This is the same exact game as 2009.

[SrMeowMeow]

You just can't improve on a near prefect game engine, now can you ... lol

Definitely imperfect, and I'm not just talking about results. The fact that the whole engine is built on predicted stats is a big flaw in my opinion. But, it is by far the best engine on the market. Which is why I play it.

[HoustonGM]

Definitely imperfect, and I'm not just talking about results. The fact that the whole engine is built on predicted stats is a big flaw in my opinion. But, it is by far the best engine on the market. Which is why I play it.

How else do you suggest generating statistics?

[cfeedback]

Okay, first I think 2010 is a great upgrade, especially for online leagues. A lot more improved than 2009 was over 2008. I have faith that the some of the lingering issues with player development will be fixed.

Definitely imperfect, and I'm not just talking about results. The fact that the whole engine is built on predicted stats is a big flaw in my opinion. But, it is by far the best engine on the market. Which is why I play it.

How else do you suggest generating statistics?

I think an engine based more on general abilities rather than specific stats would be an improvement. Right now you can make a guy with 20 predicted triples and 60 speed and he'll get a ton of triples?

A better way to generate said player would be to have something like "gap power" + speed = more doubles/triples.

Somewhat related would be what type of hitter the player is in general. Example: a dead pull hitter who would accordingly get more HRs in a stadium with a short porch on the side he's pulling the ball to.

[HoustonGM]

I think an engine based more on general abilities rather than specific stats would be an improvement. Right now you can make a guy with 20 predicted steals and 60 speed and he'll get a ton of triples?

He won't get a ton of triples if he isn't predicted too. And, also, he likely won't have a 60 speed with 20 predicted steals unless you predict him to ground into a ton of double plays, have terrible range, and get caught stealing a ton.

A better way to generate said player would be to have something like "gap power" + speed = more doubles/triples.

I don't see how that's better than specifically allocating how good a player is at hitting doubles, how good he is at triples, how good he is at SB, etc.

Somewhat related would be what type of hitter the player is in general. Example: a dead pull hitter who would accordingly get more HRs in a stadium with a short porch on the side he's pulling the ball to.

This is something I'd be interested in seeing, but I think that it can be implemented without abandoning predicted stats.

[Jeffy25]

This is something I'd be interested in seeing, but I think that it can be implemented without abandoning predicted stats.

I agree, a question was raised in the manager thread. I wish the animation would sometimes play out better, that you could put on the big Papi Bonds shift etc....And knowing if a hitter was a spray hitter etc...i tell ya, it's confusing if you have someone who you know in real life is a dead pull left handed hitter and he is getting half of his hits to the third base side

[cfeedback]

He won't get a ton of triples if he isn't predicted too. And, also, he likely won't have a 60 speed with 20 predicted steals unless you predict him to ground into a ton of double plays, have terrible range, and get caught stealing a ton.

Oh, I typo'd it...I mean 20 predicted triples not steals lol..

Basically I'm saying the engine now will allow you to have a slow 1B type that hits a ton of triples, and that's not particularly realistic?

Gap power would indicate a guy who hits more doubles/triples, while HR power is a separate rating. How likely a hitter with high "gap power" gets those doubles and triples would be based largely on his base running speed and also the range of the opposing outfielders.

Prince Fielder might have a high gap power (in addition to HR power) but not leg out many triples, where the exact same player with higher speed gets accordingly more triples. I can't really explain why I think this is better, just seems more realistic.

[cfeedback]

One other random thought on predicted stats. I think that the "eye" rating should be split up somehow. Higher predicted Ks for a hitter lowers his eye rating, but striking out isn't really that bad for a batter. Yes he may fail to advance a runner, but a strikeout also can't lead to a double play. Eye should be a reflection of how many pitches a hitter sees on average in an at bat.

Plenty of great hitters walk a ton and strike out a ton as well, and usually they're damn good hitters. Three true outcomes and all that, ya know.

cfeedback = honorary member of the Rob Deer Fan Club

[dps]

Then buy it every 3 years. Or 5 years. Whatever makes you happy. Why come on and hurl accusations and attitude?

I basically buy it every other year, though that's more of a budget thing than a complaint about the product. If I still had the same financial situation that I had back when I first discovered Baseball Mogul (1998, IIRC), I'd buy the new edition every year.

My biggest fear is always that they will change it too much, not that they won't change it enough.

[HoustonGM]

Basically I'm saying the engine now will allow you to have a slow 1B type that hits a ton of triples, and that's not particularly realistic?

That the game doesn't create such players on its own makes it not really a problem.

Gap power would indicate a guy who hits more doubles/triples, while HR power is a separate rating. How likely a hitter with high "gap power" gets those doubles and triples would be based largely on his base running speed and also the range of the opposing outfielders.

Prince Fielder might have a high gap power (in addition to HR power) but not leg out many triples, where the exact same player with higher speed gets accordingly more triples. I can't really explain why I think this is better, just seems more realistic.

And given that the players Mogul generates generally DO follow these things you're talking about, I don't see why the current system is a problem. You can create unrealistic players, sure, but you could do that with any system that you come up with, so long as editing players is allowed (which it needs to be).

One other random thought on predicted stats. I think that the "eye" rating should be split up somehow. Higher predicted Ks for a hitter lowers his eye rating, but striking out isn't really that bad for a batter. Yes he may fail to advance a runner, but a strikeout also can't lead to a double play. Eye should be a reflection of how many pitches a hitter sees on average in an at bat.

Plenty of great hitters walk a ton and strike out a ton as well, and usually they're damn good hitters. Three true outcomes and all that, ya know.

That's a disagreement with what the "Eye" rating stands for, then. You'll notice, though, that strikeouts are a minor factor. Adam Dunn still has a 90 eye. Anyway, just because it lowers the eye rating doesn't mean that the player will play worse. The predicted stats determine how he plays, not the ratings.

[Lex Logan]

Exactly. Whenever you choose to work, that's fine. But the end result should be something satisfactory. He created a good, entertaining game years ago, but what I am saying is that he has done little to improve upon it.

He's done a lot to improve it. Unhappily, however, the game has become worse in at least one key area -- player development. Having given up on BM 2009 and reverted to BM 2006, I miss a LOT of things from 2K9. And there are several improvements in 2K10 I'd like to have. But I'm not paying for a version I won't enjoy playing, and I did not enjoy 2K9.

[SrMeowMeow]

How else do you suggest generating statistics?

Sorry, I forgot about my post. The way I think baseball players actually work, simplistically, is that they have two components: raw ability, and approach. A quick example: say you have average power. If you swing for the fences all the time, you might hit 15-20 homers and 20 doubles and strike out often. If you shorten your swing and hit line drives, maybe you'll hit 45 doubles and 10 homers, and strike out much less. I would much rather see an engine based on this than predicted stats. You'd get much more organic results. As in, player A swings at a pitch. They have X chance of making solid contact. If they make solid contact, depending on their approach, they have d e and f chances of hitting a grounder, a line drive, and a fly ball. It goes a certain distance based on power and luck. The fielder gets a good or bad jump depending on how good he is at reading the ball, he covers as much ground as his speed allows, he has a chance to misplay the ball depending on how good his glovework is, etc. With an engine like this, if a player hits a fly ball 310 feet, depending on what direction it goes and where the fences are, it will simply clear or not clear the fence. So, to conclude, that is how else I would suggest generating statistics. Obviously, I don't know the details of Mogul's code, and I'm not going to come up with a flawless baseball simulation engine in a three minute forum post. But the insight into Mogul's architecture that Edit Player grants - that predicted stats drive the ratings and not the other way around - reveals a flaw in the design, in my opinion. It still works, and works pretty well, and is way too entrenched to change now, but it has problems.

My favorite example is contact rating and strikeouts. In the real world, low strikeout totals are a great predictor of future success because putting the ball in play at all, even if it's hit badly or awkwardly, gives the player a chance at a base hit. A groundout isn't better than a strikeout, and is often worse, but a groundball is better than a strikeout. However, in Mogul, if you lower predicted strikeouts, contact rating will go up, and you end up with players who hit .270 with a 95 contact rating. They strike out 50 times a season, but they're apparently hitting into an ungodly number of outs. While there are players with naturally suppressed BABIP (also known as bad hitters), the predicted stats engine doesn't care what the ball in play is. It rolls the dice, and strikeouts and hits are both relatively unlikely after everything else has been factored in, so hey, a pop up. If you had an organic engine, the ball in play would go to a place and sometimes in that place it finds a fielder and sometimes it doesn't.

EDIT: My last objection to the current engine is that it's just unrealistic. In real life, a hitter gets a hit depending on how good he is. His power and contact skills are not derived from his production. This may seem like a technicality, but it's really not. A truly good engine would simulate the process as well as the result. It's like the cognition problem: would you consider a computer (or robot) sentient if it literally had been programmed with every possible situation, and could therefore respond naturally and "sentiently" in any situation? I mean, if we're just looking at season stats, as a simplification, I could create a simulation engine pretty easily that took predicted stats for the first year, randomized them a little bit, and spat them out as "end of year stats". There's no way to tell me that it's unrealistic, because the results fall within believable parameters. But it's clearly not a perfect simulation, or even a simulation at all. If you extrapolate that up, I think you can see why I don't like the argument that "well, it works, and it's pretty complex as it is, so it's probably good".

[HoustonGM]

I don't have much to say regarding the predicted stats think except that I'm perfectly fine with the way the game is now and don't see how "your way" would be any different. The game bases its simulation on the predicted stats, not the ratings. The ratings are just for evaluation.

My favorite example is contact rating and strikeouts. In the real world, low strikeout totals are a great predictor of future success because putting the ball in play at all, even if it's hit badly or awkwardly, gives the player a chance at a base hit. A groundout isn't better than a strikeout, and is often worse, but a groundball is better than a strikeout.

I disagree with the part I bolded. I've seen no evidence regarding that. I don't believe strikeout rate is a good indicator one way or the other of future success or failure.

However, in Mogul, if you lower predicted strikeouts, contact rating will go up, and you end up with players who hit .270 with a 95 contact rating. They strike out 50 times a season, but they're apparently hitting into an ungodly number of outs.

Since this thing is purely user-initiated and doesn't exist on its own (because Mogul, for the most part, creates realistic players), I don't think it's that big of a deal...at least not something that warrants a complete rewrite of the rating system.

If you had an organic engine, the ball in play would go to a place and sometimes in that place it finds a fielder and sometimes it doesn't.

I'm pretty sure that the game does do something like this, although only Clay can confirm.

[SrMeowMeow]

Well, I added a last paragraph, probably while you were replying, so tell me what you think about that. As for strikeout rate, I don't know what to tell you. When I say predictor of future success, I'm talking about in the minor leagues, which obviously doesn't apply in Mogul. And obviously editing players doesn't come up in normal play, but it does reveal things about the engine. The game will happily produce 90+ contact, .270 or lower batting average, low K players on its own.

Finally, I'm not sure why you're even arguing. I mean, I think we agree. I already said that I think the engine is the best on the market, just that it's not perfect. I think you might be needlessly defensive on this one. I'm glad you're fine with the current system, and I am too. I don't think that should cut off debate on its flaws, though...

[HoustonGM]

The game will happily produce 90+ contact, .270 or lower batting average, low K players on its own.

I've never seen such players.

Finally, I'm not sure why you're even arguing. I mean, I think we agree. I already said that I think the engine is the best on the market, just that it's not perfect. I think you might be needlessly defensive on this one. I'm glad you're fine with the current system, and I am too. I don't think that should cut off debate on its flaws, though...

A couple of things.

a) I'm not arguing. Sorry if you felt like I was. Just discussing this because I'm trying to better understand where you're coming from.
b) I don't want to cut-off debate on its flaws.
c) I see what you're saying. I don't think an overhaul of the system is needed though. What could be done is add another level of "sanity checks" to the batters, to "force" the predicted BABIP to be realistic. The corresponding check for pitchers is not without its issues, but I think that the linked issue could be fixed and ported over to the batter side as well and take care of the issue you're mentioning (which, I agree, is a legitimate flaw).

[SrMeowMeow]

Well, all I'm arguing with is the idea that it's perfect. I don't disagree that it's good. I just don't like the predicted-stats premise.

Also, here's two players I pulled off a random file: