Batter's Box Interactive Magazine Batter's Box Interactive Magazine Batter's Box Interactive Magazine
Vernon Wells took a solid step towards a productive major league career in 2002. He played regularly, handled centrefield with flair and hit for more power than might have been expected. But he didn't walk much, so there is the inevitable concern in sabrmetric circles about his development as a hitter. What can we expect from Vernon in the upcoming years?

To address this question, I needed to identify a comparable group of young ballplayers to find clues about what the future has in store for Vernon. I approached it systematically, paring down the players based on (what I think are) reasonable criteria. First, I looked at all major league position players born 1969 though 1978 and noted their career numbers though their age-24 season (for 1969 players, 1993 is their age-24 season; for the 1978 players it is 2002). It's usually not a good idea to go too far back with age-comparisons - in some cases, what results is a comparison to what players did in very different eras. I chose 400 plus PA as my threshold, since that seemed like a reasonable sample, but still allowed a fairly large number of players to be included in the study. There were 130 such players.

To get a realistic comparison group, I chose to focus on the outfielders. I looked at players who played at least 70 games in the outfield through their age-24 season. That cut the group down to 54, but two of these - Darin Erstad and Cliff Floyd - played less than 60% of their games in the outfield, so I scratched them from the list.

Just as I had done in an earlier comparison of Felipe Lopez to Miguel Tejada, I broke down batting skill into 4 components - for each of the remaining players. These components were: ball in play average, walks per opportunity, strikeouts per opportunity, and manifested power (please see previous articles for a more thorough discussion of these rate stats). I looked at the absolute difference between Vernon's rates and each of the 51 other outfielders in the comparison group and summed these differences. The larger the resulting number, the less similar the hitter was/is to Vernon.

Before we start, it must be noted that Wells was born in December and therefore is at a slight disadvantage in this study. I could have done an age-23 comparison (in effect pretending that Vernon was born in 1979) - this would have created a slight bias in the other direction. I think I'll save that for another day.

Here are the 20 most similar hitters (outfielders) to Vernon Wells:
   name                    PA    BIP avg   walk  strikeout power old/young  similarity
Vernon Wells 845 0.300 0.043 0.140 0.147 0.056 0.000
Jeffrey Hammonds 575 0.309 0.046 0.148 0.149 0.069 0.035
Jose Guillen 1798 0.301 0.042 0.173 0.134 0.074 0.050
Carlos Lee 1136 0.322 0.044 0.146 0.159 0.085 0.065
Marc Newfield 666 0.286 0.050 0.160 0.145 0.015 0.065
Ruben Mateo 590 0.306 0.049 0.187 0.147 0.058 0.071
Terrence Long 634 0.306 0.066 0.125 0.145 0.025 0.077
Magglio Ordonez 650 0.295 0.046 0.094 0.126 0.063 0.081
Jermaine Dye 819 0.293 0.042 0.201 0.137 0.054 0.088
Shawn Green 910 0.308 0.054 0.168 0.177 0.021 0.096
Garret Anderson 1055 0.331 0.036 0.144 0.131 0.151 0.098
Juan Encarnacion 1341 0.319 0.039 0.184 0.161 0.089 0.104
Torii Hunter 441 0.296 0.062 0.177 0.116 0.042 0.115
Raul Mondesi 1125 0.323 0.040 0.170 0.183 0.074 0.119
Johnny Damon 1296 0.305 0.063 0.121 0.094 0.079 0.122
Mark Kotsay 1215 0.289 0.050 0.098 0.102 0.067 0.123
Jose Herrera 423 0.309 0.060 0.166 0.097 0.091 0.128
Todd Hollandsworth 958 0.319 0.067 0.191 0.137 0.058 0.146
Carlos Beltran 1879 0.329 0.070 0.173 0.150 0.057 0.147
Rondell White 1076 0.334 0.073 0.162 0.149 0.062 0.154
Ken Griffey Jr 3113 0.317 0.079 0.133 0.189 -0.024 0.154
average of 20 1085 0.310 0.054 0.156 0.141 0.060 0.062

The old/young column is an estimate of where a player lies on the old player/young player skills spectrum, with negative numbers placing a hitter at the old end of the spectrum. Bill James theorized that as a player ages, some skills develop while others decline. Speed is the quintessential young player's skill. Hitting for average is also a young player's skill, while walking and hitting for power are old player's skill. James hypothesized that a player with a preponderence of young player's skills has more development left in him (take heart Garret Anderson fans) - all other things being equal. Note that the estimate given here does not incorporate a speed rating - it is based only on BIP average, walks, and power.

The vast majority of these players were on the young player's side of the spectrum, mostly because they didn't walk much.

To whittle down the comparison group a little more, I scratched the 5 Dominican-born players (Guillen, Mateo, Encarnacion, Mondesi, and Herrera). Birthdates of Dominican players are notoriously suspect and the presence of these players is liable to skew any age-based study. That left 15, with a gap between Kotsay and Hollandsworth. I could have included Kotsay, but I went with the marginally better comp in Damon. The resulting group has 8 US-born players and 4 left-handed batters (Anderson, Damon, Green, Long).

Here are the average component stats for the comparison group:

BIP 0.305 Walks 0.051 Strikeouts 0.148 Power 0.138 (old/young .060; similarity .044)

I then looked at what each of the 10 had achieved in their age 25 to age 27 seasons (3 seasons). In two cases (Long and Lee), their age-27 season is 2003; in one case, the player never got to his age-27 season (Marc Newfield).
   name			PA	BIP avg	walks	k rate	power	old/young (under25)
Johnny Damon 2111 0.319 0.085 0.092 0.138 0.020 (.122)
Jermaine Dye 2027 0.327 0.084 0.164 0.199 -0.024 (.054)
Magglio Ordonez 2013 0.304 0.082 0.099 0.185 -0.049 (.063)
Garret Anderson 1980 0.324 0.036 0.118 0.129 0.137 (.151)
Shawn Green 1856 0.323 0.078 0.194 0.222 -0.041 (.021)
Torii Hunter 1565 0.314 0.050 0.199 0.191 0.027 (.042)
Terrence Long (to 26) 1327 0.294 0.066 0.152 0.130 0.018 (.025)
Carlos Lee (to 26) 1181 0.273 0.092 0.134 0.182 -0.129 (.085)
Jeffrey Hammonds 1022 0.288 0.091 0.179 0.169 -0.085 (.069)
Marc Newfield 385 0.269 0.084 0.146 0.078 -0.019 (.015)
Composite (years 25-27) ... 0.304 0.075 0.148 0.162 -0.015 ...
Composite (under 25) ... 0.305 0.051 0.148 0.138 0.060 ...

8 of the 10 played regularly throughout their age-25 to age-27 seasons. The exceptions were Jeffrey Hammonds, who suffered one injury after another, and Marc Newfield, who was never able to hold a regular big league job. His speed numbers were much worse than the other guys, suggesting that one reason he flamed out was that he was not able to handle the outfield defensively.

Most hitters improve their walk rates at the major league level. The comp-group raised their walk and power rates, while their K rates and BIP average remained stable. 6 of the 10 (Damon, Dye, Green, Hammonds, Lee, Ordonez) increased BOTH their power and walk rates. I believe that the relationship between walks and power is symbiotic: a touch more patience at the plate results in better picthes to hit, which enhances the power stats, which in turn discourages the pitchers from throwing as many strikes.

Garret Anderson's walk and power rates were unchanged, while T.Long saw a slight regression in his numbers. Newfield's numbers are of little use due to his limited PAs. That leaves Torii Hunter as the only player whose walk and power rates went in opposite directions. Torii's power surge was more impressive than anyone else's, but his walk rate dropped to the 2nd lowest in the group.

Every one of the 10 shifted towards the old end of the skills spectrum: Carlos Lee and Jeffrey Hammonds aged the most. The most improved hitter of the group was Jermaine Dye. He started out below group average in 3 categories and dead average in power. In his 25-27 years, he improved in all 4 areas (Damon was the other to do it; Magglio would have except for an uptick in strikeout rate). Dye moved to above group average in 3 of the 4 categories in his 25-27 years.

So what can we surmise about the immediate future of Toronto's centrefielder?

1) This is all guesswork. Every one of these 10 players developed (or didn't) as a hitter in a different way.
2) Vernon's low walk rate is not at all uncommon for a player his age and should not be a concern unless he fails to nudge it upwards over the next few years.
3) His power is likely to increase, though he will very likely not become an upper-tier power hitter

For the optimists among us, it is worthwhile reflecting on Jermaine Dye. As a young ballplayer, he showed less power and a worse K/W rate than Vernon Wells. Now he's an all-star caliber rightfielder, who could probably handle centrefield. Dye represents what Vernon could become if he works very hard and avoids serious injury.

What can we expect from Vernon Wells? | 6 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.
Coach - Wednesday, February 12 2003 @ 10:09 AM EST (#81657) #
...the relationship between walks and power is symbiotic

Definitely. If you get yourself out chasing 58-foot curves (see Mondesi, Raul) why would anyone throw you a fastball strike? Conversely, there's nothing like a 2-0 count; from there, your OPS should skyrocket. Unless you peek into the LF stands instead of keeping your head down, and swing right through hittable pitches.

This is all guesswork.

Too modest. Some of it is research, which I'm not inclined to do myself and therefore appreciate all the more. Some of it is interpretation, and I defer to those, like Robert, who have more experience and greater understanding of the data's significance. The only part that's guesswork is making predictions -- anything can happen, and it usually does.

I've mentioned this before (and so has Carlos Tosca; we both watch the Jays a lot) -- Vernon Wells hit an incredible number of "atom balls" last year. Frozen ropes that made OF duck while they caught them, certain doubles stolen by corner IF, singles speared by leaping middle IF; you name it, he hit 'em. His .300 average on balls in play is "due" to improve, based only on the observations of a couple of dugout rats.

When a study like this agrees with the instincts and "anecdotal evidence" some of us rely on, or when Keith Law and Tony LaCava recommend the same player, that's a very good sign. I like the Dye projection, and also think he resembles Garret Anderson, though V's a step-and-a-half quicker than either, making him more of a plus in CF.
robertdudek - Wednesday, February 12 2003 @ 11:30 AM EST (#81658) #
Just for fun, I "moved" Venon Wells' birthday back a month and compared him to outfielders' career number up to and including age 23. I moved the minimum PA down to 300 top include a few more players.

There were 83 qualifiers, of which 36 were outfielders. Jose Cruz Jr was the least similar hitting OF. Once again I dropped the Dominicans; after that, 4 new outfielders entered the 10 most similar list, replacing those who did not have the requisite PA.

The newbies were Ken Griffey Jr, Mark Kotsay, Carlos Beltran, and Curtis Goodwin (Damon, Hammonds, Anderson, Lee, Green, and Dye were the holdovers). This comparison group posted a BIP of .315, walk rate of .048, strikeout rate of .150, and a power rating of .139 (Vernon's numbers were .300, .043, .140, .147).

Whereas Vernon was a half-year younger than the age-24 comparison group, he was almost 8 months older than the age-23 group.

9 of the 10 increased their walk rates (exception: Garret Anderson) for an overall increase of 60% in the age-24-to-26 years. 6 of 10 gained power (3 remained flat, while Anderson's actually decreased). The comp-group's power rating climbed about 16%. Once again, the strikeout rate was virtually unchanged overall (though the BIP rate took a little bit of a hit).

Compare that to the 47% rise in walk rate and 18% increase in power for the age-24 comparison group in their follow-up years (age 25-27).

Splitting the difference, we can expect a 53% rise in Vernon's walk rate and his power rating to increase about 17% over the next 3 years.
Dave Till - Wednesday, February 12 2003 @ 02:52 PM EST (#81659) #
I wonder whether anyone has ever attempted to correlate a player's age of physical maturity with his growth and development curve? On average, players peak at age 27, but some peak much earlier, and some keep improving into their thirties. Not knowing any more than I do, I'd think that somebody who was shaving regularly at 13 would be less likely to be effective in his 30's than somebody who didn't start growing until his late teens. Do good scouts track this info when analyzing a prospect?

As for Wells: I noticed the "atom balls" too. I also noticed that he seemed to improve as the season went on, which always looks more impressive than just starting off good and staying there (as Hinske did). By late summer, Tom Cheek was claiming that many scouts were impressed by Wells, if that's useful information.

Robert's article made one point clear: forecasting is far from an exact science. When a player's ten best comps include Ken Griffey Jr. and Curtis Goodwin, it's safe to say that Wells's career could go in any of a number of directions.
robertdudek - Wednesday, February 12 2003 @ 03:10 PM EST (#81660) #
With the year-23 sub-study, I had to reach further to get 10 players. The average difference in the age-24 comparison group was .084 and in the age-23 it was .113. Goodwin had more speed, but much less power (than anyone else in the study). Griffey was better at everything, which adds up to a much better hitter overall.

I think the comparison group average is much more likely to tell us what we should EXPECT from Vernon. He could fail to develop (like Long) or exceed expectations (like Dye or Ordonez), but he will not become Curtis Goodwin or Ken Griffey.
_Justin B. - Wednesday, February 12 2003 @ 04:51 PM EST (#81661) #
When you look at the first table, there is a noticeable difference in the accumulated at bats of the various players up to the age cutoff. With such a finite pool of players from which to draw comparisons you couldn't really control for factors such as where experience has been garnered, but I'm curious as to if there have been any studies comparing the virtues of college experience, more seasons spent in the minors, or on the job training.
Gerry - Wednesday, February 12 2003 @ 10:29 PM EST (#81662) #
I think this excellent study takes us to (today's) limit of statistical analysis. We can see from the list above that players with close similarity scores can end up as a superstar, or out of baseball within a few years. Some players have a stronger drive to succeed. Some get a few good breaks and build on them. Some have the right manager for them, or the right mentor, among the coaching staff. None of these can show through the numbers. The numbers show the potential of the player. The unknown is his ability to realize his potential.

Vernon Wells could develop like a Shaun Green or a Johnny Damon into an all-star. Or he could be Jeffrey Hammonds. It will be fun in 2003 to see how Vernon develops.
What can we expect from Vernon Wells? | 6 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.