Bunting for hit

[Bledsoe]

Being able to bunt for a hit seems VERY successful, especially if you have a guy who's good at it. It seems to be successful ~70% of the time with guys who are good at it (In fact, I just did it three straight times in a game when I had no one on base - all successful). It seems like this should be good 20-30%, not this high

[Bledsoe]

And the computer's "bunt for hit" that prompted this post just cost me a playoff game. So...yeah, please fix.

[RchW]

What is the right number?

[Charlie Hough]

Sorry, this doesn't cut it.

What is your overall sample size? Have you done large-scale testing? Otherwise this is a moot point. Three consecutive instances in a game and one other case in isolation do not constitute a trend or anything of statistical significance.

You've also failed to account for the defensive positioning that was used and the relative defensive ratings of the players involved.

Also, in order to draw any conclusions, you'll need to examine statistical evidence of the actual rates of bunting success in MLB, both historically and in the contemporary game.

When you start sentences with "it seems", then we're headed nowhere.

Sorry if this seems harsh, but there is a flood of posts like this every week, and those posts claim unrealistic results in OOTP without offering any real evidence or support. This game has a long history of being programmed to generate highly realistic outcomes that match historical and contemporary MLB statistics. And the game achieves this very successfully.

If you suspect that there is a legitimate problem, then you've got to do appropriate testing and provide sufficient numbers to prove your case. If you can, then it will almost certainly be addressed.

[Bledsoe]

This is pretty harsh, especially towards someone who is a long-time (2001) fan of the franchise and just wanted to point out what I perceived to be a small flaw in an otherwise great game. Your contention that I must do in-depth statistical analysis to mention a potential bug is ridiculous. This is the job of the game creators/developers.

I am just a fan (not to mention purchaser) that is trying to help improve the overall product. As someone just trying to help, I certainly think "It seems..." is a perfectly valid way to mention something that may be off (What if you were a mechanic and someone brought a car to you and said "It seems my brakes are going"?) I'm not mentioning something that happened *only* three times and saying "Well, this MUST be a serious flaw". This has been the case since I started playing the game the day it was released. While I agree that anecdotal eveidence isn't as strong as sound statistical evidence, it also cannot be completely discounted or disregarded.

I am not saying I am 100% certain I am right; in fact it could be a statistical anomaly I experienced, and that's fine. I just wanted to raise an issue for those parties who have the ability to address, to do so. I don't think any developer who cared about their product would reject this feedback. If they came back and said "Hey, we checked this out, and we don't see a problem," then case closed.

Again, I'm just trying to help.

[SteveP]

I studied this at some length a couple of years ago and found that bunt for hit success in OOTP was about half of what it should be according to actual stats (which don't exist for the entire MLB history, so one has to assume this is fairly constant). So I set the config file for this to 200.

I also set the overall bunting success rate to 170, to compensate for the fact that too many sac bunts end up with double plays or outs on the lead runner, again in comparison to actual stats.

FWIW

[Charlie Hough]

Originally Posted by SteveP

I studied this at some length a couple of years ago and found that bunt for hit success in OOTP was about half of what it should be according to actual stats (which don't exist for the entire MLB history, so one has to assume this is fairly constant). So I set the config file for this to 200.

I also set the overall bunting success rate to 170, to compensate for the fact that too many sac bunts end up with double plays or outs on the lead runner, again in comparison to actual stats.

FWIW

This is a beautiful example of what I'm talking about. Large scale testing will often show the complete opposite of what people intuitively believe is happening in the game. As your testing indicated, the success rates for bunting may actually be too low in the game rather than too high.

As for Bledsoe, I'm sure your intentions are constructive, but hunches are NOT a productive way to report a potential bug. However, this isn't a personal criticism. This is just an epidemic that has been sweeping the forums for years, and your post finally moved me to express the frustration that it brings about.

Human beings are highly prone to bias, bad reasoning, and unreliability. So, in any form of software development, a problem has to be repeatable, duplicated on multiple installations, and statistically significant to warrant reviewing it.

If the developer goes off on a wild goose chase to examine every claim based on anecdotal evidence and minimal sample sizes, then we're never going to see new versions released or new features developed. All the developer's time will be spent on dispelling unfounded claims.

That's why some of the others in the community try to step in and ask people to do more research, be more sure that there is a problem, and satisfy some basic criteria before we draw any conclusions. If we have multiple users reporting the same issue and able to repeat it and confirm it with large-scale testing, then we'll be the first ones to support reporting it as a bug. And this will help get it addressed much faster and more efficiently.

There is simply no way that any developer has the time or resources to investigate every claim that is posted without sufficient evidence. So if you want your concerns addressed, there are better ways to do it.

So, apologies for you being the unwitting victim of a rant, but we're seeing similar postings constantly, and we need to figure out a way to steer things in a more sensible direction. This is perhaps something that warrants a stickied thread on all the forums to remind people about the criteria we need to satisfy to help us quickly determine if there is a real problem and how to fix it.

[Groucho Marx]

Originally Posted by Bledsoe

...[snip]Your contention that I must do in-depth statistical analysis to mention a potential bug is ridiculous. This is the job of the game creators/developers. [snip]...

Anyone in the software business in, in general, and the gaming sofware business, in particular, will tell you this.

The bulk of playtesting takes place with the end user. Most game companies have limited staff and a finite amount of time to playtest their product. They rely on the end user a lot more than most folkks think. This may or may not be the case with this company. I don't know.

The More You Know...

*little star/rainbow graphic*

[RchW]

Originally Posted by SteveP

I studied this at some length a couple of years ago and found that bunt for hit success in OOTP was about half of what it should be according to actual stats (which don't exist for the entire MLB history, so one has to assume this is fairly constant). So I set the config file for this to 200.

I also set the overall bunting success rate to 170, to compensate for the fact that too many sac bunts end up with double plays or outs on the lead runner, again in comparison to actual stats.

FWIW

Can you point out where bunts for hits are found in OOTP? I checked everything including the csv files and can't find it.

[Jamie]

Just so we all know.....what is the correct sample size so end users have creditability with their game experiences and observations?

[RchW]

Originally Posted by Jamie

Just so we all know.....what is the correct sample size so end users have creditability with their game experiences and observations?

I don't know but it needs to be more than a few games. IRL the last 5 season bunts for hits occurred at the following rate:

2007 every 8.86 games
2008 every 8.72 games
2009 every 8.83 games
2010 every 8.61 games
2011 every 7.99 games

That's 1-2 per day in a 30 team league so I'd want a seasons worth minimum before changing my settings. YMMV of course.

I can't find BFH anywhere in OOTP so if anyone can tell me I can check a few seasons "in game".

[SteveP]

Originally Posted by RchW

Can you point out where bunts for hits are found in OOTP? I checked everything including the csv files and can't find it.

If I understand the question correctly ...

There are a lot of stats in OOTP that you can't get by looking at reports or screens. You have to gather them the hard way. I did that for a bunch of stuff using an entire OOTP season (historical), and comparing the results to real life stats. As a result, I learned to do a variety of adjustments to game settings. In a few cases I was able to convince Markus to make changes (he can get stats from the game that we can't get ).

Hope that answers the question.

[RchW]

Originally Posted by SteveP

If I understand the question correctly ...

There are a lot of stats in OOTP that you can't get by looking at reports or screens. You have to gather them the hard way. I did that for a bunch of stuff using an entire OOTP season (historical), and comparing the results to real life stats. As a result, I learned to do a variety of adjustments to game settings. In a few cases I was able to convince Markus to make changes (he can get stats from the game that we can't get ).

Hope that answers the question.

OK, please tell me the hard way. Game logs?

[SteveP]

Originally Posted by RchW

OK, please tell me the hard way. Game logs?

Yes, for the most part. I also gathered data from historical game logs as well (from the same historical season) when I could not get the stats directly.