Too many two out inning rallies by AI!! (Comeback AI Discussion)

[EnigmaNemesis]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

I think closers are just as good if not better then starters (some anyway). I just think people are getting in trouble with their middle relief guys. Typically these pitcher's don't stack up with starter's because of their lack of consistency.

I also think many people get overconfident with their closers, and leave a lot of meaty fastballs too much in the zone. I have been guilty of this on several occasions. Sometimes it works, and I strike out the side on 10 pitches, and sometimes I pay with a HR shot to tie the game.

[phillyfan23]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

What sliders have u adjusted if any

still playing on my legend sliders from back in the demo ....no adjustments....

but i do think it is important to find your level of play....it's going to seem that there might be comeback logic if your sliders aren't balanced IMO.

[kehlis]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

Except that Moneyball has nothing to do with clutch. Chalk another one up for incorrect use of the word "Moneyball", a business strategy.

What?

Throughout the book Moneyball, clutch hitting was discussed at great lengths. Not sure how you could have taken my statement any other way.

[Bigtonyclark]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

What?

Throughout the book Moneyball, clutch hitting was discussed at great lengths. Not sure how you could have taken my statement any other way.

Moneyball is a business strategy based on the ability to find inefficiencies in a certain market. It can be applied in any business. In this case, the inefficiencies were found through the use of sabermetrics. Believing in Moneyball is not believing in sabermetrics or the lack of clutch hitting or anything.

Plus, Lewis contradicts himself when he talks about how Beane could easily jack them out during BP with an empty stadium, but failed to do so in a game, pressure situation.

[kehlis]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

Moneyball is a business strategy based on the ability to find inefficiencies in a certain market. It can be applied in any business. In this case, the inefficiencies were found through the use of sabermetrics. Believing in Moneyball is not believing in sabermetrics or the lack of clutch hitting or anything.

Plus, Lewis contradicts himself when he talks about how Beane could easily jack them out during BP with an empty stadium, but failed to do so in a game, pressure situation.

Okay, I apologize for not specifying that I was simply referring to the fact that he read the book "Moneyball"...

I hope you are not offended and again, I am sorry I did not specify.

I hope you feel better now.

[swaldo]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

Here is an update of the data from the tests I have been working on. (My original post showing games 1-15 are on page 43 of this thread.)

The numbers below are from 30 CPU vs CPU games on Hall of Fame level. Each team in both the NL and AL played a game twice. Default everything but I did turn ump variation off so all balls and strikes are called perfectly. Why CPU vs CPU? By testing in this matter nobody can accuse me of playing the CPU poorly.

I then compared this data to 30 real <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com</st1:date>June 1, 2008 and <st1:date Year="2008" Day="3" Month="6" ls="trans">June 3, 2008</st1:date>. Why June? I figure both batters and hitters are in good form by then. Also, all NL and AL teams played games on these dates as well.

These tests took around 30 hours for my PS3 to churn out the results (they are not quick sim games.) I wanted results as accurate as possible so I wanted to use ball physics instead of texed sim results. I first selected human vs CPU then when the game started I switched my controller to CPU. In regards to pitchers I chose a mix of good, fair and poor to replicate real life.<O</O

Before I get to the stats I want to make clear of what the definition of a comeback is. A comeback is when a team that is losing comes back to tie or take the lead at any given time. For example, here is a fictitional box score.

101005000
110040000

In the above example there were two 1 run comebacks and one 4 run comeback.

So far in my 30 game 'The Show' test here is the breakdown:

1 run CB: 16
2 run CB: 11
3 run CB: 5<O</O
5 run CB: 1<O</O
6 run CB: 1
TOTAL COMEBACKS: 34<O</O
TOTAL COMEBACK RUNS: 64

CB's in last 3 innings: 10
CB's in 9th inning: 3

Total runs: 284 (4.73 AVG per team)
Total hits: 567 (9.45 AVG per team)
Total Errors: 28 (0.47 AVG per team)

Total Runs - first 3 innings: 75
Total Runs - middle innings: 114
Total Runs - last 3 innings: 87
Extra Inning runs - 8

<st1:stockticker>REAL </st1:stockticker><st1:stockticker>MLB</st1:stockticker> GAMES ON <st1:date Year="2008" Day="1" Month="6" ls="trans">JUNE 1 & 3, 2008 </st1:date>BREAKDOWN:

1 run CB: 24
2 run CB: 8
3 run CB: 3
4 run CB: 1
TOTAL COMEBACKS: 36<O</O
TOTAL COMEBACK RUNS: 53

CB's in last 3 innings: 5
CB's in 9th inning: 2

Total runs: 288 (4.8 AVG per team)
Total hits: 531 (8.85 AVG per team)
Total Errors: 36 (0.6 AVG per team)

Total Runs - first 3 innings: 104
Total Runs - middle innings: 103
Total Runs - last 3 innings: 74
Extra Inning runs - 7

Run distribution (How many times a certain amount of runs was scored in an inning (The Show : Real <st1:stockticker>MLB</st1:stockticker>):

1 run - 72 : 84
2 runs - 57 : 39
3 runs - 19 : 18
4 runs - 5 : 7
5 runs - 3 : 3
6 runs - 0 : 2
7 runs - 0 : 0
8 runs - 0 : 1
9 runs - 0 : 1

In conclusion comebacks seem about right – although total comeback runs are higher but there was still a tiny bit more offense in The Show. Other than 1-2 more hits in The Show games the stats match almost perfectly. Here are the numbers that interest me - breaking down runs scored by inning here is the results for each: <O</O

The Show: 75,114,87
Real <st1:stockticker>MLB</st1:stockticker>: 104,103,74<O</O
<O</O

The Show last 6 innings runs scored: 201<O</O
<st1:stockticker>MLB</st1:stockticker> First 6 innings runs scored: 207<O</O
<O</O

In the real <st1:stockticker>MLB</st1:stockticker> runs were generally scored early. The Show results are opposite - runs generally were scored later. Also, the one weakness that I saw was run distribution. There were a lot of 1-2 run innings in The Show, but the <st1:stockticker>MLB</st1:stockticker> at times scored in bunches. There were less low scoring games and less high scoring games in The Show. Instead, there were a lot of mid-range scores. I will post a separate thread soon breaking this down better along with a ton of other stats. <O</O

[swaldo]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

Also, if you don't get anything from this at least take down the comeback results which you can later compare to your personal games. That way you can see if things are out of whack compared to real life probabilities of a team coming from behind.

[ComfortablyLomb]

Re: Too many two out inning rallies by AI!! (Comeback AI Discussion)

Just, lol. There are a lot of reasons that some guys can excel as starters or relievers but not be as good in the reverse role. Maybe it's a guy who has just two major league pitches and nothing else that works. He may be very useful once through the order but asking him to get guys out two or three times may be asking too much. Or maybe it's a guy who throws a bunch of different pitches but nothing particularly special. He'll likely be able to give batters a different look each time through the order but he's not going to overpower anyone. Considering how hard it is to find decent starting pitchers a team is going to hope that guy can get them five or six each time out rather than keeping him as a mediocre reliever.

Regarding Papelbon, he wouldn't be pitching the same way as starter. He couldn't be throwing every fastball with maximum effort and he couldn't be relying on it as much either. He has a useful splitter in that it plays well off his fastball and a very mediocre to poor slider. If his fastball has to be toned down a notch then I don't think either of those other two pitches are as effective.

The bottom line? I think a pitcher who ends up as a reliever GENERALLY has been converted into one because he wasn't good enough to be a starter. There are obvious and numerous exceptions but most of the time that's just how it is.

[Tyrant8RDFL]

For those that question the comebacks and AI play late in the game.

I have seen time and time again members getting fustrated with the way the game plays in the late innings. All the comebacks, and how the CPU has a great eye wont swing at anything close and everything they hit late gets rocked for HR's.

I just wanted to pass on some adjustments I made in the game. These adjustments have worked very well for me so Im passing on.

Here are the slight slider adjustments I made that to me eliminated all that crazyness.

Move the Pitchers Consistantcy slider 3 to the right.
Move base running 2 to the right.
Fielding errors and throwing errors 1 to the left.

PLay on Allstar and let it ride. Also if you change levels, just adjust Pitchers consistantcy.

The game plays a ton better this way, and no more bull crap comebacks in the 8th inning,and the perfect eagel eye all of the sudden from the hitters. Dont get me wrong you make bad pitches you will pay sometimes, but everything feels alot more balanced and realsitic this way.

Hope this helps. Peace

[Jackdog]

Re: For those that question the comebacks and AI play late in the game.

I have seen time and time again members getting fustrated with the way the game plays in the late innings. All the comebacks, and how the CPU has a great eye wont swing at anything close and everything they hit late gets rocked for HR's.

I just wanted to pass on some adjustments I made in the game. These adjustments have worked very well for me so Im passing on.

Here are the slight slider adjustments I made that to me eliminated all that crazyness.

Move the Pitchers Consistantcy slider 3 to the right.
Move base running 2 to the right.
Fielding errors and throwing errors 1 to the left.

PLay on Allstar and let it ride. Also if you change levels, just adjust Pitchers consistantcy.

The game plays a ton better this way, and no more bull crap comebacks in the 8th inning,and the perfect eagel eye all of the sudden from the hitters. Dont get me wrong you make bad pitches you will pay sometimes, but everything feels alot more balanced and realsitic this way.

Hope this helps. Peace

Great suggestion.

[TheEdgeOfSoul]

Re: For those that question the comebacks and AI play late in the game.

Are you saying that both the CPU and User sliders be adjusted accordingly or is it one or the other? Also, when you're saying "move 2 to the right", what does that mean exactly? I'd like to implement these but I just wanted some clarification.

I've been victimized by the ridiculous comeback ability that the CPU brings every time I get a lead. I'm hoping that these sliders can help.

Thanks in advance!

[chuckm1961]

Re: For those that question the comebacks and AI play late in the game.

So what you're telling us is that the "AI comeback" will stop if you make pitchers somewhat more consistent in hitting their targets, make players run the bases a bit faster, and slightly reduce the number of errors?

Sorry but that makes no sense whatsoever.

[moemoe24]

Re: For those that question the comebacks and AI play late in the game.

I have seen time and time again members getting fustrated with the way the game plays in the late innings. All the comebacks, and how the CPU has a great eye wont swing at anything close and everything they hit late gets rocked for HR's.

I just wanted to pass on some adjustments I made in the game. These adjustments have worked very well for me so Im passing on.

Here are the slight slider adjustments I made that to me eliminated all that crazyness.

Move the Pitchers Consistantcy slider 3 to the right.
Move base running 2 to the right.
Fielding errors and throwing errors 1 to the left.

PLay on Allstar and let it ride. Also if you change levels, just adjust Pitchers consistantcy.

The game plays a ton better this way, and no more bull crap comebacks in the 8th inning,and the perfect eagel eye all of the sudden from the hitters. Dont get me wrong you make bad pitches you will pay sometimes, but everything feels alot more balanced and realsitic this way.

Hope this helps. Peace

Hope this doesn't sound smart assy cause I really do welcome any suggestion and appreciate you posting this, but how does moving baserunning up help the comebacks?? I can see how the pitch consistency would help, but I don't get the baserunning adjustment. Also, what if you're already playing with slider adjustments?? Are you saying 3 clicks up from wherever you have your slider set??

[Phoenixmgs]

Re: For those that question the comebacks and AI play late in the game.

So what you're telling us is that the "AI comeback" will stop if you make pitchers somewhat more consistent in hitting their targets, make players run the bases a bit faster, and slightly reduce the number of errors?

Sorry but that makes no sense whatsoever.

I agree with that 100%

[statum71]

Re: For those that question the comebacks and AI play late in the game.

Actually this thread makes me feel a lot better. I thought it was just me.

If the game is close coming down the stretch, I can rest assured the CPU is gonna do something great late in the game. Leaving me wondering where I messed up.