Quick question about GB% and velocity.

JesseTerry · 08-02-2007, 09:27 PM

Do these effect a pitchers performance? I've had some people say they play no part, and was curious to get an official ruling in the matter.

injury log · 08-03-2007, 04:05 AM

My impression is that velocity has no effect. This impression comes from redsoxford's experiment, posted here (post no. 48):

http://www.ootpdevelopments.com/boar...=148281&page=3

That's a great thread anyway, so you might find it interesting to read some of the background posts. This doesn't mean Velocity is completely arbitrary; it does seem to be correlated with Stuff, so high-Velo pitchers are often very good. It just isn't the Velo that makes them good; they're good because they have good Stuff, and high Velo is just a byproduct of that.

GB% should certainly have some effect in-game- it should determine how many groundballs are hit in play-by-play, for example. The question is whether it has the intended effect. Pitchers with higher GB% are supposed to give up more hits, according to the stats projections in the Editor screen. RonCo has said that in studies, this particular effect doesn't show up in the game, or at least not nearly to the degree predicted by the stats projection. I can't find the link to the thread where I read this, so hopefully he'll correct me if I'm misrepresenting what he wrote. GB% is not cosmetic, in any case, as far as I understand. In particular, because pitcher performance arises from interaction between the pitcher and the defense behind him, a pitcher with high GB% should do much better with a rangey infield, and a pitcher with low GB% should do better with a rangey OF, if you're looking at ERA, or at hits allowed.

RonCo · 08-03-2007, 06:02 AM

GB% does influence outcomes as high GB% pitchers will give up fewer HR and more hits on balls in play. However, GB% is an odd creature in that it's only directly applied to outs, so the numbers are a little odd for me to get my head around.

Velocity may, or may not be directly tied to the outcome engine. Mathematical statistics over a league-wide data sample say that it is a statistically significant contrbutor to outcomes. However, correlation does not require causation, so it could be an indirect influence by driving Stuff or other ratings higher. Only Markus knows for sure. From the user's standpoint, I would not recommend ignoring it.

JesseTerry · 08-03-2007, 09:14 AM

Very, very interesting guys, thanks for the info.

injury log · 08-03-2007, 01:21 PM

RonCo- would it be correct to assume that you are not convinced by the results of redsoxford's experiment with velocity? He normalized a complete league, so every pitcher and batter had the same ratings, with one exception: each team's staff had a different velocity. There were no substantial differences in pitching performances among the teams, over a complete season. I find this to be compelling evidence that velocity alone doesn't affect performance, and I'd be interested to know why these results might not be as conclusive as I find them. Of course, velocity may have other effects, e.g. on pitcher development, which may be very important.

Curtis · 08-03-2007, 05:31 PM

injury log, my understanding is that velocity helps to determine stuff. If redsoxford gave every pitcher the same stuff, and just varied velocity, then he shouldn't see any impact on game statistics (because the impact would've been negated by his editing the stuff ratings). Did I misunderstand the procedure redsoxford followed? (It certainly wouldn't be the first time.…)

injury log · 08-03-2007, 06:59 PM

Yes, I think you've described redsoxford's results perfectly (so I probably didn't explain it very well!). He created an entire league where every player had average ratings in all categories, then only changed velocity. One team had 95-97 mph velocity pitchers, one had 80-82 mph, etc. There wasn't any significant difference in performance, and the group with 85-87 mph actually had the most strikeouts (by a small margin). The thread I linked above has the full details, along with several other interesting ratings tests redsoxford ran.

Now what does seem to be true is that when OOTP creates a fictional player, Velocity is at least loosely correlated with some ratings (Stuff, certainly, and perhaps others). So high-Velo pitchers do tend to be better if you let OOTP create your players, but if you create your own (e.g. by making your own roster set), how you set velocity won't make any difference to performance. Or at least that's the conclusion I've drawn from redsoxford's work. I'm also interested to know whether pitchers with different velocities follow different development curves- not an issue I've seen anyone comment on yet, and a difficult one to test quickly. Someday, perhaps I'll try...

Curtis · 08-03-2007, 07:48 PM

Quote:

Originally Posted by injury log

I'm also interested to know whether pitchers with different velocities follow different development curves- not an issue I've seen anyone comment on yet, and a difficult one to test quickly. Someday, perhaps I'll try...

This was stated to be true for 2006 by Markus, but I don't know that it remains true in 2007. I SEEM to recall that when they told us that velocity and stuff are now linked (which they weren't in 2006) they also said something about velocity and development either being delinked or having their linkage loosened. Have you tried asking RonCo? He seems to be the stat man for the beta team and would likely have as good a grasp of it as anyone.

RonCo · 08-04-2007, 12:25 PM

Quote:

Originally Posted by injury log

RonCo- would it be correct to assume that you are not convinced by the results of redsoxford's experiment with velocity? He normalized a complete league, so every pitcher and batter had the same ratings, with one exception: each team's staff had a different velocity. There were no substantial differences in pitching performances among the teams, over a complete season. I find this to be compelling evidence that velocity alone doesn't affect performance, and I'd be interested to know why these results might not be as conclusive as I find them. Of course, velocity may have other effects, e.g. on pitcher development, which may be very important.

If I were forced to bet my house, I would say that velocity probably is an influence on (a) stuff rating and (b) stuff development, which would mean it is an indirect influence on K/BF rate. That's a good interpretation of the intersection of redsoxford's experiment and my statistical correlation study. However, I admit I'm not 100% bought into that at this point. I consider it a theory good enough to make decisions by.

At the root of my indecision is that pitching ratings influence on the three pitching skills (K/PA, BB/PA, HR/PA) are apparently intertwined enough to make a solid call difficult. In addition, I know Markus was at one point specifically working to tie velocity to k-rate, but I'm not certain he ever got comfortable with the results.

Either way, the end result is that the v2007 engine values velocity higher than the v2006 or earlier versions, so it's a step in the right direction.

RonCo · 08-04-2007, 12:43 PM

BTW, another piece of data for this interpretation: In several test leagues I've looked at, the correlation between a pitcher's velocity and his overall stuff is in the range of .8, which is very strong. In my latest test league, a regression of velocity to stuff also shows strong statistical significance with the linear equation of stuff = (velocity * 6.87) - 2.89

So I'm growing more comfortable with the idea that velocity is a very strong influence on stuff.

Charlie Hough · 08-05-2007, 12:54 AM

The results of experiments don't matter to me at all. I have the results of gameplay to back up my view. Velocity and GB% are vital in OOTP. I have always assembled teams based on getting as many pitchers as possible with higher velocity and high GB%. Obviously you still want pitchers with good stuff, movement and control. But you would be a fool not to take a higher velocity pitcher with high GB% when all else is equal.

In fact, I tend to defer significantly to GB% when deciding on pitchers. If they have average to very good stuff, movement and control, they will be tough to beat with high GB%. When you assemble a pitching staff with these guidelines in mind, your team will be very good every time. I have never had a pitching staff fail me when drafting, trading, or signing players based on this view.

injury log · 08-05-2007, 06:46 AM

You can't extrapolate universal conclusions from your own game experience unless you've played many seasons of the game with a wide variety of settings. What works well in a 1940s style league might not in a modern-day league, or in a league based in Vanuatu with low PCMs. I find a conclusion like this a bit simplistic:

Quote:

Originally Posted by Charlie Hough

But you would be a fool not to take a higher velocity pitcher with high GB% when all else is equal.

If you had bad infield defense and phenomenal outfied defense, I don't think a high GB% would be desirable. Lower GB% pitchers also give up fewer hits, the tradeoff being fewer DPs; in some circumstances that tradeoff will be beneficial, in others not. The value of experiments, properly conducted, is that they can tell us what ought to be true in all game situations.

injury log · 08-05-2007, 12:26 PM

Quote:

Originally Posted by RonCo

If I were forced to bet my house, I would say that velocity probably is an influence on (a) stuff rating and (b) stuff development, which would mean it is an indirect influence on K/BF rate. That's a good interpretation of the intersection of redsoxford's experiment and my statistical correlation study. However, I admit I'm not 100% bought into that at this point.

...

Thanks RonCo- always interesting to read your thoughts on these issues. I wonder if the link between Velocity and development could be tested as follows:

-create a league where all players have overall ratings of 100 in all categories;
-give all pitchers 200 potentials in all categories;
-give each team a different velocity;
-track development over a few years, then run test a few more times to get a decent sample size.

Perhaps I'll give that a shot if you think the results would be conclusive.

JesseTerry · 08-05-2007, 09:22 PM

Great stuff guys, a ton more than I expected.

Charlie Hough · 08-06-2007, 01:46 AM

Quote:

Originally Posted by injury log

You can't extrapolate universal conclusions from your own game experience unless you've played many seasons of the game with a wide variety of settings. What works well in a 1940s style league might not in a modern-day league, or in a league based in Vanuatu with low PCMs. I find a conclusion like this a bit simplistic:

I don't need to tinker with countless settings to see that pitchers with high velocity consistently strike out more batters than those who don't. I'm virtually certain that I have seen a posting from Markus and/or something in the documentation that confirms this.

This is the case in real life, and it's the same in OOTP. If settings could somehow influence this fundamental fact of the game, then OOTP would be a poor excuse for a baseball simulation.

Coming from experience with a lot of historical leagues throughout different eras, and even some fictional leagues with unusual settings, the modifiers and league totals for different eras are not going to change this fact. They also won't change the fact that pitchers with higher GB% keep batted balls on the ground more often, which means more double plays, more force outs, and fewer chances for doubles, triples, and home runs.

I don't care what kind of infield defense you have. It's plain common sense that you will be better off with higher percentages of ground balls. A baseball team would never, under any circumstances, be better off with more balls being hit in the air. A groundball is never going to be a home run, a shot into the gap, or just about anything else that typically produces the quickest path to big scores and a high powered offense.

Even the worst infield could never commit enough errors, give up enough grounders down the lines, or otherwise allow enough hits to make it more worthwhile to have a pitching staff with a strong tendency to give up batted balls into the air. If you have any remaining doubt, just imagine that you could have a pitching staff filled with guys that have a GB% of 100. Now, how often do you think your opponents will be able to hit a three-run home run or a two-run double off the wall? Granted, a 100% GB tendency for a pitcher may not completely nullify a batter's power and gap ratings, but it's going to have a huge impact on the results.

I cannot fathom how anyone could even question this or feel a need to run extensive experiments. If Markus had not programmed these two facts of reality into the game, then OOTP would not be a baseball simulation. Fortunately, he HAS programmed them into the game, and OOTP behaves much like real life in these areas. I think that sometimes people get too caught up in statistically analyzing the game engine output, when the best way to get a feel for how it works is to spend a lot of time as a GM and field manager.

Instead of worrying about whether velocity has any relation to the stuff rating, try managing a few hundred games and paying attention to the impact of a 95 mph pitcher on strikeouts. You'll eventually learn that, if you really need a strikeout or a double play to get out of a jam, you're better off with a pitcher with 3 stuff, 70% GB and a 97 mph fastball than you would be with a pitcher with 7 stuff, 45% GB, and an 89-90 mph heater.

mrbill · 08-06-2007, 03:36 AM

That's quite a rant with many undefended (and undefendable?) statements. If I play in Petco park, I take a fly ball pitcher any day. Fast outfield, thick air, and Chris Young leads in ERA.

http://www.hardballtimes.com/main/ar...-the-grounder/

AFAIK, GB% has very little influence on total run expectation IRL. You basically don't pick a pitcher based on his GB% unless you have specific circumstances, like the park you play in is not average. Yes, if you need a double play, bring out a ground ball pitcher. But he's more likely to give up a hit in general, and less likely to give up a homerun. And, most pitchers all range in the 18-20% for line drives (which is removed from the OOTP model), so the chances of a double are about the same across the board, maybe slightly higher for a ground ball pitcher.

Also, velocity is correlated with strikeouts in OOTP, but the compounded effect doesn't necessarily mean I want 98 MPH + 3 Stuff over 90 MPH + 7 Stuff, not without running the numbers to see what actually comes out of it.

In fact, knowing anything means running the numbers, not just managing a few hundred games. Remember, the difference between a .270 hitter and a .320 hitter is one hit every 5 or 6 games, so you're basically not going to notice the difference unless you 'run the numbers' and look at a much longer timeframe.

RonCo · 08-06-2007, 10:23 AM

Once again we find that what we "know" we really don't know. These are the things that make baseball the beautiful game that it is.

RonCo · 08-06-2007, 10:28 AM

Also, I should note that OOTPv2007 is the first version to have ever had any correlation between velocity and strikeout rate. There also remain a large number of people--beta testers among them--who vehemently argue _against_ tying velocity to a pitcher's k-rate, despite the logs and logs of data that show this is one of the stronger correlations in all of real baseball.

So, studying the game is always advisable.

Charlie Hough · 08-06-2007, 06:24 PM

I give you the submarine pitcher as the simple proof that ground ball percentage is critical and has played a huge role in baseball history. Submarine pitchers have traditionally specialized in throwing sinker balls and in keeping the ball down low. Until recent years (when there have been several sub-par submarine pitchers that are deviating wildly from the proven submarine motions of the past), these pitchers always had higher ground ball totals and were often employed as closers because they kept the ball from getting in the air.

Take a look at the career stats of Ted Abernathy, Dan Quisenberry, Kent Tekulve, Terry Leach, Elden Auker, Mike Henneman, and virtually all of the submarine pitchers that appeared in MLB from around the 1920s to the 1980s. You'll see the bottom line impact of ground ball percentages.

Auker gave up 1 HR for every 15.2 IP in his career, and he was a starter. He is the worst of the lot, yet rarely gave up home runs. Now let's look at the remaining pitchers. Tekulve gave up 1 HR for every 22.8 IP. He also had a career ERA of 2.85. Abernathy gave up 1 HR for every 16.4 IP . Quisenberry gave up 1 HR for every 17.7 IP with a 2.76 career ERA. Terry Leach gave up just 1 HR per 18.42 IP with a career ERA of 3.15. Henneman gave up 1 HR per 15.6 IP with a 3.21 career ERA. Every one of these pitchers except for Auker had outstanding career ERAs.

So what is the common denominator, and what made them highly effective pitchers that were often used as closers? They kept the ball low and had some of the highest groundball percentages in the majors. Their tendency to get ground ball outs and double plays made them ideal choices to come in with the game on the line and runners on base. Managers could trust that they would not surrendor a game-winning home run or a huge shot into the gap that could get all the runners home.

OOTP reflects this reality in that these pitchers will be created with higher ground ball percentages in historical play. And, when they perform in the game, they tend to have low ERAs, low HRs allowed, and are generally among the best relievers. What goes for reality in this case goes for OOTP as well.

By the way, the same real life statistical tendencies hold for pitchers that have relied heavily on the sinker ball during their careers. For further data and charts, see this article on the sinker ball. Notice the graph depicting the batted ball results. Even LINE DRIVES are relatively infrequent when pitchers throw a sinker.

injury log · 08-07-2007, 10:10 AM

No one will reasonably dispute that there are good groundball pitchers, just as no one can reasonably dispute there are good flyball pitchers. Johan Santana, Jon Papelbon, BJ Ryan and Joel Zumaya all had GB/FB ratios significantly below league average last year, just to take some of the most dominant of last season's performances, and these are hardly isolated cases- look at the last three years of Pedro Martinez's career, or, as Mr Bill points out, this year's ERA leader Chris Young who has an absurdly high rate of flyballs. I doubt it's coincidence that he's playing in PetCo park with Mike Cameron in centerfield. Their ERAs are all the more impressive when you account for the fact that groundball pitchers give up more unearned runs than flyball pitchers- groundballs lead to more errors (see Derek Lowe's unearned run total from 2004-2005 to see an extreme example). In any case, this isn't germane to the original question. All things equal (defense, park factors, etc), groundball pitchers are generally better than flyball pitchers- I haven't said otherwise- but all things aren't always equal in baseball, which is one of the many reasons the game is interesting.

As for Velocity in OOTP, one doesn't need to run extensive tests to judge what effect it has on strikeouts. If anyone is interested in testing this, find two pitchers in your league with similar Stuff and very different Velocity, and compare K rate. Find two pitchers with similar Velocity and very different Stuff, and again compare K rate. If you have a significant sample size, this will demonstrate which factor has a greater influence on strikeouts. In my leagues, I don't find any significant correspondence between Velocity and K rate, Stuff being equal.

It is of crucial importance to those designing rosters from scratch to know which OOTP ratings have an important influence on performance, and which do not. It's also important to analyse these questions in order to improve the pitching model in future versions of the game. While I agree with others that Velocity should influence K rate, I have yet to see any evidence that it does in OOTP, independent of other factors. I've certainly seen evidence, in my leagues and from others' studies, that Velocity and Stuff are correlated: when OOTP creates (or imports) a pitcher with high Stuff (or K-rate), the pitcher will generally also have high Velocity. So high-Velocity pitchers in OOTP tend to strike more people out, not because they have high Velocity, but because they have high Stuff. This correlation is strong enough that it may appear in-game that Velocity helps to determine K-rate- it certainly seemed that way to me for a long time- but I'm sure I don't need to point out that correlation and causation aren't the same thing.

08-02-2007, 09:27 PM	#1
JesseTerry Minors (Double A) Join Date: Jan 2007 Posts: 174	Quick question about GB% and velocity. Do these effect a pitchers performance? I've had some people say they play no part, and was curious to get an official ruling in the matter.

08-03-2007, 04:05 AM	#2
injury log Hall Of Famer Join Date: Apr 2007 Location: Toronto Posts: 9,162	My impression is that velocity has no effect. This impression comes from redsoxford's experiment, posted here (post no. 48): http://www.ootpdevelopments.com/boar...=148281&page=3 That's a great thread anyway, so you might find it interesting to read some of the background posts. This doesn't mean Velocity is completely arbitrary; it does seem to be correlated with Stuff, so high-Velo pitchers are often very good. It just isn't the Velo that makes them good; they're good because they have good Stuff, and high Velo is just a byproduct of that. GB% should certainly have some effect in-game- it should determine how many groundballs are hit in play-by-play, for example. The question is whether it has the intended effect. Pitchers with higher GB% are supposed to give up more hits, according to the stats projections in the Editor screen. RonCo has said that in studies, this particular effect doesn't show up in the game, or at least not nearly to the degree predicted by the stats projection. I can't find the link to the thread where I read this, so hopefully he'll correct me if I'm misrepresenting what he wrote. GB% is not cosmetic, in any case, as far as I understand. In particular, because pitcher performance arises from interaction between the pitcher and the defense behind him, a pitcher with high GB% should do much better with a rangey infield, and a pitcher with low GB% should do better with a rangey OF, if you're looking at ERA, or at hits allowed. Last edited by injury log; 08-03-2007 at 04:08 AM.

08-06-2007, 03:36 AM	#16
mrbill All Star Reserve Join Date: Jul 2003 Location: LA, CA Posts: 982	That's quite a rant with many undefended (and undefendable?) statements. If I play in Petco park, I take a fly ball pitcher any day. Fast outfield, thick air, and Chris Young leads in ERA. http://www.hardballtimes.com/main/ar...-the-grounder/ AFAIK, GB% has very little influence on total run expectation IRL. You basically don't pick a pitcher based on his GB% unless you have specific circumstances, like the park you play in is not average. Yes, if you need a double play, bring out a ground ball pitcher. But he's more likely to give up a hit in general, and less likely to give up a homerun. And, most pitchers all range in the 18-20% for line drives (which is removed from the OOTP model), so the chances of a double are about the same across the board, maybe slightly higher for a ground ball pitcher. Also, velocity is correlated with strikeouts in OOTP, but the compounded effect doesn't necessarily mean I want 98 MPH + 3 Stuff over 90 MPH + 7 Stuff, not without running the numbers to see what actually comes out of it. In fact, knowing anything means running the numbers, not just managing a few hundred games. Remember, the difference between a .270 hitter and a .320 hitter is one hit every 5 or 6 games, so you're basically not going to notice the difference unless you 'run the numbers' and look at a much longer timeframe. __________________ UBL - Best Online League Evar! - Los Angeles Dodgers: 25 seasons, 13 NL West titles, 4 WC, 8 NL Titles, 5-time Champs LBB v5 league (retired) - Detroit Tigers/Commish: 19 seasons, 18 straight AL Central titles, 2006, 2008, 2014, 2015 Champs! NGBL v6 league (dead) - Texas Rangers: 10 seasons, 4 AL South titles, 2 Wild Cards, one WS app Last edited by mrbill; 08-06-2007 at 03:42 AM.

08-06-2007, 06:24 PM	#19
Charlie Hough Hall Of Famer Join Date: May 2006 Posts: 3,625	I give you the submarine pitcher as the simple proof that ground ball percentage is critical and has played a huge role in baseball history. Submarine pitchers have traditionally specialized in throwing sinker balls and in keeping the ball down low. Until recent years (when there have been several sub-par submarine pitchers that are deviating wildly from the proven submarine motions of the past), these pitchers always had higher ground ball totals and were often employed as closers because they kept the ball from getting in the air. Take a look at the career stats of Ted Abernathy, Dan Quisenberry, Kent Tekulve, Terry Leach, Elden Auker, Mike Henneman, and virtually all of the submarine pitchers that appeared in MLB from around the 1920s to the 1980s. You'll see the bottom line impact of ground ball percentages. Auker gave up 1 HR for every 15.2 IP in his career, and he was a starter. He is the worst of the lot, yet rarely gave up home runs. Now let's look at the remaining pitchers. Tekulve gave up 1 HR for every 22.8 IP. He also had a career ERA of 2.85. Abernathy gave up 1 HR for every 16.4 IP . Quisenberry gave up 1 HR for every 17.7 IP with a 2.76 career ERA. Terry Leach gave up just 1 HR per 18.42 IP with a career ERA of 3.15. Henneman gave up 1 HR per 15.6 IP with a 3.21 career ERA. Every one of these pitchers except for Auker had outstanding career ERAs. So what is the common denominator, and what made them highly effective pitchers that were often used as closers? They kept the ball low and had some of the highest groundball percentages in the majors. Their tendency to get ground ball outs and double plays made them ideal choices to come in with the game on the line and runners on base. Managers could trust that they would not surrendor a game-winning home run or a huge shot into the gap that could get all the runners home. OOTP reflects this reality in that these pitchers will be created with higher ground ball percentages in historical play. And, when they perform in the game, they tend to have low ERAs, low HRs allowed, and are generally among the best relievers. What goes for reality in this case goes for OOTP as well. By the way, the same real life statistical tendencies hold for pitchers that have relied heavily on the sinker ball during their careers. For further data and charts, see this article on the sinker ball. Notice the graph depicting the batted ball results. Even LINE DRIVES are relatively infrequent when pitchers throw a sinker. Last edited by Charlie Hough; 08-06-2007 at 06:26 PM.

08-03-2007, 06:02 AM	#3
RonCo Hall Of Famer Join Date: Aug 2003 Posts: 9,499	GB% does influence outcomes as high GB% pitchers will give up fewer HR and more hits on balls in play. However, GB% is an odd creature in that it's only directly applied to outs, so the numbers are a little odd for me to get my head around. Velocity may, or may not be directly tied to the outcome engine. Mathematical statistics over a league-wide data sample say that it is a statistically significant contrbutor to outcomes. However, correlation does not require causation, so it could be an indirect influence by driving Stuff or other ratings higher. Only Markus knows for sure. From the user's standpoint, I would not recommend ignoring it.

08-03-2007, 09:14 AM	#4
JesseTerry Minors (Double A) Join Date: Jan 2007 Posts: 174	Very, very interesting guys, thanks for the info.

08-03-2007, 01:21 PM	#5
injury log Hall Of Famer Join Date: Apr 2007 Location: Toronto Posts: 9,162	RonCo- would it be correct to assume that you are not convinced by the results of redsoxford's experiment with velocity? He normalized a complete league, so every pitcher and batter had the same ratings, with one exception: each team's staff had a different velocity. There were no substantial differences in pitching performances among the teams, over a complete season. I find this to be compelling evidence that velocity alone doesn't affect performance, and I'd be interested to know why these results might not be as conclusive as I find them. Of course, velocity may have other effects, e.g. on pitcher development, which may be very important.

08-03-2007, 05:31 PM	#6
Curtis Hall Of Famer Join Date: Jul 2006 Location: Watertown, New York Posts: 4,567	injury log, my understanding is that velocity helps to determine stuff. If redsoxford gave every pitcher the same stuff, and just varied velocity, then he shouldn't see any impact on game statistics (because the impact would've been negated by his editing the stuff ratings). Did I misunderstand the procedure redsoxford followed? (It certainly wouldn't be the first time.…)

08-03-2007, 06:59 PM	#7
injury log Hall Of Famer Join Date: Apr 2007 Location: Toronto Posts: 9,162	Yes, I think you've described redsoxford's results perfectly (so I probably didn't explain it very well!). He created an entire league where every player had average ratings in all categories, then only changed velocity. One team had 95-97 mph velocity pitchers, one had 80-82 mph, etc. There wasn't any significant difference in performance, and the group with 85-87 mph actually had the most strikeouts (by a small margin). The thread I linked above has the full details, along with several other interesting ratings tests redsoxford ran. Now what does seem to be true is that when OOTP creates a fictional player, Velocity is at least loosely correlated with some ratings (Stuff, certainly, and perhaps others). So high-Velo pitchers do tend to be better if you let OOTP create your players, but if you create your own (e.g. by making your own roster set), how you set velocity won't make any difference to performance. Or at least that's the conclusion I've drawn from redsoxford's work. I'm also interested to know whether pitchers with different velocities follow different development curves- not an issue I've seen anyone comment on yet, and a difficult one to test quickly. Someday, perhaps I'll try...

08-04-2007, 12:43 PM	#10
RonCo Hall Of Famer Join Date: Aug 2003 Posts: 9,499	BTW, another piece of data for this interpretation: In several test leagues I've looked at, the correlation between a pitcher's velocity and his overall stuff is in the range of .8, which is very strong. In my latest test league, a regression of velocity to stuff also shows strong statistical significance with the linear equation of stuff = (velocity * 6.87) - 2.89 So I'm growing more comfortable with the idea that velocity is a very strong influence on stuff.

08-05-2007, 12:54 AM	#11
Charlie Hough Hall Of Famer Join Date: May 2006 Posts: 3,625	The results of experiments don't matter to me at all. I have the results of gameplay to back up my view. Velocity and GB% are vital in OOTP. I have always assembled teams based on getting as many pitchers as possible with higher velocity and high GB%. Obviously you still want pitchers with good stuff, movement and control. But you would be a fool not to take a higher velocity pitcher with high GB% when all else is equal. In fact, I tend to defer significantly to GB% when deciding on pitchers. If they have average to very good stuff, movement and control, they will be tough to beat with high GB%. When you assemble a pitching staff with these guidelines in mind, your team will be very good every time. I have never had a pitching staff fail me when drafting, trading, or signing players based on this view.

08-05-2007, 09:22 PM	#14
JesseTerry Minors (Double A) Join Date: Jan 2007 Posts: 174	Great stuff guys, a ton more than I expected.

08-06-2007, 10:23 AM	#17
RonCo Hall Of Famer Join Date: Aug 2003 Posts: 9,499	Once again we find that what we "know" we really don't know. These are the things that make baseball the beautiful game that it is.

08-06-2007, 10:28 AM	#18
RonCo Hall Of Famer Join Date: Aug 2003 Posts: 9,499	Also, I should note that OOTPv2007 is the first version to have ever had any correlation between velocity and strikeout rate. There also remain a large number of people--beta testers among them--who vehemently argue _against_ tying velocity to a pitcher's k-rate, despite the logs and logs of data that show this is one of the stronger correlations in all of real baseball. So, studying the game is always advisable.

08-07-2007, 10:10 AM	#20
injury log Hall Of Famer Join Date: Apr 2007 Location: Toronto Posts: 9,162	No one will reasonably dispute that there are good groundball pitchers, just as no one can reasonably dispute there are good flyball pitchers. Johan Santana, Jon Papelbon, BJ Ryan and Joel Zumaya all had GB/FB ratios significantly below league average last year, just to take some of the most dominant of last season's performances, and these are hardly isolated cases- look at the last three years of Pedro Martinez's career, or, as Mr Bill points out, this year's ERA leader Chris Young who has an absurdly high rate of flyballs. I doubt it's coincidence that he's playing in PetCo park with Mike Cameron in centerfield. Their ERAs are all the more impressive when you account for the fact that groundball pitchers give up more unearned runs than flyball pitchers- groundballs lead to more errors (see Derek Lowe's unearned run total from 2004-2005 to see an extreme example). In any case, this isn't germane to the original question. All things equal (defense, park factors, etc), groundball pitchers are generally better than flyball pitchers- I haven't said otherwise- but all things aren't always equal in baseball, which is one of the many reasons the game is interesting. As for Velocity in OOTP, one doesn't need to run extensive tests to judge what effect it has on strikeouts. If anyone is interested in testing this, find two pitchers in your league with similar Stuff and very different Velocity, and compare K rate. Find two pitchers with similar Velocity and very different Stuff, and again compare K rate. If you have a significant sample size, this will demonstrate which factor has a greater influence on strikeouts. In my leagues, I don't find any significant correspondence between Velocity and K rate, Stuff being equal. It is of crucial importance to those designing rosters from scratch to know which OOTP ratings have an important influence on performance, and which do not. It's also important to analyse these questions in order to improve the pitching model in future versions of the game. While I agree with others that Velocity should influence K rate, I have yet to see any evidence that it does in OOTP, independent of other factors. I've certainly seen evidence, in my leagues and from others' studies, that Velocity and Stuff are correlated: when OOTP creates (or imports) a pitcher with high Stuff (or K-rate), the pitcher will generally also have high Velocity. So high-Velocity pitchers in OOTP tend to strike more people out, not because they have high Velocity, but because they have high Stuff. This correlation is strong enough that it may appear in-game that Velocity helps to determine K-rate- it certainly seemed that way to me for a long time- but I'm sure I don't need to point out that correlation and causation aren't the same thing.