FAQ and Forum on Advanced Stats

[ptatc]

To try and make this as short as possible, bWAR tells more about past performance, but fWAR tells more about future performances. It's more of a true talent indicator. Like for pitching, fWAR uses Fielding Independent Pitching numbers to figure in the calculation, whereas bWAR uses runs, and team wins. You get more outliers it seems with bWAR than fWAR. I think the large majority of the Sabr community would say that fWAR is the better number to use overall.

After doing research on many of the stats this is one that bothers me the most, I think.

 

If I understand it correctly (which is highly doubtful) FIP is predicated on the fact that pitchers can only control 3 factors: HR, BB and K. I think there is valid reasons to say that some pitchers can control the number of groundballs and thus control the game a little more. I realize it's not totally in the pitchers control because the fielders need to make plays. However, if the number of groundballs is increased the number of runs scored would decrease.

 

There needs to be more stats focusing on runs scored versus runs allowed. I know there are some but there should be more focus on runs due to the fact that nothing else really matters in winning a game.

[witesoxfan]

After doing research on many of the stats this is one that bothers me the most, I think.

 

If I understand it correctly (which is highly doubtful) FIP is predicated on the fact that pitchers can only control 3 factors: HR, BB and K. I think there is valid reasons to say that some pitchers can control the number of groundballs and thus control the game a little more. I realize it's not totally in the pitchers control because the fielders need to make plays. However, if the number of groundballs is increased the number of runs scored would decrease.

 

There needs to be more stats focusing on runs scored versus runs allowed. I know there are some but there should be more focus on runs due to the fact that nothing else really matters in winning a game.

 

That's pretty close. I think they basically use those 3 because most of the time, there is absolutely no way a fielder can catch that ball. What FIP suggests is that a pitcher can force ground balls exactly to where he wants the hitter to hit it (or infield flies or regular fly balls), but if he has statues that can only field what hits them and then only throw playing behind him, he can still give up an absolute ton of hits. FIP just tries to isolate the difference between a pitcher having a good fielding infield and a poor fielding infield while also taking lucky seeing eye singles out of the equation. Sabermaticians do realize that putting the ball in play, but weakly, is a very good thing.

 

Regarding runs scored, I'm curious to know what you are looking for specifically? There was actually breakthrough research done regarding run expectancy quite a few years ago that suggested that the sacrifice bunt is a poor decision in almost all situations except in a situation where you are playing for 1 run (either to tie or win), there's a runner on 2nd, and 0 outs. It can be 1st and 2nd too with no outs, but in all other situations, bunting simply works against the likelihood that you'll score a run. Was there anything else you had in mind on that?

 

 

[ptatc]

That's pretty close. I think they basically use those 3 because most of the time, there is absolutely no way a fielder can catch that ball. What FIP suggests is that a pitcher can force ground balls exactly to where he wants the hitter to hit it (or infield flies or regular fly balls), but if he has statues that can only field what hits them and then only throw playing behind him, he can still give up an absolute ton of hits. FIP just tries to isolate the difference between a pitcher having a good fielding infield and a poor fielding infield while also taking lucky seeing eye singles out of the equation. Sabermaticians do realize that putting the ball in play, but weakly, is a very good thing.

 

Regarding runs scored, I'm curious to know what you are looking for specifically? There was actually breakthrough research done regarding run expectancy quite a few years ago that suggested that the sacrifice bunt is a poor decision in almost all situations except in a situation where you are playing for 1 run (either to tie or win), there's a runner on 2nd, and 0 outs. It can be 1st and 2nd too with no outs, but in all other situations, bunting simply works against the likelihood that you'll score a run. Was there anything else you had in mind on that?

I know that the pitcher can't control the fielders. However, FIP is being used frequently for pitchers performance. I think that a combination using a GB and line drive percentage maybe along with the others may be able to show more of the pitchers influence on the game rather than only BB, K, and HR which is the only pure control variables.

 

For the runs scored, I know I'm in the minority but I'm less concerned about an individuals performance than how the individual impacts a win or loss. The performance aspect is for arbitration and salaries. A players impact on winning a given game is more what I would be interested in. The only way I can see to start breaking it down is how he impacts scoring runs at bat and prevents runs on defense. WAR doesn't really do it because that really looks at how his performance is better than another player of lesser performance. The runs prevented stat seems to make sense on defense and OPS and OPS+ kind of start it for offense but there are still too many variables not taken into account for it to do too much.

[Eminor3rd]

Stats like WAR are a guidepost, I get that, but should not be held as an end all be all of player production. If I am evaluating a player to come to my team, WAR is the last thing I would use as a GM. WAR is more like the preseason top 25 in college football, it is a measuring stick that calculates success in a vacuum but as an evaluation/scouting tool is not very useful.

 

If you make this claim, you are then also claiming the same of every statistic from a player's past. So if you really mean to say, "If I am evaluating a player to come to me team, numbers are the last thing I would use as a GM," then I can see your argument. Otherwise, I don't think it makes sense. All the "noise" you mentioned is precisely what makes context-dependent statistics LESS dependable for evaluating future performance.

[Eminor3rd]

After doing research on many of the stats this is one that bothers me the most, I think.

 

If I understand it correctly (which is highly doubtful) FIP is predicated on the fact that pitchers can only control 3 factors: HR, BB and K. I think there is valid reasons to say that some pitchers can control the number of groundballs and thus control the game a little more. I realize it's not totally in the pitchers control because the fielders need to make plays. However, if the number of groundballs is increased the number of runs scored would decrease.

 

You're mostly right, but there's an important detail missing about the efficacy of the assumption. The reason that FIP is based on those three things is because those are what Voros McCracken found to be statistically reliable predictors of future performance when controlling for other variables, thus implying that they are a product of pitcher "true talent." Isn't because someone just decided that those metrics made sense intuitively, like some people think.

[Eminor3rd]

I know that the pitcher can't control the fielders. However, FIP is being used frequently for pitchers performance. I think that a combination using a GB and line drive percentage maybe along with the others may be able to show more of the pitchers influence on the game rather than only BB, K, and HR which is the only pure control variables.

 

For the runs scored, I know I'm in the minority but I'm less concerned about an individuals performance than how the individual impacts a win or loss. The performance aspect is for arbitration and salaries. A players impact on winning a given game is more what I would be interested in. The only way I can see to start breaking it down is how he impacts scoring runs at bat and prevents runs on defense. WAR doesn't really do it because that really looks at how his performance is better than another player of lesser performance. The runs prevented stat seems to make sense on defense and OPS and OPS+ kind of start it for offense but there are still too many variables not taken into account for it to do too much.

 

That makes intuitive sense, but research has shown no significant year-to-year correlation between LD rates at all. GB rate is less reliable than you'd think, but is more consistent than LD -- the problem is that it cannot be considered an inherently good or bad thing because it doesn't fit into the equation of linear weights. In other words, it is a class of event, but it isn't a final outcome. Therefore, it cannot hold a run value. A GB can become one of several types of hits or outs, and the odds of each event occurring are entirely dependent on defense, chance, and the ability of the hitter -- all context-dependent and infinitely variable.

Edited February 21, 2014 by Eminor3rd

[ScottyDo]

That makes intuitive sense, but research has shown no significant year-to-year correlation between LD rates at all. GB rate is less reliable than you'd think, but is more consistent -- the problem is that it cannot be considered an inherently good or bad thing because it doesn't fit into the equation of linear weights. In other words, it is a class of event, but it isn't a final outcome. Therefore, it cannot hold a run value. A GB can become one of several types of hits or outs, and the odds of each event occurring are entirely dependent on defense, chance, and the ability of the hitter -- all context-dependent and infinitely variable.

Are there any attempts to use velocity of the ball off the bat, or vector somehow? Ball-off-bat speeds are available data, right? I have to imagine they would correlate negatively with offensive production somehow.

[Eminor3rd]

Are there any attempts to use velocity of the ball off the bat, or vector somehow? Ball-off-bat speeds are available data, right? I have to imagine they would correlate negatively with offensive production somehow.

 

That's a good question. I've heard people refer to off-the-bat speed, so I assume it's being recorded. Whether or not it's freely available and, if it is, whether someone has done research with it I don't know.

[harfman77]

If you make this claim, you are then also claiming the same of every statistic from a player's past. So if you really mean to say, "If I am evaluating a player to come to me team, numbers are the last thing I would use as a GM," then I can see your argument. Otherwise, I don't think it makes sense. All the "noise" you mentioned is precisely what makes context-dependent statistics LESS dependable for evaluating future performance.

 

Thats crazy, so because WAR is a bad indicator, every stat is bad incator, talk about throwing the baby out with the bath water. There are a ton of numbers to use to indicate and predict performance, WAR is not a strong indicator stat, more of a picture of what happened in a given set of circumstances. The Dodgers should trade Kershaw straight up for Carlos Gomez because they add the same value to the team? Both had an 8.4 WAR, so they should be equally valuable, but that is dumb. If I use WAR as an evaluation tool, I should make that deal, especially given the contracts involved, but common sense says hell no.

 

If you were building the 2007 team and used WAR from 2006 to build the team, they would be expected to win 90 games, except they lost 90 games. WAR gives me no context as to why or where they regressed.

 

If I am evaluating hitters I am looking at P/AB, OBP, BABIP, K/BB as indicators as to how a player has performed and using some common sense to evaluate whether those numbers are sustainable or outliers. Those are factors that a batter can control. Defensive stats are pretty clunky to me, there are defensive stats that say players I can see with my eyes aren't good, are good and vice versa. WAR is OK to talk about when saying Willie Mays is better than Mike Trout as it gives you some semblance of a tool to evaluate players over their career, but as a tool to build a team, it is not useful.

[Eminor3rd]

Thats crazy, so because WAR is a bad indicator, every stat is bad incator, talk about throwing the baby out with the bath water. There are a ton of numbers to use to indicate and predict performance, WAR is not a strong indicator stat, more of a picture of what happened in a given set of circumstances. The Dodgers should trade Kershaw straight up for Carlos Gomez because they add the same value to the team? Both had an 8.4 WAR, so they should be equally valuable, but that is dumb. If I use WAR as an evaluation tool, I should make that deal, especially given the contracts involved, but common sense says hell no.

 

No. Why would you not consider your team's construction when making such a decision? Why would you not consider team control and cost?

 

If you were building the 2007 team and used WAR from 2006 to build the team, they would be expected to win 90 games, except they lost 90 games. WAR gives me no context as to why or where they regressed.

 

Who expects it to? We re talking about evaluating a player to be added to your team.

 

If I am evaluating hitters I am looking at P/AB, OBP, BABIP, K/BB as indicators as to how a player has performed and using some common sense to evaluate whether those numbers are sustainable or outliers. Those are factors that a batter can control. Defensive stats are pretty clunky to me, there are defensive stats that say players I can see with my eyes aren't good, are good and vice versa. WAR is OK to talk about when saying Willie Mays is better than Mike Trout as it gives you some semblance of a tool to evaluate players over their career, but as a tool to build a team, it is not useful.

 

So you think pitches per at bat is more important than WAR or its components? OBP is captured in wOBA, but is then added proportionately to TB statistics to give you a more complete picture, and then league adjusted into wRC+, which is fed by wRAA. Dozens of studies have shown BABIP to be among the least controllable factors by both hitters and pitchers. Not entirely random, but a poor predictor of itself in all but the most extreme cases. K/BB is less important than OBP.

 

Other than pitches per at bat, all the things you mentioned are captured in WAR components. You don't have use WAR itself unless you're comparing positions, but why on Earth would you not use wRC+, for example?

[harfman77]

No. Why would you not consider your team's construction when making such a decision? Why would you not consider team control and cost?

 

So because the Dodgers have too many OF's, that's why its a bad deal? No, its because Gomez is an OK player and Kershaw is on his way to being an all time great. Control and cost kind of cancel each other out in that scenario as Gomez has $8.5M AAV over the next three seasons.

 

Who expects it to? We re talking about evaluating a player to be added to your team.

 

As constructed with the moves KW made, and the players that were added and retained, the players on the 2007 team had a 2006 WAR of 89.4, but won only 72 games.

 

So you think pitches per at bat is more important than WAR or its components? OBP is captured in wOBA, but is then added proportionately to TB statistics to give you a more complete picture, and then league adjusted into wRC+, which is fed by wRAA. Dozens of studies have shown BABIP to be among the least controllable factors by both hitters and pitchers. Not entirely random, but a poor predictor of itself in all but the most extreme cases. K/BB is less important than OBP.

 

Other than pitches per at bat, all the things you mentioned are captured in WAR components. You don't have use WAR itself unless you're comparing positions, but why on Earth would you not use wRC+, for example?

 

The point that BABIP is uncontrollable tells me how reliable the other stats, like OBP are, if it is artificially inflated, than I know it is likely to regress, also if it is abnormally low, I know that a player is likely to improve. Pitches per at bat tells me what kind of eye that my batters have, how well they can work a count, and helps me to get to a bullpen quicker, where I know it is more likely that my team can score more runs. K/BB indicates the quality of AB's when coupled with P/AB. I can tell if a guy is a flailer or has a legitimately good feel for the strike zone. These are also all environmental independent stats.

 

wOBA and wRC+ are good in concept, though I don't understand how they appropriately weight a performance in the Cell against a performance in Petco to have enough faith in the number to be reliable. A HR in the cell is a fly ball at Petco, but the weighted numbers wont tell me that, it will tell me that the player that hit the ball at the Cell is superior to the one that hit the same ball in San Diego. I have not idea how they could account for that. I understand that they use multipliers and park factors and such, but when you are applying park factor to an out, I would assume that it is then displayed as an out everywhere. Sure your HR at Petco could count as 1.4 HR's in Chicago, but it doesn't work the other way.

 

 

 

[Eminor3rd]

The point that BABIP is uncontrollable tells me how reliable the other stats, like OBP are, if it is artificially inflated, than I know it is likely to regress, also if it is abnormally low, I know that a player is likely to improve. Pitches per at bat tells me what kind of eye that my batters have, how well they can work a count, and helps me to get to a bullpen quicker, where I know it is more likely that my team can score more runs. K/BB indicates the quality of AB's when coupled with P/AB. I can tell if a guy is a flailer or has a legitimately good feel for the strike zone. These are also all environmental independent stats.

 

wOBA and wRC+ are good in concept, though I don't understand how they appropriately weight a performance in the Cell against a performance in Petco to have enough faith in the number to be reliable. A HR in the cell is a fly ball at Petco, but the weighted numbers wont tell me that, it will tell me that the player that hit the ball at the Cell is superior to the one that hit the same ball in San Diego. I have not idea how they could account for that. I understand that they use multipliers and park factors and such, but when you are applying park factor to an out, I would assume that it is then displayed as an out everywhere. Sure your HR at Petco could count as 1.4 HR's in Chicago, but it doesn't work the other way.

 

I think I understand what you're saying now. You want precision for the sake of determining "fit" in terms of a desired player profile. That makes sense. You threw me off that trail when you were talking about the interchangeability of Kershaw and Carlos Gomez but I get it now.

 

Yeah, I'd look outside the WAR family of numbers for type as well. I usually like the Pitch F/X stats for that stuff you mentioned, though, O-swing%, Z-swing%, etc. Pitches per at bat is something that can be demanded by coaches or determined by situation -- I think see how they recognize and treat pitch is a better indicator of "true talent."

[Jake]

Even though things like WAR are going to be far less influenced by things like luck than are simpler metrics, they are not projections or predictions.

 

And, of course, even projections are only so helpful: http://www.fangraphs.com/blogs/the-white-s...ng-projections/

 

But also, please just keep in mind that projections are not predictions. They are a snapshot of what we think a team’s median true talent level might be, and it should be understood that there’s a pretty sizable margin for error based on things that projection systems simply can’t forecast, and also the errors that come from having imperfect information or imperfect calculations. The standard deviation in wins from the good forecasting systems are somewhere in the range of eight wins, meaning that a team forecast for 77 wins could reasonably be expected to win anywhere from 69 to 85 games without it saying anything about the model having a breakdown. There are simply too many uncontrollable and unpredictable variables to get so precise with our preseason forecasts, and simply because of how bell curves work, there are always going to be teams that fall at the tail end of the distribution, making the forecasts look wildly wrong in retrospect.

 

Most years, there’s one team that beats its forecast by 15 wins, sometimes even 20. Last year, we had two, with Baltimore and Oakland winning far more than anyone expected. Pretty much any team forecast for north of 70 wins has some chance at making the playoffs if the stars align. A 77 win forecast for the White Sox — from any projection system — shouldn’t be taken as a death knell for their season, especially with what we know about their apparent ability to keep their pitchers healthy. But, it’s still worth knowing the consensus projections for the White Sox this year have them as something like the 10th best team in the American League. That doesn’t mean that they can’t win, but it does mean that it’s less likely that they’re going to make the playoffs than most of their competitors.

[ptatc]

That makes intuitive sense, but research has shown no significant year-to-year correlation between LD rates at all. GB rate is less reliable than you'd think, but is more consistent than LD -- the problem is that it cannot be considered an inherently good or bad thing because it doesn't fit into the equation of linear weights. In other words, it is a class of event, but it isn't a final outcome. Therefore, it cannot hold a run value. A GB can become one of several types of hits or outs, and the odds of each event occurring are entirely dependent on defense, chance, and the ability of the hitter -- all context-dependent and infinitely variable.

To what are they trying to correltate LD Or GB? I'm looking for some type of correlation with winning games. I know they aren't consistent from year to year but niether are wins. Canthey use a pearson or rho correlation to find GB with wins in a given year? I don't have the answers and I cant find much from anyone else either.

[ptatc]

That makes intuitive sense, but research has shown no significant year-to-year correlation between LD rates at all. GB rate is less reliable than you'd think, but is more consistent than LD -- the problem is that it cannot be considered an inherently good or bad thing because it doesn't fit into the equation of linear weights. In other words, it is a class of event, but it isn't a final outcome. Therefore, it cannot hold a run value. A GB can become one of several types of hits or outs, and the odds of each event occurring are entirely dependent on defense, chance, and the ability of the hitter -- all context-dependent and infinitely variable.

I understand the GB is dependent on the defense, however has it shown that the GB rate has an influence on runs scored. There cannot be a direct correlation due to the variability of events however is there a place that lists the GB% and runs scored. This summer this will be a project to research.

[Eminor3rd]

To what are they trying to correltate LD Or GB? I'm looking for some type of correlation with winning games. I know they aren't consistent from year to year but niether are wins. Canthey use a pearson or rho correlation to find GB with wins in a given year? I don't have the answers and I cant find much from anyone else either.

 

Well, LD rate correlates very highly with high BABIP, which correlates highly with successful hitting metrics in general. So you can't link LD rate directly to wins, but you can link it directly to good hitting and then link good hitting directly to wins.

 

The reason it's important to find year-to-year correlations is to help you evaluate if a hitter's success is sustainable. Again, it's trying to boil down metrics to representatives of "true talent," which should be less likely to disappear.

[Eminor3rd]

I understand the GB is dependent on the defense, however has it shown that the GB rate has an influence on runs scored. There cannot be a direct correlation due to the variability of events however is there a place that lists the GB% and runs scored. This summer this will be a project to research.

 

I second that this would be interesting research.

[cabiness42]

Even though things like WAR are going to be far less influenced by things like luck than are simpler metrics, they are not projections or predictions.

 

And, of course, even projections are only so helpful: http://www.fangraphs.com/blogs/the-white-s...ng-projections/

 

From that same article:

 

Chicago has only had 14 pitchers throw at least 100 innings as a starter for them over the last eight years, and of those 14, only three — Orlando Hernandez, Clayton Richard, and Philip Humber — could be described as below average Major League starters during their time in Chicago. That’s remarkable. Even other teams that have focused heavily on pitching during this run have ended up giving long runs to lousy pitchers, simply due to the fact that pitchers break down, and teams either live through terrible performances trying to get them fixed or have terrible replacements come up from the minors.

 

The White Sox simply haven’t had that problem. They might not have had a rotation fronted by Roy Halladay or Cliff Lee, but they also didn’t give 265 innings to Adam Eaton and his 136 ERA-. While most of the analysis about a pitching rotation’s strength focuses on how good the first few starters are, the contributions of the guys at the back end can make a huge difference as well. And, no team has gotten more value from their back-end starters than the White Sox, primarily because they’ve been able to keep them healthy and avoid the roller coaster of minor league fill-ins that most teams inevitably have to endure.

 

Bold sections highlighted, just in case you happen to know anybody that still thinks giving up a draft pick to give Ervin Santana 4/40+ is a good idea.

[ptatc]

Well, LD rate correlates very highly with high BABIP, which correlates highly with successful hitting metrics in general. So you can't link LD rate directly to wins, but you can link it directly to good hitting and then link good hitting directly to wins.

 

The reason it's important to find year-to-year correlations is to help you evaluate if a hitter's success is sustainable. Again, it's trying to boil down metrics to representatives of "true talent," which should be less likely to disappear.

I was looking at it from the pitching aspect of GB% not the hitting. Still seems to be that good pitching beats good hitting especially in the post season. If LD rate is correlated to BABIP, it makes sense that a high GB% or lower LD rate would be correlated with better pitching and poorer hitting. I need to find the numbers though.

[witesoxfan]

I was looking at it from the pitching aspect of GB% not the hitting. Still seems to be that good pitching beats good hitting especially in the post season. If LD rate is correlated to BABIP, it makes sense that a high GB% or lower LD rate would be correlated with better pitching and poorer hitting. I need to find the numbers though.

 

My hypothesis would be that you'd see some slight correlation to GB% and run prevention, but it wouldn't be significant. Your best groundballers (as SP) since the statistic has been kept have been Brandon Webb, Derek Lowe, Jake Westbrook, Chien-Ming Wang, Tim Hudson, Aaron Cook, Zach Day, Fausto Carmona/Roberto Hernandez, Kirk Saarloos, and Justin Masterson. That's a pretty mixed bag. The one thing I notice when looking over those pitchers is that the most successful ones have been those that have been able to accrue some strikeouts.

 

http://www.fangraphs.com/leaders.aspx?pos=...=0&sort=6,d

 

You can mess around with that a little bit. One thing's for certain - you don't make it starting if you are throwing flyballs unless you pitch in a big ballpark. Chris Young and Jered Weaver are really the only starters to find some success.

Edited February 26, 2014 by witesoxfan

[Eminor3rd]

I was looking at it from the pitching aspect of GB% not the hitting. Still seems to be that good pitching beats good hitting especially in the post season. If LD rate is correlated to BABIP, it makes sense that a high GB% or lower LD rate would be correlated with better pitching and poorer hitting. I need to find the numbers though.

 

Keep in mind that high GB% correlates strongly with higher BABIP, so a GB pitcher will be expected to give up more hits but fewer homers. The addition of hits can be mitigated, of course, with a better infield defense. Theoretically, for GB% to be considered a thing to seek, you'd have to know you had a good infield, and, ideally, you'd have to play in a park where homers are a threat to be avoided at all costs.

 

Historically, FB-heavy pitchers tend to be inconsistent because there doesn't appear to be much of any year-to-year correlation in HR/FB%. If you see a guy who looks amazing or horrible out of the blue, there's a good chance he had a big shift in his HR/FB% that will probably not holdover to the next season. There's been a lot of content recently online about using this as a factor to identify buy-low regression candidates.

 

I can't find it unfortunately, but I read an article a couple years ago about how Mark Buerhle relied more on the defense of his teammates than any other pitcher at the time. This was derived from looking at GB%, BABIP, LOB%, and a bunch of other stuff that basically pointed to him being successful when positive things happened in the field of play -- partially from defenders sucking up a way above average amount of ground balls and working well with him to control the running game. It makes intuitive sense, of course, because he doesn't strike a lot of guys out. He's been an awesome pitcher, but he may have been significantly less awesome playing in front of a different team.

[ptatc]

My hypothesis would be that you'd see some slight correlation to GB% and run prevention, but it wouldn't be significant. Your best groundballers (as SP) since the statistic has been kept have been Brandon Webb, Derek Lowe, Jake Westbrook, Chien-Ming Wang, Tim Hudson, Aaron Cook, Zach Day, Fausto Carmona/Roberto Hernandez, Kirk Saarloos, and Justin Masterson. That's a pretty mixed bag. The one thing I notice when looking over those pitchers is that the most successful ones have been those that have been able to accrue some strikeouts.

 

http://www.fangraphs.com/leaders.aspx?pos=...=0&sort=6,d

 

You can mess around with that a little bit. One thing's for certain - you don't make it starting if you are throwing flyballs unless you pitch in a big ballpark. Chris Young and Jered Weaver are really the only starters to find some success.

I know that GB% alone isn't a key factor. Maybe it's K/BB or K% and GB%. I think the best place to start is to look at the pitchers who gave up the least runs and do a step wise linear regression to see what factors pop up.

 

[witesoxfan]

I know that GB% alone isn't a key factor. Maybe it's K/BB or K% and GB%. I think the best place to start is to look at the pitchers who gave up the least runs and do a step wise linear regression to see what factors pop up.

 

I think they've done quite a bit of this and it's where they derive the statistics from. So many of these variables show such little significance towards the runs scored with such high variability that it's impossible to say that one is more important than the other, but it essentially follows basic guidelines and sounds logic.

 

1) A strikeout is the best guarantee for an out.

2) A walk is never an out.

3) If the batter hits it, you prefer it to be on the ground because it's almost never a home run and the batter will usually not advance beyond 1B.

4) A flyball is preferable to a line drive, but those can be dangerous.

5) You do not ever want to give up line drives.

 

Your ideal pitcher is one who has good command, gets a fair amount of strike outs, and keeps the ball on the ground.

 

I'd argue that the best pitcher - starter or reliever - of the modern era is Mariano Rivera. His numbers:

 

8.22 K/9

2.01 BB/9

52.5% GB%

 

Those can be supplemented, and you can be great without 1 of them, but I generally think you need at least two of them to be a great pitcher.

[ptatc]

I think they've done quite a bit of this and it's where they derive the statistics from. So many of these variables show such little significance towards the runs scored with such high variability that it's impossible to say that one is more important than the other, but it essentially follows basic guidelines and sounds logic.

 

1) A strikeout is the best guarantee for an out.

2) A walk is never an out.

3) If the batter hits it, you prefer it to be on the ground because it's almost never a home run and the batter will usually not advance beyond 1B.

4) A flyball is preferable to a line drive, but those can be dangerous.

5) You do not ever want to give up line drives.

 

Your ideal pitcher is one who has good command, gets a fair amount of strike outs, and keeps the ball on the ground.

 

I'd argue that the best pitcher - starter or reliever - of the modern era is Mariano Rivera. His numbers:

 

8.22 K/9

2.01 BB/9

52.5% GB%

 

Those can be supplemented, and you can be great without 1 of them, but I generally think you need at least two of them to be a great pitcher.

That makes sense. So if you run the step wise regression you should get something like K is 30% of the variability, low walk rate is 15% and GB% is 10%. I'm not saying those are the numbers but just for example. I'd like to see that information compared to runs. I'm going to use the fangraphs info and runs the data over break and see what comes up.

[Eminor3rd]

I know that GB% alone isn't a key factor. Maybe it's K/BB or K% and GB%. I think the best place to start is to look at the pitchers who gave up the least runs and do a step wise linear regression to see what factors pop up.

 

Well, K and BB are two of the pillars of FIP. That's not to say GB isn't important, it's just that it's context-dependent and thus a relative factor. DIPS theory is essentially trying to give you as complete a picture of the pitcher alone as it directly contributes to runs saved. Other important metrics are derived statistics and are critical for seeing how a player fits into your own situation, but aren't useful in comparing players in an absolute sense.

Edited February 26, 2014 by Eminor3rd