Thursday, June 25, 2009

Pitching and the DEFENSIVE factor

Don't get too offended by this comment, because it's not meant to be derogatory to the entire pitching profession, but nobody really cares how well a pitcher fields his position when they're giving out contracts. That does not, however, mean that a pitcher does not have a DEFENSIVE component within his PEVA Pitching Rating Grade, it's just not the traditional kind. When we're talking about DEFENSIVE factors for a Pitcher Grade, it's not about Errors, Putouts, and Assists. It's about Strikeouts, Walks, and Home Runs Allowed.

Those are three stats in the pitcher's line where he has the most control, where he can defend the outcome, and where his defenders and team don't play much of a part. So in the pitcher venacular for PEVA DEFENSIVE components, it is an average of those two component factors (Strikeout to Walk Ratio and HR Allowed per 9 Innings Pitcher), along the same MAX/AVE/MIN scale, that counts.

But isn't that unfairly giving credit to a strikeout pitcher?

No. It gives credit to a pitcher who has control compared to how many he strikes out. One example, Greg Maddux. Most years Maddux would lead the league in SO/W Ratio despite not being a pure strikeout pitcher. (He did strikeout his fair share, BTW, even though his fastball didn't sit in the 90s)

Why should you give so much credit to a pitcher with a low HR/9IP ratio?

Look at Curt Schilling, he'd often give up solo homer or two, but that would be about it. That's true, but since Schilling often was near the top in SO/W ratio (thus being the reason they were only solo homers), his lower HR/9IP factor did not injure the overall PEVA Pitching Rating.

Does this or any other pitching factor play a moderating factor like Run Production does to a position player or even Field Value for that matter?

No. But it all works out in the wash. And that wash is how a pitcher, whether relief or starting, whether playing for a good team or a bad one, is valued when comparing his PEVA to contract terms. And that's where RAVE and EXPEQ comes in, ... but we won't bore you with those Explaining RAVE blog series right now. We'll clue you in on those at a later date. Pennant races are starting to heat up and the All-Star game is coming soon.

Saturday, June 20, 2009

Field Value - The Most Controversial Component

Well, let's just get right into it. In this 5th installment of Explaining PEVA, we're going to try and tackle the questions surrounding Field Value, the DEFENSIVE component for Position Players, and how it works in context with the other five categories.

Is Field Value equal to the other components of the index (i.e. the 2 DURABILITY Factors and the three DEPENDENT and INDEPENDENT production factors discussued earlier) or does it have special value like that of Run Production?

Field Value is one of the six components and in its original incarnation counted the same as any of the other five factor components, however, it became apparent, perhaps one third of the way through the development process, that that would not be accurate. As most sabermetrics fans or proponents know, the position you play has an awful lot to do with your value, and should be weighted differently. But then another question popped into play.

Okay, so a 1st baseman with a gold glove year isn't nearly as valuable as a Shortstop with a gold glove, but what about a 1st Basemen who's pretty bad in the field, but nobody cares because of how valuable he is with the bat? How would the Field Value of that player be adjusted to reflect his true value to a team?

That's where the RPR (Run Production Factor) adjustment comes into play, raising the level of a poor fielder with exceptional production in the overall PEVA factor. While he actual Field Value will remain what it is, when calculating the overall PEVA Player Rating, it will be moderated according to the level of RPR.

Okay, I'm lost. How is the Field Value calculated in the first place, and which positions does the system value the most highly?

Field Value is calculated in the same manner as the other factors, placing a top factor for the player with the MAX value in a category and grading all players between that MAX, the AVERAGE, and the MIN. The MIN is not 0, but a MINIMUM grade we have determined as appropriate within the PEVA system. Unlike other categories, the MAX and AVE change with each position, and the final PEVA Field Value is calculated using the Field Value for each position a player plays and weighted to how many innings a player plays that position.

The system weights the MAX in this way: Catcher 2.10, Shortstop 1.75, Third Base/Outfield 1.70, Second Base 1.50, 1st Base 1.40. The value for a pitcher can be as high as 1.00, depending on the amount of Innings Pitched, but not his value in the field.

I understand why Catcher and Shortstop would be that high, but why is an Outfielder MAX higher than Second Base?

For an outfielder to reach the 1.70 figure, he would most likely have to be a Gold Glove level centerfielder. We don't really know why the Second Base figure calculated with such a low MAX, but suffice it to say, that is how Payroll values the position with a top level fielding 2nd baseman. And you know, just from a subjective point of view. 2nd base is the position many infielders default to in their careers, but it's unlikely that a poor fielder ends up in center field. He may end up in left, but not center.

But what are the components that make up those MAX/AVE/MIN factors for each position? Do you get into Range Factors and newer Zone Ratings, etc? We do use Range Factor due to its objective stats, and even though we think Zone Ratings and other such data is valuable, we have not included it in the calculation for two reasons; its subjectivity and the fact that such data does not have a long historic past for comparing older era players.

Field Value Component Factors (Position, Stat, Weight)

Catcher -
IP/GP 25%, Fielding PCT 25%, Range Factor 25%,Caught Stealing % 25%

Infield - IP/GP 33%, Fielding PCT 33%, Range Factor 33%
Outfield - IP/GP 25%, Fielding PCT 25%, Range Factor 25%
Assists Per 9 IP (or Games Played) 25%

Note 1: Prior to 2000, using Games Played.
Note 2: For Catchers, the Caught Stealing Percentage used from 1960-2006. Prior to 1960, other factors reflect 33%.

So there you have it, an overview of the most controversial aspect in the PEVA factor universe. But while it is controversial, and requires that deviation away from a static max, it is important, and allows a light hitting shortstop with spectacular defensive abilities to warrant the salaries they earn while allowing that lunk hands 1st baseman with the RBI potential to warrant the multi-year $100 million dollar contract.

Wednesday, June 17, 2009

Sosa and the PED story

PED. Performance enhancing drugs. Well, now, after all the suspicions and denials by Mr. Sosa, his name has surfaced as one of the 104 players who tested positive in 2003, and if that's true, is another one of the star players of this steroid era to have bitten the infield dirt, or outfield grass in Sammy's case. Hall of Fame. Gone. Reputation. Sullied. Still rich as all get out though, ... isn't that just great.

The thing is, Mr. Sosa was not a special player even with the PED assistance. I know. I know. He hit all those home runs. 609 dingers. And knocked in all those runs. 1682 RBI is one heck of a total. Sure is. Well, okay, he was pretty good. But pretty good in ranking #82 all-time in Total PEVA Player Rating @ 232.726. That's just below Brooks Robinson and only a couple ahead of Dale Murphy, Tim Raines, Edgar Martinez, and Jimmy Wynn. The point being, Sammy was only in the "maybe they'll get in category" when you look at the totality of his career, including the fact of the parks he played in, the fact he didn't really play defense, or get on-base that much with a 0.344 career OBP. Oh, but he could smile. And he could lie. And he could make money. To bad Sammy doesn't speak English well enough to lie to Congress on his own, otherwise he could ask for membership.

And if all these allegations of PED use are actually true, and you start to factor in the assistance they probably gave him in reaching those totals. Que lastima. Just look at Sammy's stat line. If you accept the fact that he was a PED user in 2003, and that the home run chase between he and McGwire was a PED chase in 1998, then that covers the six best, and only great years, in Sammy's career.

Let's do some quick math deductions.

1996 - 36 HR, 119 RBI, 0.268 Ave. - 16.064 PEVA Rating
1997 - 40 HR, 100 RBI, 0.273 Ave. - 9.113 PEVA
1998 - 36 HR, 119 RBI, 0.251 Ave. - 13.226 PEVA

Ave. Three Years Prior
37.3 HR, 112.7 RBI, 0.263 Ave. - 12.801 PEVA

Ave. Probable Steroid Years (1998-2003)
55.3 HR, 134.7 RBI, 0.302 Ave. - 24.155 PEVA

2004 - 35 HR, 80 RBI, 0.253 Ave. - 6.719 PEVA
2005 - 14 HR, 45 RBI, 0.221 Ave. - 2.214 PEVA
2007 - 21 HR, 92 RBI, 0.252 Ave. - 3.971 PEVA

Ave. Three Years After (2004, 2005, 2007)
23.3 HR, 72.3 RBI, 0.243 Ave. - 4.358 PEVA

So here's what we'll deduce. Sammy Sosa may have been the player who most benefited, in his stats, from the PED era, if the allegations are true. Even if you say that his performance in those six years would have equaled that of his previous three years, that means a loss of ...

108 HR, 132 RBI, and 68.124 PEVA.

And that leaves Sammy with 501 HR, 1550 HR, and 164.602. And what player neighborhood would that be in ... #218 Best Player according to PEVA Ranking, a drop from that #82, just behind Robin Ventura, Roy White, and ahead of Joe Start, Tommy Leach, and Lave Cross.

But that would likely be giving Sammy too much credit, when you look at what happened in his career after the 2003 season. If you say his performance during that six year stretch would have equaled that average year of his previous three and three years after, that means a loss of ...

150 HR, 253.2 RBI, and 93.453 PEVA.

And that leaves Sammy with 459 HR, 1429 HR, and 139.273. And what neighborhood would he be playing in in that more likely scenario ... #307 PEVA Best Player, just behind Cy Williams and Cy Seymour, and ahead of Ken Caminiti, Roger Maris, Denny Lyons, and Garret Anderson. Do you see Hall of Fame next to any of those names?

But what might be the most insidious of the losses that would have occurred, is the one that would have occurred in his pocketbook. SPRO calculation of the lost earnings in the second, and more likely scenario, would be nearly $75,000,000. Yes, that's $75,000,000.

Geez, even in baseball, that's what most would call real money.

Monday, June 15, 2009

Explaining PEVA - #4 Independent Production

Just what is INDEPENDENT PRODUCTION in the terms of a baseball player? Well, it is those categories or stats that rely predominantly on the player's own abilities, and are less dependent on team fortunes. Baseball is one of the few sports, due to the unique nature of the pitcher versus batter dynamics that has categories of production that are nearly independent of their teammates, while still being in a team game. For a pitcher, we're using Earned Run Average and Walks Plus Hits Over 9 Innings, WHIP9. Yes, this is the traditional WHIP category extended to a value over 9 innings pitched. For a position player, we're using On Base Percentage and Slugging Percentage.

The pitching categories are simple standard categories in the PEVA index, measuring their factor values from the relationship between the MAX/AVE/MIN for the season.

For the batting categories of OBP and SLG, it is not so simple, due to their relationship to the average and how some very special players (very few players reach this level in a decade) who rise above a multiple of the average. These players are given special factors above the normal scale when utilizing the OBP or SLG PEVA factor in the calculation of the overall PEVA Player Rating Grade. How special can a player be on that scale? Up to 50% higher than the max.

In a previous installment of Explaining PEVA, you said that OBP and SLG factors could also be modified by a stellar Run Production Factor, isn't that true? Absolutely. For players that produce runs, above and beyond the logic of their OBP or SLG factors, the factors for those categories are adjusted in the calculation to account for that ability. For example, Phillies shortstop Jimmy Rollins rarely gets on base for a leadoff hitter and really doesn't slug as high as most people think (with some exception for his MVP season of 2006), but he scores and produces runs. This nets him a contract above what most traditional believers in pure OPS think warranted, and while this is an extension of his DEPENDENT production stats and not his INDEPENDENT ones, it is how he has extra values. Now one could argue whether his value would be this high on a team without run producers such as Chase Utley and Ryan Howard batting behind him, but the point is moot when considering Payroll or even Player Rating values, although a player with his statistical profile will likely not be valued as high in an index such as PEVA as some would think, despite the adjustment given for Run Production.

So OBP and SLG factors become very important for the stellar player, especially ones such as Babe Ruth, or even Barry Bonds during those questionable years, when their OBP and SLG went off the scale compared to the average for the season. It is also important to all other players, including those with profiles such as Rollins or other run producers who benefit from good players around them.

Wednesday, June 10, 2009

Explaining PEVA - #3 W, SV, SV+W

Let's talk about those pitchers. Past the pitching durability stats (Games, Games Started, and Innings Pitched), two of the remaining four categories that go into the PEVA Player Rating index and Player Grades are pretty straightforward (ERA and WHIP9) in how they fit into the system. One, the dependent category of pitching wins, saves, and wins plus saves, is not that much harder to apply, although it is divided into those three sub-categories.

So when you say subcategories, do you mean to say that Wins is a subcategory worth 1/3 of the value, with Saves and Wins Plus Saves each worth the same?

Short answer would be no. These subcategories are independent of each other, and the pitcher receives a PEVA factor value based on the highest value of the three on the scale from MAX to AVE. to MIN. So if a pitcher has a higher factor of wins compared to the MAX/AVE/MIN, then he does for saves compared to MAX/AVE/MIN, or wins plus saves compared to MAX/AVE/MIN, then that's the factor he gets.

Why so complicated?

It's really not that complicated. A starting pitcher needs to be based on how he performs in the most important DEPENDENT category in the game for a starting pitcher, and that would be WINS. But a relief pitcher, particularly a closer whose value per payroll has risen in some true measure to the rise in SAVES, needs a DEPENDENT category that is important to his role. For the pitcher who is neither a true starter or true closer, i.e. spends the majority of the year in neither role, it is unfair to base his DEPENDENT category factor on either by themselves, therefore the use of WINS PLUS SAVES.

What does the term DEPENDENT category mean? It means that the total of wins or saves has as much to do with the team as it does the individual player. There is only one DEPENDENT Production category in use for pitchers (W/S/W+SV) as there is for batters (Run Production).

USE/DURABILITY - Games, Games Started, Innings Pitched (Pitching); Games, Plate Appearances (Batting)
DEPENDENT PRODUCTION - Wins/Saves/Wins Plus Saves (Pitching); Run Production (Batting)
INDEPENDENT PRODUCTION - ERA, Walks/Hits Per 9-WHIP9 (Pitching); On Base Percentage, Slugging Percentage (Batting)
DEFENSIVE - HR per 9 IP/SO to W Ratio (Pitching); Field Factor (Position Players/Batters)

We'll discuss how they fit in in a later installment of Explaining PEVA.

Friday, June 5, 2009

Explaining PEVA - #2 Run Production

Thought it was about time we delved a bit further into the PEVA Major League Baseball Player's Rating system as the 2009 baseball season moves further and we begin to see player value in view again. Back in February, we discussed the girth of the system in our article on Ben Sheets (see Durability and Why the Baseball Evaluation System Values It So Much. The two factors that go into that girth provide 1/3 of the player's value, both for a pitcher or position player. This is true no matter the era or season the player played in, too.

But, oh, that's not very sexy. Girth. Isn't that something only sumo players and offensive lineman find attractive. That's a good point. But the point here is, the PEVA system may gain it's weight from those unsexy girth stats such as Games, Games Started, Plate Appearances, and Innings Pitched, but the alluring stats flesh out the system, and begin to tell us who the stars of the game truly are, not only for a season, but for their careers.

And the first one we'll discuss for Position Players and Batters is likely the most important of them all, ... Run Production. How do we define Run Production? Runs and Runs Batted In combined. No, this is not the Runs Created Stat that many use, where they delete Home Runs from the number so as not to count it twice. We count it twice, as a batter who can not only score that run, but knock it in is more valuable than the other player dependent on a Run Producer, too. There might be some disagreement on our take on this, but after those 5,000 hours of developing the system (and an initial thought to use Runs Created instead), Run Production became the first dependent production stat used in the PEVA index.

Like the value placed on the Durability Stats, the Run Production factor of a player is measured on a scale from the MAXIMUM for the year in question vs. the AVERAGE for the year in question. A player below the average is weighted down to a percentage of the value, based on the same scale. After calculation, this factor becomes one of the six multiples that will make up the PEVA Rating (Scaled between 0-64).

But is Run Production Only as Valuable as Innings Pitched or Plate Appearances, that doesn't seem right?

No, it isn't. Run Production is the Factor which plays a part in other factors as well, which will be discussed further in another Explaining PEVA installment. But for now, it affects the remaining three... Field Factor, On Base Percentage, and Slugging Percentage. While all three of those categories have their own inherent factor, they are modified by Run Production if Run Production reaches a certain threshold and the factor for those categories fall below the Run Production factor.

So, for the position player, you can say that Run Production becomes the most important factor of the six categories, although for most players, it is only equal to the others. However, in special circumstances, depending on the type of player and their value to the team, it rises in importance to the PEVA index and to the way players have been paid.