/cdn.vox-cdn.com/uploads/chorus_image/image/27084585/20130909_kdl_ar7_075.0.jpg)
It is that terrible point in the off-season when nothing is happening and all I want is baseball conversation. So why not make our own? Ryan has started his prospect review series, and I thought I would add to it with some data. I played a role in tweaking the PTM system with Jay a long time ago and I thought I would roll out a streamlined iteration of it here. This is not a prospect ranking. It is simply an attempt to put the performance of our positional prospects in a comparative context, and a context that includes a consideration of the level, age, and position of each player. That's it.
Here is the basic logic:
- Quantify performance (relative to league)
- Factor in age (relative to league)
- Factor in position
- Weight the overall score across the past three seasons (if 3 seasons of data are available)
So here is how I did the above.
Performance - I decided to evaluate performance as a combination of OBP and SLG relative to league averages. Going along with the relative value of these two metrics at the major league level, OBP was weighted at 1.8x that of SLG. This is not necessarily the right way to project prospects, but again, that is not really the goal. More technically, difference between performance and league average were put on a logarithmic scale so as to make the distribution of values symmetrical around the league average value. So in other words:
Performance = 1.8[ln(OBP)-ln(avgOBP)]+[ln(SLG)-ln(avgSLG)]
Age - If you look at a player who performed 10% better than league average in both OBP and SLG, you end up with a value of about .25. I used this as a standard to develop age adjustments (pretty arbitrarily, though it is easy to increase or decrease these values accordingly). I made 1-year equal to 5%, so a player 2 years younger than league average got a bump in their "score" equal to +10% of league average. I capped this at a 3-year advantage, so a guy like Lindor, who was almost 5 years below league average age at AA, only received a 3-year bump (+15%).
Position - I used the fairly standard positional adjustment spread that can be found at places like fangraphs. I again somewhat arbitrarily scaled this adjustment to a similar scale as age. Catcher, the most defensively valuable position, got scored as +10%. First baseman/DH got the opposite, -10%. The other positions lined up in between, shortstop and 2B/3B/CF positive, corner OF spots negative.
3-year weighting - I weighted each season's data progressively, up to as many as three years in the past. If only one season was available, obviously that season counted for everything. If two seasons of data were available, I weighted the most recent season double that of the season two years ago. If three seasons of data were available, they were weighted 3:2:1.
A few other caveats. I did not use any short-season data. This means that no one without a full season at class A or higher got counted (so no Clint Frazier). I did not factor in playing time, aside from not including seasons that constituted fewer than 50 games played. Finally, for players promoted mid-season, I weighted each partial season accordingly.
I'm not terribly tied to the scale of my age and position adjustments, or the weighting of OBP vs. SLG. If someone wants to make a good argument for why I should change them, I'm happy to listen.
Some background league average data
Level | 2013 | 2012 | 2011 |
AAA (age) | 26.9 | 27.5 | 26.8 |
OBP | 328 | 328 | 329 |
SLG | 384 | 389 | 400 |
AA (age) | 24.4 | 24.4 | 24.3 |
OBP | 330 | 330 | 329 |
SLG | 387 | 392 | 395 |
A+ (age) | 22.8 | 22.6 | 22.8 |
OBP | 332 | 327 | 322 |
SLG | 380 | 390 | 379 |
A (age) | 21.3 | 21.6 | 21.6 |
OBP | 328 | 326 | 323 |
SLG | 375 | 379 | 370 |
These data were the baseline league average data off of which individual year-by-year performance scores were calculated.
One thing I like about this approach is that you can consider the relative importance of each component for a player's overall performance score. Some players, you will see below, have essentially league average performance, but because they are young and/or play a defensively challenging position, have a high score. Others have a strong statistical performance, but because they aren't young and are defensively limited, their overall score is severely reduced. For a guy already in AAA, that might not matter as much, because you might decide performance is the main thing that matters that close to the majors. Or not. Other guys might be putting up decent scores, but still waiting to actually break out on a performance level.
Again, this is not valuable as a way of actually assessing the worth of these players as prospects or their likely success. But this is a convenient way to put their actual performance into perspective given league average data. And that, in and of itself, can be pretty useful. I started with positional players because, frankly, they are a lot easier. Pitchers present a much more significant challenge and one I don't feel all that confident in addressing (though I might try at some point).
The list
- Francisco Lindor (SS) composite 0.755 [2013 - 0.833, 2012 - 0.598, 2011 - n/a]: Lindor gets the max age bump each season, plus the bump for playing SS. Even without those bumps, however, his statistical performance in 2013 was extremely good, elevated by his fantastic OBP at both Carolina and Akron.
- Jose Ramirez (2B/SS/3B) composite 0.523 [2013 - 0.344, 2012 - 0.880, 2011 - n/a]: I didn't factor in Ramirez's MLB time in September, but you can see why I suggested Ramirez's offensive ceiling might be higher than we saw in 2013. Ramirez had a spectacular (maybe BIP inflated) offensive season in Lake County in 2012.
- Jake Lowery (C) composite 0.521 [2013 - 0.625, 2012 - 0.312, 2011 - n/a]: This is who everyone expected, right? Lowery gets a big bump for being catcher (though I did not give him the full positional adjustment), but his 2013 offensive season was impressive in Akron...especially for a catcher, especially given his age.
- Luigi Rodriguez (CF) composite 0.509 [2013 - 0.547, 2012 - 0.434, 2011 - n/a]: I didn't ding Rodriguez for time lost to injury, though he barely qualified. When he has played, though, he was very good this past season.
- Joe Wendle (2B) composite 0.495 [2013 - 0.495, 2012 - n/a, 2011 - n/a]: Wendle has a negative age adjustment that more than erases the slight bump he gets for playing 2B. But Wendle had the best individual offensive performance of any prospect last season.
- Eric Haase (C) composite 0.474 [2013 - 0.474, 2012 - n/a, 2011 - n/a]: Oh, to be young and a catcher. Maybe my catcher adjustment is too severe...or maybe not. Haase was a 20-year old catcher putting up good numbers in his first full-season exposure.
- Tony Wolters (C/2B) composite 0.410 [2013 - 0.466, 2012 - 0.296, 2011 - n/a]: Another catcher (sort of...I didn't give him the whole bump in 2013). Wolters will be another interesting guy to watch this season. (If he had played enough to be eligible, Alex Monsalve, another young catcher, would have slotted in here)
- Logan Vick (CF) composite 0.338 [2013 - 0.338, 2012 - n/a, 2011 - n/a]: This one surprised me. Vick was old for Lake County last year, and only age average in Carolina, but he put up a great OBP at both stops. He'll need to continue to perform well if he wants to keep his status, though.
- Ronny Rodriguez (SS/3B) composite 0.282 [2013 - 0.220, 2012 - 0.392, 2011 - 0.247]: Rodriguez had a down season this year, but even a year ago his numbers are brought down by his well below-average OBP. This score is all about age and position.
- Tyler Naquin (CF/OF) composite 0.279 [2013 - 0.279, 2012 - n/a, 2011 - n/a]: Our favorite example. Naquin was slightly young and gets a little positional bump, but in general, his offensive performance was not incredible.
Six more:
Dorssys Paulino (0.199), Carlos Moncrief (0.196), Jesus Aguilar (0.189), Giovanny Urshela (0.149), Jordan Smith (0.088), Erik Gonzalez (0.058).
Not eligible due to playing time/level: Alex Monsalve, Levon Washington, Nellie Rodriguez, Francisco Mejia, Clint Frazier, etc.
So there it is. Not surprisingly, the list is almost entirely middle of the diamond talent. Lindor is well ahead of the curve, followed by a cluster of four or five B/B- guys, then a set of guys laid out behind them.
.