Checking Out OOTP Again

[DaddyO]

I bought 2006, 2007, 2008 and 2009, but not 2010.

Now that DMB is REALLY going kaput this time, I'm looking at options. I have a newer computer now with a better monitor, and Skunkle has encouraged me to give this product another chance.

I am a historical replayer.

In looking at the online manual, I don't see an option to calculate a player's potential ratings based on a user-selected range of years or a rolling 3-year average (my wish-list approach). Is this correct?

Since the game DOES ship with a normalized historical database, how doable is it to customize that so you end up with my wish list approach? I can use Excel. Can I export the database to Excel, modify it, then import it back into the database?

Thanks for your help in answering these questions.

[skunkle333]

You will get some good help here...DaddyO...

Have a good day...

[DaddyO]

Thanks Skunkle. Hey, when you get a chance, and if you are willing, send me a screen print with your family's faces in the game. That sounds like a hoot!

[Charlie Hough]

Actually, if you have 3-year recalc turned on, the game bases potential ratings on that same 3-year window. It actually overrides the setting for basing potential on remaining years of career, peak seasons, etc. OOTP has been set up like this since the 3-year recalc was established.

However, Markus is adding a new option in the next patch so that the 3-year recalc won't override the potential settings.

One thing you can't do, however, is select your own range of years for potentials.

[DaddyO]

Charlie, thanks. I was just reading your thread on the problems with recalc and potential ratings. I'll follow this issue closely.

Part of me wonders if the whole thing is a matter of trying to push a square peg through a round hole, the square peg being historical replay and the round hole being a game designed to function best with fictional play.

I'll have to decide which way to go, since I don't really have time or energy for both.

I'll just experiment until I can decide.

[DaddyO]

How well does the 3-year recalc Historical Option work?

Personally I think single season replays often suffer from the "Outlier of an Outlier" syndrome. By in 1979. Willie McCovey hit 18 in 1964. Players can woefully underperform or overperform for a single season in average, or any other statistic as well. That's OK. You're replaying a single season, after all. But the trouble is that simulation randomness also introduces outliers. This amplifies the outlier effect already existing in real life. So in your 1979 replay it's not out of the question that Lopes would hit 35 home runs, or even more. McCovey might hit 11 home runs in your replay. By multiplying the outlier effect you come up with even more extreme outliers that don't really reflect real life very well.

The best way to limit this problem as I see it is to smooth out a player's single season baseline to an average of surrounding years. How many surrounding years is best is a question that could be answered more than one way depending on the user's need, but for me to make the best compromise between the need to smooth out outliers and the need to retain the flavor of a slice of time in a player's career you would use a 3-year period. The replay itself will add some randomness so you don't know for sure just how a player will perform, but you have a good idea what to expect. All this mirrors the position of management in real life (except you are using the following year as well as the previous one).

I'm thinking of creating my first Historical League starting in 1966. I'll see how this works for myself, but I'd be interested in the opinion of others as I launch into it.

[DaddyO]

I'd like to add another question that's not clear to me:

If I select Neutralized Stats, which I am inclined to do, this assumes the simulation takes into account park effects, league effect, etc. That way, if a player plays in his historical park the park effects will move him back to his historical performance baseline. If there are no effects influencing the simulation, then you've neutralized stats but left out the effects that makes this necessary in the first place. You've reverse engineered the stats, but you haven't re-engineered them back in when you run the simulation.

Does OOTP do this?

One more:

If I change the team abbreviation on the historical league from the one supplied by the Lahman Database, does that screw things up now or later? On my stats I prefer the abbreviation WAS instead of WS2, CHC instead of CHN, etc.

[Spritze]

Originally Posted by DaddyO

I'd like to add another question that's not clear to me:

If I select Neutralized Stats, which I am inclined to do, this assumes the simulation takes into account park effects, league effect, etc. That way, if a player plays in his historical park the park effects will move him back to his historical performance baseline. If there are no effects influencing the simulation, then you've neutralized stats but left out the effects that makes this necessary in the first place. You've reverse engineered the stats, but you haven't re-engineered them back in when you run the simulation.

Does OOTP do this?

Yes

Originally Posted by DaddyO

One more:

If I change the team abbreviation on the historical league from the one supplied by the Lahman Database, does that screw things up now or later? On my stats I prefer the abbreviation WAS instead of WS2, CHC instead of CHN, etc.

It screws things up now AND later. So you get a dual screw things up with one change.

[DaddyO]

Thanks Spritze. It's very helpful to get direct answers to questions.

[Charlie Hough]

By the way, you can choose a 3-year recalc or a 5-year recalc. So at least you have that much flexibility.

OOTP also automatically adjusts players who have unusually high or low statistics but had relatively few at bats or innings pitched. You can also change these settings to your liking in terms of how much they're adjusted and what the usage thresholds are.

Overall OOTP does a pretty solid job keeping things reasonable and fairly accurate, especially when using a 3-year recalc. For the closest thing to single-season performance, the obvious choice is a 1-year recalc. But these other settings like the player adjustments should prevent the outlier effect that you're worried about. It probably won't be perfect, but the instances of major deviation from real life performances are fairly rare.

Once Markus straightens out the issue over recalc's effect on potentials, I think OOTP is going to be a very complete historical simulation engine. Obviously there can and still will be improvements like possibly adding automatic real life transactions or getting late 19th century sims closer to real life conditions. But to me OOTP is now very close to being the ideal historical career sim.

[DaddyO]

Thanks CH.

[TheKenoshaKid]

Originally Posted by DaddyO

How well does the 3-year recalc Historical
The best way to limit this problem as I see it is to smooth out a player's single season baseline to an average of surrounding years.

Oh wow, I just posted a topic today about this. Guess I didn't look hard enough, since it was right in front of me the whole time

I have a question, though. How do I "smooth out a player's single season baseline"? I'm assuming that means decreasing the weight of the current year's stats and increasing the weight of the previous and future years. So 3 years in a row with weights "20-60-20" would turn into something like "25-50-25". I don't know how to do this though.

I hope I explained all this good enough. Kinda hard to keep it short when my problem is so specific

[Charlie Hough]

You can't assign weights to each season in a 3-year or 5-year recalc. You can double the weight of the current season, but that's your only option. You can't set weights for the previous or subsequent season.

[DaddyO]

Originally Posted by TheKenoshaKid

Oh wow, I just posted a topic today about this. Guess I didn't look hard enough, since it was right in front of me the whole time

I have a question, though. How do I "smooth out a player's single season baseline"? I'm assuming that means decreasing the weight of the current year's stats and increasing the weight of the previous and future years. So 3 years in a row with weights "20-60-20" would turn into something like "25-50-25". I don't know how to do this though.

I hope I explained all this good enough. Kinda hard to keep it short when my problem is so specific

Well, the way I was conceiving of the 3-year recalc turns out to be slightly wrong, I think. It is not the year before, the current year, and the year after, it's the two years prior and the current year. At least I think this is correct.

But whatever, the very process of using 3 years instead of one accomplishes the smoothing by itself. Take the Lopes example I gave above. He hits 28 in 1979. But in '78 he hit 17, and in '77 he hit 11. 3-year recalc would calculate his baseline at 56 (11+17+28) divided by 3, end result is 18.7 HRs as a baseline, much "smoother" as an average than the 28 he hit that year.

The option I see in OOTP is to double the weight of the current year, which would mean something approximately like this: 84 (11+17+28+28) divided by 4, end result is 21 HRs as a baseline. That means less of the "smoothing" effect I was talking about.

Hope this helps.

[TheKenoshaKid]

Originally Posted by DaddyO

Well, the way I was conceiving of the 3-year recalc turns out to be slightly wrong, I think. It is not the year before, the current year, and the year after, it's the two years prior and the current year. At least I think this is correct.

But whatever, the very process of using 3 years instead of one accomplishes the smoothing by itself. Take the Lopes example I gave above. He hits 28 in 1979. But in '78 he hit 17, and in '77 he hit 11. 3-year recalc would calculate his baseline at 56 (11+17+28) divided by 3, end result is 18.7 HRs as a baseline, much "smoother" as an average than the 28 he hit that year.

The option I see in OOTP is to double the weight of the current year, which would mean something approximately like this: 84 (11+17+28+28) divided by 4, end result is 21 HRs as a baseline. That means less of the "smoothing" effect I was talking about.

Hope this helps.

Yeah, actually that's exactly what I was looking for. I think I made the same mistake you did.

Thanks! all cleared up now (i hope)

[Elysian Fields]

Originally Posted by Spritze

Yes

It screws things up now AND later. So you get a dual screw things up with one change.

How does it screw things up again?

[skunkle333]

Originally Posted by DaddyO

Thanks Skunkle. Hey, when you get a chance, and if you are willing, send me a screen print with your family's faces in the game. That sounds like a hoot!

Oops... I went and checked the thread you are referring to and I guess I was allowing my goal to interfere with reality. Probably a mixture of when I did something like this a few years back with Action and Strat. You can't do personnel with them but I did it with player pics.

So back to my Oops. I actually haven't put the pics in yet with my family. Too much other work to do. Anyway...if I can remember...I'll try to capture a screen and send it to you when I accomplish the goal.

Take care...and make some wise decisions about gaming down the road!

[Charlie Hough]

Originally Posted by DaddyO

Well, the way I was conceiving of the 3-year recalc turns out to be slightly wrong, I think. It is not the year before, the current year, and the year after, it's the two years prior and the current year. At least I think this is correct.

Actually, you had it right the first time. It's the season before, the current season, and the year after.

[StyxNCa]

Originally Posted by DaddyO

Now that DMB is REALLY going kaput this time, I'm looking at options. I have a newer computer now with a better monitor, and Skunkle has encouraged me to give this product another chance.

Quite a few historical replayers are using Action! Baseball for that since DMB is apparently on the way out and OOTP hasn't been exactly what they wanted for that. Have you checked into Action?

You didn't mention if you replay history or individual seasons but Action is only good if you like to play individual seasons. Most of us who use Action for replaying seasons still use OOTP for fictional play.