Stats-Based CPU Sliders [MLB 12 Version]

[bullpen]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Hi!
I'm a manage only player.
I've played all of spring training with your set and everything was realistic.
Then I started regular season...
After 6 games my era is 7.80.
My team (Toronto Blue Jays ) is hitting for. 325 (with 14 homeruns).
My last 3 games ended 12-5, 12-5 and 19-5 (againt Boston ).
I know 6 games is a small sample but That's alot of offense.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]



Toronto's doing very mediocre in my season. :P

Fenway is a very hitter friendly park (at least in the game), so those games can happen...

[DarkSith777]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Is it safe to say that lowering foul frequency will increase swings and misses?

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Is it safe to say that lowering foul frequency will increase swings and misses?

It seems it does... At Foul Frequency = 5, the swing and miss rate was 17.9% (in 231 games). At 4, it's 18.6% (in 90 games).

It might be within the "margin" of error given the latter is only after 90 games, but it makes sense from the description of the slider.

[bubs3141]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Hi!
I'm a manage only player.
I've played all of spring training with your set and everything was realistic.
Then I started regular season...
After 6 games my era is 7.80.
My team (Toronto Blue Jays ) is hitting for. 325 (with 14 homeruns).
My last 3 games ended 12-5, 12-5 and 19-5 (againt Boston ).
I know 6 games is a small sample but That's alot of offense.

ist weird but I have very similar results with the yankees. Very high scoring games every time. And my starters only can go 3 to 4 innings before having had given up 6 or 7 runs.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

<a href="http://www.flickr.com/photos/78426415@N02/6918943500/" title="2012-04-10-140757 by baseball17k, on Flickr"><img src="http://farm6.staticflickr.com/5322/6918943500_7808343339.jpg" width="500" height="283" alt="2012-04-10-140757"></a>

The Yanks are obviously a good offensive team and doing well in my season as well, but not to the extent that every game is a blowout.

[kingdevin]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Nomo

Curious about your theory re: CPU v CPU and difficulty levels.
Also I play with fouls at zero. I don't see a signif difference in results

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Nomo

Curious about your theory re: CPU v CPU and difficulty levels.
Also I play with fouls at zero. I don't see a signif difference in results

Foul Freq. does reduce the number of fouls, but its effect on swing & miss may not be as large as I thought... The numbers that I cited above, you could say it's within the margin of error. But then where does all the borderline fouls go? It must be poor hits or swing & miss right...?

Anyways, my theory about the difficulty levels is that those things probably are added in the "user interface" layer.

Because we know from experience CPU's strategies (or execution I should say) against us differ at different difficulty levels, some mechanism must be adjusting that part of the game. Since Brian@SCEA says that his AI is not aware of the opponents' difficulty level, it is likely not done at the AI level. Then where can it be? I think it's the layer that exists in the HUM vs. CPU game, but not CPU vs. CPU.

What I mean by the above is that if the game (not CPU/AI) detects that you a user is hitting at Rookie level (for example) in which we know CPU pitcher throws more strikes than he does at an AI-neutral difficulty level (i.e., the setting in which AI fully goes by its strategy), then the game "overrides" the AI's strategy and let the CPU pitcher throw more strikes at early counts.

So strictly speaking, Brian's AI isn't aware of difficulty levels. In CPU vs. CPU games, the difficulty levels of one team against another aren't even considered, so AI fully makes and executes it's strategy. In HUM vs. CPU game, there is a layer that gets inserted that applies the HUM difficulty levels to adjust the game accordingly... there, AI is sort of "rigged" to play to user's difficulty level.

Does that make sense? This is totally my wild guess.

[kingdevin]

Re: Stats-Based CPU Sliders [MLB 12 Version]

That makes sense

[Bobhead]

Re: Stats-Based CPU Sliders [MLB 12 Version]

It makes sense to me also, although I have to wonder why they went that route. You'd think it would have been easier to just link strategy and decisions directly to the difficulty level, ie: a Veteran AI, an All-Star AI, etc...

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

It makes sense to me also, although I have to wonder why they went that route. You'd think it would have been easier to just link strategy and decisions directly to the difficulty level, ie: a Veteran AI, an All-Star AI, etc...

Well that's just my own speculation so we really don't know how the difference in difficulty levels are implemented...

I think making AIs for different levels is actually less efficient, since you basically have to spend almost equal amount of time on each difficulty.

And It's a game that AI is trying to win (without cheating), so the dev should be spending time making the best AI first and foremost. Then if there's a need for adjusting to gamers who aren't up to that challenge, then you only work on the parts that need to be changed, rather than reinventing the wheel for all difficulty levels.

[kingdevin]

Re: Stats-Based CPU Sliders [MLB 12 Version]

That makes sense

Even with that logic, throughout all the years I've watched CPU, hof seemed the best setting to avoid an excess of 1 to 2 run games and late comebk

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Even with that logic, throughout all the years I've watched CPU, hof seemed the best setting to avoid an excess of 1 to 2 run games and late comebk

In the end I'm not here to convince anyone otherwise... I just collect information by reading and searching this forum to make my own judgement about things, and just share what I think is more reliable information with the like-minded people.

My opinion hasn't changed about difficulty levels after running games with different combinations. The stats just don't show anything different.

But that basically also means I can even suggest you to use the difficulty setting that you think plays the best game... For you it's Hall of Fame, for someone else, it's All Star. I don't think that makes any difference, but if everyone finds whatever makes them happier, that's the point of gaming!

So by all means, play at HoF!

[kingdevin]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Have you found by increasing starters stamina that the big 6 inn rallies have somewhat subsided

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Have you found by increasing starters stamina that the big 6 inn rallies have somewhat subsided

My latest conclusion is that letting starters have more stamina *delays* the onset of rally (see the green line in the figure):

http://www.operationsports.com/forum...post2043550784

probably because starters would have more available pitches before they get exhausted and pulled. So in that sense increasing stamina helps to reduce the 6th inning rallies.