Stats-Based CPU Sliders [MLB 12 Version]

[nomo17k]

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

The thing that's so cool about your sliders... Is how the fielders and the ball physics are interacting.
As you know, I play CPU vs CPU and I play (watch) the full MLB schedule.
It took me all day Saturday and most of the afternoon Sunday to get through the early Cards/Marlins night game, and the six or seven Openers the next day.
I gotta tell ya,
I saw inside the park HRs, dropped balls in the outfield, double steals, collisions at home plate, a runner advancing from first to home on a single, a complete game,passed balls, wild pitches, and just a lot of really cool ballhandling mishaps, balls dropping out of an outfielders glove when he backed into the wall too hard, numerous times when the outfielder kicked the ball or got the ball tangled up between his feet.
And I didnt see any of this before the patch and your sliders.
I must admit I didnt get serious with my season until the Patch on Friday.
But what I was trying to say was that on my Opening Day, I saw something new almost every inning. It was exciting and wide open baseball.

And yet, the scoring was realistic.

So I'm recommending to everybody, these are the sliders you want.

Thanks for your kind words... but it really is the developers who deserve all the credits. The sliders presented in this thread is just one (of many) ways of adjusting the game to play well.

If anything, the reason why (you feel) the play variety has increased is mostly likely because of slowing down the game overall a bit (via Fielder Run Speed, Arm Strength, etc.), and increasing the Fielder Error as you mentioned earlier. I think doing that tends to increase "borderline" plays (as opposed to fielders making proper plays more often), so that those "rare" events/animations get triggered a bit more.

But the increase in play variety over the years is indeed staggering! Becoming more and more natural and organitc every year.

[steviegolfballs]

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

The slider sare doing well in my season, but I'm concerned about the hook being at 4. I just saw Ubaldo Jimenez go 163 pitches in an 11-1 win over Toronto. What are your stats for moving the hook to 4?

[nomo17k]

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

The slider sare doing well in my season, but I'm concerned about the hook being at 4. I just saw Ubaldo Jimenez go 163 pitches in an 11-1 win over Toronto. What are your stats for moving the hook to 4?

I can see that happening... In that kind of blowout game, the AI manager tends to leave the starter in to go for a CG, even at a very high pitch count. My guess for a justification for such a move is that (in an old-school thinking) a CG is a cool and prestigious thing, so as long as the game isn't on the line, the manager's giving the starter a chance. But with very low energy, the starter is also going in a dangerous territory....

As far as the stats are concerned, with Starter Stamina 7 and Manager Hook 4, pitcher hook is fine (3.87 pitchers used per game compared to 3.88 which is the MLB ave), the total pitch count is fine (146.8 vs 146.3 per game), and starter pitch count is also okay (98.6 vs 96.5 per game). The starters pitched more innings on average (6.23 vs. 5.96 IP), but that's probably partly due to Pitcher Control = 10, which made pitchers overall perform better with less hits and walks issued...

The only thing that's wildly off is CG, which is pretty much consistently more than twice the MLB ave, even before I increased Starter Stamina. So it's a manager AI thing, I think.

It comes down to what you prefer.

[ralphieboy11]

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

I think manager hook and starter stamina are tough to find the right mix. Right now I have it at Starter Stamina=7 and Manager Hook=4. For the majority of games this is perfect. The problem seems to be when one team gets a big lead. The AI will leave the starter in to try and get the complete game, like your game with Jimenez. If the Indians had a one or two run lead, the AI would have realistically turned to the bullpen as Jimenez began to tire.

[nomo17k]

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

I think manager hook and starter stamina are tough to find the right mix. Right now I have it at Starter Stamina=7 and Manager Hook=4. For the majority of games this is perfect. The problem seems to be when one team gets a big lead. The AI will leave the starter in to try and get the complete game, like your game with Jimenez. If the Indians had a one or two run lead, the AI would have realistically turned to the bullpen as Jimenez began to tire.

This is my observation as well.

[shadia147]

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

Just shifted PC to 7.
I notice that the batter makes contact on every swing.
No swings and misses.
Do you see this as well?

[nomo17k]

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

Just shifted PC to 7.
I notice that the batter makes contact on every swing.
No swings and misses.
Do you see this as well?

No, if anything swing & miss rate will be a tiny bit higher with Pitcher Control = 7, since the pitches will be to good spots where batters tend to chase. I wouldn't worry about it for now..

[Bobhead]

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

Apologies if you've already explained this, but what exactly caused you to reduce Manager Hook in the first place?

[nomo17k]

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

Apologies if you've already explained this, but what exactly caused you to reduce Manager Hook in the first place?

It's not a very essential thing to do... I've been keeping track of how many pitchers the CPU manager uses in a game as well as the pitch counts by starters and relievers. Manager Hook at 4 produces the best matching numbers to those stats. But then again I'm mostly going by things that I can quantify (like things above), so just because the numbers work out doesn't mean the game actually play realistically.

I'm not a micro manager, so I like seeing a bit less quick hooks by CPU manager, and that certainly may be biasing that particular move as well. Again, it's mostly a preference thing.

[CubFan23]

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

I have noticed a very high pitch count in my franchise but mostly by my team(Cubs). Volsted had 105 in the 5th, and Maholm had 50+ in the 3rd. I know these are not top notch starters but neither were the opposing pitchers and thier pitch count seemed normal. The CPU takes alot and fouls alot off it seems, more than my team. Is it just me?

[nomo17k]

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

I have noticed a very high pitch count in my franchise but mostly by my team(Cubs). Volsted had 105 in the 5th, and Maholm had 50+ in the 3rd. I know these are not top notch starters but neither were the opposing pitchers and thier pitch count seemed normal. The CPU takes alot and fouls alot off it seems, more than my team. Is it just me?

Attached is a small sample of starter inning pitches, and pitch counts for starter (PC1) and relievers (PC[2345678]).

[ralphieboy11]

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

It really is hard to get a grasp as to what drives the AI to a pitcher change.

I have Starter stamina at 7 and manager hook at 4.

I just watched a great cpu vs. cpu game with the Rays and Yankees that went to extra innings. Earlier in the top of the 7th, the Rays pulled David Price out of the game for reliever Wade Davis. This was after the shortstop for the Rays committed a two base throwing error to start the inning. To that point Price had thrown 81 pitches and had given up 1 run on 3 hits.

Obviously we've figured out there is a huge difference between what the AI will do in close games compared to blow outs. Others have witnessed complete games and over 150 pitches from a starter with these same settings.

This move to the bullpen by Tampa seemed a little early and drastic, especially when you've got an ace like Price on the mound.

I don't really mind the variety, but it would be nice to find out what makes the AI decide to make certain moves.

[nomo17k]

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

It really is hard to get a grasp as to what drives the AI to a pitcher change.

I have Starter stamina at 7 and manager hook at 4.

I just watched a great cpu vs. cpu game with the Rays and Yankees that went to extra innings. Earlier in the top of the 7th, the Rays pulled David Price out of the game for reliever Wade Davis. This was after the shortstop for the Rays committed a two base throwing error to start the inning. To that point Price had thrown 81 pitches and had given up 1 run on 3 hits.

Obviously we've figured out there is a huge difference between what the AI will do in close games compared to blow outs. Others have witnessed complete games and over 150 pitches from a starter with these same settings.

This move to the bullpen by Tampa seemed a little early and drastic, especially when you've got an ace like Price on the mound.

I don't really mind the variety, but it would be nice to find out what makes the AI decide to make certain moves.

I just responded to your post in the other thread...

I think the gist is in the bold statement... AI tends to weigh game situations much more than the quality and the status of the pitcher itself.

Yet another reason managerial tendencies would be a great addition!!

[ralphieboy11]

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

I think you are right about the AI weighing the game situation. I think I remember somewhere where I read lefty-righty stuff was not programmed into the AI's decision making. While it may not have been intentionally, it seems like it still creeps its way in there. Davis was brought in to face a right handed batter here, while Price is a lefty.

Through all this, I just want to point out these are minor complaints. With the work of the devs, and your sliders here, I was just on the edge of my seat for every pitch of a game being played between little computer generated men.

[nomo17k]

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

I think you are right about the AI weighing the game situation. I think I remember somewhere where I read lefty-righty stuff was not programmed into the AI's decision making. While it may not have been intentionally, it seems like it still creeps its way in there. Davis was brought in to face a right handed batter here, while Price is a lefty.

Through all this, I just want to point out these are minor complaints. With the work of the devs, and your sliders here, I was just on the edge of my seat for every pitch of a game being played between little computer generated men.

I think what Knight said (that the devs said) was that the lefty-right split *stats* up to that point in the season is used to decide which reliever to bring in. The game keeps track of the stats (presumably batting average) so that certainly can be done.

So to that extent the platoon split may be considered by AI. The issue of course is that especially early in a season such stats may not be very stable due to limited pitching appearances. And it's not a way a typical real-life manager handles his bullpen, so I think it's a sort of thing that should be improved in future.