Monday, December 20, 2010

Albert Pujols, Omar Infante and the 2011 Prospect Evaluations

How many Omar Infantes would it take to equal one Albert Pujols?


This year is sort of an anniversary for us. It was ten years ago this off-season that I developed what I called ‘Peak Performance Analysis’ and began incorporating it into my prospect evaluations. At the time it was a relatively crude system that used an ‘aging’ formula on normalized Minor League data to project a player’s performance over their 27, 28 and 29yo seasons. At the time, I was the lead writer for a now-defunct prospect website and one of my readers was starting a prospect web-site and published the entire 27yo projections for nearly 500 players using this formula. I have no idea if that site is still around, but I was having a discussion the other day that caused me to go back and look at that ten year old data.

As regular readers are aware, the foundational philosophy for our prospect analysis is the notion of what value a player is likely to produce over his career. While it isn’t quite this straightforward, the concept of expected career Wins Above Replacement (WAR) is the metric we use to operationalize ‘value’. This is relevant, because the discussion I was having had to do with Albert Pujols and Omar Infante—both players that we rated higher than anyone else on our Top 300 list ten years ago (Diamond Futures’ Retrospective: Our 2001 Top 100). Essentially the discussion centered around these players of the same age and their relative trade values. If Albert Pujols is going to produce approximately 8.0 WAR in each of the next few seasons and Omar Infante is going to produce 2.0 WAR, then are four Omar Infantes equal to one Albert Pujols? Before I get to the discussion, remember that prior to the 2001 season, Pujols had just finished his 21yo season with 490 professional ABs and only 14 of them above A-ball…Infante had finished his 19yo season and had then amassed 404 ABs—with none above A-ball. Here were our age 27-29yo projections for them (with actual 27-29yo results for Pujols and 26-28yo results for Infante):

Albert Pujols
Age: 22
Rank: #9
AB – 1623 (1657)
R – 310 (323)
H – 537 (558)
2B – 130 (127)
3B – 6 (2)
HR – 111 (116)
RBI – 384 (354)
SB – 5 (25)
BB – 234 (318)
K – 200 (176)
AVG - .331 (.337)
OBP - .416 (.444)
SLG - .623 (.626)
OPS – 1.039 (1.069)
WAR – 26.8 (27.1)

Omar Infante
Age: 20
Rank: #79
AB – 1155 (991)
H – 181 (134)
2B – 55 (48)
3B – 6 (7)
HR – 13 (13)
RBI – 140 (114)
SB – 15 (9)
BB – 96 (70)
K – 152 (134)
AVG - .304 (.309)
OBP - .357 (.354)
SLG - .395 (.411)
OPS - .752 (.765)
WAR – 6.9 (5.8)

Both players, given their experience levels—and thereby the associated risk to achieve these projections—had similar beta values. While Infante is a bit light in the playing time component, it is quite probable that once his 26yo season is replaced with his 29yo season, after this upcoming year, that both of these projections will turn out to be remarkably accurate. But the key thing is that consistent 4:1 WAR ratio between the two players. Once again, are four Omar Infantes equal to one Albert Pujols?

Looking at the question in a different way, who was the better player Howard Johnson, with 225 HRs, two all-star appearances and three MVP top 10 finishes; or Mark Grudzielanik and his 90 career HRs? Career WAR values 25.0 to 24.3. How about Carlos Lee, with 331 HRs and four all-star appearances or Bernard Gilkey with 118 HRs and no all-star games? Career WAR values 20.5 to 21.4. Therein lies the problem with a WAR approach to prospect analysis. Our contention is that there exists no number of Omar Infantes that would ever be of equal value to Albert Pujols, because 'Pujols-like' upside can only be valued against players with similar upside—which are obviously extremely rare.

Despite being able to objectively quantify most every aspect of our prospect analysis, we still have not been able to satisfactorily develop a single objective measure of prospect value. While you can read about our approach here Do-It-Yourself - Understanding Performance Evaluation and here This Week's Mailbag - Prospect Rankings Questions, in the end we objectively evaluate a player along two distinct dimensions—a player’s expected peak performance value and the certainty of them achieving it. Then we have to take these two objective measures and subjectively determine a value for the combination of the two. Without a lengthy discussion of vector analysis and vector magnitude the problem may not be entirely clear. However, where it relates to our prospect analysis is that we give more relative weight to a player’s upside than we do the certainty of them reaching it. In the end we produce a distribution of expected career WAR values, that lean more heavily to upside. So keep that in mind when you read our prospect rankings over the next two months. The other thing to keep in mind is that, as demonstrated above, we are very comfortable in projecting younger players at the lower levels of the minors. As you can see in this three part analysis of our last year rankings (How Did We Do? Trying to Make a Quantitative Assessment of a Subjective Topic ; How Did We Do? Trying to Make a Quantitative Assessment of a Subjective Topic (part II) ; How Did We Do? Trying to Make a Quantitative Assessment of a Subjective Topic (part III) ) it gives us a leg up on the competition. The result is that while you will read about players like Fabio Martinez-Mesa, Adrian Salcedo and Oswaldo Arcia in other places this year, we were already talking about them last year.

As to what you can expect in regards to our prospect lists, beginning tomorrow we will be posting approximately four teams per week. We will do this in reverse order of organizational strength—just like last year. So over the next two months, you will receive our team-by-team lists for all thirty teams, culminating with our top 300 list in mid-February. Our grading scale is unchanged. We begin with a list of over 5000 Minor League players and winnow that list down to the 2000 players or so that we truly consider to be legitimate Major League prospects. The top 1% of Minor League players earn a grade of ‘A’. The next 1% earn a grade of ‘A-‘ and the third 1% earn a ‘B+’. Prospects that fall in the 4th thru 6th percentile earn a grade of ‘B’ and the 7th thru 10th percentile earn a ‘B-‘.

There are some changes from last year. First, due to our compressed time frame this year, our focus is on the Prospect eGuide (more on the guide in the coming weeks). On the site, you will get write-ups for the top twelve prospects for each team plus a listing of any other players that are considerations for the top 300 list. You will have to purchase the eGuide for more in-depth analysis. Second, we are adding more retrospective analysis on each team’s prospect ‘risers’ and ‘fallers’ from last year’s list. Finally, although you will have to purchase the guide to really take advantage of it, one of the biggest changes to the guide this year is some new metrics that we have developed on ‘organizational’ performance. We hope you will enjoy the content.

3 comments:

  1. great stuff. just wondering, If you just give A's to the top 1% then it is not much different than just ranking them which you already do. However, if you only give A's to the players that deserve an A, (ie. some years there may be 20 "A" prospects while other years there may be 25) then the reader may be able to see that while the 30th overall prospect in 2010 was an A, the 30th overall in 2011 is not as good and is thus an A-

    ReplyDelete
  2. Drew...excellent points. The only problem that I have with that logic is that it assumes that we know as much about evaluating prospect talent in 2005 as we do in 2011. Our processes are continually refined and hopefully improved each year as we find better ways to do things. Perhaps one day we will get to a point where I feel that we have it all figured out and can do just that--but we aren't there yet :)

    ReplyDelete
  3. Wow... your site is great and is filled with valuable information... Your site really helped me to find what info I was looking for and you really know your subject. I will recommend your blog to all of my single friends.

    Online single man from www.talkativetimes.com - Date singles online for free

    ReplyDelete