One Pitch vs Pitch by Pitch

[greengold01]

Anybody find different results using either mode while managing? I spent half a season with the Angels using Pitch by Pitch and it was a pitchers battle almost every game. Considering I spent half that time with my top 3 starters on the DL (and considering the Angels offense) that was surprising.

I just finished half a season with the Diamondbacks (whose offense is less than stellar) and I found much more realistic results. 4.7 runs/game, I have seen two suspensions, multiple outfielder errors (none while using pitch by pitch,) better steal percentages, even smarter bullpen substitutions by the A.I.

I'll test this more, half a season isn't exactly a telling time table but have you guys found anything similar?

[phenom]

The only thing I've found with pitch-by-pitch is that two-out rallies are the norm, not the exception.

[greengold01]

I've noticed that too. Countless times I've gotten two quick outs, given up a double (or even better, a throwing error and the runner ends up on second), a stolen base, and a wild pitch would bring in the tying/go ahead run. Aggrevating when your best hitter is a catcher with a rubber arm and you keep losing 2-1 games due to this.

[RchW]

I'm not sure how one can make any definitive case for different results. How can you tell, for example, that the results wouldn't have been the same if you used one pitch?

Game 1-81 cannot be compared to game 82-162. The only way I can think of to try and compare would be to alternate each game and peruse the box scores and game logs to identify trends based on which pitch mode was used. Even then, different pitchers, some different players and different teams could fog the outcome.

I just finished half a season with the Diamondbacks (whose offense is less than stellar) and I found much more realistic results. 4.7 runs/game, I have seen two suspensions, multiple outfielder errors (none while using pitch by pitch,) better steal percentages, even smarter bullpen substitutions by the A.I.

Help me understand. Do you just control the pitch mode and let the AI do the rest?

Was the half season (that gave more realistic results) all one pitch?

When did you do the pitch by pitch to compare?

How many games of each?

I sim all my regular season games. I'm sure I could find a block of games with errors in bunches and then another block with none. Suspensions are an artifact that appears to be random IMO. Not sure there is any connection to pitch mode.

The 2 out rallies idea should be easy (though time consuming) to check. Just check the game logs vs the pitch mode. You'd still need a bucket full of games to identify a definitive trend.

I'm not trying to poo poo the thought but unless you have a significant number of results that repeatedly shows a trend, I'm not sure how any conclusion can be reached.

[Malleus Dei]

Originally Posted by RchW

unless you have a significant number of results that repeatedly shows a trend, I'm not sure how any conclusion can be reached.

Well put.

[SteveP]

If you are calling plays in both cases, you could get different results. In pitch-by-pitch, you are able to decide exactly what play you want called at each pitch count. In one-pitch, if you call a play you are turning over some control to the AI as to the execution of that play during the course of the AB. For example, if you call for an SH, the AI may execute it when there are 2 strikes, even if it's your firm policy never to do that when you are calling plays on a pitch-by-pitch basis.

If you are only watching the game and the AI is in total control of play-calling, there will not be (there can't be) any difference if you are in one mode or the other.

[highandoutside]

This topic is really important to me since I play a lot of pitch by pitch. Does anyone have a sense if the statistical behavior of the players is affected by the mode selected?

[Killing Time]

A "sense" won't cut it. To answer this someone would have to gather hard data.

[OakDragon]

While what RchW and SteveP say is basically true, greengold's impressions are only to be expected. The game output can either be controlled by stats or by human decisions... you can't have it both ways overall. OOTP is controlled by stats (so that a .260 hitter averages .260 and so on). If you're pitch-by-pitching everything the game engine has to have a way to even out the stats after awhile, hence the two out rallies. I'm sure there are as many hitting droughts where needed in such cases, too. Since in pitch-by-pitch mode there are more opportunities to control strategy (hit-n-run, steal, playing the infield in, etc.) these effects will only be more noticeable.

Anyway, that seems to me like a reasonable explanation for those observations.

[RonCo]

Unless something changed in X, OOTP does nothing to ensure a .260 hitter ends up a .260 hitter. There is no "evening out of stats." Everything is controlled by (the grossly mis-named) league totals--which are applied on the basis of a single AB, not on an annual basis. It's possible that adjustments for count could be causing some oddities, but I doubt it. I had done a study of the "damage" caused by always taking to 2-strikes, and found no degradation in batting averages...and in fact wound up with the opposite, probably because my hitters were facing tired pitchers more often.

I've come to the conclusion that pitch-by-pitch mode is essentially a "cheat" to help the human win. It's not designed that way, but it's a huge advantage to the human as best as I can tell both because of these design elements (results are fundamentally designed to occur on an at-bat basis, not a per-pitch basis) and because the in-game AI will always be behind a reasonably knowledgable human.

[OakDragon]

Well, I admit my "either or" statement was extreme, but I guess I don't understand something in that case. I don't see how setting strategy on a pitch by pitch basis can do anything but give the human player an extreme advantage (assuming the human uses good strategy) when compared to outcomes on a per play basis, unless pitch by pitch mode has no effect at all. The game would have to do something to compensate for this in order for the stats to come out anywhere near reasonable... otherwise you'd end up basically with an arcade-style baseball game. Anyone remember Home Run on the Atari 2600? You could end up with games that had scores of like 163-0, with 120 HRs in a game. Admittedly, there's no way OOTP could come out that extreme, but you get the idea.

[SteveP]

1. 80% of what goes on in a baseball game is the pitcher-batter duel. OOTP abstracts this part of the game so that the OOTPer has little or no influence on this duel, regardless of what mode you are playing. If you want to see every foul ball, you can. Or you can see what happens on the 3-2 pitch, after the pitcher has already thrown 11 pitches to this batter. Your choice.

2. If you call the plays, the decisions you get to make on a pitch-by-pitch are things like exactly when you want to SBA, or whether or not to SH with 2 strikes, or to take a pitch on a 2-2 count. I don't do this, because it makes for a very slow game -- I prefer to watch from the stands.

3. You are smarter than the AI. You will always be smarter than the AI. You don't need to be given any advantages in that direction. If you call your own plays on a pitch-by-pitch basis, you have that many more opportunities during a game to out-think the AI. It would make a difference.

4. If you don't call the plays, it doesn't matter whether you watch the game pitch-by-pitch, one-pitch, sim the whole game with one click, sim ten games with a click, etc. Everything happens exactly the same way (only slower or faster).

5. I don't know who to reconcile the League Total/League Total Modifier scheme, which is based on ABs, with a resolution calculation which is based on individual pitches. All I am saying is that it has nothing to do with what mode you play the game. It works the same way in any mode you choose. That .260 batter isn't going to be affected by your choice, unless you order him to SH every time he comes to bat, or some such thing.

[RonCo]

Originally Posted by OakDragon

Well, I admit my "either or" statement was extreme, but I guess I don't understand something in that case. I don't see how setting strategy on a pitch by pitch basis can do anything but give the human player an extreme advantage (assuming the human uses good strategy) when compared to outcomes on a per play basis, unless pitch by pitch mode has no effect at all. The game would have to do something to compensate for this in order for the stats to come out anywhere near reasonable... otherwise you'd end up basically with an arcade-style baseball game. Anyone remember Home Run on the Atari 2600? You could end up with games that had scores of like 163-0, with 120 HRs in a game. Admittedly, there's no way OOTP could come out that extreme, but you get the idea.

Markus suggested at one point that the game made some adjustments in possible outcomes based on count, but my testing suggested those might be of fairly limited impact.

I actually played several games out taking every pitch I possibly could, even with two strikes. It resulted in a lot of offense for my team and a ton of tired pitchers for the AI opponent.

[RonCo]

Just a 1-game data point...take it for what it's worth.

I just pulled up my one on-going league in v9 and played a game under the rule that I auto-simmed the opponents time at bat, and played pitch-by-pitch while my team batted. In addition, I made a rule that said I would take every pitch I could until my hitter either walked or got two strikes. Then I would swing away. No special strategies. Just take or swing.

Now, my team is pretty good...but I don't think it's ever lost a game under this structure. Today was no different. We were the visiting team.

Results:
--------
Us 12
Them 3

Our pitchers: 137 pitches (starter lasted 6 innings and 103 pitches)
Their pitchers: 212 pitches (starter lasted 4 innings and 109 pitches, five other pitchers appeared, two with over 40 pitches thrown)

My team's hitters:

1) drew 14 walks
2) had 8 total hits, two homers and two doubles
3) struck out 9 times

We average 11 hits, 2 homers, 1.5 doubles, 4 walks, and 6.5 strikeouts

[Killing Time]

That's one data point. We all know what drawing conclusions from one data point is worth.

[RonCo]

Originally Posted by Killing Time

That's one data point. We all know what drawing conclusions from one data point is worth.

If you know anything about my work, I've got a lot more data behind me than this one game...but here's one more data point--the next game in the schedule:

Same house rules:


Results:
--------
Us 13
Them 7

Our pitchers: 160 pitches (starter lasted 9 innings as the AI let him throw 51 in the last inning and give up 3 runs)
Their pitchers: 238 pitches (starter lasted 4.2 innings and 107 pitches, five other pitchers appeared, all with over 21 pitches thrown)

My team's hitters:

1) drew 7 walks
2) had 12 total hits, three homers and one double
3) struck out 8 times

We average 11 hits, 2 homers, 1.5 doubles, 4 walks, and 6.5 strikeouts

This was the first game in a series, so now the opponent's bullpen is shot.

[Afino]

Originally Posted by RonCo

Our pitchers: 160 pitches (starter lasted 9 innings as the AI let him throw 51 in the last inning and give up 3 runs)

It's neither here nor there, but that upsets me as well.

Apparently AI only uses the pen if you sim the rest of the game, not half-innings or innings. Can't stand it.

[Malleus Dei]

Originally Posted by Afino

Apparently AI only uses the pen if you sim the rest of the game, not half-innings or innings. Can't stand it.

I sure hope that's not true.

[RonCo]

Yes, that appears to be true.

[RchW]

How could that have been missed.