Stats-Based CPU Sliders [MLB 15 Version]

[nomo17k]

This post is a placeholder for now. Once the slider set converges to something acceptable, this post will be updated with the "currently recommended" set.

... got a late start this time (busy with work, etc.), but for those who are not familiar with this set, the goal of Stats-Based CPU Sliders is to find a set of CPU-related sliders for (1) statistical accuracy in view of recent MLB trends, and (2) aesthetic of watching games in CPU vs. CPU (or MoM) mode, in that order. As such, HUM-related sliders won't be touched and no attempts will be made to take into human skill gap, difficulty levels, user interface and such.

A similar thread for the past game is here:

http://www.operationsports.com/forum...4-version.html


Following the trend of sharing through the "cloud," the main spreadsheet can be accessed in Google Sheets.




Currently recommended set (0524):

Contact: 5
Power: 5
Timing: 5
Foul Frequency: 4
Solid Hit: 6
Starter Stamina: 7
Reliever Stamina: 5
Pitcher Control: 5
Pitcher Consistency: 4
Strike Frequency: 6
Manager Hook: 4
Pickoffs: 5
Pitch Speed: 10
Fielding Errors In: 7
Fielding Errors Out: 2
Throwing Errors In: 3
Throwing Errors Out: 5
Fielder Run Speed: 4
Fielder Reaction: 4
Fielder Arm Strength In: 3
Fielder Arm Strength Out: 5
BR Speed: 5
BR Steal Ability: 10
BR Steal Frequency: 5
Wind: 5
Injury Frequency: 5

[rjackson]

I am so happy to see you here! Can't wait to see your results...I've been doing everything straight out of the box and this is probably the best iteration of that ever but not hardly as scientific as what you do. There is a Brian something (maybe BrianU?) that mentioned he was using your spreadsheet and accumulating some data just FYI. He mentioned that he hoped to give you a head start with it or help in case you came back...

[BrianU]

The king is back!

I did run about 50 total CPU v CPU tests spread over various settings as did other people I spoke with and found the default sliders very solid this year. My biggest issues and findings were:

- Very high GIDP/BIP and GIDP per game despite accurate batted ball profile.
- Slightly lower than average first pitch strike %
- Errors seem to need the same tweaks as last year i.e. fielding errors being lower than normal and throwing errors higher, out of the box.

One area I think may need a look at is base runner speed. Just from the eye test I saw many instances of balls just in front of home plate being turned into hits. Most of the slider makers here have gone ahead and decided to turn it down to 4 for much the same reason. I remember reading your base runner speed test last year I believe showing that 5 was the most realistic speed, I think with the changes to base running animations and speed differentials a new test might be necessary.

Also based on eye test I saw a lot of cases where short throws from the outfield or from the cutoff man would pull the 3rd and 2nd basemen off the bag when I felt they shouldn't have been. I'm talking pulling the 3rd baseman into foul territory. Bobhead mentioned that lowering throwing errors could help and I did see improvement. I made a bug report about this issue because others agreed it didn't look right.

I used Pitch speed 10 for all of my tests and it looks initially the problem of last year where steal % was 45%~ on that setting is resolved. It was back to 70% from what I saw in limited testing.

My testing method was not as extensive as Nomo's and I am not as well read on the sliders on what to change to get the desired effects statistically. I can't wait to read your results and findings.

[nomo17k]

Quote: Originally Posted by rjackson

I am so happy to see you here! Can't wait to see your results...I've been doing everything straight out of the box and this is probably the best iteration of that ever but not hardly as scientific as what you do. There is a Brian something (maybe BrianU?) that mentioned he was using your spreadsheet and accumulating some data just FYI. He mentioned that he hoped to give you a head start with it or help in case you came back...

Yes there are so many Brians talked about in the forum that OS might as well ban anyone who's not named Brian or Brianna...

Quote: Originally Posted by BrianU

The king is back! I did run about 50 total CPU v CPU tests spread over various settings as did other people I spoke with and found the default sliders very solid this year. My biggest issues and findings were: - Very high GIDP/BIP and GIDP per game despite accurate batted ball profile. - Slightly lower than average first pitch strike % - Errors seem to need the same tweaks as last year i.e. fielding errors being lower than normal and throwing errors higher, out of the box. One area I think may need a look at is base runner speed. Just from the eye test I saw many instances of balls just in front of home plate being turned into hits. Most of the slider makers here have gone ahead and decided to turn it down to 4 for much the same reason. I remember reading your base runner speed test last year I believe showing that 5 was the most realistic speed, I think with the changes to base running animations and speed differentials a new test might be necessary. Also based on eye test I saw a lot of cases where short throws from the outfield or from the cutoff man would pull the 3rd and 2nd basemen off the bag when I felt they shouldn't have been. I'm talking pulling the 3rd baseman into foul territory. Bobhead mentioned that lowering throwing errors could help and I did see improvement. I made a bug report about this issue because others agreed it didn't look right. I used Pitch speed 10 for all of my tests and it looks initially the problem of last year where steal % was 45%~ on that setting is resolved. It was back to 70% from what I saw in limited testing. My testing method was not as extensive as Nomo's and I am not as well read on the sliders on what to change to get the desired effects statistically. I can't wait to read your results and findings.

Thanks for the heads-up. I'm only a dozen games or so in so cannot really compare what I'm seeing to yours meaningfully yet. Offense has been pretty anemic (like last year) for my set so far, but they are all pitched by aces so this could either mean a good thing (i.e., aces pitch like aces) or I still might see low BABIP like I did last year.

Just out of curiosity, by "various settings," what have you varied?

[Aensland]

I've managed to record a very small sample of a few stats for comparisons. It's 15 games played, opening day schedule but with starters being rotated so 1-5 were used three times.

2014 RL MLB stats

4.07 runs per game
8.56 hits per game
1.67 2B per game
0.17 3B per game
0.86 home runs per game
2.88 BB per game
7.7 SO per game
.251 team BA
0.74 GIDP per game

Default MLB 2015 Slider Settings results:
4.3 runs per game
8.6 hits per game
1.13 2B per game
0.3 3B per game
0.86 home runs per game
3.26 BB per game
6.43 SO per game
.248 team BA
1.2 GIDP per game

-BA was a bit low but the runs scored seemed pretty good for default IMO. As already mentioned, there were too many double plays in the game. This small sample I've recorded doesn't reflect it well because I've already had other games (that wasn't tested) where teams would turn 4+ double plays.

My next slider set below was adjusted to fix that, plus increase extra base hits.

CPU Solid Hits - 6
Fielding Errors Infield - 7
Fielding Errors Outfield - 7
Throwing Errors Infield - 4
Throwing Errors Outfield - 4
Fielder Run Speed - 4
Fielder Arm Strength Infield - 4
Fielder Arm Strength Outfield - 4

results:
4.26 runs per game
9.23 hits per game
1.53 2B per game
0.3 3B per game
0.83 home runs per game
3.1 BB per game
6.73 SO per game
.255 team BA
0.76 GIDP per game

-Hits and doubles increased but oddly enough, the runs per game decreased. Home runs were down as well, which could be part of the reason for the runs scored dropoff. The slower OF speed allowed for more balls to drop in the gaps, but that with lower OF arm strength also resulted in a cheap inside the park HR at Comerica.

-Aces actually pitched like aces. In past MLBs I felt aces (and especially closers) had a tendency to blow up too often.

But again, this just a small sample, so over the course of 75 games to a full season, results could be quite different. Anyways, I'm looking forward to seeing more extensive results from you guys.

[Heroesandvillains]

Welcome back, Nomo. I can't wait to get your opinion on a couple pressing issues I've been facing lately.

1. Is mayo a great condiment on a burger, or is it the greatest condiment on a burger?

2. If mayo isn't available for frech fries, what's a suitable replacement for my dipping needs? Or do we just trust the grease and salt to the trick and go dip-free?

3. I can't wait to see what you come up with this year. My errors/SB sliders are lonely and might possibly need to be touched.

I ALMOST thought you wouldn't be doing this again this year. I hope all is well both with The Show and at home and look forward to following this thread again for 15.

[capa]

Quote: Originally Posted by Aensland

I've managed to record a very small sample of a few stats for comparisons. It's 15 games played, opening day schedule but with starters being rotated so 1-5 were used three times. 2014 RL MLB stats 4.07 runs per game 8.56 hits per game 1.67 2B per game 0.17 3B per game 0.86 home runs per game 2.88 BB per game 7.7 SO per game .251 team BA 0.74 GIDP per game Default MLB 2015 Slider Settings results: 4.3 runs per game 8.6 hits per game 1.13 2B per game 0.3 3B per game 0.86 home runs per game 3.26 BB per game 6.43 SO per game .248 team BA 1.2 GIDP per game -BA was a bit low but the runs scored seemed pretty good for default IMO. As already mentioned, there were too many double plays in the game. This small sample I've recorded doesn't reflect it well because I've already had other games (that wasn't tested) where teams would turn 4+ double plays. My next slider set below was adjusted to fix that, plus increase extra base hits. CPU Solid Hits - 6 Fielding Errors Infield - 7 Fielding Errors Outfield - 7 Throwing Errors Infield - 4 Throwing Errors Outfield - 4 Fielder Run Speed - 4 Fielder Arm Strength Infield - 4 Fielder Arm Strength Outfield - 4 results: 4.26 runs per game 9.23 hits per game 1.53 2B per game 0.3 3B per game 0.83 home runs per game 3.1 BB per game 6.73 SO per game .255 team BA 0.76 GIDP per game -Hits and doubles increased but oddly enough, the runs per game decreased. Home runs were down as well, which could be part of the reason for the runs scored dropoff. The slower OF speed allowed for more balls to drop in the gaps, but that with lower OF arm strength also resulted in a cheap inside the park HR at Comerica. -Aces actually pitched like aces. In past MLBs I felt aces (and especially closers) had a tendency to blow up too often. But again, this just a small sample, so over the course of 75 games to a full season, results could be quite different. Anyways, I'm looking forward to seeing more extensive results from you guys.

What skill level? All star? HOF?

C

[RoyceDa59]

Glad you are back nomo as well as like many other have stated I always enjoy your stat and slider analysis for this game.