Update + Another DIPS Discussion...

[Gastric ReFlux]

Originally Posted by 9lives

But they're not equal. Every pitcher has a different arsenal of pitches and every pitch jumps off the bat differently when contact is made. If you include non-major league pitchers the variance is greater on balls in play from one pitcher to the next...but without the crappy pitchers it still exists.

Do the same test of 1000 balls in play with Pedro vs. John Halama.

That test has been done in the form of McCracken's and Tippet's studies.

[obaslg]

Originally Posted by rcbuss

Great job, Markus, and thanks for making a great game even more realistic. That's what a lot of us thought that OOTP6 was missing. Finally, proof that the potential for pitcher control of BABIP in OOTP6 exists. It stands to reason that pitchers who do well in DIPS categories also tend to do better in tougher-to-measure, non-DIPS skills (e.g., BABIP), though there are exceptions.

Tom Tippett's article (http://www.diamond-mind.com/articles/ipavg2.htm) says as much (emphasis mine) -- hopefully, this is the last time I trot this one out:



:
:
:

I didn't read the original - only your excerpt - but I see some problems here. The first is the idea that the BABIP "may" vary around a number other than zero - couldn't he have just tested that?

The second problem is his conclusion about better pitchers - those with longer careers - have better BABIP. There's nothing surprising about the fact that a pitcher with a better BABIP would have a lower ERA and therefore a longer career. The question is whether he is causing the good BABIP or it's just a matter of luck. The very low year-to-year correlation indicates to me that it's almost all luck.

[Sweed]

Originally Posted by Markus Heinsohn

True... See, before that, you could have pitchers with an opponents AVG of .220 and just 3.5 K/9, who would put those numbers up regularly. K's didn't matter at all, it was just another out. That has changed with the DIPS engine, and it produces stats that are far more true to life than the OOTP 5 engine. And like Henry said, 90+% is based on McCrackens work

IMHO this is working great in ootp and that is what counts. It is producing very realistic stats and should make people on both sides of the dips argument happy.

[IatricSB]

Originally Posted by Markus Heinsohn

...before that, you could have pitchers with an opponents AVG of .220 and just 3.5 K/9, who would put those numbers up regularly.

Addie Joss.... 3.55 K/9 for his career, guestimate on career average ~.213 (based on his 7.30 H/9 career average)

[IatricSB]

Originally Posted by Sweed

IMHO this is working great in ootp and that is what counts. It is producing very realistic stats and should make people on both sides of the dips argument happy.

But not all (some who sim historical leagues). Wish I had time to do a complete league study of the bell curve distribution of ERAs in a dead ball league as compared to a present day league (comparing real life spread versus OOTP spread). Then at least I could either become a believer in it or else see that it verifies what I've perceived.

[matskralc]

Originally Posted by IatricSB

Addie Joss.... 3.55 K/9 for his career, guestimate on career average ~.213 (based on his 7.30 H/9 career average)

Addie Joss is what you would call an "exception". Pre-OOTP 6, low K/9 pitchers exhibiting low batting averages against were not exactly exceptions.

[sebastian0622]

First, thanks for the update Markus! I am able to breathe a little easier knowing that BABIP is, in some way, incorporated into the OOTP engine. Although, from my studies, BABIP does not rely on stuff. The lowest 10 BABIPs in my expanded study of 66 current pitchers shows the following:

1. T.Percival
2. B.Zito
3. B.Wagner
4. E.Gagne
5. J.Washburn
6. Hoffman
7. K. Wood
8. Franklin
9. B.Webb
10. T.Wakefield

While many have good stuff, many others who have good stuff are lower on the list (Prior and Oswalt are in the bottom 10). Without getting too deep into the DIPS argument on this thread, I will just say that BABIP doesn't appear to be a composite of any of the DIPS categories. I have unsuccessfully tried to correlate BABIP to all three of the DIPS statistics, both individually and weighted with variations of Kevin Harlow's DIPS-based game scoring. In future incarnations of OOTP, I believe BABIP should be a seperate rating with the option of being hidden for those nonbelievers out there :-D

[rcbuss]

Originally Posted by obaslg

I didn't read the original - only your excerpt - but I see some problems here. The first is the idea that the BABIP "may" vary around a number other than zero - couldn't he have just tested that?

I'm assuming that that number other than zero in Tippett's article would be the career IPvsAvg or IPvsTm for each pitcher. Of course, each player has improvement and diminution of skills over the course of a career, but the career numbers are a good guide.

Originally Posted by obaslg

The second problem is his conclusion about better pitchers - those with longer careers - have better BABIP. There's nothing surprising about the fact that a pitcher with a better BABIP would have a lower ERA and therefore a longer career. The question is whether he is causing the good BABIP or it's just a matter of luck. The very low year-to-year correlation indicates to me that it's almost all luck.

I think there is an element of selection bias here, as there should be. Inferior, 4-A type pitchers get weeded out early on in their careers, while pitchers with DIPS skills also end up exhibiting some BABIP skill, as well, and go on to longer career. I don't think that it's coincidence that major league pitchers, when grouped by career length, show a positive correlation between BABIP and DIPS skills. The difference is small, and the correlation low, but it is present. And with the number of BFP being looked at (over 2,500,000), the chances of this being due to luck are small.

It appears that BABIP is a skill, as are BB and HR avoidance, and getting Ks. The relative importance of each of the four skills depends on the individual pitcher. Take a look at the Charlie Hough graph in the article for a good example of this. Without BABIP, he wouldn't have lasted nearly as long as he had.

[sebastian0622]

very well said rcbuss.

If you haven't read the Tippett study, in ITS ENTIRETY, stop, don't post, don't pass go, and instead go read it now. The same can be said if you haven't read both of Voros McCracken's releases on the subject. You can learn a lot more from reading those studies than asking questions on this board. Then, you can judge for yourself.

[dougaiton]

Originally Posted by obaslg

The second problem is his conclusion about better pitchers - those with longer careers - have better BABIP. There's nothing surprising about the fact that a pitcher with a better BABIP would have a lower ERA and therefore a longer career. The question is whether he is causing the good BABIP or it's just a matter of luck. The very low year-to-year correlation indicates to me that it's almost all luck.

We've all danced this one a million times, but you're quite right obaslg.

It's worth saying again that a lot of people wholeheartedly agree with Tippett's results, but think he drew conclusions that weren't necessarily there. Moreover, his conclusions aren't weighted - although he has 10 points talking about how pitchers can modify BABIP, he has one point which says that it isn't really a consistent skill, and another that says it ain't a patch on HR/K/BB.

Once again, Tippett's great. But it's worth remembering that like all analysts, his conclusions aren't always as sound as his methodology.

[dougaiton]

Originally Posted by sebastian0622

very well said rcbuss.

If you haven't read the Tippett study, in ITS ENTIRETY, stop, don't post, don't pass go, and instead go read it now. The same can be said if you haven't read both of Voros McCracken's releases on the subject. You can learn a lot more from reading those studies than asking questions on this board. Then, you can judge for yourself.

Very good point - our chances of adding anything new to the discussion is pretty unlikely. Rather, we can hopefully arrive at a synthesis of what's written and proceed accordingly!

[Bluenoser]

Originally Posted by Markus Heinsohn

Now, what's going on with OOTP 6? Well, next up is the update, which unlike a patch, will add some more features along with the bug fixing. H2H is the primary feature that is planned for that update, but lots of small useful things will be part of it as well. There is no release date for the update yet, but I hope to have it all wrapped up before the MLB playoffs start. I may release a patch prior to that, since I will concentrate on bug-fixing first before putting more features in.

Everyone else seems quite excited by this post, and while I do appreciate the status update, I have to ask. Why is it we're still waiting for something (H2H) that was supposed to be added within a few weeks of game release? I won't go so far as to liken it to false advertising, but when a game developer makes a post stating that he'll release the game early and something that was supposed to follow shortly after will now be added "hopefully" by Oct, I have to question that. I, like some others, paid for the game for many reasons, one of them being led to believe that H2H would follow soon afterwords.

I said it before and I'll say it again. Financial Co-Efficient = H2H. It's not going to happen.

You can prove me wrong by releasing it, not by feeding me more words that it's coming soon, maybe in a couple more months.



Now I'm quite sure I'll get flamed for this negative post, but by all means, have at it. It won't bother me in the least.

[MLBCOOTP]

yeah i have to agree with Bruce on this one. Not that i dont appreicate everything that Marcus has done for the OOTP series, but it just seems like there is more focus on OOTP7 at the moment than fixing the problems with the current game.

not that i dont think ootp6 isnt good to begin with, i like it. but there are still some known issues from what i have read, and i think those should get proirity over the release of a new version of a game that isnt working 100% correctly.

and sorry if i am also negative in my post. just trying to voice my thoughts.

[sporr]

Originally Posted by BruceM

Everyone else seems quite excited by this post, and while I do appreciate the status update, I have to ask. Why is it we're still waiting for something (H2H) that was supposed to be added within a few weeks of game release? I won't go so far as to liken it to false advertising, but when a game developer makes a post stating that he'll release the game early and something that was supposed to follow shortly after will now be added "hopefully" by Oct, I have to question that. I, like some others, paid for the game for many reasons, one of them being led to believe that H2H would follow soon afterwords.

I said it before and I'll say it again. Financial Co-Efficient = H2H. It's not going to happen.

You can prove me wrong by releasing it, not by feeding me more words that it's coming soon, maybe in a couple more months.



Now I'm quite sure I'll get flamed for this negative post, but by all means, have at it. It won't bother me in the least.

Nobody said that H2H would be included within a few weeks of the game's release. That was only true if the game wouldn't have been released when it was.

[Bluenoser]

Originally Posted by sporr

Nobody said that H2H would be included within a few weeks of the game's release. That was only true if the game wouldn't have been released when it was.

Yes, Markus did say it would be out within a few weeks. That was the whole point made by Markus when he started the topic of early release. The only thing he had left to do was a few small things and add the H2H. He said if he released early, that H2H would be added within a couple of weeks. He even started a poll on it. The vast majority voted for early release and said they could live without H2H for the few weeks.

[Deitrus]

[sebastian0622]

I think it is important that we don't bash Markus at every thread. He is very well aware of the H2H problem. I am a little pissed about it too. However, we don't want Markus to feel he can't post anywhere about any topic before he completes H2H. Regardless of the H2H situation, I was glad to read this post.

[Bluenoser]

Originally Posted by sebastian0622

I think it is important that we don't bash Markus at every thread. He is very well aware of the H2H problem. I am a little pissed about it too. However, we don't want Markus to feel he can't post anywhere about any topic before he completes H2H. Regardless of the H2H situation, I was glad to read this post.

Please enlighten me as to where in my post I "bashed" Markus. I kept it very simple and to the point without being rude, inflamatory, or bashing. In fact I opened my thread by mentioning my appreciation for the update. I'm not pissed, I'm just asking a fair question.

[sebastian0622]

I'm sorry, let me rephrase:

I think it is important that we don't ask Markus about H2H in every thread. He is very well aware of the H2H problem. I am a little pissed about it too. However, we don't want Markus to feel he can't post anywhere about any topic before he completes H2H. Regardless of the H2H situation, I was glad to read this post.

[sporr]

Originally Posted by BruceM

Yes, Markus did say it would be out within a few weeks. That was the whole point made by Markus when he started the topic of early release. The only thing he had left to do was a few small things and add the H2H. He said if he released early, that H2H would be added within a couple of weeks. He even started a poll on it. The vast majority voted for early release and said they could live without H2H for the few weeks.

"OK, today I have been thinking about a possibility to get the game released sooner. The only real big feature that needs to be put in is the internet head-to-head mode. It is already 50% implemented, but I guess getting it done 100% and testing it properly will take 1 week.

So, the question is, I could leave this out of the initial release, and concentrate on it fully for the first big update, which usually is 2-3 weeks after release time.

What do you think, would you mind if I left this feature out of the initial release to get the game out sooner?"

I added the bolding. Please note that H2H had been 50% implemented at the time. However, when the game was released early, it was 0% implemented. Maybe not exactly starting from scratch, but we can't assume that the same amount of time would be needed to get it back in there. As it is, it is still planned for the 1st big update. I don't disagree that it has taken longer than expected to get this feature in the game, but I argue with those who say it should have been out a few weeks after the game was released.