Re: Stats-Based CPU Sliders [MLB 16 Version]

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

For steals, do you think we'll have to go to 3 FB pitch speed? That wouldn't impact anything else (except the visuals)?

I don't know what they did between MLB15 and 16 but the steals are just...

I'm just trying to see which BR Steal Ability (for the current Pitch Speed sliders of 4) brings SB% closest to the real-life 71 - 72%.

But basically as I pointed earlier in the thread, SB/CS is one of the hardest to nail down, not because it requires a set of tricky adjustments, but because relevant data only slowly accumulate and the stats don't stabilize as fast as other events which happen much more frequently.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I am sure that there is a joke here.

Dee somehow thought he's strong enough to hit without protection with the help of [...].

Disappointed but also kinda surprised a speedster like him went for them.

[KBLover]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I'm just trying to see which BR Steal Ability (for the current Pitch Speed sliders of 4) brings SB% closest to the real-life 71 - 72%.

But basically as I pointed earlier in the thread, SB/CS is one of the hardest to nail down, not because it requires a set of tricky adjustments, but because relevant data only slowly accumulate and the stats don't stabilize as fast as other events which happen much more frequently.

Just wondering because my steal rates are still...ugh.

Makes me not even want to try to steal...but of course the CPU will do it anyway with people with 40 Steal Ability...

Be even 60, which I assume is about the average rating (i.e. what a player average in an ability is rated) should then be around irl average.

I don't know, maybe I'll just have to pretend it's a universe where 55-60% is "average" instead of 70-75%.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

Just wondering because my steal rates are still...ugh.

Makes me not even want to try to steal...but of course the CPU will do it anyway with people with 40 Steal Ability...

Be even 60, which I assume is about the average rating (i.e. what a player average in an ability is rated) should then be around irl average.

I don't know, maybe I'll just have to pretend it's a universe where 55-60% is "average" instead of 70-75%.

Is your overall steal success rate higher than 71% or below?

(Though the samples were still small), I saw with Pitch Speeds at 4, BR Steal Ability 10 overshot and BR Steal Ability 8 undershot that target %, so I'm just trying to be hopefully within a ballpark with setting it to 9 right now.

But then my current run of the game with 9, I have 1 SB with 4 CS in 3 games... Steal is very situational and unless you can set up an isolated test environment, it takes time to get useful numbers.

Just need to be patient.

[KBLover]

Re: Stats-Based CPU Sliders [MLB 16 Version]

Is your overall steal success rate higher than 71% or below?

(Though the samples were still small), I saw with Pitch Speeds at 4, BR Steal Ability 10 overshot and BR Steal Ability 8 undershot that target %, so I'm just trying to be hopefully within a ballpark with setting it to 9 right now.

But then my current run of the game with 9, I have 1 SB with 4 CS in 3 games... Steal is very situational and unless you can set up an isolated test environment, it takes time to get useful numbers.

Just need to be patient.

Pretty significantly below. If I'm counting right, I'm at about 60% overall.

The typical scenario for a CS goes like this:

I take one extra lead. The runner gets a good jump. He's completely out of the base runner window sometimes before the pitcher releases his pitch.

Then when the game pulls out to show the throw, I look at the base runner diagram and he's only about 50-55% the way to 2nd base. I don't understand that - good jump, 80 speed, 70 steal ability, and it's like he ran only about 10 feet since he left the runner window.

Sometimes I see the runner moving slowly, then suddenly speed up like he's going "oh yeah, I want to steal" and then of course it's too late.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

Pretty significantly below. If I'm counting right, I'm at about 60% overall.

The typical scenario for a CS goes like this:

I take one extra lead. The runner gets a good jump. He's completely out of the base runner window sometimes before the pitcher releases his pitch.

Then when the game pulls out to show the throw, I look at the base runner diagram and he's only about 50-55% the way to 2nd base. I don't understand that - good jump, 80 speed, 70 steal ability, and it's like he ran only about 10 feet since he left the runner window.

Sometimes I see the runner moving slowly, then suddenly speed up like he's going "oh yeah, I want to steal" and then of course it's too late.

If you just want to see SB% ever go above 70% (even just for testing, to see it can ever happen), I'd try Pitch Speeds at 4, BR Speed 5, BR Steal Ability 10 for a while.

So far that's the only settings with which I could consistently be above 70% steal success rate for a relatively extended time (still just 35 games though).

Also, if you are actually playing and playing above All-star, pitch speed slightly goes up... so the Pitch Speed sliders may need adjustments.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

Given that basically all my bug reports have just vanished, I think it's probably useless at this point to waste time filing bug reports, but I created a bug report on steal success rate anyways:

http://theshownation.com/bug_reports/13799

Please upvote if you care about this type of thing.



#### Copy of bug report 13799

Subject

Steal success rate down (game balance issue)


Description

I have been getting a lot of test CPU vs. CPU games to make (slider) adjustments to make the game stats becoming closer to recent MLB numbers, and it has come to attention that steal success rate (with default sliders) is quite a bit lower than real-life MLB number (of 71 - 72%).

The OS thread is here: http://www.operationsports.com/forum...6-version.html and this is not an isolated issue for myself, but most if not all people who have been looking at steal success rates report similar trend of SB% being lower than desired.

The numbers and slider settings are tallied in the Google Sheet accessible from the cited forum (it is here: https://docs.google.com/spreadsheets...6ff7683#gid=13), but in short, with default sliders and only Pitch Speed maxed out at 10 (*), the steal success rate about 49% (SB 51, CS 53).

At Pitch Speed sliders maxed out, I could only consistently get SB% of around 50% even with BR Steal Success slider set to 10 (out of 10).

At Pitch Speed sliders default (5 out of 10) and BR Steal Success rate maxed out at 10, I could get SB% of 64% (SB 115, CS 65), which is still significantly down from the MLB ave of 71 - 72%.

The only setting where I could get a realistic SB % close to 71 - 72% so far is when (1) Pitch Speed sliders are on or blow 4 (out of 10) and (2) BR Steal Ability is at 10. (Assuming the rest of sliders are at default). Then I seem to be able to overshoot SB% of 71%, although that's from only 35 games or so worth of stats.

SB% being off in this way is unfortunate because, in order to get a realistic SB%, users are forced to play with Pitch Speed lower than even the default slider settings. If we wish to challenge ourselves with higher pitch speed, we have to sacrifice realistic steal success rate.

(*) I know that increased Pitch Speed sliders reduce steal success rate, but I am adjusting sliders for CPU vs. CPU game watching and Manage-only mode, so it is my personal preference to adjust sliders so that things stay realistic with max pitch speed, which is visually more authentic.



Steps to Reproduce

Run a bunch of CPU vs. CPU games (or gameplay settings close to keep the rating-driven environment), adjust relevant sliders as described in Description. Observe SB and CS, compute success rate SB / (SB + CS).

[tessl]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I'm curious what you have umpires set to and whether you are leaving the setting on default dynamic difficulty. I know in the past difficulty didn't matter but as soon as I changed everything to veteran I started seeing more offense. I also have balls and strikes set to perfect.

The problem with stolen bases is the tag animation which is also causing a very high percentage of runners attempting to advance on a wild pitch to be thrown out and both of those things are contributing to low run totals. If they would simply put the tag animation back to what it was last year I could live with the game.

Another issue is morale which is having a major impact on in-game attributes. I haven't simmed to the off season but I hear morale expectations change during the off season which then will change the way the game plays on the field as attributes change. For players who have personal performance and/or team performance as an expectation their attributes are dynamic during the season. I haven't seen a detailed explanation of progression but it seems to be at least somewhat performance based this year. My point is the game is more dynamic this year and adjusting sliders might be the equivalent of chasing your tail.

[KBLover]

Re: Stats-Based CPU Sliders [MLB 16 Version]

You have my upvote.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I'm curious what you have umpires set to and whether you are leaving the setting on default dynamic difficulty. I know in the past difficulty didn't matter but as soon as I changed everything to veteran I started seeing more offense. I also have balls and strikes set to perfect.

The problem with stolen bases is the tag animation which is also causing a very high percentage of runners attempting to advance on a wild pitch to be thrown out and both of those things are contributing to low run totals. If they would simply put the tag animation back to what it was last year I could live with the game.

Another issue is morale which is having a major impact on in-game attributes. I haven't simmed to the off season but I hear morale expectations change during the off season which then will change the way the game plays on the field as attributes change. For players who have personal performance and/or team performance as an expectation their attributes are dynamic during the season. I haven't seen a detailed explanation of progression but it seems to be at least somewhat performance based this year. My point is the game is more dynamic this year and adjusting sliders might be the equivalent of chasing your tail.

Variable umpires is used but there has been no evidence to believe the use of variable strike zone affecting the overall balance of the game. I keep track of home plate umpires for all games so if it is a factor it should be noticeable.

Difficulty level if it affects CPU vs. CPU game, that might be new or simply a bug. I set them to All-star across the board but I've never seen difficulty level affecting CPU vs. CPU games. If difficulty level affects this and it is not a bug, then you should be able to see the effect quickly using something like Beginner Mode, in which pitchers should pitch everything down the middle.

For stolen base, the quick tag animation is one of the factors, but I don't think that the only issue. For one thing you can influence steal success rate by using SB Steal Ability slider, but the slider effect is just not big enough unfortunately. So if the slider has a bigger dynamic range for the effect, we should be able to at least correct the issue.

I think the reason why steal has been difficult to balance properly even for the developers is that doing it correctly involves multiple aspects of the game... from fielding animations (pitching delivery, catcher, middle infielders, etc.), pitch speed, throw speed, base running speed, manager AI, etc., and since it's not just one person balancing this aspect of the game, it's not hard to imagine coordinating all the changes across different parts of the game can be tough. Even if one area is trying to do it perfectly, if the other parties are slightly off, then it's enough to throw off the overall game balance.

All these progression issues are the issue with the season/franchise engine (which admittedly has not been very good in my opinion) and not that of the gameplay engine. The gameplay just takes in the player attributes when the game starts, so the gamepley should be pretty agnostic of what's going on outside the game being played.

You have my upvote.

Thank you for upvoting.

[Monster6968Rehab]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I have a few questions.Now that you have the beta slider set what type of things are you testing besides stolen base percentages? Are you still trying to improve on the beta set? Is the beta slider set in the slider vault or is there another version of a slider set coming up? Thanks for the work your doing I really appreciate it.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I have a few questions.Now that you have the beta slider set what type of things are you testing besides stolen base percentages? Are you still trying to improve on the beta set? Is the beta slider set in the slider vault or is there another version of a slider set coming up? Thanks for the work your doing I really appreciate it.

My interests mostly are (1) to see what the best compromise is for getting realistic SB% and (2) a bunch of potential aesthetic adjustments (which hopefully would also fine tune certain aspects better) without making the game balance off from the beta.

So in a sense I'm still trying to see where fine tuning could happen, but trying not to deviate much from the beta set, so that I could effectively "validate" the beta set... i.e., my current set is basically identical to the beta set in the OP, with (hopefully) only a few aesthetic adjustments which I don't expect to change the game balance much if at all, so that I'm kinda trying to shoot two birds with a PS4.

I will probably upload a set to the vault at some point when all is done. But given that it probably only matters to a grand total of half a dozen CPU game watchers and MoM people, and those people are probably already here in this thread, I don't really see the point. The rest of people wouldn't use it for the purpose it was created, so I don't know what the point is...

[Monster6968Rehab]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I hope since there is only like half a dozen people that really enjoy your slider sets that you feel a sense of accomplishment from what you do. Are you in a year to year save for franchise? If so do you make these slider sets for your own enjoyment as well. I bet you do but just curious how your franchise is going. Do you plan or intend to continue making these slider sets for the future?

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 16 Version]

I hope since there is only like half a dozen people that really enjoy your slider sets that you feel a sense of accomplishment from what you do. Are you in a year to year save for franchise? If so do you make these slider sets for your own enjoyment as well. I bet you do but just curious how your franchise is going. Do you plan or intend to continue making these slider sets for the future?

Oh man I think people shouldn't be doing this kind of thing just to feel a sense of accomplishment... there are better things we can do for that. In the end it really doesn't matter what goes on in an anonymous internet forum like this one. Perhaps due to anonymity, so many people are blunt, rude, abrasive, enjoy too much of meaningless back and forth just to prove frivolous point, etc., etc., it's quite depressing to be honest.

I'm in it mostly for sharing enjoyment of the game with others who do so similarly, because I enjoy this game so much myself. Most baseball simulation games that I know aren't that good to worth putting in this type of time and effort or even appreciate, but The Show so far has stood that test for me, so I'm hoping by sharing some potentially useful info with like-minded people, something meaningful comes out that helps the game moving forward in ways that improve its authenticity as a baseball simulation. That's pretty much the only reason that I hang around here.

But at the same time I'm not sure if I can keep doing this kind of slider testing. Nobody knows what his life would be like in a year, so it is possible that I simply won't have the time for it... it's just a general concern for everyone. (It's possible I'll be on a mission to Mars without PS4, for example.)

Sad to say though I haven't really invested much of my time in franchise mode lately, as for my taste the franchise mode engine has required too much baby-sitting to keep things realistic for me... I'd much rather play OOTP for team management type of experience.

[KBLover]

Re: Stats-Based CPU Sliders [MLB 16 Version]

The rest of people wouldn't use it for the purpose it was created, so I don't know what the point is...

Well, I'm one of those not using it for the strict purpose of CPU vs CPU type gameplay...

Sorry for abusing your slider work. Please don't hit me with one of your forkballs, Nomo!

But you're the only one (I see here) that's showing what good setting for the CPU are in terms of realistic event rates like contact%, OSwing%, etc. while still getting good "standard stats" and gameplay. I've gotten some interesting and good games since using your set for my CPU settings and setting my pitching to what you have so that meshes up (I've gone to 3 consistency since I saw you try it).

I feel like I'm getting a good baseball challenge, not just video game challenge, which is why I signed up for the two least user input modes around. Not sure why I don't just go be a MoM player...lol

So if they send you to Mars, accept only on condition that they'll attach some solar panels to your PS4, computer, and monitor.