Stats-based sliders for CPU

[nomo17k]

Re: Stats-based sliders for CPU

If you're tracking it anyway, what the heck? Right?

Okay, I played a CPU vs. CPU game today, while hanging out with family. We all loved it.

I vote that you and Daiyoung get together, like TNK and Psychobulk, and make a CPU/MOM set...

For completely selfish reasons, ofcourse. I LOVE watching games!

It's good several people doing different things; it helps to test different slider sets at once.

I just opened up my own thread since I wasn't sure the CPU vs CPU thread guys would comeback and do the same thing this year. Seeing how activity level is high in the CPU thread, maybe independent thread wasn't necessary after all, which is good thing.

[nomo17k]

Re: Stats-based sliders for CPU

Just a quick update for my current slider set, which is everything at default (5) except:

Power: 3
Foul Frequency: 4
Strike Frequency: 3
Fielding Error: 10
Throwing Error: 4
BR Steal Ability: 2
BR Steal Frequency: 8

The stats after 30 games looks very good but unfortunately except walks for reasons I don't really understand; maybe a couple tests I did with 15 games weren't enough, which is the reason I ran 30 games this time.

I'm quite amazed how those less quoted stats like swung % and miss % come out so close to real life numbers!! The person(s) designing this part of the game must have really really studied well how pitcher-batter interaction works. SCEA should definitely give the person a substantial raise, haha.

But walks are not coming as frequently, which I think is caused by the CPU being only slightly more aggressive (maybe 0 - 1% in swung %) than real life hitters. That's probably why we need to increase the balls that hitters cannot ever hit by decreasing pitcher consistency or strike frequency slider.

So I just want to increase walks. I have two choices, using either strike freq or pitcher consistency. I was thinking of lowering pitcher consistency, but in the CPU thread it was mostly shown that it works, so instead I decided to lower strike freq by one while keeping other offense sliders. Hopefully I see some effect by the strike freq slider, if any, on overall offense stats as well.

So my next set is everything at default (5) except:

Power: 3
Foul Frequency: 4
Strike Frequency: 2
Fielding Error: 10
Throwing Error: 4
BR Steal Ability: 10
BR Steal Frequency: 8

Comment: I've been lowering steal ability by a significant margin, but all I see is only a slight decrease in steal success rate, and that got me thinking maybe I'm not moving that particular slider in the right direction. Last year's CPU thread, it appeared increasing steal ability decrease success rate for some reason. So for the next couple testing runs I'll be maxing out steal ability slider to see if this affect steal success rate in the desired direction. Just to see...

[DaiYoung]

Re: Stats-based sliders for CPU

Good to see the fielding errors slider having some effect Nomo. I've raised it to 8 but it hasn't made much difference thus far (only 17 games into testing), looks like I should go higher.

As for SB/CS. Using all my data (228 games) my franchise is running at 74% steal success rate and 0.57 steal attempts per game, pretty much in line with MLB numbers. The only change I've made during that time was raising SB Attempts to 8 after about 60 games.

My latest test involves editing all catchers to 99 Blocking and 99 Reaction to reduce WP. Very early days but it looks reasonably promising and doesn't seem to have any nasty side effects.

[Jgainsey]

Re: Stats-based sliders for CPU

Good to see the fielding errors slider having some effect Nomo. I've raised it to 8 but it hasn't made much difference thus far (only 17 games into testing), looks like I should go higher.

As for SB/CS. Using all my data (228 games) my franchise is running at 74% steal success rate and 0.57 steal attempts per game, pretty much in line with MLB numbers. The only change I've made during that time was raising SB Attempts to 8 after about 60 games.

My latest test involves editing all catchers to 99 Blocking and 99 Reaction to reduce WP. Very early days but it looks reasonably promising and doesn't seem to have any nasty side effects.

DaiYoung,

What do you have the steal ability slider at?

[nomo17k]

Re: Stats-based sliders for CPU

DaiYoung,

What do you have the steal ability slider at?

I think Dai's keeping the steal ability at default.

[nomo17k]

Re: Stats-based sliders for CPU

Good to see the fielding errors slider having some effect Nomo. I've raised it to 8 but it hasn't made much difference thus far (only 17 games into testing), looks like I should go higher.

As for SB/CS. Using all my data (228 games) my franchise is running at 74% steal success rate and 0.57 steal attempts per game, pretty much in line with MLB numbers. The only change I've made during that time was raising SB Attempts to 8 after about 60 games.

My latest test involves editing all catchers to 99 Blocking and 99 Reaction to reduce WP. Very early days but it looks reasonably promising and doesn't seem to have any nasty side effects.

Yes, I think the field error freq slider does have a desired effect (i.e., increasing fielding errors). What may not be so cool is that it appears quite some fielding errors are of "dropping routine fly balls" kind, which should happen but rarely. So we may be able to increase the stats by the slider adjustment, but people may end up feeling, you know, outfielders shouldn't keep botching plays like that! You know what I mean? But I like there are some fielding errors happening though... Errors are still rares events and they often come in bunch, so we have to keep racking up stats.

For SB/CS, that's one thing I was puzzled by your stats compared to mine. Yours is pretty dead on, but mine has consistently been off. I say small sample size, and with the data I had with default sliders, maybe the truth lies somewhere between. (Un)fortunately, varying the steal ability slider has had only marginal effect on stats; I think still worth doing what I'm testing with the steal ability slider in case we end up needing to adjust one way or other, we should at least know which way to move.

Editing catchers hopefully is something of final resort though... I think we should request SCEA to revisit the WP issues, either in the bug or glitch thread? Who knows it might not be that hard for them to adjust... We need slightly more HBP and about 2 - 3 times less WPs.

Yes I'm starting to get a bit worried how much WPs affect games in tight situations.

[DaiYoung]

Re: Stats-based sliders for CPU

Yes, I think the field error freq slider does have a desired effect (i.e., increasing fielding errors). What may not be so cool is that it appears quite some fielding errors are of "dropping routine fly balls" kind, which should happen but rarely. So we may be able to increase the stats by the slider adjustment, but people may end up feeling, you know, outfielders shouldn't keep botching plays like that! You know what I mean? But I like there are some fielding errors happening though... Errors are still rares events and they often come in bunch, so we have to keep racking up stats.

For SB/CS, that's one thing I was puzzled by your stats compared to mine. Yours is pretty dead on, but mine has consistently been off. I say small sample size, and with the data I had with default sliders, maybe the truth lies somewhere between. (Un)fortunately, varying the steal ability slider has had only marginal effect on stats; I think still worth doing what I'm testing with the steal ability slider in case we end up needing to adjust one way or other, we should at least know which way to move.

Editing catchers hopefully is something of final resort though... I think we should request SCEA to revisit the WP issues, either in the bug or glitch thread? Who knows it might not be that hard for them to adjust... We need slightly more HBP and about 2 - 3 times less WPs.

Yes I'm starting to get a bit worried how much WPs affect games in tight situations.

Our SB/CS discrepancy is odd. I suppose it shows that even doing hundreds of tests sometimes isn't enough.

I really didn't want to get into editing players this year but other than a patched solution I figured it was the only way to try and curb the large amount of WP. The good thing about editing catchers is that there are only 65 or so players to alter (I only play MLB games) so it's not too time consuming.

The preliminary results of my editing are looking promising. 20 games into testing and there have been 14 WP.

[Heroesandvillains]

Re: Stats-based sliders for CPU

Keep up the good work guys!!!

Nomo, the outfielder dropping balls issue. Where was the fielding errors slider at then? 10?

Are you guys sure you're seen enough errant throws with throwing errors at 4.

Kind of off topic, but outfielders should make a large chunk of their errors via faulty throws. It's too bad SCEA doesn't give us infielder and outfielder sliders independently. Sure, at 4, infielders numbers/overall throwing errors may be fine...but are outfielders pretty much bullet proof with throwing errors at 4?

[DaiYoung]

Re: Stats-based sliders for CPU

I still have throwing errors at 5, heroesandvillains. Seems about right, I'd say the majority of outfield errors are still of the errant throw variety. To be honest, it's just nice to see ANY outfield errors. I think in last year's game I saw maybe three the whole time I was playing.

[nomo17k]

Re: Stats-based sliders for CPU

Our SB/CS discrepancy is odd. I suppose it shows that even doing hundreds of tests sometimes isn't enough.

I really didn't want to get into editing players this year but other than a patched solution I figured it was the only way to try and curb the large amount of WP. The good thing about editing catchers is that there are only 65 or so players to alter (I only play MLB games) so it's not too time consuming.

The preliminary results of my editing are looking promising. 20 games into testing and there have been 14 WP.

Yup "E" is not ideal but if doing so affects the # of WPs, would be good to know regardless. Thanks for testing that!!

[nomo17k]

Re: Stats-based sliders for CPU

Nomo, the outfielder dropping balls issue. Where was the fielding errors slider at then? 10?

Are you guys sure you're seen enough errant throws with throwing errors at 4.

Kind of off topic, but outfielders should make a large chunk of their errors via faulty throws. It's too bad SCEA doesn't give us infielder and outfielder sliders independently. Sure, at 4, infielders numbers/overall throwing errors may be fine...but are outfielders pretty much bullet proof with throwing errors at 4?

For the last stats I posted, I was using fielding error of 10. I'm not sure outfielders dropping routine flyballs are due to increased occurrences of fielding error or due to the slider reduces their fielding ability so that they are more likely to make poor plays like those...

Throwing errors are still abundant at 4. It kinda makes me nuts throwing errors from outfielders tend to be routine throws to cutoff men turned errant in no tight situations. Also fielders in general make too many throws to bases when they should know they have no chance to get the runner out. But that's a couple of wishlist items I guess.

However I definitely like the variety of plays SCEA has added though! It's really a pleasure to watch games this year! Quite some improvements over MLB 10.

I agree with you on the last point. Should be ways to independently control error frequencies for in and outfielders.

[nomo17k]

Re: Stats-based sliders for CPU

I still have throwing errors at 5, heroesandvillains. Seems about right, I'd say the majority of outfield errors are still of the errant throw variety. To be honest, it's just nice to see ANY outfield errors. I think in last year's game I saw maybe three the whole time I was playing.

With fielding errors maxed out, I started seeing a lot more errors by outfielders.

But SCEA added quite a variation in plays in outfield that won't be recorded as errors, but still sorts of misplays. I like those misjudges made by outfielders that end in extra bases. Quite human I'd say.

[DaiYoung]

Re: Stats-based sliders for CPU

With fielding errors maxed out, I started seeing a lot more errors by outfielders.

But SCEA added quite a variation in plays in outfield that won't be recorded as errors, but still sorts of misplays. I like those misjudges made by outfielders that end in extra bases. Quite human I'd say.

Totally agree. A really nice surprise that they spent so much time on making outfield play more realistic. True to life, also, in that most times those sort of misjudgements are not scored as errors in MLB.

[DaiYoung]

Re: Stats-based sliders for CPU

How are you dealing with RA Dickey and Tim Wakefield in your franchise Nomo? Reckon I'm going to DFA both of them, lest they sweep the Cy Young's.

[nomo17k]

Re: Stats-based sliders for CPU

How are you dealing with RA Dickey and Tim Wakefield in your franchise Nomo? Reckon I'm going to DFA both of them, lest they sweep the Cy Young's.

I haven't started real runs on any game mode, but for sliders I just removed the games they pitched from the stats. If I control over their appearances then I just skip Dickey's start and force Wakefield not used out of the pen. But DFA is a good idea.

Hopefully SCEA patches the knuckleball glitch before starting real runs.