Jump to content

Some Post-patch Info On Ai Agression


soxfan01605

Recommended Posts

Hey guys. I posted this over at Operation Sports. Since IE now ties into the ratings properly, ratings edits now mean more in terms of gameplay results. I

I did some tests with the walk rating to see if I could now get the AI aggression down more (I wasn't really satisfied with what I saw right after the patch).

Also, I've listed the rating ranges that correspond to the IE grades to help guys maybe get an idea of how to tie the in.

Not sure if this will help anyone or not, but I thought I'd share for anyone who either might want to know or to help out potential roster makers get a jump on tweaks to better tie in with IE...

"First, IE reports now adjust properly based on ratings changes made by the user. It works accross the board for every rating and every player. This is important because what I've found is preliminary evidence that the rating level does indeed effect AI swing tendency now. This is good news because it's not as locked down to hot/cold zones as I first believed.

What I did was gather the ratings ranges for each corresponding letter grade for the IE Report. Then, I tested the same lineup (Red Sox) 1-9, 3 times at each IE grade level for walks ONLY. All other ratings were left at default for each given batter (for variance) and this was all done on DEFAULT PRO.

***Bear in mind that this is only VERY preliminary and further, more detailed testing over a MUCH greater sample using more teams, etc would be needed to get a more accurate range.***

I just wanted to share some early results as they at least seem to indicate that the walk rating does indeed effect IE ratings now:

The numbers are averaged on 3 separate cycles through the lineup 1-9. Same Pitcher, but I restarted each game after I got through the lineup one time (so basically, 1-9 in 3 separate games, but the same pitcher). I also re-did this using all 3-0 counts as well.

[b]"A+" Rating: On average, 7/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 4/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 8/9 took a FB right down the pipe, regardless of hot/cold zone.

"A" Rating: On average, 6/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 4/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 8/9 took a FB right down the pipe, regardless of hot/cold zone.

"A-" Rating: On average, 6/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 3/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 8/9 took a FB right down the pipe, regardless of hot/cold zone.

"B+" Rating: On average, 6/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 3/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 7/9 took a FB right down the pipe, regardless of hot/cold zone.

"B" Rating: On average, 5/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 2/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 8/9 took a FB right down the pipe, regardless of hot/cold zone.

"B-" Rating: On average, 5/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 3/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 7/9 took a FB right down the pipe, regardless of hot/cold zone.

"C+" Rating: On average, 5/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 2/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 7/9 took a FB right down the pipe, regardless of hot/cold zone.

"C-" Rating: On average, 4/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 2/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 6/9 took a FB right down the pipe, regardless of hot/cold zone.

"D+" Rating: On average, 3/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 1/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 6/9 took a FB right down the pipe, regardless of hot/cold zone.

"D" Rating: On average, 2/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 0/9 took two straight FB right down the pipe.

On average in 3-0 Counts, 5/9 took a FB right down the pipe, regardless of hot/cold zone.

"F" Rating: On average, 2/9 took a 1st pitch FB right down the pipe, regardless of hot/cold zone. 0/9 took two straight FB right down the pipe. (I did have 1 instance of someone taking 2 straight FB down the pipe.)

On average in 3-0 Counts, 5/9 took a FB right down the pipe, regardless of hot/cold zone.

********************************************************************************

*********************************************************************************

******************************

As I said, nothing conclusive, but I was happy to see a steady downward progression as the ratings were lowered. This should mean that we have another reliable way to further improve the AI's plate presence in-game. Not sure how it would effect sim stats (though it would likely lead to higher walk totals I'd imagine).

Other items of note:

-Every time I've tried to groove a 3-0 FB down the pipe with the bases loaded to a batter with a "C-" or better, the pitch was taken (as it should be)...with the exception of one player, Dustin Pedroia...lol. 2/3 times he swung away.

There was a bit more variance in players "D+" and below, but for the most part the pitch was still taken. This should prove that the patch (whatever you think of it's full value) works as advertised. This almost NEVER happened pre-patch.

-This test was done with IE influence at it's default though (meaning maxed) and each player has individual swing tendencies by pitch count. So that could essentially factor in as a randomizer to go along with the ratings.

-On one of my 3-0 tests, I grooved two straight FB's to Bay and he took them both. I then tried to paint the corner with one and he took that as well. So I actually got the AI to take 6 straight pitches for a K. There's hope yet!

-I also played one game against my Red Sox with maxed out walk ratings and it was VERY noticeable how much more patient the AI was. It was actually a good game and I actually had to pull my SP due to pitch count for the first time.

I would say ratings will be a big factor in getting this to work. I'm not sure what can be accomplished via sliders since IE is already maxed at default and I'm not sure what a couple do.

Also not saying you should have to do so much editing to get a good game of baseball, but the point is that it seems if you want to, you can likely tweak the ratings more to realistic satisfaction now.

********************************************************************************

*********************************************************************************

*********************************

For those dedicated enough, here's a chart to show what ratings range corresponds to each IE letter grade:

A+: 93 and up

A : 88-92

A-: 84-87

B+: 81-83

B : 76-80

B-: 71-75

C+: 68-70

C : 63-67

C-: 58-62

D+: 55-57

D : 50-54

F : 49 and under

Hopefully this will help guys out with tweaking not only walk ratings, but other IE ratings as well.

Sorry for the long post, but I know some people have still been les than satisfied with the AI tweaks in the patch. Now that IE actaully works properly (so it seems, anyway), this may help people find what they're looking for. Early returns have been promising for me, so I shared.

Good luck!"

Link to comment
Share on other sites

Thanks Soxfan for taking the time to do that research. B)

My very limited time with the patch says the AI now swings at 95% of anything trown at the strike zone whereas before they swing 99.9% of the time. I have found that it's not as easy to pick off runners so that's good. Al in all i'm still having fun with this game as I also have The Show for the PS3 and PSP. MLB 2K9 is a lot of fun and i'm looking forward to seeing what VC will do with MLB 2K10.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...