Stats-Based CPU Sliders [MLB 12 Version]

[nomo17k]

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

I've played a few games, and Crawford and Ellsbury don't even draw throws. But I get the whole see it for myself thing.

Now 15 games in, and 26 SB and 2 CS. So I think it's definitely reversed. At this point I'm just trying to see where SB success % stays with that slider at 0. Right now it's 93%, with default it was about 68%. Just want to know how many clicks are needed to get it to 72%.

I'm also +1 with BR Speed, so that's also a factor for the very high success rate right now.

[tbbucsfan001]

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

Sidenote: Am I implementing the HR/H stat right on this sheet?

[nomo17k]

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

Sidenote: Am I implementing the HR/H stat right on this sheet?

Where does this come from? I thought the game keeps track of HR and AB against RHP and LHP separately, no?

[tbbucsfan001]

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

Where does this come from? I thought the game keeps track of HR and AB against RHP and LHP separately, no?

Your formulas. I have the ZiPS Projections at the left and I don't know if I have the HR/H and Power Ratings right.

[nomo17k]

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

Your formulas. I have the ZiPS Projections at the left and I don't know if I have the HR/H and Power Ratings right.

Oh I see... thanks for noticing that! I now see what you were asking (hopefully). I remember it's this:

http://www.operationsports.com/forum...post2042364723

I just made that "generic" platoon split formula to deal with missing split info... since I was going to use the Lahman database which doesn't keep vs LHP/RHP info. My goal was to come up with something for creating classic players because using the same ratings against R/LHP would be kinda boring. So if you have the split stats for H and HR, you may want to use that directly instead of that formula.

I think the way you applied the formula is most correct, except I don't understand why you used different "r" (ratio of ABs vs LHPs to RHPs) and PLT% (platoon advantage %... something you may have to estimate your own for HR ratings.) for RHB and LHB. You should use the same number for r and PLT% for both.

Well, cool spreadsheet in any case!!

[nomo17k]

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

Got 25 games in with the "test" set. The BR Steal Ability slider is most likely reversed.

- BR Steal Ability at 5, I saw 117 SB and 56 CS (67.6% success rate).
- BR Steal Ability at 0 (also BR Speed was at 6... slightly making steal easier, but hopefully not too much), I saw 45 SB and 7 CS (86.5% success rate).

So, one click down in BR Steal Ability very roughly corresponds to 3.8% increase in steal success %. So probably BR Steal Ability at 4 would do it.

I'm pretty happy with what I'm seeing, and the stats that are kinda still off (e.g., GIDP, WPs, Chase %) are something that's not very easy or impossible to correct with the given slider options.

For now, I'll treat as my first working release, and will see if any issues come up, hopefully minor. And things like stamina/manager/injury related stuff, I need to see if there's any way to improve.

One thing I notice is that pitchers tend to go for CG more than IRL (about twice more frequent). Pitch counts seem to be okay, but this is perhaps related to sometimes quirky pitcher hooks that CPU makes.


### Mar. 23 set for the stats shown ###

Contact: 4
Power: 5
Timing: 6
Foul Frequency: 4
Solid Hits: 5
Starter Stamina: 5
Reliever Stamina: 5
Pitcher Control: 5
Pitcher Consistency: 4
Strike Frequency: 5
Manager Hook: 5
Pickoffs: 5
Pitch Speed: 10
Fielding Errors: 7
Throwing Errors: 5
Fielder Run Speed: 4
Fielder Reaction: 4
Fielder Arm Strength: 3
BR Speed: 6
BR Steal Ability: 0
BR Steal Frequency: 8
Wind: 5
Injury Frequency: 5

[nomo17k]

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

The only change is BR Steal Ability at 4 now.

Mar. 25 set

Contact: 4
Power: 5
Timing: 6
Foul Frequency: 4
Solid Hits: 5
Starter Stamina: 5
Reliever Stamina: 5
Pitcher Control: 5
Pitcher Consistency: 4
Strike Frequency: 5
Manager Hook: 5
Pickoffs: 5
Pitch Speed: 10
Fielding Errors: 7
Throwing Errors: 5
Fielder Run Speed: 4
Fielder Reaction: 4
Fielder Arm Strength: 3
BR Speed: 6
BR Steal Ability: 4
BR Steal Frequency: 8
Wind: 5
Injury Frequency: 5

[floydpinkster]

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

Hey nomo. This is great stuff. But i'm wondering... your last set had foul freq. 4, fielder react. 4, and runner speed 6. did you abandon these changes? Why?
THanks Bud

[nomo17k]

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

Hey nomo. This is great stuff. But i'm wondering... your last set had foul freq. 4, fielder react. 4, and runner speed 6. did you abandon these changes? Why?
THanks Bud

because..... I made a mistake!!! I just copied and pasted wrong stuff... thanks for noticing.

[floydpinkster]

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

Cool cool. I see you made the changes but base runner speed still down too5 from 6... is that what we're going with?

[nomo17k]

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

Cool cool. I see you made the changes but base runner speed still down too5 from 6... is that what we're going with?

This is what happens if you haven't gotten laid for too long. Was a mistake again... fixed. Thanks!!!

[floydpinkster]

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

LOL well we need ya on your "A GAME" so get out there and get yourself some. LOL
K Thanks NOMO.

[NAFBUC]

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

It's totally fine to chime in. Sorry if I sounded aggressive (it's just the nature of how I write).

But I'm clearly taking a lot of time before making any sort of conclusions, and I really cannot offer good suggestions when you are too quick about making your judgement.

Bouncing off ideas is totally welcome, but as I keep repeating, it's just absolutely meaningless to make any sort of conclusions only after a few games after you change something. I try not to make suggestions like that, and I won't appreciate any feedback like that either, since it only adds to confusions and meaningless back and forth about nothing.

I am in a league and we have some members who want to make slider adjustments after the first couple of games. Good thing the veteran players realize the importance of waiting until we have sufficient data to warrant slider adjustments.

[nomo17k]

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

Some people might already know that I'm basically here at OS to discuss MLB the Show and *love*. In fact I hope to find the love of my life by hanging around at OS, where a lot of gorgeous ladies hang around as well.

I've done plenty of the former, but what about the latter? So, what to do on a Sunday evening after weekend work (of course!), when you don't seem to find enough love in your life?

I test Pitcher Stamina slider and try to understand how it works.

What I wanted to see is how many pitches that a pitcher can throw with a given Pitcher Stamina rating and slider.

I used Classic pitching with max effort, aiming every 4SFB (didn't test with other pitches) well outside the zone (up and away), and recorded the pitch count when I saw nothing on the Energy bar in the game. The test subject was Mike Pelfrey, since he seemed mediocre enough. I varied his Stamina rating over 0 to 99 at Starter Pitcher Stamina slider at 0, 5, or 10.

The results are plotted in the attached figure.

On the x-axis, that's Pelfrey's Stamina rating. On y-axis, that's how many pitches he threw before exhausting energy completely. The marker color corresponds to (Starting) Pitcher Stamina slider value. So with Stamina rating at 99 and the slider at 10, Pelfrey was able to throw 178 pitches before exhausting his energy.

I also tested how much change that I see if I use minimum effort on Classic. Those are indicated by a bit transparent markers. Pelfrey with above example could threw 187 pitches that time. Only about 10 pitches more. So apparently varying effort doesn't preserve/waste energy as much as we may think. An additional test with Pelfrey with rating at 0 and slider at 0, he threw 56 pitches (with max effort, he threw 52... not a huge change).

One trend I see in the plot is that the Stamina slider doesn't necessarily work linearly, meaning increasing the slider twice as much doesn't lead to twice more available pitches (after subtracting baseline). If I take data at Stamina ratings at 25 and 75 for example, I expect to see some sort of curves than straight lines.

To see if the slider works the same way with the relief pitchers, I did one test using Frank Francisco, with Stamina rating at 0 and Reliever Stamina slider at 0. The result was the same, so given this very limited testing, I suppose that Reliever Stamina slider works the same way, but it only applies to relivers.

I hope I'm on to something and eventually find love by doing this. Something tells me that I'm doing the right thing.

[Heroesandvillains]

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

How many pitches is Pelfrey able to throw at his default stamina rating with a default stamina slider?

What's the highest amount of pitches he's ever thrown and/or, how many does he average per start (in recent years)?

I only ask because I'm thinking of testing the stamina slider myself and was wondering if maybe SCEA programmed the game to burn out starting pitchers too quickly? Or rather, if they program the ratings hit to occur too soon, or, make their ratings take too much of a hit when exhaustion happens in general.