Category Archives: Player Evaluation

Statistical Analysis: wOBA


I had planned on researching and writing up weighted On Base Percentage, but I found after doing that research that I couldn’t write it up any better than some of my reading. So instead here’s what I’ve found, and they’re definitely some excellent reads.

From The Book: Tom Tango’s article about wOBA

From Fangraphs: Dave Cameron’s writeup about wOBA

From Yahoo: Alex Remington’s writeup about wOBA

From SBNation.Com’s Bless You Boys: Saber 101: wOBA

From Saberlibrary: wOBA

The long and short of wOBA is this: It takes into account both OBP and slugging percentage when evaluating a hitter and his ability to provide value to his team. It also works with the league average player as well. The key is that this helps to value properly players who get on  base a lot and also hit for non-home run power. A player who hits doubles and triples but very few homeruns still has a lot of value to a team.

Statistical Analysis: Runs Created


Another statistics that we can use to help us look at a hitter’s value is Runs Created.

Here’s the formula (courtesy of Wikipedia)

(H + BB – CS + HBP – GIDP) * (TB * (.26 * (BB – IBB + HBP)) + (.52 * (SH + SF + SB))
——————————————————————————————————
(AB + BB + HBP + SH + SF)

Let’s take a look at the 2009 season of the NL MVP, Albert Pujols

H: 186 BB: 115 CS: 4 HBP: 9 GIDP: 23
TB: 374 IBB: 44 SH: 0 SF: 8 SB: 16
AB: 568

So, this calculates as follows:

(186 + 115 – 4 + 9 – 23) * (374 + (.26 * (115 – 44 + 9)) + (.52 * (0 + 8 + 16))
——————————————————————————————————
(568 + 115 + 9 + 0 + 8 )

Or: (283 * (374 + 20.8 + 12.48))/700, Or 165 Runs Created

Now, this number by itself does not really help us to translate his performance per se. Tomorrow I will be going into what we can use Runs Created in conjunction with so that it becomes more useful to us.

Scouting: What is the 20-80 Scale?


The 20-80 scale is a fairly standardized way of evaluating players. Generally a player is given two scores for each category in which they are rated. The first score is how the player currently rates in that particular area, and the second is what their potential is believed to be at their peak.

The scale itself, as best defined, from Kevin Goldstein of Baseball Prospectus:

A score of 50 is major-league average, 60 is above-average (also referred to as “plus”), and 70 is among the best (“plus-plus”). 80 is top of the charts, and not a score that gets thrown around liberally. 80s in any category are rare, and the scoring system is definitely a strong curve that regresses to around 50 at the major league level, but lower as you move down. Very few players have a 50 score or higher for every tool. Just being average across the board is quite an accomplishment.

I actually don’t have a lot to say about this particular topic that hasn’t already been said (and better), so I’ll leave you with some other articles you can read about the topic.

Statistical Analysis: The Fallacy of Small Sample Sizes


Looking at statistics is one key way for us to help evaluate a player and their contributions to a baseball team. However, when looking at these statistics, it is critical that the sample size is sufficiently large. Let’s take a look at some numbers, and see what conclusions we can draw from them.

Player A: .300 batting average, 3 HR, 10 RBI
Player B: .300 batting average, 3 HR, 10 RBI

Without any other information, these appear to be essentially the same in terms of production, right? At least until we look at some more details.

Player A: .300 BA (6/20), 3 HR, 10 RBI
Player B: .300 BA (90/300), 3 HR, 10 RBI

Knowing what kind of sample size we are looking at can help us to determine better about a player. In the example above, the sample size for Player A is simply too small to judge effectively without further information. In his case, we would want either a full season worth of statistics or some other statistics from previous seasons or levels.

Player B helps us to draw a few more conclusions. In 300 at bats, a player is likely to be portraying a majority of the skills he has. The fact that he hit .300 over that span can give an evaluator at least some confidence in their ability to do so in the future. They can also infer that the player is not a particularly powerful hitter at this point, as evidenced by only 3 HR and 10 RBI. While there are other factors outside of that player’s control, the sample size is large enough that some reasonable conclusions can be drawn.

The poster child for small sample sizes is Chris Shelton. Back in 2006, Shelton had fallen into the first base job for the Tigers, after having Carlos Pena perform poorly and end up being released. He had never played a full season at the majors, and he got off to a hot start in the month of April.

April: .326/.404/.783, 10 HR, 20 RBI in 104 plate appearances

Unfortunately, that was easily the best month he had of the season.

May: .286/.340/.363, 1 HR, 8 RBI in 100 plate appearances
June: .205/.286/.364, 4 HR, 9 RBI in 97 plate appearances
July: .289/.344/.386, 1 HR, 9 RBI in 90 plate appearances

He missed all of August and a majority of September due to injuries. I have to imagine that there were more than a few fantasy owners that want some trades involving him back. Looking at his minor league statistics to that point, the most home runs he had hit in any single season was 21, so the players looking at the 10 homers in a month and thinking he could hit 40 were rudely awakened.

Conclusions

The key to remember when doing any statistical analysis is to look at the period of time you are drawing your statistics from, and determining if it is a relevant period of time. It’s an unfair judge of a player and their abilities to look at too small of a time period.

Statistical Analysis: Why ERA is a Poor Measure of Success


For as long as can be remembered, ERA has been considered to be a good measure of the success of a pitcher. But unfortunately it really can be a misleading metric.

Earned run average is calculated in the amount of earned runs per 9 innings pitched.

ERA = ER * 9 / IP

The fallacy with ERA lies in the theory that a pitcher controls everything involved with the statistic. The definition of an earned run is a run scored that is not as a result of an error. Unearned runs, which are not counted in ERA, are based on whether a player in the field was charged with an error or not. There really isn’t a particularly consistent guideline for what is an error, either. It is one of those “I know it when I see it” things unfortunately.

Of larger concern however, is the theory that unearned runs don’t matter. While an unearned run may not be the pitcher’s fault, it is not a particularly fair judge of a pitcher either. Since ERA only measures the Earned Runs and not other stats such as hits or walks, a pitcher can very easily post a low ERA and not pitch particularly effectively.

Statistical Analysis: FB/GB/LD Rates


Looking at the end results of an at bat can be misleading sometimes. Not all singles are created equally. This is where batted ball data can come into play.

FB (Fly Ball) Rate: The percentage of batted balls which are considered flyballs. To me, this is judged as any ball that a fielder can take time and get undet to make a play on.

GB (Ground Ball) Rate: The percentage of batted balls considered ground balls. This one is pretty self-explanatory, as it is any batted ball which hits the ground prior to having the opportunity for a fielder to make a play. It also must hit the ground in the infield.

LD (Line Drive) Rate: The percentage of batted balls considered to be line drives. These are generally the most solidly hit balls, and have a much better chance of resulting in a positive outcome for the hitter.

What can these ratios tell us?

Generally, the higher a player’s LD ratio is, the more likely they are to have success as a hitter. Since line drives are generally the best hit balls, they are more likely to lead to success.

Home run hitters tend to have higher LD and FB rates,and speedy hitters tend to be more successful if they have higher GB rates, as this tends to allow them to use their speed to their advantage. Let’s look at some examples:

Aaron Hill hit 36 HR in 2009, easily doubling his previous career high of 17, set in 2007.

2007: .291, 17 HR, .459 SLG – 20.8% LD, 40.3% GB, 38.9% FB
2009: .286, 36 HR, .499 SLG – 19.6% LD, 39.5% GB, 41% FB

The key to notice with Hill is that his FB% went up by 2% from the 2007 season. While this is not a substantial enough jump to completely show the home run total as a fluke, it is something to watch with him as his previous full seasons he had not hit above 36% flyballs.

Ben Zobrist had a breakout season in 2009, hitting 27 HR, also more than double his previous high of 12 in 2008 (albeit in reduced playing time).

2009: .292, 27 HR, .543 SLG, 20% LD, 41.5% GB, 38.5% FB

For Zobrist, it was his line drive % which tells us his story. In 2008, Zobrist had only 13.5% line drives, while he had been at around 20% or higher previously in his career. Since arriving in the Majors (albeit briefly) in 2006, he had been reducing his GB rate, and seeing improvement in both FB and LD as a result. By reducing his GB%, he was making more solid contact and seeing improved results as a by-product.

How can we use this information?

Realistically, this data is best used in conjunction with BABIP to help determine the likelihood a player will repeat the statistics they had in the previous season.

What about using these stats for pitchers?

Pitchers have, realistically, the opposite goal and can see what kind of success they are having based on how low their LD% is. Odds are that if hitters are not hitting line drives, they are not making good contact, and as a result are less likely to reach safely.

Statistical Analysis: What is BABIP?


A term we hear a lot about when it comes to statistical analysis is BABIP. But what is BABIP, and how can it be used to help determine things about a player?

I had planned on writing about what I know about BABIP, but after reading this article, it is apparent I know less than I thought I did. So instead, I’m just going to suggest that if you are interested in this, you read this article by Tristan Cockcroft from ESPN’s Fantasy Draft Kit. It’s an excellent primer on the topic, and I’m not sure I honestly could say it any better or really add that much to it.

My biggest thing with BABIP is that it, like nearly all other statistics, don’t tell the whole story, but can help to give you an idea of something to look at when making distinctions about players. Let’s look at the top 5 and bottom 5 players in BABIP, and see what (if anything specific) we can discern about them.

Leaders from Fangraphs.com for 2009 season:

David Wright (.394 BABIP) – Wright hit .307 last season, which is well within his range of performance in the past. His BABIP was a bit higher than it had been in the past, previously ranging from .321 to .356 in the previous 4 seasons. With him, I am not entirely sure I am concerned about this, as he also saw a substantial drop in his home run total, and as a result would have had more balls in play. 10 additional homeruns (to bring his total for the season to 20) would have dropped his BABIP to .376, which while still higher than any previous season, would have been closer to the range he had hit in previously.

Ichiro Suzuki (.384 BABIP) – Ichiro hit .352 in 2009, which was one of his highest averages in his career. For him though, his range of BABIPs since arriving in 2001 leads me to believe that while he may have been slightly lucky, it was not that far out of his range of expected numbers. His BABIP range since 2001: .316 to .399

Hanley Ramirez (.379 BABIP) – Hanley hit for his best batting average to date, posting a .342 batting average in 2009. He had hit over .300 in each of the previous 2 seasons, and posted BABIPs from .329 to .353. There may be a little luck in there, but I think it would require more investigation into Hanley specifically to see if there is something else in there that helped him in 2009.

Joe Mauer (.373 BABIP) – Mauer had his best season to date, posting a .365 batting average. Mauer has never posted a batting average for a full season under .293, and as a result his BABIP has also never dropped below .319 during that time. While .373 was his highest BABIP so far, he has consistently posted BABIPs over .320, and seems unlikely to see a particularly large regression in 2010.

Joey Votto (.372 BABIP) – Votto completed only his 2nd full season in 2009, and improved his batting average by 25 points from his 2008 season (.297 to .322). However, his BABIP spiked from .328 to .372. While I don’t necessarily think he’s going to drop off the face of the earth for batting average, I could see him potentially regressing back closer to .300 than to .320. Another player who may have other factors playing into his statistics that aren’t as easily seen.

Bottom 5 (Regulars only):

Ian Kinsler (.241 BABIP) – Kinsler hit .253 for the season overall, and had posted BABIPs of .304, .279, and .334 in his previous 3 seasons. For me, I would think that he was probably at least a bit unlucky last season, as he was 30 points below his career low of .279. When he returns to the field of play, I could see a rebound back toward the .270 range for his batting average as his BABIP gets closer to the .300 range he had been averaging.

Carlos Pena (.250 BABIP) – Pena hit .227, which was low even for him. In his previous two full seasons, his BABIP had been .297 and .298 respectively. The part that isn’t seen as easily is the fact that in those two seasons, his batting average was .282 and .247. So while he also appears to have been a bit unlucky, it is hard to discern which of the two batting averages is more likely to be the one you’ll get from him.

Jimmy Rollins (.251 BABIP) – Rollins hit .250 last season, which was down from his career average of .275. Over the 6 seasons prior to 2009, his BABIP had never been below .281, and his batting average below .261. This is a player who has shown for the most part a specific range of what can be expected out of him, and should rebound this season as well.

Yuniesky Betancourt (.256 BABIP) – Betancourt hit ,245 last season, but had previously shown himself to be able to hit .280-.290. During those same seasons, his BABIP had been between .289 and .308. Another player who may have been a bit unlucky last season at the plate.

Aubrey Huff (.260 BABIP) – Huff hit .241 last season, well off of his career average of .282. His BABIP range doesn’t necessarily tell us everything about him though, as he has varied widely over the previous 8 seasons, ranging from .267 to .315.

Overall, I think that BABIP can help you to look at whether a player may be due for a regression or an improvement, but it needs to be taken with a grain of salt. Over time, players will generally stay within a specific range of performance, and if that performance is high, BABIP may not tell you anything of use.

Scouting: What is a 5-tool Prospect?


Whenever top prospects are discussed, the term “5-Tool Prospect” seems to come up often with the best of them. But what exactly are they talking about when the term is used?

The 5 Tools

Hit for Average – Essentially whether or not a player can hit and reach base on a consistent basis.

Hit for Power – This is not necessarily limited to home run power, but can also include the ability to hit doubles and triples as well.

Run – Not always necessarily viewed as the speed of a player, although this seems most common. This can also include a players ability to be a good baserunner, including taking extra bases, and not getting caught in rundowns, etc.

Throw – This one is pretty self explanatory. This is the player’s ability to throw, both distance as well as velocity (quickness really), and accuracy.

Field – Another one that’s self-explanatory. This one helps to gauge the player’s ability to make the plays required for their position. This can also include their ability to position themselves, both before the play and during the play as well.

Generally, a lot of the top hitting prospects are considered to be 5 tool players. In an ideal world, you would want a player at every position that does all of these things well. Obviously, that would come with a price to be sure. Some recent examples of 5-tool prospects include new Braves RF Jason Heyward, Diamondbacks RF Justin Upton, and Rangers 1B Justin Smoak.

Now, the 5 tools also don’t tell the whole story, as players are being evaluated on some other topics as well, many of which help to tell the story of the 5 tools. These can include a player’s ability to draw a walk, their ability to hit to all fields, and their range out in the field.

Statistical Analysis: The Three True Outcomes


The hardest part of trying to evaluate players statistically is to discern what is in their control, and what is beyond their control. A pitcher has essentially no control over what happens to a ball once the hitter put the ball in play. He is at the mercy of his fielders, their positioning, their throwing strength, their abilities, et cetera. A hitter is at these same mercies when the ball is hit as well. So how do you judge what a player can do independent of the situation?

The Three True Outcomes help to explain what is independent and what is not. The Three True Outcomes are a strikeout, a walk, or a homerun. But why are these considered independent?

For pitchers:

Strikeout – For a strikeout, a pitcher is essentially reliant on himself only. He doesn’t require any fielders to make the out (other than the catcher to catch the ball), no throws are required, and are not affected by the positioning of the fielders.

Walk – For a pitcher, the same logic holds for a walk that would for a strikeout. The fielders have no affect on a walk, and only whether the pitcher can locate his pitches will cause or avoid a walk.

Homerun – This one is slightly less independent, as it is reliant upon the hitter to cause the outcome. Again, no fielders are generally able to make plays on homeruns (most of them anyway), and as a result, the outcome is viewed as a direct result of the pitcher’s ability to locate his pitches properly.

For hitters:

Strikeout – For hitters, the logic is essentially the same. A hitter has the option of whether or not to swing the bat, and as such is in direct control of his ability to avoid a strikeout.

Walk – For hitters, their ability to discern balls and strikes is directly related to his own skills. If he can effectively judge the strike zone so that he can swing at good pitches and take poor ones, it also stands to reason that he is in direct control of his ability to draw a walk.

Homerun – This rule applies for homeruns which clear the fence (as opposed to inside-the-park homeruns). Because any ball hit over the fence cannot be affected by any of the fielders, the home run is also considered to be a direct result of the hitter.

What do the Three True Outcomes Tell Us?

Generally, most players do not fall directly into the Three True Outcomes. Players tend to hit a lot of balls into the field, whether they be ground balls, fly balls, or line drives. That said, you can discern a lot about a player based on their results within those three categories. The players who are generally considered to be the most valuable can (for hitters) compile walks and homeruns, while avoid strikeouts. For pitchers, they generally can compile strikeouts and avoid walks and homeruns.

What can we use to look at players regarding the Three True Outcomes?

The most common statistics to help measure the Three True Outcomes are:

  • Strikeouts-to-Walks (ratio, pitchers)
  • Walks-to-Strikeouts (ratio, hitters)
  • Strikeouts per 9 innings pitched (ratio, pitchers only)
  • Walks per 9 innings pitched (ratio, pitchers only)
  • Home Runs Allowed per 9 innings pitched (ratio, pitchers only)

There are some other statistics that can be used to help discern other information regarding this specific stats, which we will go into another time.

Strikeouts-to-Walks: This is simply comparing the strikeouts a player has versus the amount of walks they have. Having a high strikeout-to-walk ratio can illustrate one of three things: Either the pitcher strikeouts a lot of batters, he doesn’t walk very many batters, or both. Ideally, this is a number that pitchers want to be higher.

2009′s League Leaders in Strikeout to Walk Ratio, along with their Cy Young voting finish:

  • National League: Dan Haren (5.868) – 5th place
  • National League Cy Young Winner: Tim Lincecum (3.838) – 7th place in NL
  • American League: Roy Halladay (5.943) – 5th place
  • American League Cy Young Winner: Zack Greinke (4.745) – 2nd place in AL

While not a perfect judge of pitching success, it can help to tell you a lot about the quality of a pitcher and his ability to control his pitches. The better the pitcher’s control, the more likely he is to get good outcomes when pitching.

Walks-to-Strikeouts: The reverse of strikeouts-to-walks, the more walks a hitter draws, the more chances he has to score a run. Strikeouts have the opposite effect, so they are to be avoided as much as possible. A higher walk to strikeout rate generally bodes well for a hitter. This is essentially a judgment on a hitter’s ability to judge the strike zone.

2009′s League Leaders, along with the MVPs finish in the category:

  • National League Leader (and MVP as well): Albert Pujols – 1.80 walks per strikeout
  • American League Leader – Dustin Pedroia – 1.64 walks per strikeout
  • American League MVP – Joe Mauer – 1.21 walks per strikeout – 2nd in AL.

Generally, players who draw walks and avoid strikeouts are going to show more success than the average player who does not.

Strikeouts per 9 innings and Walks per 9 innings pitched: These both help to tell the story of the strikeout-to-walk ratio. Take these two examples:

Pitcher A: Strikeout to Walk Ratio of 4.0
Pitcher B: Strikeout to Walk Ratio of 4.0

They both seem to be equal, right? What happens when we add some more information to them?

Pitcher A:  Strikeout to Walk Ratio of 4.0, 9 IP, 4 K, 1 BB – Strikeouts per 9 Ratio of 4, Walks Per 9 Ratio of 1.
Pitcher B: Strikeout to Walk Ratio of 4.0, 9 IP, 12 K, 3 BB – Strikeouts per 9 Ratio of 12, Walks per 9 Ratio of 3.

These now tell us two very different stories. Pitcher A is much more reliant upon is fielders, as they were responsible for getting 8 more outs than for Pitcher B. Ideally, you want a strikeout/9 ratio that is higher, with a walk/9 ratio that is lower. They don’t tell the whole story by themselves either, but help to paint a broad picture when combined with strikeout-to-walk ratio.

Home Runs per 9 innings: This is simply a metric of how many home runs a pitcher allows. The reason it is important to look at how many innings pitched is to help judge a player against his counterparts. When taken in context with the amount of total homeruns, this can help to give an indication if a pitcher is unlucky in terms of home runs allowed, or if it is a trend to be monitored. The theory being that if the ball is kept in the park, there’s a better chance that a fielder will be able to make a play on it and potentially get an out.

2009 Leaders in HR/9:

  • Chris Carpenter – 0.33 (2nd in Cy Young voting)
  • Clayton Kershaw – 0.37
  • Tim Lincecum – 0.40 (NL Cy Young winner)
  • Zack Greinke – 0.43 (AL Cy Young winner)

Does this really tell us anything of use?

Alone, it does not necessarily help. But when looking at these statistics, they can help to give us a basic idea of whether or not a pitcher will have long term success or not. Players who are completely at the whim of the players in the field become more reliant on luck than players who are not.

So players who only do these three things should be the best automatically, right?

Unfortunately, no. There is clearly more to a baseball game than just strikeouts, walks, and home runs. What this gives us is a base to look at what skills a player has, and will help us to look at other things that the player has done to get a good idea of what they are capable of doing in the future. There are a few hitters who are known as “Three True Outcome Hitters”, as they tend to do these three things most often. Adam Dunn and Jack Cust are two players who are well known for being TTO hitters. But looking at their stats indicates that these players are clearly still different:

2009 stats for both players:

  • Adam Dunn: 17% walk rate, 32% strikeout rate, 0.66 K/BB ratio, .267 batting average, .398 on base %, 38 HR
  • Jack Cust: 15% walk rate, 36% strikeout rate, 0.50 K/BB ratio, .240 batting average, .356 on base %, 25 HR

Both players ended up with one of the Three True Outcomes in over half of their plate appearances in 2009, but clearly based on these numbers Adam Dunn had the better season overall. What the Three True Outcomes cannot tell us is anything involved with what happens in the field of play. All it can do is to give us a starting point for what is within that player’s control.

Conclusions

Three True Outcomes is just a starting point. You can discern a lot about a player from their ability to either limit or achieve these outcomes in the game, but it’s not going to tell you the whole story. What they can help to tell you is how well a player can judge the strike zone, which is still a critical skill for both hitters and pitchers. They can help us to get context on what information that some of the other statistics may not necessarily show. As a general rule, the players who can perform best in these three categories will have continued success at the Major League level.

Scouting and Statistical Analysis: What is the Main Goal?


Before getting into ways that players are analyzed using scouting tools and statistical analysis, I figured it was important to discuss what the point is of doing either.

Baseball is no different than any other business, in that they are looking for the best return on investment. On an organizational level, the best possible return is a championship. All other decisions are based off of this goal. The difference for many teams is based on the timing of this goal. For teams like the Yankees and Red Sox, this goal is aimed for every single season. For others, like the Nationals and Marlins, it may not be every year that they will compete for a championship, so they need to operate such that they will improve each season in order to compete down the line.

So where do scouting and statistical analysis come into play with this goal? Simply put, you need to evaluate your players, so that you know what to expect in terms of quality of performance. Quality baseball players are a finite group, as are jobs with Major League teams. So each team is looking to get the most from each roster spot they have, toward the end goal of a championship.

For years, scouting was viewed as the only tool in player evaluation. You didn’t need to see what players had done previously, you just had to SEE a player in a game or practice and could make judgments on him and what his potential was. This was the only way to judge players in a lot of teams’ minds.

Over time, fans and teams both began looking for other ways to evaluate players. Baseball has always had a large quantity of statistics available, but a lot of them didn’t tell the whole story. Over time, statistical analysis of the outcomes on the field slowly crept into Major League front offices, as teams learned that there were more to be learned besides what could be seen. The largest motivating factor behind this became the amount of money being spent on the product on the field. With payrolls upwards of $100 million per year, and players making upwards of $5 million per year in many cases, it was in the organizations’ best interest to ensure there was as little risk involved with players as possible.

There is no right answer to which works better. Both are equally important for an organization to use, as both can help a team to evaluate their players more effectively.

My Knowledge Level

The point of writing about these topics, for myself anyway, was to learn more about both topics. Currently, I have a much larger knowledge base in the statistical analysis side of player evaluation (at least in comparison to scouting). I do not consider myself to be an expert on either topic, but am hoping to learn about them, and in turn help you to learn about them as well.

In the coming days and weeks, I hope to learn quite a bit, and just continue to add to my knowledge about the great game. So, to repeat, the main goal of both scouting and statistical analysis is to help us to evaluate players and organizations, toward the end of winning championships.