Poise, Poise, and Poise again ;)

[thomamon]

Well, I am dissapointed of the loss of Poise too... I thought it was going to help the quick drop as ratings decline and making a great pitcher, bad.

I can get over this very quickly if that problem is fixed, I don't want great players turning bad over night. I think it should vary by player, let some stay great longer and so on.

TM

[Mr.Opinion]

Originally posted by thomamon
Well, I am dissapointed of the loss of Poise too... I thought it was going to help the quick drop as ratings decline and making a great pitcher, bad.

I can get over this very quickly if that problem is fixed, I don't want great players turning bad over night. I think it should vary by player, let some stay great longer and so on.

TM

I don't know that it should have been reflected as poise...maybe a confidence rating or something, where it's how easily a player gets shaken after throwing a homerun ball or after he strikeouts out the side. I know that that definately has an effect on a pitcher performance...you have to think about the mental aspects of the game as well...

[DHS23]

Originally posted by sporr
A pitcher like Pedro Martinez has great control but still hits a lot of batters for other reasons.

In that case, we need a "Hate for Opposing Players" rating.

[Dark Horse]

That truly is good news.

[skkrrt]

I think Velocity should have a bigger role. Influence stuff in some way (harder to get "good wood" on a ball traveling at 100MPH) and maybe even HR....

just a thought.

[danwaechter]

I really like Carltons idea of having a 'Focus' type rating which decides the HBP, balks, and WP. Even if it were just an A-E rating, I think it is more realistic to combine those areas, and focus pretty much sums it up in my opinion.
As for poise, I coul;dnt care less. I agree that people made too big of a deal out of it before they even knew what it was.

[crackpott]

Originally posted by Markus Heinsohn
Here is the final list of pitcher ratings:
Stuff (heavily influences Ks)
Movement (Influences HRs)
Control (Influences BBs, and slightly Ks and HRs)
Velocity (Slightly influences Ks)

Endurance, Hold Runners, HBP, Balks and WPs.

Cheers,
Markus

Well, among the lovefest I thought I'd come up with my constructive criticism so to speak.

For one, I don't understand from a baseball point the difference between "stuff" and "movement". They seem to be synonyms for each other.

My other concern is the weighting of each category. I would argue that velocity should have a slightly greater role (maybe not as much as the others, but it should influence K's more than "slightly", and control's influence on homerun's should be minimal. My concern is that the control rating is going to be far too important in the grand scheme of things.

Also Markus, I read in the other thread that you're trying to make players import from V5 as smooth as possible, and I'm wondering what idea you have to convert V5 ratings to V6? ie. should online leagues be actually worrying about current players K ratings, and playing down their avoiding hits rating?

[Carlton]

The only thing I STILL don't comprehend is Control, Stuff and velocity influence K's...why 3 for one aspect? Hence why I asked why is Velocity even important if Stuff is the heavy K influence?

Wouldn't Stuff be better suited in as someone said...slowing down the decline of a pitcher? In theory a P that has better stuff could hang on a bit longer even when his velocity and control wanes a bit...or maybe it could be influential in FB/GB%?

and how will this effect the import from Lahman?

As for Poise...my stance was that I thought it would be purely for the sake of the CPU AI sticking with that pitcher over a younger pitcher with similiar ratings. Normally with the AI as it is in 5 a blue star 5 P would supplant a 3 gold star Vet...even if their ratings are almost identical...that also leads to being able to take that talented 3 star P out of the CPU teams hands because it "thinks" the 5 Blue Star is more valuable...

[crackpott]

Originally posted by Carlton
The only thing I STILL don't comprehend is Control, Stuff and velocity influence K's...why 3 for one aspect? Hence why I asked why is Velocity even important if Stuff is the heavy K influence?

Wouldn't Stuff be better suited in as someone said...slowing down the decline of a pitcher? In theory a P that has better stuff could hang on a bit longer even when his velocity and control wanes a bit...or maybe it could be influential in FB/GB%?

and how will this effect the import from Lahman?

As for Poise...my stance was that I thought it would be purely for the sake of the CPU AI sticking with that pitcher over a younger pitcher with similiar ratings. Normally with the AI as it is in 5 a blue star 5 P would supplant a 3 gold star Vet...even if their ratings are almost identical...that also leads to being able to take that talented 3 star P out of the CPU teams hands because it "thinks" the 5 Blue Star is more valuable...

re: all those things influencing K's... it's that way cause it makes sense.

The movement/stuff (I still don't see a difference) on a pitch will naturally cause more stikeouts. So will a faster pitch that a batter can't catch up to. And, if one is able to pinpoint where the pitch is going (control) you're more likely to strike out a batter with one where you just "freeze" them so to speak.

[attackemu]

Originally posted by matskralc
Thanks, Markus! Keep up the good work!

I was gonna say that!

[GODHAN]

Originally posted by rcbuss
Echoing my thoughts in the other thread, a BABIP rating (as opposed to "poise") wouldn't be a bad idea, however keeping this rating "hidden" would make sense, given that this is something about a pitcher that is often only known retrospectively. See excerpt from my post in that thread below:

"We know now that many pitchers have the distinct ability to prevent hits on balls in play over the course of a career (see Tom Tippett's excellent article on this at http://www.diamond-mind.com/articles/ipavg2.htm), however this is something that is often known only retrospectively about a pitcher, after he's pitched several seasons. For this reason, keeping this rating "hidden" makes sense."

If it is hidden, how to u set the rating in the roster sets? Also, does it change over time, or are players doomed randomly from the very beginning??

[UD411Flyer]

Markus my friend, you were sick, and thought of all of this....You've got Poise!

[Killebrew]

Good stuff Markus, way to go.

[UD411Flyer]

I also believe that this will influence online trading of pitchers. I would like to see if this could have any "mixing" to older pitchers, and help them last longer than 1 week of talent an rating hits.

[GODHAN]

Originally posted by crackpott
The movement/stuff (I still don't see a difference) on a pitch will naturally cause more stikeouts.

Yeah, I always thought "stuff" just referred to a pitchers velocity and movement. I have to again say that "pitch placement" or "keeping ball down" is a rating to include. Or have it like a regular baseball video game, where you have a bunch of pitches, and u have a %mastery of each pitch. Actually, that isn't a great idea...

How about tendency to pitch inside... either as a "TEAM STRATEGY" rating, or a player's personal rating. Then if a guy pitches in too often, he gets charged at, and injured more often..

[rcbuss]

Originally posted by GODHAN
If it is hidden, how to u set the rating in the roster sets? Also, does it change over time, or are players doomed randomly from the very beginning??

I don't have the answer to the first question. I suppose that there could be an option for setting the initial value for each pitcher in a roster set, but then have the value "hidden" when it came to league play. I'd think that the rating could change over time, but, for the reasons stated below, it would be very hard to determine if the actual skill changed over time (because of the low year-to-year correlation for individual pitchers) without having the pitchers in question face at least a couple thousand batters.

Tippett's article (http://www.diamond-mind.com/articles/ipavg2.htm) shows a strong correlation between career BABIP and career length -- BABIP appears to be a distinct skill for pitchers, just as BB, K and HR rates are.

Unlike other objective measures of pitching ability (BB, K, and HR rates), however, BABIP's relatively weak correlation year-to-year for individual pitchers is low, so it can often vacillate wildly, from season to season, about a pitcher's "true" ability of preventing hits on balls in play (BABIP). Oftentimes, the signal gets lost in the noise. It is for that reason that I would advocate the rating's being hidden. There is simply no way to know what a pitcher's "true skill" is in BABIP until he's faced several hundred to a couple thousand batters. Keep in mind the selection bias, as well. Many pitcher's who demonstrate poor skill in terms of BABIP in their first few hundred BFPs may simply not get another chance to establish themselves. Of course, many of these pitchers are also subpar when it comes to other rate stats, as well.

That being said, there are exceptions. Tippett's article shows that Jamie Moyer appears to have developed his skill for BABIP in mid-career (perhaps analogous to a sudden ratings change at a distinct point in time). Andy Pettitte apparently never developed the skill, instead managing to have a long career on the basis of his other rate stats. Same for Tommy John.

In a sample of 351 pitchers with at least 6000 career balls in play, more than 12% of them posted results that would happen less than 1% of the time by chance. And that understates the case, too, because you get to keep pitching if you're that much better than the league, but you usually don't make it to 6000 balls in play if you're that much worse than the league. If one end of the distribution hadn't been truncated by job losses, approximately 20% of those pitchers would have fallen outside the range that can be explained by chance.

It is primarily for this reason that BABIP should be a distinct rating for pitchers in OOTP, just as BB, K and HR ratings are. Perhaps the proposed name "Poise" gave the wrong connotation, but if it was meant to reflect BABIP (batting average on balls in play) for individual pitchers, then I believe that Markus was on the right track.

[dougaiton]

Very good post, but a few issues...

Remember that Tippett also says that 'their influence over in-play hit rates is weaker than their influence over walk and strikeout rates", and that's understating. There is a very weak correlation - as might be expected, trends might be spotted in anything. It wouldn't surprise me if any variable performed differently to what we might expect in as skewed a distribution as major league pitchers. Howeverr, it is the strength of the correlation that matters - you could show me that certain pitchers always performed differently on a Monday than a Tuesday, but unless you could show me how and why I wouldn't include it in the game.

The other thing with BABIP is, if it is truly a skill, then why is there such a weak correlation? There are equally weak correlations for day/night splits - should these be included in the game? Anything that cannot be predicted nor modelled isn't a skill but a fluctuation. There may be something that causes these fluctuations to be more or less than we might expect, but in most cases it can't be judged what that something is.


I'd also dispute that job losses truncate the study. What they in fact do is remove ptichers from the equation who might well normalise within the range of chance. As they currently stand as outliers who may well post unexpected results, then their inclusion would be more likely to suggest that BABIP matters less, not more.

Hope that made some sense...

[rcbuss]

One could say that the vast majority of major league pitchers show little to no ability to affect the result of balls put into play (using IPAvg vs Lg, vs Tm, as per Tippett's article).

But how do you account for players like Walter Johnson, who, in addition to putting up great BB, K and HR stats, was off the charts as far as BABIP were concerned?

Granted, Johnson (and perhaps Charlie Hough) are perhaps the most extreme of outliers, but an accurate model of a universe of big league pitching should account for the fact that pitchers good enough to have extremely long careers (BFP>10000) yielded a BABIP of 0.003 (vs Tm) less than the average over major league history (and 0.004 vs Lg). This accounts for nearly a quarter (23.5%) of all BFP in the history of major league baseball. From Tippett's article, I'd think that this would be highly significant from a statistical standpoint, and should be accounted for.

In addition, I'd think that the extreme performance of the BFP>10000 set skews the averages for all of the other groups; if you removed that subset, the pitchers with 7000-9999 BFP would then separate themselves from the rest, and so on.

If "BABIP" were to have been a 1-10 rating for BABIP vs team/league (and if it were to be called "Poise", let's say) this is how I would rate these subgroups of pitchers, along with approximate number in each grouping (BF divided by midrange value in each group -- exceptions: I used 200 for the 0-999 group, and 15000 for the 10000+ group):

Career BF BF % BF vsLg vsTm N "Poise"
0001-0999 401,138 3.6% 0.017 0.015 2006 0
1000-1999 931,981 8.4% 0.006 0.004 0621 1
2000-2999 1,105,712 10.0% 0.002 0.001 0442 2-3
3000-3999 1,179,916 10.7% 0.000 0.000 0337 4-6
4000-4999 906,271 8.2% 0.000 0.001 0201 4-6
5000-5999 920,680 8.3% 0.000 0.000 0167 4-6
6000-6999 647,553 5.9% -0.001 -0.001 0100 6
7000-7999 843,937 7.7% -0.002 -0.001 0113 7-8
8000-8999 716,200 6.5% -0.002 -0.002 0084 7-8
9000-9999 788,532 7.1% -0.002 -0.001 0083 7-8
10000+ 2,589,409 23.5% -0.004 -0.003 0173 9-10
Total/Avg 11,031,329 100.0% 0.000 0.000 4327 5

[dougaiton]

Sorry, I've only got a minute before I'm back working, so I need to keep this brief.

I had similar concerns earlier (months ago) concerning the fact that a DIPS system would not affect the stats that were produced but rather just 'how' they were generated. Thus, I wondered whether it was worth risking statistical accuracy for the few extreme examples, in exchange for a more realistic way of generating those stats.

In the end, I decided that Markus would almost definetly program around the outliers, and have some way of helping extreme examples (perhaps a hidden option on one in every 100) compensating (for who random chance doesn't alone). I slept a lot better after that!

Apologies for the brevity. Your post deserves more - and I agree that these issues must be raised!

[Brian the Stunt Bum]

This is kind of a shame in my mind, I was hoping this rating would be a great tool for older guys to still be effective when their abilities started to decline and a way for superskilled young kids to still get shelled now and then.