Historical replay accuracy

[Siv]

Originally Posted by Carlton_Willey

Siv makes great points and leads to an addition to all the current sports sims I play--OOTP, FOF, and FM, mostly--that I'd love to see implemented in some form.

I'd like an indication of the die roll and whether or not it was a lucky one. These games have lots of die rolls for each event, I think and not one 'handle pull' but I'd still like to see a comment about luck when an almost-off-the-chart event occurred.

A lot of the perception of bad design would go away if the player could see the die rolls. I play board games, and its improbable to roll three sixes three times in a row, but I've done it and seen it done. When this happens we don't blame the game, we blame the dice

Perhaps a 'luck factor' column added to individual totals at the end of the season so we could determine the impact of the die rolls on the player's performance. It could be based on standard deviation, and done broadly. One, two or three STDs up or down, from expected would be a useful indicator for the die roll impact.

I started this thread thinking I was secure in my belief of the historical replay, but I kind of had an epiphany and now feel I may need a baseball game like strat, where they use each season and not a combined one. Hmmm thanks guys more money to spend i guess.

[knockahoma]

Wow I haven't thought of Strat O Matic in years! I would love to see the dice roll and cards, man that was awesome to play. i would sit for hours playing the season I had (1979) and keeping all the stats... I honestly thought they were out of business.

Not hardly. . But speaking of Historicals, I guess that's what I'd like. Feeling the highest respect for both companies, I would like to see OOTP have STRAT's season accuracy as the base-line, and from there, we can fiddle with options, until our baseball world is as wacky as we please.

It would be the perfect blend. And, good news for Markus. I would think he's got a head-start on the historical companies.

I think OOTP developing its own historical database would be a quantum leap forward for the game. But, the AI needs to understand that managers put different value on the glove at different positions, and in different lineups.


Gosh. Painting a masterpiece is so easy when you slop a big brush over a wall and tell the next guy, " Now, just paint some detail and we got the Mona Lisa!"

[StyxNCa]

Originally Posted by RonCo

Nice discussion. Most people, and I mean this nicely, have no real idea how random "random' is.

And some people have no idea just how significant an "insignificant" number like 7% can be in a sports environment.

[damientheomen3]

7% can mean the world in sports...

[jbergey22]

Originally Posted by StyxNCa

And some people have no idea just how significant an "insignificant" number like 7% can be in a sports environment.

No offense Styx but RonCo understands this stuff. You may not want to believe what he has to say but he is right.

OOTP is trying to recreate history without wrapping it in a box. Im sure Ron could answer this but what are the odds all 30 teams would finish in the exact position in the standings(assuming everything is equal)? 1 in a million?

What are the odds that Ichiro hits 350 one year and 300 the next? How does Andrew Jones as one of the premier talents hit 50+ homers followed by a complete collapse?

You want something that isnt possible nor realistic. Strat o Matic takes the random aspect out of the game to give accurate results if that makes sense. A baseball simulation in order to be realistic has to have the random/luck factor in it.

Your argument is that Out of the Park isnt an accurate historical sim which is not true. The results wont be as close to what they were as you might like but that is part of baseball. That is why the 91 World Series matched 2 teams that finished last the year before.(Do you think they really improved that much or would you guess they were a bit unlucky the prior year and lucky during their World Series year?)

When the World Series is done every year its very rarely I actually feel the best team(throughout the season) actually won the world series. Its a skill game where 2 lucky bounces can determine the winner and the loser.

If you want a replay sim that is fine but to say OOTP doesnt do a good job with historical for the reasons you give is incorrect.

Whats that comment from Bull Durham? Two lucky flares a week is the difference between a .220 hitter and a .300 hitter?

Replay sims and historical sims ARE different.

[RonCo]

Styx -- dude ... feel free to wail away at me all you want. I have no energy for someone who openly admits to merely playing games, and will not engage in detailed discussion with you on anything. But I'll just say that in all this talk of being "insignificant" you're taking me way, way, way out of context.

[BaseballMan]

Where do you draw the lines with accuracy?
To me the results are better since recalc.
I know that Ruth has the ability to hit 700 hr but i dont know for sure that he will. However i do know that Jose Canseco wont and Ozzie Smith should never ever even come close to 700.
I no longer try to replay for the exact results of real life. Oh sure i wouldnt mind having the real rosters and trades but even with injurys off it wouldnt bother me if the Orioles finished 5th in ERA. It might bother me if they constantly finished 10th or higher in ERA.
But for me its a waste of time to try and duplicate what happened in real life.
I got baseball encyclopedias for that. I prefer to give the players their abilities and think of it has a 2nd chance. But even in a 2nd chance Ozzie Smith should not have the ability to hit 300 hrs.
But if a game could duplicate the exact results of the real thing how much fun would it be? Then i suppose we would be discussing that the game isnt inaccurate enough.
I had a replay in which Ty Cobb hit .367 but had 4,400 hits. Should i be concerned that he had 200 more hits? Or does it have to be .367 with 4,200
for it to be realistic?

[knockahoma]

Where do you draw the lines with accuracy?

As a business venture, I think it would be extremely beneficial to draw it at the top (match at least where Diamond Mind and Strat perform).

Then have a game engine that acts like an old stereo knob, where you turn the accuracy up, or down, according to the player's taste. That would be the ultimate for the player and the company. And again, I think Marcus has a real head-start over Strat and Diamond Mind for the perfect blend of history and fantasy. He's got the fantasy part down and he's got a very respectable historical engine.

But, I don't think OOTP gets to historical "top form" with just the Lahman database.

[RonCo]

If you want to get more precise than the examples above, you have to remove randomness from the equations.

[Cryomaniac]

Originally Posted by RonCo

If you want to get more precise than the examples above, you have to remove randomness from the equations.

And that defeats the object.

[knockahoma]

If you want to get more precise than the examples above, you have to remove randomness from the equations.

And that defeats the object.

But, randomness exists within STRAT and Diamond Mind.

I think the general consensus is that those two are better historical "replays". To me, since OOTP has ventured into the historical replay arena, why not aim at matching what those two companies can do? Of course, it would be important that the user can "dial down" the replay if he wants more variation.

Is that do-able with OOTP's engine? I'm guessing it is. It appears that STRAT and Diamond Mind's advantage comes from better player information. That's why I suggested to reach the next level in historical replay OOTP would need more than the Lahman database.

I'd never really looked at Diamond Mind until this past week. I can tell you that the developer's over there see the possibilities of a top-notch replay with the ability to dial down into more variation. You can now edit and create players in particular seasons, thereby creating your own fictional league.

I really appreciate OOTP innovation, lay-out and options. But, as this thread proves with SIV, a few others and myself, there's money to be made, or lost, on the historical aspect.

I just think Marcus has an advantageous lead on the perfect blend. I'd like to see him win the competition, cuz it's coming.

[BigRed75]

Originally Posted by Cryomaniac

And that defeats the object.

And makes the game quite a bit more boring.

For my historical needs, I'm more concerned with not having real-life transactions and lineups. Ending up with trades that didn't happen and superstars buried on the bench by the computer frustrates me a lot more than statisical variances that are within the realm of possibility.

The difference between a .300 hitter and a .250 hitter (in 600 AB) is 30 hits. That's 1 1/3 a week over the course of a 22-week season. As far as I'm concerned, the statistical side of things is fine for historical play.

[knockahoma]

The difference between a .300 hitter and a .250 hitter (in 600 AB) is 30 hits. That's 1 1/3 a week over the course of a 22-week season. As far as I'm concerned, the statistical side of things is fine for historical play.

I think it's probably pretty decent as a computer roll of the dice. It's what gets lost in the translation from lahman to the engine that may be a big part of the problem.

I have played 1974 and 1976 dozens of times. Phil Niekro constantly gets rated as a number 4 or 5 starter, constantly comes in with ERAs around 4.75 or 5.00, usually goes 7-19, or something. Similarly, the LA Dodgers staff is translated into mediocre fare, usually landing them in 5th, or 6th place. Mark Belanger is translated into an average fielding shortstop who hits .230, rather than one of the best fielders ever, thereby making him very expendable to the Orioles.

Over and over again the players are mistranslated. I don't mind different paths for Phil Niekro. But, in the first season, I'd like him to be at least a distant relative to the real deal.

So, the approximation to history is already pretty loose from the get-go.

Again, Diamond Mind is now including the ability to edit players and create players for that fictional market, working downward from a top-notch historical premise. I think Marcus has the lead in making the perfect blend of historical and fictional, but it appears the historical boys are expanding into fictional territory.

I like OOTP very much and will probably be spending more dollars on it. But, like Siv and others, I really want a top-notch historical replay.

It would be nice to see it all in the same package with all OOTP's great bells and whistles included.

I also wonder how much better the AI would do with more accurate representations of historical players? Would the dumb trades diminish? Would Mark Belanger be valued as a top-notch defensive shortstop instead of being an Oriole cast-off?

Lahman alone leaves a lot to be desired for any sim engine.

[Cryomaniac]

Originally Posted by knockahoma

I think it's probably pretty decent as a computer roll of the dice. It's what gets lost in the translation from lahman to the engine that may be a big part of the problem.

I have played 1974 and 1976 dozens of times. Phil Niekro constantly gets rated as a number 4 or 5 starter, constantly comes in with ERAs around 4.75 or 5.00, usually goes 7-19, or something. Similarly, the LA Dodgers staff is translated into mediocre fare, usually landing them in 5th, or 6th place. Mark Belanger is translated into an average fielding shortstop who hits .230, rather than one of the best fielders ever, thereby making him very expendable to the Orioles.

Over and over again the players are mistranslated. I don't mind different paths for Phil Niekro. But, in the first season, I'd like him to be at least a distant relative to the real deal.

So, the approximation to history is already pretty loose from the get-go.

Again, Diamond Mind is now including the ability to edit players and create players for that fictional market, working downward from a top-notch historical premise. I think Marcus has the lead in making the perfect blend of historical and fictional, but it appears the historical boys are expanding into fictional territory.

I like OOTP very much and will probably be spending more dollars on it. But, like Siv and others, I really want a top-notch historical replay.

It would be nice to see it all in the same package with all OOTP's great bells and whistles included.

I also wonder how much better the AI would do with more accurate representations of historical players? Would the dumb trades diminish? Would Mark Belanger be valued as a top-notch defensive shortstop instead of being an Oriole cast-off?

Lahman alone leaves a lot to be desired for any sim engine.

As someone said earlier, maybe OOTP needs to develop their own historical database (Including minors if you don't mind it taking years! )

[PhillieFever]

Originally Posted by actionjackson

Just for clarification, I'm talking about long-term sims as well.

I'm sorry for getting back so late,I was talking about ballpark factors.I use the default DB most times when I play.

[knockahoma]

As someone said earlier, maybe OOTP needs to develop their own historical database (Including minors if you don't mind it taking years! )

Yeah, that was me. I think it may be the only way to get to the higher level. For me, it's not so much that Phil Niekro's career path goes south. It's that the import named Phil Niekro was a representation of the player in name only.

But, I'm just spit-balling at targets I can barely see. I was looking at some pitchers over the weekend from the 74 and 76 season and wondering what OOTP saw in Lahman that made Phil Niekro a bad pitcher 18 out of 20 seasons. It wasn't just that Phil "developed" into a bad pitcher. He COMES IN rated as a lower rotation pitcher. I'll have to do a study on whether the 1 year, 3 year, or career stats import has any effect on this.

Strangely, this seems to happen to a lot of low strikeout and walk pitchers who performed well a number of seasons. I'll look at their homer numbers, K's and walks. Several had similar numbers there. I hate to go open that old can of worms about BABIP, but that's what seemed out of line and the reason the pitchers had such high ERAs, thereby under-performing. I just didn't see anything else. The rest of the numbers seemed reasonably real-life. But that was just checking in on several disappointing pitchers, not anything close to a study. Maybe I'll do one and bring it here for comment.

Regarding fielding, it's clear Belanger's numbers don't match the perception of him-- yet, I'd be inclined to suspect that's due to the faulty nature of Range Factor numbers, rather than Belanger was over-rated and merely an average shortstop in range.

Again, I like so much in OOTP, I don't mind editing. I find it fun. But, I think I'm missing some pieces on how to manipulate the engine to my satisfaction.

And definitely, I cannot find a way to make the Orioles appreciate a guy like Belanger, even after editing him to superlative shortstop. I thought of raising the Orioles appreciation for defense, but that's global. Thinking of their line up in 74 that might work out alright. But, it doesn't work for other teams which batted sluggers while keeping a good glove at the pivotal SS position. So, I dunno...

[PhillieFever]

Here;s my take on historical replays as far as OOTP goes.Its a good game,good not great,if you want to play through an era of baseball.Playing one season with OOTP is just a waste of time,the engine just stinks for one year replays.But,if you change some of the default settings,sim one season with trades on very low and injuries off then take over a team you really can have some fun with the game .First off I recommend these settings:

DB-default
L/R splits-Off
Ballpark factors-Off
recalc-3 years(anything else it seems kills the results)
retire according to history-On
Player development-disabled(another must)
strategy-default
Financials-No free agency prior to 76,everything else default
Trades-very low
Injuries-low/very low with fatigue on high

By doing this,I've regularly gotten good results,not perfect mid you,but good believable results.Combine that with the fact that you can play through multiple seasons without season disks and I'm sold.

[NHSOXFAN]

Originally Posted by knockahoma

Again, Diamond Mind is now including the ability to edit players and create players for that fictional market, working downward from a top-notch historical premise. I think Marcus has the lead in making the perfect blend of historical and fictional, but it appears the historical boys are expanding into fictional territory.

It is important to remember that, up until a few months ago, it was unclear whether Diamond Mind was going to continue to support and develop it's PC game. It's important to keep that in mind moving forward. They did not produce a "projection disk" this season for the first time since 1999.

As for their fictional offering, I'm curious as to how that would work. If you create a fictional player on the 2008 season disk, do you have to recreate them all again each subsequent year?

[Siv]

Well I think I will try OOTP X, since I preordered and do some replays that way for now. Thanks for the info on DMB.

[pstrickert]

IMO, the batting and pitching models work quite well for single-season replays (i.e., they produce highly credible results). The fielding model, however, is not up to par. It is no surprise, then, that (in my experience) teams that excel in both pitching and defense do not "translate" adequately into the OOTP world. Markus indicated that he'll work on the fielding model in OOTP11. In the meantime, it would be wonderful if someone created a utility to make fielding modifications easier. Don't know if that'll happen, though.