individual player is a real unit (you can see that for yourself), and since these unitary but mysterious players function at the same level on which the game of baseball is played, it makes sense to try to find effective ways of identifying which players are most valauble in producing winning baseball. At least, that seems to be what many stats are meant to do. But what role do such stats play in OOTP? As much as we all love the game, it is -- ahem! -- only a model, and the players are, sad to say, mere lines in a database. So on the one hand, it's perfectly comprehensible why we would want to have stat outputs on something like L/R batting averages, because we can have faith that the game engine takes L/R batting into consideration when it does it cranking. Thus the L/R averages are "real" in the sense of representing something that the game itself factors into the equation, so to speak. But there are other stats in OOTP that appear to be purely cosmetic --BA in May versus August, for example. The numbers are there in the database, you can "do the math", as they say, but it is not so obvious (not to me, anyway) that the game engine actually considers month-to-month fluctuations in its calculation of results. So here's a question -- when Markus decides to include something like WAR in a forthcoming version of the game, does its inclusion mean that: a) he's just going to dig out some numbers and produce a new column in the already crowded stat output, or b) he is going to add yet another modifier to the calculation of the basic outcome of any single play? In other words, which stats approaches are important enough to warrant inclusion as factors in how the game produces outcomes? just wondering . . . . Last edited by thbroman; 02-10-2010 at 02:42 PM.

02-10-2010, 05:27 PM	#290
actionjackson Hall Of Famer Join Date: Apr 2009 Location: Toronto, ON Posts: 6,179	General question that anyone can jump in on, with a rather lengthy setup/introduction. I'm working on creating a custom Lahman database that will present more of a challenge for a historical dynasty by creating more of a "Fog of War" and level the playing field between me and the AI, editing only the master.csv file. My aim is to take a bunch of randomly generated names courtesy of the random name generator site I mentioned above and plug them into the nameFirst and nameLast columns. Before doing this I think I would blank out the birthState, birthCity (so that whatever Babe Ruth is called I won't be able to figure out that it's him by MD, Baltimore in those two columns), nameNote, nameGiven, and nameNick columns, so I won't be influenced by a George Herman (though I can't recall if given names show up anywhere in the game), here or a "The Bambino" there. I would leave everything else alone. ...Now (finally) for the question. Would writing over the nameFirst and nameLast columns and blanking out the other 5 in the master.csv file affect the functioning of the game? Or does it all just key off of the lahmanID and playerID columns and whatever's in these 7 columns doesn't matter much in terms of game performance? Marcus, Spritze and any other database geniuses, please feel free to jump in. Thanks in advance for your response(s). Last edited by actionjackson; 02-10-2010 at 05:32 PM.

02-10-2010, 06:47 PM	#292
fhomess Hall Of Famer Join Date: Nov 2002 Posts: 3,693	The different files in the database key off of the lahmanID to match players to their stats. I'm pretty sure that the game loads them into its internal database and then doesn't look at the files again until it's time to load the next batch of players. In short, changing the names shouldn't affect anything other than the names they import with. __________________ StatsLab- PHP/MySQL based utilities for Online Leagues Baseball Cards - Full list of known templates and documentation on card development.

02-10-2010, 06:55 PM	#293
fhomess Hall Of Famer Join Date: Nov 2002 Posts: 3,693	On the issue of advanced/sabermetric stats, most of the basic ones are easily calculated from the traditional stats, so they don't require any additional database space to store. Things like basic wOBA and BP's new SIERA can be calculated on the fly. Other stats that require park and/or league adjustments are a bit harder to do as they require calculating something else first, so they might require a bit of database space being taken up. As far as the addition of markUsZR to provide some better defensive statistics, I do hope that the underlying data will be available in the data dump somehow. Speaking of the data dump, I've got numerous suggestions from my work with StatsLab that would be nice to have in there. Some examples: 1. RBOE - reached base on error isn't in there 2. Game Logs aren't reproducable. Would allow for things like WPA to be calculated for the more statistically inclined, but would also allow for evaluation of key game events/external game summaries. 3. Stint - there's currently no way to tell the order a player played for different teams during a single season. 4. Transactions - there's no record of transactions/trades (kind of goes with #3, but provides more) __________________ StatsLab- PHP/MySQL based utilities for Online Leagues Baseball Cards - Full list of known templates and documentation on card development.

02-10-2010, 09:03 PM	#299
bubba22plus All Star Reserve Join Date: Sep 2008 Location: Equestria Posts: 808	Just read this after a long break from OOTP. All i can say is: Why can't i invite Free Agents to Spring Training yet? Joking Aside... Nice list, looking forward to it. __________________ OOTP Resident Brony